Split (1)

train · 901k rows

article_title stringlengths 1 1.44k	abstract stringlengths 10 22.8k	contributor stringlengths 29 38.3k	journal stringlengths 3 246	full_text stringlengths 22 2.53M
Left Ventricular Pacing In Patients With Congestive Heart Failure	<p>Cardiac resynchronisation therapy (CRT) using biventricular (BIV) pacing has proved its effectiveness to correct myocardial asynchrony and improve clinical status of patients with severe congestive heart failure (CHF) and widened QRS. Despite a different effect on left ventricular electrical dispersion, left univentricular (LV) pacing is able to achieve the same mechanical synchronisation as BIV pacing in experimental studies and in humans. This results in clinical benefits of LV pacing at mid-term follow-up, with significant improvement in functional class, quality of life and exercise tolerance at the same extent as those observed with BIV stimulation in non randomised studies. Furthermore these benefits are obtained at lesser costs and with conventional dual-chamber devices. However, LV pacing has to be compared to BIV pacing in randomised trials before being definitely considered as a cost-effective alternative to BIV pacing.</p>	<contrib contrib-type="author"><name><surname>Etienne</surname><given-names>Yves</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Fatemi</surname><given-names>Marjaneh</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Blanc</surname><given-names>Jean-Jacques</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Despite advances in drug treatment, congestive heart failure (CHF) remains a major health care problem associated with a poor quality of life and a high mortality rate. During the past decade, cardiac resynchronisation therapy (CRT) using biventricular (BIV) pacing emerged as a promising technique improving quality of life, exercise tolerance and mortality [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R3">3</xref>], and is now an admitted therapy in a selected population of patients with widened QRS and severe CHF despite optimal pharmacological therapy [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R5">5</xref>]. It has been well documented that intraventricular conduction disturbances, namely left bundle branch block, induce left ventricular (LV) activation delay leading to discoordinated myocardial contraction and, consecutively, to deleterious effects on myocardial systolic function, diastolic filling time and mitral regurgitation [<xref ref-type="bibr" rid="R6">6</xref>]. Finally this dyssynchronised LV function exaggerates clinical symptoms of CHF. The rationale for CRT is based upon the hypothesis that, by correcting inter and intra-ventricular asynchrony, BIV pacing improves LV function and favourably affects clinical condition and prognosis of patients with severe CHF and prolonged QRS. However it has never been demonstrated that electrical resynchronisation induced by BIV pacing is a sine qua non condition to achieve better mechanical coordination as, in experimental data, and despite electrical dispersion, LV pacing does as well as BIV pacing on mechanical synchronisation with similar hemodynamic benefit [<xref ref-type="bibr" rid="R7">7</xref>].</p><p>Many other arguments support the concept that LV univentricular pacing by reversing intraventricular dysynchrony is sufficient in humans to improve LV function and clinical status at the same extent and at a lower cost-effective ratio than BIV pacing and this is in accordance with the fact that intraventricular asynchrony seems more relevant that interventricular asynchrony to predict prognosis in patients with dilated cardiomyopathy [<xref ref-type="bibr" rid="R8">8</xref>].</p></sec><sec sec-type="" id="s2"><title>Hemodynamic effects of pacing</title><p>The first clinical series with CRT were performed empirically using BIV pacing [<xref ref-type="bibr" rid="R9">9</xref>]. However, since this period , several studies in humans have reported similar hemodynamic benefit using either LV or BIV pacing during acute studies, showing a decrease in pulmonary pressures, an increase in cardiac outpout, systemic blood pressure, and dP/dt when compared to baseline or to right ventricular pacing in patients with severe CHF and intraventricular conduction delays [<xref ref-type="bibr" rid="R10">10</xref>-<xref ref-type="bibr" rid="R12">12</xref>];. Some authors reported that LV stimulation had even a greater effect on LV dP/dt than BIV [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R12">12</xref>]. This hemodynamic benefit was independent of atrio-ventricular conduction as the same effects were observed in patients in sinus rhythm or atrial fibrillation [<xref ref-type="bibr" rid="R13">13</xref>]. Furthermore, improvement in myocardial contractility induced by LV pacing was obtained with the same extent of reduced energetic cost for the myocardium than with BIV pacing [<xref ref-type="bibr" rid="R14">14</xref>]. Finally, a similar reduction in mitral regurgitation was observed by echocardiography using the two pacing modes with a significant decrease by 30 to 50 % in mitral jet area or mitral regurgitant orifice area in non-randomised studies [<xref ref-type="bibr" rid="R15">15</xref>-<xref ref-type="bibr" rid="R17">17</xref>], despite less striking effects upon LV reverse remodeling during LV versus BIV pacing in some studies [<xref ref-type="bibr" rid="R15">15</xref>-<xref ref-type="bibr" rid="R17">17</xref>].</p></sec><sec sec-type="" id="s3"><title>Clinical benefits</title><p>Whether or not these hemodynamic benefits result in sustained improvement of clinical status during permanent LV stimulation has not been extensively investigated until now. However, some studies reported significant improvement in both exercise tolerance and quality of life at mid-term follow-up during chronic LV pacing [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>] or failed to demonstrate any superiority of BIV pacing over LV pacing on clinical data[<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R20">20</xref>]. Aurricchio et al compared 3 months of active LV pacing to 3 months of inactive pacing in 86 patients with severe CHF and wide QRS duration. They observed a significant increase in peak V0<sub>2</sub> (2.46 ml/mn/kg), in 6 minute walk distance (47 m) and a decrease in quality of life score (8.1) during the active pacing period in the group of patients with the longest QRS duration (above 150 ms) [<xref ref-type="bibr" rid="R18">18</xref>]. In another study comparing 2 active periods (4 weeks each) of uniLV and BIV pacing to baseline in a cross-over design, the same authors reported similar improvement in functional class, quality of life score, and exercise tolerance in the 2 pacing modes [<xref ref-type="bibr" rid="R20">20</xref>]. Our personal studies lead to the same conclusions: by comparing to baseline, we found after 12 months of LV pacing in 22 patients with severe CHF, sinus rhythm and left bundle branch block a significant improvement in NYHA functional class by 40%, exercise tolerance (6 minute walk distance by 30 % and peak V0<sub>2</sub> by 26 %) and norepinephrine level by 37% [<xref ref-type="bibr" rid="R19">19</xref>]. Again and in agreement with the previous studies, a similar improvement in functional class and exercise performance was found in 2 groups of patients (BIV: 12 pts - LV: 14 pts) after 6 months follow-up in a non-randomised study, although, as expected, a significant decrease in QRS duration (-19 ms) was showed only in the BIV pacing group [<xref ref-type="bibr" rid="R17">17</xref>].</p></sec><sec sec-type="" id="s4"><title>LV pacing and ventricular asynchrony</title><p>Are these clinical effects correlated with improvement in myocardial synchrony? This is an important issue on a physiopathological point of view and for the selection of candidates for LV pacing. In an interesting experimental study using a canine model of cardiac failure and left bundle branch block, Leclercq et al examined mechanical and electrical synchrony during right atrial, LV and BIV stimulation; as expected, electrical dispersion decreased during BIV pacing whereas it increased during LV pacing; however, despite this opposite action upon electrical phenomena, the same improvement in mechanical coordination within the LV, measured by tagged magnetic resonance imaging, was observed with LV and BIV pacing, and this was correlated with improvement in hemodynamics (25 % increase in dP/dt and aortic pulse pressure) [<xref ref-type="bibr" rid="R7">7</xref>]. Similarly, in humans, LV dyssynchrony measured either by echocardiographic phase analysis, echo-contrast, or tissue Doppler imaging was improved significantly and almost at the same extent during LV and BIV pacing [<xref ref-type="bibr" rid="R16">16</xref>,<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R22">22</xref>]. The decrease in the septal to lateral contraction delay during LV pacing was particularly correlated with an increase in LV ejection fraction or dP/dt [<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R22">22</xref>] .</p></sec><sec sec-type="" id="s5"><title>Cost effectiveness</title><p>By comparison to BIV pacing, LV pacing offers several technical and economical advantages : 1) if insertion of a third lead (the right ventricular one) is not the most difficult part of the operative procedure, it could prolong significantly its duration and the X-ray exposure; 2) the presence of 3 leads instead of 2 induces certainly more adverse events and increases the risk of venous thrombosis or mechanical complications; 3) more importantly, uniLV pacing needs only a less complex to program and less often subject to dysfunction conventional dual-chamber pacemaker. Finally, BIV pacing is more expensive than LV pacing as it implies implantation of a supplementary right ventricular lead and of a specific and more costly device: in Europe, the extra cost could be approximately evaluated at 30 % and this must be taken in account in the context of the growing “epidemic” of heart failure consecutive to demographic ageing and of the control of medical expenses in many countries.</p><p>So, at this time, there are no arguments in favour of the hypothesis that, for patients with severe CHF and enlarged QRS duration, BIV pacing would be significantly more effective than LV pacing in terms of mechanical coordination, hemodynamic and clinical benefits. However, our knowledge about the effects of LV pacing still suffers some limitations: the present studies are only observational and non-randomised. Furthermore, it seems, as shown during acute hemodynamic studies, that in some patients LV pacing would not be as effective as BIV pacing (but the opposite is true also… ), and this could perhaps be the case in patients with a dysfunctional right venricle but no study have focused on this topic at this time. Finally and unlike BIV pacing [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R23">23</xref>], no study have examined the effects of LV pacing on mortality and morbidity, and whether or not LV pacing improves mortality in severe CHF remains to be demonstrated. However, the ongoing trials comparing LV to BIV pacing are not designed to show that one of these pacing modalities is better than the other but just equivalent… an argument to think that they are really similar.</p></sec>
Biological Pacemakers	<p>Genetically engineered pacemakers could be a possible alternative to implantable electronic devices for the treatment of bradyarrhythmias. The strategies include upregulation of beta adrenergic receptors, conversion of myocytes into pacemaker cells and stem cell therapy. Pacemaker activity in adult ventricular myocytes is normally repressed by the inward rectifier potassium current (I<sub>K1</sub>). The I<sub>K1</sub> current is encoded by the Kir2 gene family. Use of a negative construct that suppresses current when expressed with wild-type Kir2.1 is an experimental approach for genesis of genetic pacemaker. Hyperpolarisation activated cyclic nucleotide gated (HCN) channels which generate If current, the pacemaker current of heart can be delivered to heart by using stem cell therapy approach and viral vectors. The unresolved issues include longevity and stability of pacemaker genes, limitations involved in adenoviral and stem cell therapy and creation of genetic pacemakers which can compete with the electronic units.</p>	<contrib contrib-type="author"><name><surname>Rajesh</surname><given-names>G</given-names></name><degrees>MBBS, MD, DNB, DM</degrees></contrib><contrib contrib-type="author"><name><surname>Francis</surname><given-names>Johnson</given-names></name><degrees>MBBS, MD, DM, FCSI</degrees></contrib><aff> Department of Cardiology, Medical College Hospital, Calicut, Kerala, India</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Implantable electronic pacemakers remain the treatment of choice for high degree atrioventricular blocks and sinus node dysfunction. The shortcomings of electronic pacemakers include limited battery life, need for lead implantation into heart and lack of response to autonomic and physiologic demands on the heart. Molecular approaches to the development of a biological pacemaker are a conceptually attractive alternate treatment modality for heart blocks. The approaches attempted to provide such pacemaker function include up regulation of β<sub>2</sub> adrenergic receptors [<xref ref-type="bibr" rid="R1">1</xref>], down regulation of K+ current I<sub>K1</sub> [<xref ref-type="bibr" rid="R2">2</xref>] and over expression of HCN2 (hyperpolarisation activated cyclic nucleotide gated) channels the molecular correlate of the endogenous cardiac pacemaker current I<sub>f</sub> [<xref ref-type="bibr" rid="R3">3</xref>]. The genetic treatment can be applied to heart by plasmid injection, use of viral constructs or stem cell therapy [<xref ref-type="bibr" rid="R4">4</xref>] [<xref ref-type="bibr" rid="R5">5</xref>].</p></sec><sec sec-type="" id="s2"><title>Molecular targets for genesis of biological pacemaker</title><sec id="s2a"><title>β<sub>2</sub> adrenergic receptors</title><p>The sinus node has a higher density of β adrenergic receptors (βAR) compared with surrounding atrium [<xref ref-type="bibr" rid="R1">1</xref>]. This density of βAR and its regulation of If current suggest that increases in the density of βAR in the vicinity of the sinus node may lead to an increase in heart rate. The up regulation of β<sub>2</sub> adrenergic receptors can be achieved by plasmid injection into heart. It was noted that after injection of plasmids in porcine right atrium heart rates were 50% faster than those of controls. One potential limitation of this strategy is that the diseased endogenous cardiac pacemaker mechanisms are left intact and the β<sub>2</sub> receptor is used as a nonspecific stimulator of heart rate. It can influence other catecholamine sensitive channels also.</p></sec><sec id="s2b"><title>HCN channel and I<sub>f</sub> current</title><p>Action potential of pacing cells is unique in that they have a slow depolarizing phase, rendering them spontaneously active [<xref ref-type="bibr" rid="R6">6</xref>]. The depolarization involves interaction between HCN channels and L & T type calcium channels. The modification of these channels is a therapeutic target.</p><p>HCN channels generate I<sub>f</sub> current which contribute to genesis of pacemaker activity. I<sub>f</sub> channel is activated on membrane hyperpolarisation rather than on depolarization [<xref ref-type="bibr" rid="R7">7</xref>]. It has four fold selectivity for K+ than Na+. The typical features of I<sub>f</sub> current include activation by hyperpolarized membrane potential, conduction of Na+ and K+, modulation by cyclic adenosine monophospate (CAMP) and blockade by cesium (Cs+) [<xref ref-type="bibr" rid="R8">8</xref>]. HCN generated current also has the above features. Four different HCN genes have been identified [<xref ref-type="bibr" rid="R9">9</xref>]. HCN1 is the most rapidly acting channel, HCN4 the slowest with HCN2 and 3 possessing intermediate kinetics [<xref ref-type="bibr" rid="R10">10</xref>]. HCN1, 2 and 4 have been found to express in adult heart, HCN4 being the most highly expressed one in SA node. HCN2 expression was noted in atrium, ventricle and SA node.</p><p>HCN can be delivered to heart by adenoviral construct or using stem cells. The nucleic acids delivered by adenoviruses do not integrate into genome as they are episomal. Stem cell therapy may be more promising than viral strategy. The approach using HCN may be less problematic and proarrythmic as it incorporates the endogenous pacemaker channel gene, which selectively activates only during diastole [<xref ref-type="bibr" rid="R14">14</xref>].</p></sec><sec id="s2c"><title>Inward Rectifier Potassium Current (I<sub>K1</sub>)</title><p>I<sub>K1</sub> and other background K+ selective currents contribute to action potential depolarization and establish diastolic resting membrane potential. Down regulation of the background K+ current I<sub>K1</sub> is one of the approaches attempted to provide pacemaker function. Genetic suppression of I<sub>K1</sub> can converts quiescent myocytes into pacemaker cells.</p><p>I<sub>K1</sub> is the classical inward rectifier potassium current. Inwardly rectifying K+ channels (Kir) are responsible for stabilizing the resting membrane potential. Inward rectification is a phenomenon in which conductance of a Kir channel increases with hyperpolarisation but decreases with depolarization. Rectification in Kir channels results from voltage dependent channel block by intracellular cations [<xref ref-type="bibr" rid="R12">12</xref>]. IK1 is absent or poorly expressed in sinus and AV nodal cells. Native I<sub>K1</sub> in human ventricular myocytes is reduced by adrenergic receptor stimulation.</p><p>It was observed that a dominant negative strategy to reduce I<sub>K1</sub>, which usually maintain ventricular myocytes at negative membrane potentials, induced spontaneous impulse initiation in guinea pig heart. The inward rectifier potassium current is encoded by Kir2 gene family. Replacement of 3 amino acid residues in the pore structure of Kir2.1 creates a dominant negative construct [<xref ref-type="bibr" rid="R12">12</xref>]. Downregulation of I<sub>K1</sub> removes an important determinant of repolarisation leading to prolonged repolarisation in cells lacking this current [<xref ref-type="bibr" rid="R13">13</xref>]. This may result in excessive dispersion of repolarisation leading to theoretical risk of proarrhythmia.</p></sec></sec><sec sec-type="" id="s3"><title>Stem cell therapy</title><p>Human embryonic stem cells can be used to create pacemakers or adult mesenchymal stem cells may be used as platforms for delivery of pacemaker genes to myocardium. The advantage of these cells includes their ability to make functional gap junctions and generate spontaneous rhythms [<xref ref-type="bibr" rid="R15">15</xref>]. The approach using embryonic stem cells carry the problems of identifying appropriate cell lineages, possibility of stem cell differentiation into lines other than pacemaker cells, and potential for neoplasia. Adult mesenchymal stem cells are biologically inert vectors which can deliver genetic information to myocardium. Human mesenchymal stem cells (hMSCs) as a platform for delivery of genes into heart is a more attractive option because they can be obtained in large numbers, easily expanded in culture, capable of long term transgene expression and their administration can be autologous or via banked stores [<xref ref-type="bibr" rid="R15">15</xref>].</p></sec><sec sec-type="" id="s4"><title>Gene therapy versus stem cell therapy</title><p>In gene therapy a cardiac myocyte is converted into a pacemaker cell whereas in stem cell therapy myocytes retain their original function. An inherent problem of gene therapy is use of viruses. Replication deficient adenoviruses with little infectious potential lead to only transient improvement in pacemaker function. Retroviruses may be carcinogenic and infective.</p></sec><sec sec-type="" id="s5"><title>Important studies on biological pacemakers</title><sec id="s5a"><title>1. Molecular transfer of the human β<sub>2</sub> Adrenergic receptor cDNA</title><p>Effects of transferring the human β<sub>2</sub> adrenergic receptor were studied by Edelberg JM et al [<xref ref-type="bibr" rid="R1">1</xref>] in chronotropy studies with isolated myocytes, and transplanted as well as endogenous murine heart. Murine embryonic cardiac myocytes were transiently transfected with plasmid constructs. The total percentage of spontaneously contracting myocytes was greater in β<sub>2</sub>AR transfected cells compared with controls. Also the percentage of myocytes with chronotropic rates more than 60 beats per minute was greater in β<sub>2</sub>AR population than controls. To study the ex vivo effects of targeted expression of β<sub>2</sub>AR a murine neonatal cardiac transplantation model was used. Injection of β<sub>2</sub>AR construct increased the heart rate by 40%. These studies demonstrate that local targeting of gene expression may be a feasible modality to regulate the cardiac pacemaking activity.</p></sec><sec id="s5b"><title>2. Local expression of HCN2 in canine left atrium</title><p>Research by Jihong Qu et al [<xref ref-type="bibr" rid="R13">13</xref>] showed that HCN2 over expression provides an I<sub>f</sub> - based pacemaker current sufficient to drive the heart when injected into a localized region of atrium. Adenoviral constructs of mouse HCN2 and green fluorescent protein (GFP) or GFP alone were injected into LA, terminal studies performed 3-4 days later, myocytes examined for native and expressed pacemaker current ( I<sub>f</sub>). Spontaneous LA rhythms occurred after vagal stimulation-induced sinus arrest in 4 of 4 HCN2 + GFP dogs and 0 of 3 GFP dogs (P<0.05).</p></sec><sec id="s5c"><title>3. Biological pacemaker implanted in canine left bundle branch</title><p>Alexi N. Plotnikov et al [<xref ref-type="bibr" rid="R14">14</xref>] studied the effect of administration of the HCN2 gene to the left bundle branch system of dogs. An adenoviral construct incorporating HCN2 and green fluorescent protein (GFP) as a marker was injected via catheter under fluoroscopic control into the posterior division of the LBB. Controls were injected with an adenoviral construct of GFP alone or saline. During vagal stimulation, HCN2 injected dogs showed rhythms originating from the left ventricle, the rate of which was significantly more rapid than controls.</p></sec><sec id="s5d"><title>4. Human mesenchymal Stem Cells as a gene delivery system to create cardiac pacemaker</title><p>Potapova I et al [<xref ref-type="bibr" rid="R3">3</xref>] tested the ability of human mesenchymal stem cells to deliver a biological pacemaker to the heart. hMSCs transfected with a cardiac pacemaker gene, mHCN2, by electroporation expressed current as I<sub>f</sub>- like. They demonstrated that genetically modified hMSCs can express functional HCN2 channels in vitro and in vivo, mimicking over exression of HCN2 genes in cardiac myocytes, and represent a noval delivery system for pacemaker genes into the heart or other electrical syncytia.</p></sec></sec><sec sec-type="" id="s6"><title>Limitations of approaches to development of biological pacemaker</title><p>Use of viruses to deliver the necessary genes has inherent problems. Replication deficient adenoviruses that have little infectious potential lead to only transient improvement in pacemaker function as well as potential inflammatory responses. Retroviruses carry a risk of carcinogenicity and infectivity. Limitations of stem cell therapy include immunogenicity of cell, the potential for neoplasia, proper engineering of pure cardiac lineages and spatial non uniformity of implants. Regulating the level of expression to achieve optimal pacemaker rate is critical. Biological pacemaker needs an optimal cell mass and optimal cell-cell coupling for long term normal function. Research is ongoing to identify optimal cell numbers and coupling ratios needed to optimize the function of biological pacemakers.</p><p>A major issue is duration of efficacy of biological pacemakers. The duration of pacemaker function in approaches using viruses depend on how long the viruses and resulting protein constructs survive in the host. To ensure long term function the appropriate delivery system in which the construct is effective for long periods must be identified. What will be the longevity and stability of next generation of pacemaker genes?</p><p>The onset of pacemaker function after a pause following the last intrinsic beat is a critical factor. Can a pacemaker gene inserted into proximal conduction system create a functioning biological pacemaker which can drive the ventricle in demand mode when the sinus node signal fails? This requires proper engineering of genes. Considering the cell-cell coupling differences in gene therapy and stem cell therapy, the engineering of mutant genes will differ importantly between approaches.</p><p>The autonomic responsiveness of biological pacemakers, the ideal site for implantation, the extent of recovery of diseased sinus node and the ideal construct to be preferred remain unanswered questions. None of the studies tested whether a biological pacemaker could be engineered into the ventricular conducting system. Will the functional characteristics of biological pacemakers compete with that of electronic units available?</p></sec>
P Wave Dispersion is Increased in Pulmonary Stenosis	Could not extract abstract	<contrib contrib-type="author"><name><surname>Ozmen</surname><given-names>Namik</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Cebeci</surname><given-names>Bekir Sitki</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Kardesoglu</surname><given-names>Ejder</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Celik</surname><given-names>Turgay</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Dincturk</surname><given-names>Mehmet</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff3">‡</xref></contrib><contrib contrib-type="author"><name><surname>Demiralp</surname><given-names>Ergun</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff3">‡</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>P wave dispersion (PWD) indicates the spreading of the sinus stimulus in the atria. geneity in the depolarisation of the atria and thus causes atrial fibrillation. Stretching of the atria due to pressure and/or volume load, electrolyte imbalance or increase in sympathetic activity are the main causes that increase PWD. It is recognised that PWD and hence the probability of atrial fibrillation are increased in pathological conditions that increase the left atrial pressure such as mitral stenosis, aortic stenosis and systemic hypertension [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R3">3</xref>] or in pathological conditions that increase the right atrial pressure such as atrial septal defect [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R5">5</xref>] and COPD (Chronic Obstructive Pulmonary Disease) [<xref ref-type="bibr" rid="R6">6</xref>]. On the other hand, chronic renal insufficiency accompanied by intravascular volume load and electrolyte imbalance also increase PWD [<xref ref-type="bibr" rid="R7">7</xref>].</p><p>In atrial septal defect or COPD, the right atrial pressure increases, and the right atrium stretches and becomes more enlarged. Thus, the depolarisation duration of the right atrium is prolonged and this causes an increase in PWD. Valvular pulmonary stenosis (PS) results from the fusion of pulmonary cusps in the intrauterine period and constitutes 7% of congenital heart diseases. Pulmonary stenosis which is a relatively rare congenital heart disease causes concentric hyperthrophy of the right ventricle, increase in end-diastolic pressure and increase in right atrial pressure. This condition can be expected to lead an inrease in PWD. P wave dispersion in PS patients was not defined previously. In this study, we have investigated PWD in patients with valvular PS.</p></sec><sec sec-type="materials\|methods" id="s2"><title>Material and Methods</title><p>A total of 42 patients with valvular PS diagnosed by echocardiography performed after cardiac examination were admitted into the study. 33 apparently healthy age-matched subjects were included to the study as the control group. Healthy status was documented using physical examination, ECG, chest X-ray and echocardiography. Patients with any acquired or congenital heart disease, valvular disease, pericardial disease, bronchial asthma, COPD or any systemic disease were excluded from the study. All patients provided written consent and the study was approved by the local Ethics Committee.</p><sec id="s2a"><title>Echocardiography</title><p>The left and right ventricles and their valvular structures were assessed in detail using the standart left lateral decubitus position Vingmed system FiVe with 2.5 MHz probe (GE, Holten, Norway). In the apical four chamber image, the diameter of the right atrium (RAD) and the diameter of the left atrium (LAD) were measured at the level of the annulus of mitral and tricuspid valves in millimeters (mm), respectively defined as the distance from the lateral wall of the right atrium to the interatrial septum, and from the lateral wall of the left atrium to the interatrial septum. The pulmonary valve was structurally assessed in the parasternal short axis image. Then, the mean and peak pressure gradients over the pulmonary valve were obtained using CW Doppler. Valvular pulmonary stenosis was defined as a dysplastic or dismorphic pulmonary valve with a peak systolic pressure gradient (PG) more than 20 mm Hg and a mean pressure gradient more than 10 mm Hg.</p></sec><sec id="s2b"><title>Electrocardiography (ECG)</title><p>Twelve-lead surface ECG was obtained from all patients and controls at a paper of 50 mm /s with 1 mV/cm standardization after resting 5 minutes. Subjects were allowed to breathe freely but not to speak or cough during recording. The P wave duration was manually measured by a cardiologist blind to the study, using a magnifying lens. The onset of the P wave was defined as the junction between isoelectric line at the beginning of the P wave deflection and the offset of the P wave and the isoelectric line. The leads with the onset or offset of the P wave not clearly determined were excluded from the analysis. 2 patients whose ECGs not clear enough were also excluded from the study. This method was used in previous studies by different investigators [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R9">9</xref>]. P maximum (p max) duration and P minimum (p min) duration were measured from 12-lead ECG. Then, PWD was defined as the difference between p max and p min.</p></sec><sec id="s2c"><title>Statistical assessment</title><p>The data from the patient and control groups were compared with each other using SPSS 11.0 statistical software package (SPSS Inc., Chicago, IIinois, USA) program. PWD, Pressure gradient, right and left atrium diameters were compared with the Mann-Whitney U test, and the correlation between PWD and the other variables was seeked using the Pearson correlation co-efficient. The data were expressed as mean ± SD. P values under 0.05 were considered statistically significant.</p></sec></sec><sec id="s3"><title>Results</title><p>Forty two patients with valvar PS were included in the study; 35 were males, 7 were females and the mean age was 22.1 ± 4.2 years. The control group was comprised of 33 subjects; 23 were men, 10 were women and the mean age was 23.8 ± 3.1 years. In the PS group, only 5 patients complained of palpitations, but rare atrial premature beats were detected, and atrial fibrillation was not observed in any patient. There was not any statistical significance between groups as to the mean ages or heart rate (82.4 ± 8.5 vs 78.2 ± 7.2, p>0.05).</p><p> While the mean pressure gradient at the pulmonary valve was 43.1 ± 18.8 mmHg in the PS group, it was 8.4 ± 4.5 mmHg in the control group (p=0.01). The average right atrium diameter was 38.7 ± 3.9 mm in PS group while it was 30.2 ± 2.5 mm in control group (p=0.02). However, the average left atrium diameter was 32.2 ± 4.1 mm in PS group and 30.2 ± 2.8 mm in control group, and there was no significant difference (p=0.64). According to this, while the right atrium was significantly larger in PS patients, the left atrium diameter was not different from that of the control group (<xref ref-type="table" rid="T1">Table 1</xref>).</p><p>While p max was 107.1 ± 11.5 ms in patients with PS, it was 98.2 ± 5.1 ms in the contol group (p=0.01), and while PWD was 40.4 ± 1.2 ms in PS group, it was 27.2 ± 9.3 ms in the control group (p=0.01). Both p max and PWD values were significantly greater in the PS group. However, p min was similar in both groups (66.1±1.2 ms vs 71.2 ±4.3 msn p>0.05) (<xref ref-type="table" rid="T2">Table 2</xref>).</p><p>In the pulmonary stenosis group, PWD, p max and mean pressure gradient and the right atrium diameter were significantly greater than control group. A correlation analysis of PWD with pressure gradient and right and left atria diameters was performed; while PWD was correlated with pressure gradient (r = 0.38, p = 0.009), there was no correlation between PWD and the right and left atrium diameters ( r = - 0.05, p= 0.74 and r=0.15, p= 0.30, respectively).</p></sec><sec id="s4"><title>Discussion</title><p>The main findings of our study are as follows: 1. PWD is increased in pulmonary stenosis. 2. PWD shows correlation with the pressure gradient. 3. PWD is not greatly affected by the right atrial diameter in PS patients. The low AF incidence in PS patients may be attributed to the intervention in this patient group before PWD is further increased.</p><p>Increased PWD is an electrocardiographic marker that has been associated with the inhomogeneous and discontinuous propagation of sinus impulses. It can be defined as the difference between maximum and minimum p-wave duration. Prolongation of intra-atrial and interatrial conduction time and inhomogenous prolongation of sinus impulses are known electrophsiologic characteristics of atria prone to fibrillation [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R10">10</xref>]. Moreover, the correlation between the presence of intra-atrial conduction abnormalities and the induction of paroxysmal atrial fibrillation have been documented [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R12">12</xref>]. Also, the increase in the pressure or volume load of the atria, stretching of the atria, electrolyte imbalance and increase in sympathetic activity cause PWD to increase. Stretching and enlargement of the atria, disorganisation of the atrial muscles even if partial, and fibrosis in the atrial wall lead to the prolongation of atrial conductance and impairment of its homogeneity.</p><p>It is known that atrial conduction is impaired and PWD is prolonged in haemodialysis patients as a result of intravascular volume load and electrolyte imbalance [<xref ref-type="bibr" rid="R7">7</xref>]. Reduction of intravascular volume with diuretic use, on the other hand, decreases the duration of p max and PWD [<xref ref-type="bibr" rid="R13">13</xref>]. However, it is evident that not volume load but pressure load is involved in PS.</p><p>On the other hand, PWD and the duration of p max are increased also in aortic stenosis and mitral stenosis in which the pressure load in the left atrium is increased. Turhan et al [<xref ref-type="bibr" rid="R2">2</xref>], showed that p max and PWD were higher than in healthy subjects. However, they reported that there was no difference with respect to echocardiographical parameters. In another study by Turhan et al. again, they showed that in mitral stenosis, a prominent reduction occurred in p max, PWD, the diameter of the left atrium and the transvalvular pressure gradient with mitral baloon valvuloplasty 24 hours after and one month after the procedure [<xref ref-type="bibr" rid="R1">1</xref>]</p><p>It is observed that p max is prolonged in both the mentioned studies and our study. This is interpreted as an indicator of interatrial conduction impairment independent of the atrium diameter [<xref ref-type="bibr" rid="R14">14</xref>]. Likewise, Dilaveris [<xref ref-type="bibr" rid="R10">10</xref>] and Ishimoto et al. [<xref ref-type="bibr" rid="R15">15</xref>] reported that there was no correlation between p max and PWD and the left atrium diameter. In our study, although the right atrium diameter was increased, we showed that PWD was not correlated with the right or left atrium diameter but correlated with the pulmonary valve pressure gradient. As in the previous studies, p min did not differ between the two groups in our study either.</p><p>Guray et al. showed that p max and PWD were prolonged in patients with ASD [<xref ref-type="bibr" rid="R4">4</xref>]. In that study while PWD was correlated with Qp/Qs, the left atrium diameter was the same with the control group. However, the right atrium diameter was not mentioned in that study. In our study, the right atrium diameter was significantly greater than in the control group.</p><p>The only study related to the right atrium and right cardiac pressure loading is the study by Tukek et al. [<xref ref-type="bibr" rid="R6">6</xref>] on PWD in patients with COPD. In this study, the researchers demonstrated that PWD, p max and p min values were greater in patients with COPD compared with the values in the control group and that in the same fashion PWD was greater in patients with paroxysmal AF compared to patients without paroxysmal AF. They also stated that there was no relationship between PWD and pulmonary functions, blood gases and right and left atrium functions. Our study is similar to this study in that it investigates the right atrium pressure load and PWD. However, our patients were from a younger age group and consisted of individuals that did not have COPD or any other systemic disease.</p><p>Morover, PS is relatively rare congenital heart disease, there is not enough information on PWD in PS. That these patients are detected and treated at a young age may be the reason for this lack of information. Considered from this perspective, our study might be interesting also because it is the first of its kind.</p></sec><sec id="s5"><title>Conclusion</title><p>P wave dispersion and p max are increased in PS which is a congenital heart disease. There is a correlation between the pressure gradient that is the degree of narrowing and PWD. On the other hand, there is no correlation with the right and left atrium diameters. However, we think that studies involving larger patient populations are warranted.</p></sec>
Andersen-Tawil Syndrome	<p>Andersen-Tawil syndrome (ATS) is a rare condition consisting of ventricular arrhythmias, periodic paralysis, and dysmorphic features. In 2001, mutations in KCNJ2, which encodes the a subunit of the potassium channel Kir2.1, were identified in patients with ATS. To date, KCNJ2 is the only gene implicated in ATS, accounting for approximately 60% of cases. ATS is a unique channelopathy, and represents the first link between cardiac and skeletal muscle excitability. The arrhythmias observed in ATS are distinctive; patients may be asymptomatic, or minimally symptomatic despite a high arrhythmia burden with frequent ventricular ectopy and bidirectional ventricular tachycardia. However, patients remain at risk for life-threatening arrhythmias, including <italic>torsades de pointes</italic> and ventricular fibrillation, albeit less commonly than observed in other genetic arrhythmia syndromes. The characteristic heterogeneity at both the genotypic and phenotypic levels contribute to the continued difficulties with appropriate diagnosis, risk stratification, and effective therapy. The initial recognition of a syndromic association of clinically diverse symptoms, and the subsequent identification of the underlying molecular genetic basis of ATS has enhanced both clinical care, and our understanding of the critical function of Kir2.1 on skeletal muscle excitability and cardiac action potential.</p>	<contrib contrib-type="author"><name><surname>Smith</surname><given-names>Andrew H</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Fish</surname><given-names>Frank A</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Kannankeril</surname><given-names>Prince J</given-names></name><degrees>MD, MSCI</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Andersen-Tawil syndrome describes a rare condition consisting of ventricular arrhythmias, potassium-sensitive periodic paralysis, and developmental anomalies. [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>] While two cases of familial periodic paralysis associated with ventricular ectopy were first described in 1963, [<xref ref-type="bibr" rid="R3">3</xref>] the suggestion of a potential syndromic association was not proposed until 1971, when Ellen Andersen and colleagues described the characteristic triad in an 8 year-old dysmorphic boy who experienced episodes of muscle weakness and ventricular extrasystoles. [<xref ref-type="bibr" rid="R4">4</xref>] Nearly a quarter of a century later, 10 previously-described cases were reviewed, in addition to 4 new patients in 3 kindreds, who appeared to inherit the condition predominantly in an autosomal dominant pattern. Noting for the first time the phenotypically heterogenous nature of this disease, this was also the first report to ascribe the name of Andersen’s syndrome to this constellation of symptoms. [<xref ref-type="bibr" rid="R1">1</xref>] Subsequently, the syndrome was renamed Andersen-Tawil Syndrome (ATS) in recognition of the contributions of Dr. Rabi Tawil, who developed diagnostic criteria and refined understanding of this entity. [<xref ref-type="bibr" rid="R5">5</xref>]</p><p>Following its initial characterization, significant advances have improved our understanding of the genetic basis and molecular aspects of ATS. In the seminal paper by Plaster et al, [<xref ref-type="bibr" rid="R2">2</xref>] linkage analysis in a large kindred revealed a locus corresponding to a genetic region of over 40 cM on chromosome 17q23. With the notion that both the periodic paralyses and abnormalities in myocardial cell repolarization were associated with ion channel mutations, so-called channelopathies, the 3 ion channel genes within the linked region (KCNJ2, CACNG1, and SCN4A) were suitable candidate genes. SCN4A was known to underlie periodic paralysis without heart or developmental problems, [<xref ref-type="bibr" rid="R6">6</xref>] and the CACNG1 gene product was not detected in heart. Due to the known function and expression pattern of Kir2.1, the protein encoded by KCNJ2, it was considered an excellent candidate gene for ATS. Eight different mutations in KCNJ2 were initially identified in ATS patients, and functional characterization of 2 of the mutations revealed dominant negative effects on channel function. These findings demonstrated that ATS mutations in KCNJ2, which encodes the Kir2.1 channel α subunit, appear to markedly impair its vital role in stabilizing resting membrane potential and mediating the terminal repolarization phase of the action potential, resulting in a predisposition to ventricular arrhythmias through mechanisms discussed below.</p><p>We will review the clinical manifestations of Andersen-Tawil syndrome, including distinguishing historical, physical examination, and electrophysiologic findings. We will discuss the genetic and molecular aspects of ATS, and will address the characteristic heterogeneity at both the genotypic and phenotypic levels. Finally, we will explore various therapies devised to address the primary cause of morbidity and mortality in ATS, namely the propensity for life-threatening ventricular arrhythmias.</p></sec><sec sec-type="" id="s2"><title>Clinical Manifestations</title><p>Since the initial report of Andersen-Tawil syndrome, literature has emerged emphasizing the inconsistencies in phenotypic expression and disease severity [<xref ref-type="bibr" rid="R7">7</xref>]. Such variability has been described among kindreds, as well as among members of the same families [<xref ref-type="bibr" rid="R8">8</xref>]. This variability has made the diagnosis of ATS difficult. In fact, half of the new patients described in the initial paper proposing the term “Andersen’s syndrome” were diagnosed only after repeated evaluations [<xref ref-type="bibr" rid="R1">1</xref>]. Unpredictability is seen not only in dysmorphic characteristics, but electrophysiologic manifestations as well. Although ATS is inherited in an autosomal dominant pattern, given its heterogenous expressivity, a family history of associated characteristics may not be elicited. In fact, there is up to a 20% incidence of non-penetrance in individuals with KCNJ2 mutations. Additionally, up to 40% of patients with ATS do not have KCNJ2 mutations, further confounding the ability to utilize genetic screening as a diagnostic tool [<xref ref-type="bibr" rid="R5">5</xref>].</p><sec id="s2a"><title>Dysmorphic features</title><p>Appreciation of dysmorphic manifestations comprise one of the three fundamental elements in the evaluation of ATS. Features initially described included short stature, hypertelorism (wide-set eyes), low-set ears, palatal defects, mandibular hypoplasia, single palmar crease, crytporchidism, and slight bilateral ptosis [<xref ref-type="bibr" rid="R4">4</xref>]. In a recent series of 36 KCNJ2 mutation carriers, 78% had at least two dysmorphic features, the most common being clinodactyly (permanent medial or lateral curvature of a finger or toe, seen in 64%) and mandibular hypoplasia (44%) [<xref ref-type="bibr" rid="R8">8</xref>]. Other features described have included scoliosis, hyperthyroidism, vaginal atresia, and unilateral dysplastic kidney; structural cardiovascular anomalies such as a bicuspid aortic valve with or without associated coarctation, and valvular pulmonic stenosis have been reported as well [<xref ref-type="bibr" rid="R9">9</xref>]. Also noted in one series was that the severity of facial dysmorphism did not correlate with the severity of cardiac or skeletal muscular involvement [<xref ref-type="bibr" rid="R10">10</xref>].</p></sec><sec id="s2b"><title>Periodic Paralysis</title><p>Periodic paralysis serves as another key element to the clinical diagnosis of ATS. Historically, the age of onset of weakness is highly variable, though typically within the first two decades of life [<xref ref-type="bibr" rid="R10">10</xref>]. These episodes typically will present before any cardiac symptoms, and are usually manifested following periods of prolonged physical exertion. Examination reveals a muscular weakness that is proximal in nature, and may be associated with muscle wasting. Attacks have been demonstrated in patients with varying serum potassium levels (either hypo-, hyper-, or normokalemia), though levels have been found to remain consistent among an individual kindred.2 Potassium challenges have been found to at times either induce or even successfully treat episodic weakness [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R10">10</xref>]. Noted in one kindred was a sex-specific expressivity, with periodic paralysis appearing to segregate as an autosomal dominant trait only among males [<xref ref-type="bibr" rid="R9">9</xref>]. Additional findings have included markedly elevated serum creatine kinase levels, normal electromyogram, and tubular aggregates or minimal mypoathic changes apparent on pathologic examination of skeletal muscle biopsy [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R8">8</xref>]. </p></sec><sec id="s2c"><title>Cardiac Manifestations</title><p>True to its variability in phenotypic expression, there exists a wide range of electrophysiologic manifestations in ATS. Some patients have QT interval prolongation, while others demonstrate normal QT intervals. Prominent U waves in the anterior precordial leads are common in ATS [<xref ref-type="bibr" rid="R11">11</xref>]. Although U waves can be observed in normal individuals at low heart rates or in hypokalemic states, these are described in ATS patients at higher heart rates, suggesting that this may represent a manifestation of disease rather than a normal variant [<xref ref-type="bibr" rid="R8">8</xref>]. Associated arrhythmias range from isolated premature ventricular beats to complex ventricular ectopy and polymorphic ventricular tachycardias such as bidirectional ventricular tachycardia or less commonly <italic>torsades de pointes</italic> (see <xref ref-type="fig" rid="F1">Figure 1</xref>) [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R12">12</xref>]. Complete right bundle branch block, left bundle branch block, bifasicular block, and first-degree atrioventricular block have also been described in patients with known KCNJ2 mutations [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R11">11</xref>]. Both nonfatal cardiac arrest with documented torsades de pointes, and sudden death, thought to be less common relative to its long QT syndrome (LQTS) counterparts, have also been reported [<xref ref-type="bibr" rid="R13">136</xref>]. Sex-specific expressivity of the arrhythmia phenotype has been described in 1 family, where 13 of 16 female carriers, but no male carriers, demonstrated a history of ventricular arrhythmias [<xref ref-type="bibr" rid="R9">9</xref>]. While gender differences in cardiac repolarization are well described but not completely understood,[<xref ref-type="bibr" rid="R14">14</xref>] hormonal modulation of arrhythmia risk is suggested by our observation that arrhythmias in ATS may be quiescent during pregnancy.</p><p>The association between ATS and a prolonged corrected QT interval has recently been called into question. Citing its relationship to other disorders of myocellular repolarization, predisposition to ventricular arrhythmias, and the purported presence of a prolonged QT interval in 71% of all KCNJ2 mutation carriers, it had been suggested that ATS be classified as LQT7 [<xref ref-type="bibr" rid="R8">8</xref>]. A systematic assessment of ATS patient electrocardiograms was recently completed with conflicting results. Specifically, in ATS patients with and without KCNJ2 mutations, median corrected QT intervals were minimally longer (20 ms) than controls, remaining within the normal range (median, 440 ms); in fact, only 17% of ATS patients in this series had a QTc of over 460 ms. Additionally, it was noted that those patients possessing KCNJ2 mutations appeared to have a distinctive electrocardiographic phenotype, consisting of a prolonged terminal T wave downslope, a widened T-U wave junction, as well as biphasic and enlarged U waves; such findings were not evident in either ATS patients lacking KCNJ2 mutations or in healthy controls. Accordingly, some now recommend that patients with KCNJ2 mutations be classified as ATS1, rather than LQT7 [<xref ref-type="bibr" rid="R11">11</xref>].</p><p>A well-characterized feature of ATS is its propensity for ventricular arrhythmias. In addition to isolated premature ventricular beats, ventricular tachycardias (VT) involving a beat-to-beat variability in axis (polymorphic ventricular tachycardia), such as bidirectional ventricular tachycardia and torsades de pointes, have been described [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R15">15</xref>]. <italic>Torsades de pointes</italic> and its potential for progression to life-threatening ventricular fibrillation is a well-known complication associated with the prolongation of the plateau phase of the action potential seen in congenital and acquired long QT syndromes [<xref ref-type="bibr" rid="R16">16</xref>,<xref ref-type="bibr" rid="R17">17</xref>]. Bidirectional ventricular tachycardia is a distinct form of polymorphic VT, characteristically associated with intracellular calcium overload and seen in catechecholaminergic polymorphic ventricular tachycardia (CPVT) and digitalis toxicity [<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>]. A relationship to intracellular calcium overload and Na+/Ca2+-mediated triggering of arrhythmic activity has been proposed in ATS [<xref ref-type="bibr" rid="R8">8</xref>]. Bidirectional VT, though rare, has been described extensively in association with ATS [<xref ref-type="bibr" rid="R2">2</xref>]. Indeed, bidirectional ventricular tachycardia in the absence of other causes should prompt a thorough search for other features of ATS in the patient and kindred. In a cohort of 36 carriers of KCNJ2 mutations, a relatively low rate of syncope and cardiac arrest was described, despite a high arrhythmia burden. Ventricular arrhythmias were noted in 88% of patients, with nonsustained VT present in 65% of probands, and bidirectional VT seen in 18% of probands. Only 2 patients experienced nonfatal cardiac arrest (1 with documented <italic>torsades de pointes</italic>) and there were 4 episodes of syncope. Additionally, there were no cases of sudden cardiac death, nor was there a reported family history of sudden death among subjects [<xref ref-type="bibr" rid="R8">8</xref>]. Interestingly, the incidence of symptoms such as syncope in this cohort was less than 20%, markedly lower than the 63% and 46% reported in LQT1 and LQT2 syndromes respectively [<xref ref-type="bibr" rid="R20">20</xref>].</p><p>As ATS is a rare disorder, there exists no literature currently which correlates particular genotypes or clinical features with an assessment of risk for development of life-threatening ventricular arrhythmias and sudden death. While such reports of sudden death are rare in the literature, ATS patients remain predisposed to the development of polymorphic ventricular tachycardias such as bidirectional VT and torsades de pointes; with their potential for deterioration to ventricular fibrillation and hemodynamic collapse. These patients consequently remain at an increased risk for sudden cardiac death relative to the general population. </p></sec></sec><sec sec-type="" id="s3"><title>Diagnosis</title><p>Given the heterogenous nature of the clinical manifestations of Andersen-Tawil syndrome, a definitive diagnosis at times is elusive. Manifestation of at least two phenotypic classes (skeletal, cardiac, or developmental) has been thought to be necessary for diagnosis of ATS [<xref ref-type="bibr" rid="R8">8</xref>]. While some patients may manifest aspects consistent with all three classes of the clinical anomalies, others, (including those with known KCNJ2 mutations) may manifest only a single phenotypic classification. In these instances, some authors have suggested diagnosing such patients should they have a substantial family history [<xref ref-type="bibr" rid="R5">5</xref>]. Noting that to date, about 60% of Andersen syndrome patients are found to have mutations in Kir2.1, screening for KCNJ2 mutations also provides an increasingly valuable tool in either confirming or establishing diagnosis in a phenotypically heterogenous condition.</p></sec><sec sec-type="" id="s4"><title>Molecular manifestations</title><sec id="s4a"><title>Kir2.1</title><p>Mutations in KCNJ2, which encodes the a-subunit of Kir 2.1, an inwardly-rectifying potassium channel, have been identified as the genetic defects underlying the clinical phenotype of Andersen-Tawil syndrome [<xref ref-type="bibr" rid="R2">2</xref>]. The Kir2 family of potassium subunits is expressed in both heart and skeletal muscle, and is the first disordered channel associated with phenotypic manifestations within both skeletal and cardiac muscle [<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R22">22</xref>]. Kir2.1 subunits form a 427 amino acid protein, with 2 transmembrane domains and an extracellular pore forming loop containing the GYG amino acid sequence responsible for determining channel selectivity for potassium. The Kir2.1 channel functions as an inward rectifier, its conductance changing with voltage differences across the membrane. The term inward rectification indicates that while there is an inward flow of potassium current at hyperpolarized potentials, the same voltage gradient with depolarization produces much less outward potassium flow. At potentials more negative to - 20 mV, the Kir channels increasingly permit potassium efflux, providing a current which mediates the terminal phase of repolarization. The current provided by inward rectifiers such as the Kir2.1 channel, termed IK1, has consequently been shown to play a role in the determining cardiac excitability, and to “set” the resting potential [<xref ref-type="bibr" rid="R23">23</xref>,<xref ref-type="bibr" rid="R24">24</xref>].</p></sec><sec id="s4b"><title>Kir2.1: Molecular pathology</title><p>Four individual Kir2.1 proteins co-assemble to form a functional tetrameric channel. While injection of normal Kir2.1 cRNA into oocytes results in potassium currents with strong inward rectification, injection of mutant (D71V and R218W) cRNA into oocytes resulted in no detectable current [<xref ref-type="bibr" rid="R2">2</xref>]. These results raise 2 possibilities: the mutant proteins either failed to co-assemble, or assemble normally but impair channel function. The latter was suggested by the finding of “dominant negative” suppression of channel function; > 50% reduction in current with coexpression of normal and mutant Kir2.1 subunits. It is proposed that mutant proteins assemble with normal proteins into tetrameric channels, but the presence of 1 or more mutants in the channel is sufficient to disrupt current flow. Assuming random expression and co-assembly, only 1 out of every 16 channels would consist of 4 normal subunits, and function normally, while 15 of 16 would harbor at least 1 mutant protein. Indeed, coexpression of the D71V mutant cRNA with normal cRNA resulted in a current approximately 1/16<sup>th</sup> of normal current.</p><p>While other mutations have been described involving the external pore forming loop of Kir2.1,[<xref ref-type="bibr" rid="R7">7</xref>] mutations in KCNJ2 impairing channel interactions with phosphatidylinositol 4,5-bisphosphonate (PIP2) are thought to play a significant role as pathogenic mechanism in ATS.[<xref ref-type="bibr" rid="R25">25</xref>] PIP2 is a membrane-bound second messenger which activates Kir2.1, among other inward-rectifiers, and promotes an open-channel position [<xref ref-type="bibr" rid="R5">5</xref>]. Also, in vitro it was demonstrated that in general (and especially with the previously described R218W mutation), these mutations resulted in impaired channel- PIP2 interactions, affecting whole-cell current and marked decreases in open-channel probability [<xref ref-type="bibr" rid="R25">25</xref>]. Consequently, impairments in PIP2 - mediated channel function would result in an increase in mean closed time, as has been described, resulting in a decrease in channel activity and manifestation of disease phenotype.</p></sec><sec id="s4c"><title>Effects of impaired I<sub>K1</sub></title><p>Decelerates repolarization/prolongs action potential duration/depolarizes, destabilizes resting membrane potential</p><p>The effects of reduced I<sub>K1</sub> in a theoretical model of rabbit ventricular myocyte demonstrated a prolongation of the terminal phase of the action potential [<xref ref-type="bibr" rid="R8">8</xref>]. Reductions in extracellular potassium concentrations along with decreases in I<sub>K1</sub> were also associated with development of “terminal-phase” early after depolarizations (EADs), leading to spontaneous action potentials and providing a possible explanation for the frequent ventricular ectopy seen in patients with ATS. The term “terminal-phase” was used to distinguish the observed phenomena from the typical EAD’s seen in LQTS that arise from the plateau or early repolarization phase of the action potential. Interestingly, also noted was the sodium-calcium exchanger-dependent development of delayed after depolarizations, much like those seen with digitalis toxicity. This may underlie the risk for bidirectional VT that is often observed in ATS.</p><p>Using adenoviral gene transfer techniques, Miake and colleagues were able to further delineate the in vivo role of I<sub>K1</sub> in cardiac repolarization, studying isolated ventricular myocytes from adult guinea pig with both augmented and diminished I<sub>K1</sub> expression [<xref ref-type="bibr" rid="R26">26</xref>]. “Dominant negative” reduction in I<sub>K1</sub> was achieved by over expression of mutant Kir2.1with the GYG potassium selectivity motif replaced with AAA. This resulted in prolongation of action potential duration (APD), deceleration of phase 3 repolarization, and depolarization of the resting membrane potential (RMP). Conversely, over expression of human Kir2.1 with a consequent increase in I<sub>K1</sub> led to significant APD shortening with hyperpolarization of RMP and acceleration of phase 3 repolarization.</p><p>Clinical application of the recent understanding of the role of Kir2.1 and its corresponding current, I<sub>K1</sub>, continues to evolve. Patients with ATS and KCNJ2 mutations demonstrate a distinct change in U wave morphology, an electrocardiographic characteristic associated with frequent premature ventricular beats and nonsustained VT, potentially demonstrating an increased vulnerability to ventricular arrhythmia [<xref ref-type="bibr" rid="R11">11</xref>]. Mutation-induced reductions in I<sub>K1</sub> have been proposed to lead to a prominent U wave morphology and ventricular ectopy through an increase in transmural dispersion of repolarization, leading to action potential duration changes that vary in degree from one region of myocardium to another, and potentially serving as the substrate for reentrant arrhythmias. Interestingly, the gain-of-function mutations in KCNJ2 and resulting increases in I<sub>K1</sub> have also recently been correlated with disease. The resulting decreases in action potential durations leads to accelerated atrial and ventricular repolarization characteristic of the short QT syndrome. In the form of short QT syndrome (SQT3) associated with a KCJN2 mutation, ventricular fibrillation has been observed [<xref ref-type="bibr" rid="R27">27</xref>].</p></sec><sec id="s4d"><title>Molecular Genetics</title><p>ATS occurs sporadically, or is inherited in an autosomal dominant manner. Mutations in KCNJ2 were the first described as responsible for the phenotype of ATS. Kir2.1 subunits, encoded by KCNJ2, combine to form tetrameric channels with additional Kir2.1 subunits,[<xref ref-type="bibr" rid="R28">28</xref>] or with other subunits of the Kir2.x subfamily [<xref ref-type="bibr" rid="R29">29</xref>]. Mutations in KCNJ2 have been demonstrated to impair Kir2.1 channel function usually through a mechanism of dominant negative suppression [<xref ref-type="bibr" rid="R30">30</xref>]. The dominant-negative nature of ATS mutations has been well-described, whereby heterogenous expression of wild type Kir2.1 with any number of mutant Kir2.1 subunits results in a loss of current [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R30">30</xref>]. Also demonstrated was the ability of mutant Kir2.1 to exert dominant-negative suppressive effects upon other members of the Kir2.x subfamily, possibly serving as an explanation for the interfamilial phenotypic variations in expressivity [<xref ref-type="bibr" rid="R29">29</xref>].</p><p>Between 6-20 % of individuals with identified KCNJ2 mutations appear phenotypically normal (non-penetrant) [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R9">9</xref>]. Additionally, approximately 30-40% of patients with ATS do not harbor KCNJ2 mutations [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R8">8</xref>]. In the first demonstration of a relationship between KCJN2 mutations and ATSs, 3 of 16 families examined had no mutations in the coding regions of KCNJ2, with no apparent phenotypic differences in those with KCNJ2 mutations and those without [<xref ref-type="bibr" rid="R2">2</xref>]. Several mechanisms for this genotypic heterogeneity have been proposed, including mutations in KCNJ2 regulatory proteins, and mutations in other members of the Kir2.x subfamily, as these subunits assemble in heterotetramers to form functional channels [<xref ref-type="bibr" rid="R29">29</xref>].</p></sec></sec><sec sec-type="" id="s5"><title>Treatment</title><p>Therapy for ATS is directed toward its phenotypic manifestations which lead to significant morbidity, namely associated episodic weakness and most importantly its association with life-threatening ventricular tachyarrhythmias. From initial reports, therapies such as oral potassium supplementation, sodium restriction, spironolactone, and acetazolamide have anecdotally been shown to ameliorate symptoms of weakness [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R31">31</xref>]. We have obserevd improvement with oral contraceptive therapy, with exacerbation of weakness after discontinuation in a female patient with ATS. Unfortunately, perhaps due to the genotypic and phenotypic heterogeneity in disease, in concert with erratic and paradoxical exacerbation of symptoms with therapy, no therapeutic standards exist to date.</p><p>While the cardiac sequelae of ATS hold the greatest potential for morbidity and mortality, they remain the most problematic in devising effective medical therapeutic options. Given the phenotypic heterogeneity that complicates diagnosis, this variability also plays a role as a complicating factor in effective risk stratification. The most concerning aspect of ATS, paroxysmal ventricular tachycardia leading to syncope or ventricular fibrillation with cardiac arrest, is only infrequently inducible with invasive electrophysiologic study [<xref ref-type="bibr" rid="R12">12</xref>]. Recently, radiofrequency catheter ablation targeting premature ventricular beats, which may serve as triggers for sustained ventricular arrhythmias, has been performed in other genetic arrhythmia syndromes, including LQTS and Brugada syndrome [<xref ref-type="bibr" rid="R32">32</xref>][<xref ref-type="bibr" rid="R33">33</xref>]. This may be an attractive approach in ATS due to the frequent ventricular ectopy. Initial attempts to map ventricular ectopy in patients with ATS reveal that the ectopic beats originate from different parts of the left Purkinje network (personal communication, P. Sanders, 2005; see <xref ref-type="fig" rid="F2">Figure 2</xref>). Although to date, catheter ablation in ATS has not been reported, in a patient with arrhythmic storms, this approach could be considered.</p><p>Successful pharmacologic therapy remains elusive, lacking effectiveness in reducing the frequency of ventricular ectopy; [<xref ref-type="bibr" rid="R1">1</xref>] in fact, even from early reports such therapy has been recognized to exacerbate symptoms of weakness [<xref ref-type="bibr" rid="R1">1</xref>][<xref ref-type="bibr" rid="R3">3</xref>]. We have observed similar phenomena: improved rhythm control but worsening weakness, or improved muscle strength, but increased ventricular ectopy. Reports of antiarrhythmic drug efficacy remain anecdotal, and conflicting. While calcium channel blockers are theorized to address the proposed arrhythmogenic substrate of intracellular calcium overload, and have been shown to terminate bidirectional VT in ATS,[<xref ref-type="bibr" rid="R15">15</xref>] they have also been shown ineffective in reducing ventricular ectopy [<xref ref-type="bibr" rid="R12">12</xref>]. Similarly, amiodarone has been described as a successful therapy for VT in a patient with ATS due to the R218W KCNJ2 mutation [<xref ref-type="bibr" rid="R31">31</xref>]. The authors suggested a possible pharmacogenetic link between the R218W mutation and treatment response. However, subsequently amiodarone was implicated in inducing <italic>torsades de pointes</italic> in a different ATS patient who also harbored the R218W mutation [<xref ref-type="bibr" rid="R34">34</xref>]. While amiodarone may exacerbate early afterdepolarizations and predispose to TdP with its inhibition of I<sub>K1</sub> and I<sub>Kr</sub>, its effects in a given patient may not be predictable. Pacing in combination with nicorandil was successful in treating ventricular arrhythmias in the latter patient. Pacing, by increasing heart rate and nicorandil, by increasing outward potassium current through I<sub>KATP</sub>, resulted in attenuation of the prominent U wave, theoretically reducing transmural dispersion of repolarization and reducing risk for torsades de pointes. Beta-blockers (including propranolol, atenolol, and sotalol) and sodium channel blockers (including mexiletine, propafenone, and flecainide) have been described as ineffective in select patients [<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R31">31</xref>,<xref ref-type="bibr" rid="R34">34</xref>]. We have used beta-blockers alone, and in combination with calcium channel blockers with some success, albeit in a modest number of patients.</p><p>Aside from preventive measures, an intervention demonstrated to successfully and reliably treat hemodynamically unstable ventricular tachyarrhythmias is the implantable cardioverter-defibrillator (ICD) [<xref ref-type="bibr" rid="R35">35</xref>]. The diagnosis of ATS itself can be considered a class IIb indication for ICD placement, citing familial or inherited conditions with a high risk for ventricular tachyarrhythmias as a consideration [<xref ref-type="bibr" rid="R36">36</xref>]. A history of cardiac arrest due to ventricular fibrillation or VT would be regarded as a class I indication for ICD placement in ATS, noting that there is evidence and/or general agreement that ICD placement is both useful and effective. Although sustained VT can also be considered a class I indication for an ICD, many ATS patients are asymptomatic during episodes of sustained ventricular tachycardia [<xref ref-type="bibr" rid="R12">12</xref>]. Until further data regarding risk startification becomes available, it seems prudent to reserve ICD therapy for ATS patients with a history of cardiac arrest, syncope, or sustained rapid and/or symptomatic VT. Furthermore, ICD programming should be tailored to minimize the risk of multiple shocks for VT that may be well tolerated. Programming no therapy for VT [<xref ref-type="bibr" rid="R12">12</xref>], or anti-tachycardia pacing rather than shocks for VT have been advocated [<xref ref-type="bibr" rid="R37">37</xref>].</p></sec><sec sec-type="" id="s6"><title>Conclusions</title><p>Andersen-Tawil syndrome is a rare condition with variable expression, caused by sporadic mutations or in an autosomal dominant manner. Mutations in KCNJ2 which encode the inwardly rectifying channel Kir2.1 lead to the clinically apparent phenotype of ventricular arrhythmias, periodic paralysis, and developmental anomalies. Marked variation in expression confounds ability to diagnose, risk stratify, and effectively treat patients. ATS differs both on a molecular and phenotypic level from the long QT syndromes, and it has been suggested to abandon its determination as LQT7. Effective pharmacologic therapy for associated ventricular arrhythmias is lacking, prompting consideration of ICD placement or possibly catheter ablation in high-risk patients.</p></sec>
Cardiac Arrhythmia and Geomagnetic Activity	Could not extract abstract	<contrib contrib-type="author"><name><surname>Stoupel</surname><given-names>E</given-names></name><degrees>FESC</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Geomagnetic activity (GMA) is part of the physical environment surrounding us. In last century many studies were published discussing different links between the level of GMA and human homeostasis. They include studies on cardiac arrhythmia and sudden cardiac death (SCD). The medical community is not equivocal considering those findings. In an editorial comment introducing one recent paper in this field the Editor-in-Chief of “PACE” wrote: “…The question is whether the findings represent a new area for investigation or is a “straw man” [<xref ref-type="bibr" rid="R1">1</xref>]. In this brief review some data will be presented demonstrating some links between cardiac arrhythmic events and level of GMA.</p></sec><sec sec-type="" id="s2"><title>Geomagnetic activity</title><p>The geomagnetic field is a physical phenomenon resulting from different rotation speeds of different layers of our planet . The level of activity of the field can be affected by “geoeffective” parts of magnetic fields coming as a result of solar explosions and accompanied giant magnetic fields. A small portion of them can “disturb” the original geomagnetic field, resulting in active or stormy levels of GMA. The level of GMA is measured in Nanotesla every three hours (8 parameters in 24 hours). The six highest results (in K or integrated A indices of GMA) describe the day as Quiet (I), Unsettled (II), Active (III) or Stormy (IV) day of GMA (<xref ref-type="table" rid="T1">Table 1</xref>). In the last decades many changes in the human hemeostasis related parameters were analyzed in relation with the level of GMA [<xref ref-type="bibr" rid="R2">2</xref>-<xref ref-type="bibr" rid="R4">4</xref>]. But the level of GMA is not only a factor that can be related with some pathologic effects. The field is also a shield defending our planet from very energetic and potentially harmful space physical activity ingradients, such as cosmic rays and very closely related to them, high energy space proton flux, that can be more active on the surface of our planet when the GMA is extremely low [<xref ref-type="bibr" rid="R5">5</xref>-<xref ref-type="bibr" rid="R7">7</xref>]. For more information special cosmophysical literature must be used, or at least the Glossary of Solar-Terrestrial Terms [<xref ref-type="bibr" rid="R8">8</xref>].</p></sec><sec sec-type="" id="s3"><title>Summary of clinical data</title><sec id="s3a"><title>Forensic medicine data</title><p>Sudden cardiac deaths (n=43) with signs of coronary atherosclerosis, but without acute myocardial infarction occurred more often on days of lowest (Quiet-Io) GMA compared with higher levels of GMA.( p>0.001) [<xref ref-type="bibr" rid="R2">2</xref>].</p></sec><sec id="s3b"><title>Holter monitoring data</title><p>Hourly atrial premature complex (APC) and ventricular premature complex (VPC) number was more on days of lowest GMA [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R10">10</xref>].</p><p>Sudden deaths (n=480) in three hours after beginning of symptoms (before admission, or at the admission department) were higher on days of lowest GMA (p <0.01)[<xref ref-type="bibr" rid="R9">9</xref>]. Number of sudden deaths dropped on days of severe geomagnetic storm (very high GMA) in July 2000 - the "Bastille day event" [<xref ref-type="bibr" rid="R11">11</xref>]. Arrival of patients to the emergency department at a tertiary university hospital of patients with atrial fibrillation of new origin (n=653, 1185 days of observation) was inverse correlated with four levels of GMA (r=-0.97, p=0.02) [<xref ref-type="bibr" rid="R12">12</xref>].</p><p>Ventricular tachycardia episodes by ambulatory Holter monitoring (n=3019) registered 27% on days of low (Io-IIo) GMA and 17% on days of two higher levels (chi<sup>2</sup>=7.3, p=0.006) [<xref ref-type="bibr" rid="R13">13</xref>]. In patients admitted for acute myocardial infarction (n=14,529; 8586 men) cardiac arrhythmic events (PAF, AT,VT, VF, n=2024) were relatively higher on days of lowest GMA [<xref ref-type="bibr" rid="R14">14</xref>]. SCD that were occurring in one hour after the onset of symptoms (n=261) was more on days of lowest GMA in men < 65 years (p=0.06) and women > 65years (n=0.027). SCD in woman younger than 65 years was relatively rare [<xref ref-type="bibr" rid="R15">15</xref>].</p><p>Discharges (n=402) in 137 days of 25 patients with implanted ICD for ischemic cardiomyopathy, with impaired left ventricular systolic function, for VT , VF were inverse correlated with the level of four daily levels of GMA, (r=-0.96-0.97,p=0.03-0.04) [<xref ref-type="bibr" rid="R16">16</xref>].</p><p>The number of SCD (n=516), according to the first aid service data, was significantly correlated with monthly flux of high energy space protons (>90MeV) - a physical parameter closely related to cosmic ray activity (described according to neutron monitoring data on the earth's surface in impulse/min) and inverse related to the levels of solar and GMA [<xref ref-type="bibr" rid="R17">17</xref>].</p><p>A new experimental study in pigs has shown that in very low magnetic fields a prolongation of QT interval and changes in P and T waves were registered. Those changes are explained as a result of Calcium ion efflux from the myocytes following channel blocking or inactivation [<xref ref-type="bibr" rid="R18">18</xref>]. Also, a possibility was discussed that the electrons that are involved in the electrical activity of the heart, are changing in different levels of magnetic field activity - less active in higher magnetic field [<xref ref-type="bibr" rid="R19">19</xref>].</p><p>Now in progress are a number of studies considering the possible role of cosmic ray activity - a factor inverse related to GMA, measured by neutron activity on the earth’s surface in the pathogenesis of life threatening cardiac arrhythmias and SCD [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R21">21</xref>].</p><p>The purpose of this short review is to draw attention on possible environmental physical effects that can be an additional pathogenetic factor affecting the time distribution of events related to cardiac arrhythmia and sudden cardiac death.</p></sec></sec><sec sec-type="" id="s4"><title>Conclusion</title><p>The presented data, a part of science named Clinical Cosmobiology, has shown that temporal links exist between cardiac arrhythmia, including life threatening events, SCD and level of cosmophysical activity. Atrial fibrillation,ventricular tachycardia/fibrillation and SCD are occurring in inverse relationship to the level of GMA. It is possible that the level of GMA is a factor preventing SCD, especially in patients with damaged heart muscle. The role of factors becoming more active in low GMA, like neutron activity, are an object of further studies.</p></sec>
Changing QRS Morphology: What is the mechanism?	Could not extract abstract	<contrib contrib-type="author"><name><surname>Haghjoo</surname><given-names>Majid</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Arya</surname><given-names>Arash</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Dehghani</surname><given-names>Mohammad Reza</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Sadr-Ameli</surname><given-names>Mohammad Ali</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<p>ECG in sinus rhythm with ventricular preexcitation and changing QRS morphology was seen that was initially interpreted as the multiple accessory pathway from elsewhere. (<xref ref-type="fig" rid="F1">Figure 1A</xref>).</p><p>The following mechanisms are potentially involved in the electrogenesis of changing QRS morphology in WPW syndrome: 1) multiple accessory pathways [<xref ref-type="bibr" rid="R1">1</xref>]; 2) simultaneous occurrence of aberrant atrioventricular conduction with accessory pathway conduction [<xref ref-type="bibr" rid="R2">2</xref>]; 3) ventricular fusion of preexcited sinus impulse with ectopic impulse.</p><p>Electrophysiologic study showed short PR (75 ms) interval with wide QRS (152 ms) and negative HV (-12 ms) interval. No change in delta wave polarity was observed during HRA and CS pacing. In full preexcitation, no breakthrough was seen in the CS. During incremental ventricular pacing, atrial breakthrough site is initially recorded on the HRA catheter and then changed to distal pole of CS catheter with progressive decrease in pacing cycle length. During ventricular pacing at cycle length of 500 ms (S1), earliest atrial activity is recorded on HRA catheter.</p><p>Changing QRS could not be explained by presence of <italic>multiple APs</italic> because only right-sided AP had bidirectional conduction and no distal CS breakthrough was seen simultaneous with changing QRS morphology. The possibility of <italic>aberrant conduction</italic> is excluded by presence of negative HV interval in the beats with differing QRS morphology. No sinus cycle length variation before and after the beats with different morphologies are against the occurrence of functional LBBB. The prematurity of ventricular electrogram in His recording catheter with variable HV (H-electrogram is recorded after V-electrogram in second beat and before V-electrogram in third beat) and fixed V-RB intervals (interval from ventricular electrogram in His to the RB potential) are compatible with <italic>ventricular fusion of preexcited sinus impulse with ectopic ventricular impulse</italic> originating from parahissian area (explaining LBBB and inferior axis morphology of the beats with changing QRS) but not from the His bundle or RBB itself (because H-electrogram and RB potential is recorded after V-electrogram in the second beat with greater degree of ventricular fusion)(<xref ref-type="fig" rid="F1">Figure 1B</xref>).</p>
The Postural Tachycardia Syndrome (POTS): Pathophysiology, Diagnosis & Management	<p>Postural tachycardia syndrome (POTS), characterized by orthostatic tachycardia in the absence of orthostatic hypotension, has been the focus of increasing clinical interest over the last 15 years 1. Patients with POTS complain of symptoms of tachycardia, exercise intolerance, lightheadedness, extreme fatigue, headache and mental clouding. Patients with POTS demonstrate a heart rate increase of ≥30 bpm with prolonged standing (5-30 minutes), often have high levels of upright plasma norepinephrine (reflecting sympathetic nervous system activation), and many patients have a low blood volume. POTS can be associated with a high degree of functional disability. Therapies aimed at correcting the hypovolemia and the autonomic imbalance may help relieve the severity of the symptoms. This review outlines the present understanding of the pathophysiology, diagnosis, and management of POTS.</p>	<contrib contrib-type="author"><name><surname>Raj</surname><given-names>Satish R</given-names></name><degrees>MD MSCI</degrees></contrib><aff>Autonomic Dysfunction Center, Division of Clinical Pharmacology, Departments of Medicine & Pharmacology, Vanderbilt University, Nashville, Tennessee, USA</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Postural tachycardia syndrome (POTS), characterized by orthostatic tachycardia in the absence of orthostatic hypotension, has been the focus of increasing clinical interest over the last 15 years [<xref ref-type="bibr" rid="R1">1</xref>]. Patients with POTS complain of symptoms of tachycardia, exercise intolerance, lightheadedness, extreme fatigue, headache and mental clouding. This disorder is not new [<xref ref-type="bibr" rid="R2">2</xref>], but has gone by many different names over the last 150 years, including mitral valve prolapse syndrome, neurocirculatory asthenia, orthostatic tachycardia, and orthostatic intolerance [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R4">4</xref>]. An advantage of the name postural tachycardia syndrome (POTS) is that it focuses attention on the sympathetic activation which characterizes the disorder. This review outlines the present understanding of the pathophysiology, diagnosis, and management of POTS.</p></sec><sec sec-type="" id="s2"><title>Physiology of Upright Posture</title><p>Assumption of the upright posture requires prompt physiological adaptation to gravity. There is an instantaneous descent of ~500 ml of blood from the thorax to the lower abdomen, buttocks, and legs. In addition, there is a 10-25% shift of plasma volume out of the vasculature and into the interstitial tissue [<xref ref-type="bibr" rid="R5">5</xref>]. This shift decreases venous return to the heart, resulting in a transient decline in both arterial pressure and cardiac filling. This has the effect of reducing the pressure on the baroreceptors, triggering a compensatory sympathetic activation that results in an increase in heart rate and systemic vasoconstriction (countering the initial decline in blood pressure). Hence, assumption of upright posture results in a 10-20 beat per minute increase in heart rate, a negligible change in systolic blood pressure, and a ~5 mmHg increase in diastolic blood pressure.</p></sec><sec sec-type="" id="s3"><title>Pathophysiology of Orthostatic Dysregulation</title><p>Failure of the regulatory mechanism to respond properly may lead to either <bold><italic>orthostatic hypotension</italic></bold>, as is seen in autonomic failure, or <bold><italic>orthostatic tachycardia</italic></bold>, as is seen in POTS. Orthostatic hypotension is defined as a fall in pressure on standing of more than 20/10 mmHg. However, it is common in patients with autonomic failure for the decline to be much greater than this, which may result in loss of consciousness soon after standing. On the other hand, in POTS, blood pressure is typically maintained on standing or may even increase. Heart rate rises more than 30 bpm and symptoms reminiscent of impaired cerebral perfusion may develop.</p></sec><sec sec-type="" id="s4"><title>Clinical Presentation of Postural Tachycardia Syndrome (POTS)</title><sec id="s4a"><title>Diagnostic Criteria & Common Clinical Features</title><p>POTS is defined (<xref ref-type="table" rid="T1">Table 1</xref>) as the presence of symptoms of orthostatic intolerance for at least 6 months accompanied by a heart rate increase of at least 30 beats/min within 5-30 minutes of assuming an upright posture. This should occur in the absence of orthostatic hypotension (a fall in blood pressure >20/10 mmHg). The syndrome must occur in the absence of prolonged bed rest, medications that impair autonomic regulation (such as vasodilators, diuretics, antidepressants or anxiolytic agents), or any other chronic debilitating disorders that might cause tachycardia (such as dehydration, anemia or hyperthyroidism). It is important to recognize that this syndrome is typically disabling. Hence, the mere observation of orthostatic tachycardia is not, by itself, sufficient to make the diagnosis of POTS.</p><p>Symptoms include mental clouding (“brain fog”), blurred or tunneled vision, shortness of breath, palpitation, tremulousness, chest discomfort, headache, lightheadedness and nausea. While pre-syncope is common in these patients, only a minority (~30%) actually pass out. The chest pains are almost never due to coronary artery obstruction, but are sometimes associated with electrocardiographic changes in the inferior leads, particularly when upright [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>Many patients complain of significant exercise intolerance and extreme fatigue. Even activities of daily living, such as bathing or housework, may greatly exacerbate symptoms with resultant fatigue. This can pose significant limitations on their functional capacity.</p><p>The disorder primarily affects women of child-bearing age. The female:male ratio is 4:1. The reason for the strong female predominance is not known, but it should be noted that orthostatic tolerance is reduced in normal healthy females [<xref ref-type="bibr" rid="R7">7</xref>]. Others disorders such as autoimmune diseases and irritable bowel syndrome are seen commonly in patients with POTS, and also have higher prevalence in women.</p><p>Patients frequently report that their symptoms began following acute stressors such as pregnancy, major surgery, or a presumed viral illness, but in others cases, symptoms develop more insidiously. About 80% of female patients report an exacerbation of symptoms in the pre-menstrual phase of their ovulatory cycle (unpublished data). Gazit et al. have also reported an association between joint hypermobility and POTS [<xref ref-type="bibr" rid="R8">8</xref>]. Many patients have bowel irregularities and have been co-diagnosed with irritable bowel syndrome, and some have abnormalities of sudomotor regulation [<xref ref-type="bibr" rid="R9">9</xref>].</p></sec><sec id="s4b"><title>Psychological Profile in POTS</title><p>Patients with POTS are sometimes clinically diagnosed as having anxiety disorders such as panic disorder. Indeed, patients demonstrate elevated scores on the Beck Anxiety Inventory [<xref ref-type="bibr" rid="R10">10</xref>] (23±10 vs. 7±8; P<0.001), a commonly used instrument that quantifies the magnitude of anxiety symptoms [<xref ref-type="bibr" rid="R11">11</xref>]. Unfortunately, this questionnaire includes somatic anxiety symptoms (such as palpitation) which can result from a hyperadrenergic state such as is seen in POTS. When a newer, cognitive-based measure of anxiety (the Anxiety Sensitivity Index [<xref ref-type="bibr" rid="R12">12</xref>]) is used, there was a trend toward less anxiety in the patients with POTS than the general population (15±10 vs. 19±9; P=0.063) [<xref ref-type="bibr" rid="R11">11</xref>]. Thus, much of the anxiety attributed to patients with POTS might be due to a misinterpretation of their physical symptoms.</p><p>We did find that patients with POTS often have diminished attention and concentration compared to matched healthy volunteers [<xref ref-type="bibr" rid="R11">11</xref>]. Using the Inattention score from the Connors Adult ADHD Rating Scale [<xref ref-type="bibr" rid="R13">13</xref>], the patients with POTS scored significantly higher than did the normal control subjects.</p></sec><sec id="s4c"><title>Physical Findings in POTS</title><p>The most striking physical feature of POTS is the severe tachycardia that develops on standing from a supine position. Blood pressure and heart rate must be measured in both postures and should be taken not only immediately after standing but also at 2, 5 and 10 minutes as occasional patients have a delayed tachycardia [<xref ref-type="bibr" rid="R14">14</xref>]. Normal subjects commonly develop a transient tachycardia within the 1st minute of standing that should not be mistaken for POTS. A sustained heart rate increase ≥30 beats per minute is considered diagnostic of orthostatic tachycardia (<xref ref-type="fig" rid="F1">Figure 1</xref>). The systolic blood pressure should not fall by more than 20 mmHg, and in many cases it will actually increase with standing. Recent data suggests that there may be a significant circadian variability in the orthostatic tachycardia seen in patients with POTS [<xref ref-type="bibr" rid="R15">15</xref>]. In a cohort of 17 patients with POTS, the orthostatic tachycardia was greater in the morning than in the evening (38±4 bpm vs. 27±3 bpm; P<0.001), while there was no diurnal difference in the orthostatic change in blood pressure. These data suggest that to optimize diagnostic sensitivity, postural vital signs should be performed in the morning.</p><p>Cardiac auscultation may reveal a murmur of mitral valve prolapse, but significant mitral regurgitation is unusual. A striking physical feature of POTS is the dependant acrocyanosis that occurs in 40-50% of patients with POTS (<xref ref-type="fig" rid="F2">Figure 2</xref>). These patients experience a dark red-blue discoloration of their legs, which are cold to the touch. This can extend from the feet to above the level of the knees. The reasons underlying this phenomenon are not clear. The current data suggest that the problem is not due to increased pooling in the venous capacitance vessels, but rather due to decreased blood flow in the skin [<xref ref-type="bibr" rid="R16">16</xref>,<xref ref-type="bibr" rid="R17">17</xref>].</p></sec><sec id="s4d"><title>Laboratory Abnormalities in POTS</title><p>Some authors advocate the use head-up tilt table testing as a standardized method to assess an individual's response to a change in posture [<xref ref-type="bibr" rid="R1">1</xref>]. The patient is positioned on a standard tilt table and following baseline measurements of blood pressure and heart rate, the patient is inclined to a 70-degree head-up angle. Blood pressure and heart rate are then measured either continuously or at least every 12 minutes. The orthostatic tachycardia is often measured in a similar fashion to the standing test, with a similar threshold used to diagnose orthostatic tachycardia (an increase of ≥30 bpm) [<xref ref-type="bibr" rid="R1">1</xref>]. However, the physiology in response to passive standing on a tilt table (with the legs still) is not the same as“active standing” where the patient must support their own weight and maintain their balance. The latter requires use of the “skeletal muscle pump” and mimics real life, while the tilt table does not. For this reason Streeten et al. use similar criteria for orthostatic tachycardia (>27 bpm), but only with active standing [<xref ref-type="bibr" rid="R18">18</xref>]. In a recent study, we compared the orthostatic heart rate response of these 2 methods, and found that the tilt table test was associated with an increased orthostatic tachycardia in both patients with POTS and control subjects [<xref ref-type="bibr" rid="R19">19</xref>]. While both tests were sensitive for the diagnosis of POTS with a 30 bpm threshold for orthostatic tachycardia, the stand test had a specificity of 79% compared to only 23% for the tilt table test.</p><p>POTS patients should have only sinus tachycardia. An electrocardiogram should be done routinely to rule out the presence of an accessory bypass tract or any abnormalities of cardiac conduction. A Holter monitor might prove useful to exclude a re-entrant dysrhythmia, especially if the patient gives a history of paroxysmal tachycardia with a sudden onset and sudden offset. Other tests such as echocardiograms are only required in individual cases when there is doubt about the structural integrity of the heart.</p><p>We often measure plasma norepinephrine levels in both a supine and standing position (at least 15 minutes in each position prior to blood sampling). The supine norepinephrine is often high normal in patients with POTS, while the upright norepinephrine is usually elevated (>600 pg/ml), a reflection of the exaggerated neural sympathetic tone that is present in these patients while upright.</p><p>Tests of autonomic nervous system function typically show intact or exaggerated autonomic reflex responses. These patients often have preserved vagal function as reflected by their sinus arrhythmia ratio in response to deep breathing. They often have a vigorous pressor response to the Valsalva maneuver, with an exaggerated blood pressure recovery and overshoot both before and after release [<xref ref-type="bibr" rid="R20">20</xref>].</p><p>The blood volume is low in many patients with POTS [<xref ref-type="bibr" rid="R5">5</xref>]. This can be objectively assessed with nuclear medicine tests to directly measure either the plasma volume or the red cell volume. This knowledge may help to focus the treatment plan.</p><p>Some patients with POTS have co-existent complaints of episodic flushing. In about half of these cases there is an associated mast cell activation disorder [<xref ref-type="bibr" rid="R20">20</xref>]. This can be diagnosed by collecting urine from individual 2-4 hour voids following a severe flushing spell for determination of methylhistamines.</p></sec></sec><sec sec-type="" id="s5"><title>Differential Diagnosis</title><p>The clinical picture of POTS can be confused with pheochromocytoma because of the paroxysms of hyperadrenergic symptoms. Patients with pheochromocytoma are more likely to have symptoms while lying down than POTS patients, and often have much higher plasma norepinephrine levels. The diagnosis of pheochromocytoma is made by assessment of plasma or urinary metanephrines [<xref ref-type="bibr" rid="R21">21</xref>].</p><p>There is commonly some confusion between neurally mediated syncope and POTS. There is a clinical overlap between the 2 disorders, such that about 30% of patients with POTS also have neurally mediated syncope. Nonetheless, most patients with POTS do not faint.</p><p>Almost all patients with POTS also have associated fatigue. The reasons are not entirely clear. In some patients, but not all, the fatigue improves with pharmacological control of the orthostatic tachycardia. Some patients with POTS have symptomatic overlap with chronic fatigue syndrome.</p></sec><sec sec-type="" id="s6"><title>Pathophysiology of POTS</title><p>Tachycardia and asthenia on standing is a final common pathway of many pathophysiological processes. POTS is therefore best viewed as a syndrome rather than a disease. Many disorders with a common key clinical presentation (the orthostatic tachycardia) have been described. Over the last decade, much has been learned about specific forms or sub-types within POTS, although a simple test to categorize the individual patient remains elusive. We discuss here the common POTS phenotypes including neuropathic POTS and central hyperadrenergic POTS (<xref ref-type="fig" rid="F3">Figure 3</xref>).</p><sec id="s6a"><title>Neuropathic POTS</title><p>Considering that POTS patients have high plasma NE levels, it would seem paradoxical that a neuropathy is proposed as an underlying process. Yet some of them have a form of dysautonomia, with preferential denervation of sympathetic nerves innvervating the lower limbs [<xref ref-type="bibr" rid="R22">22</xref>-<xref ref-type="bibr" rid="R24">24</xref>]. There have been several findings consistent with this hypothesis. The results of sudomotor axon reflex testing [<xref ref-type="bibr" rid="R22">22</xref>] and galvanic skin stimulation [<xref ref-type="bibr" rid="R23">23</xref>] support this as well as skin biopsy results [<xref ref-type="bibr" rid="R25">25</xref>]. Further, these patients have been found to be hypersensitive to infusions of norepinephrine and phenylephrine into veins of the foot, despite high circulating plasma norepinephrine concentrations [<xref ref-type="bibr" rid="R24">24</xref>]. This suggests that there is a denervation hypersensitivity of the leg veins. Using a segmental norepinephrine spillover approach, Jacob et al. [<xref ref-type="bibr" rid="R26">26</xref>] demonstrated that patients with POTS had normal sympathetic neuronal norepinephrine release in their arms, but less norepinephrine release (and thus less sympathetic activation) in their lower body.</p></sec><sec id="s6b"><title>Hypovolemia & Blood Volume Regulation</title><p>Many patients with POTS have low plasma volumes [<xref ref-type="bibr" rid="R27">27</xref>,<xref ref-type="bibr" rid="R28">28</xref>], but not all. To determine if hypovolemia existed in an unselected group of POTS patients, we studied 15 patients with POTS (not selected for blood volume) and 14 control subjects [<xref ref-type="bibr" rid="R5">5</xref>]. Plasma volume was measured using <sup>131</sup>I labeled human serum albumin using a dye dilution technique, and compared to the predicted blood volume for each individual, based upon their height, weight, and gender. As can be seen in <xref ref-type="fig" rid="F4">Figure 4</xref>, the control subjects did not have a significant plasma volume deficit (0.8±2.5%). In contrast, the patients with POTS had a plasma volume deficit of 12.8±2.0% (P<0.001).</p><p>The renin-angiotensin-aldosterone system plays a key role in the neurohormonal regulation of plasma volume in humans. Plasma renin activity and angiotensin II would be expected to increase in response to hypovolemia in order to promote blood volume expansion. Angiotensin II promotes sodium and water retention directly by stimulating sodium resorption in the proximal tubules, and indirectly by stimulating aldosterone secretion.</p><p>Patients with orthostatic tachycardia who were also hypovolemic have low levels of standing plasma renin activity and aldosterone compared to normovolemic patients [<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R22">22</xref>]. This is true in both supine (190±140 pM vs. 380±230 pM; P=0.017) and upright posture (480±290 pM vs. 810±370 pM; P=0.019). One would have expected a compensatory increase in both plasma renin activity and aldosterone given the hypovolemia in these patients. This low level of plasma renin activity and aldosterone is a paradox that remains unexplained. These data suggest that abnormalities in the renin-angiotensin-aldosterone axis might have a role in the pathophysiology of POTS by contributing to hypovolemia and impaired sodium retention. Such hypovolemia could be accounted for by a neuropathic process involving the kidney. A significant modulator of renin release is the sympathetic nervous system. Thus perturbations in the renin-aldosterone system might result from partial sympathetic denervation involving the kidney.</p></sec><sec id="s6c"><title>Central Hyperadrenergic POTS</title><p>As a part of the definition, POTS is associated with a hyperadrenergic state (<xref ref-type="table" rid="T1">Table 1</xref>). In many such cases, the hyperadrenergic state is secondary to a partial dysautonomia or hypovolemia. There are some cases, however, in which the primary underlying problem seems to be excessive sympathetic discharge. These patients often have extremely high levels of upright norepinephrine. While we require the upright norepinephrine level to be >600 pg/ml for the diagnosis of POTS, the hyperadrenergic subgroup often has upright norepinephrine level >1000 pg/ml and it is occasionally >2000 pg/ml. These patients sometimes have large increases in blood pressure on standing, indicating that baroreflex buffering is somehow impaired.</p><p>Central hyperadrenergic POTS in its most florid form is much less common than neuropathic POTS, comprising only ~10% of patients. Thus therapy in these cases usually targets a decrease in sympathetic tone both centrally and peripherally.</p><p>Central sympatholytics such as methyldopa or clonidine can be used. Peripheral beta-adrenergic blockade may be better tolerated by these patients than by those with neuropathic POTS.</p></sec><sec id="s6d"><title>Norepinephrine Transporter Deficiency</title><p>A specific genetic abnormality has been identified in a kindred with hyperadrenergic POTS [<xref ref-type="bibr" rid="R30">30</xref>]. These individuals have a single point mutation in the norepinephrine transporter (NET). The resultant inability to adequately clear norepinephrine produces a state of excessive sympathetic activation in response to a variety of sympathetic stimuli. While rare, this mutation has taught us much about the importance of a functional NET.</p><p>Although functional NET mutations might be infrequent, pharmacological NET inhibition is very common. Many antidepressant and attention deficit medications work at least in part through inhibition of NET. This includes traditional drugs such as tricyclic antidepressants, and newer medications which are pure NET inhibitors (e.g. atomoxetine or reboxetine). Both we [<xref ref-type="bibr" rid="R31">31</xref>] and others [<xref ref-type="bibr" rid="R32">32</xref>] have found that pharmacological NET inhibition can recreate an orthostatic tachycardia phenotype in susceptible healthy volunteer subjects. Yohimbine, a central alpha-2 antagonist that will also increase synaptic norepinephrine, can also cause orthostatic tachycardia [<xref ref-type="bibr" rid="R33">33</xref>].</p></sec><sec id="s6e"><title>Mast Cell Activation</title><p>Some patients with POTS have co-existent mast cell activation. These patients have episodic flushing and abnormal increases in urine methylhistamine (the primary urinary metabolite of histamine) [<xref ref-type="bibr" rid="R20">20</xref>]. Methylhistamine should ideally be measured in 2 hour aliquots at the time of a flushing episode and not just in a random 24 hour period. Other associated symptoms include shortness of breath, headache, lightheadedness, excessive diuresis, and gastrointestinal symptoms such as diarrhea, nausea, and vomiting. Flushing can be triggered by long-term standing, exercise, premenstrual cycle, meals, and sexual intercourse. These patients often have a hyperadrenergic response to posture, with both orthostatic tachycardia and hypertension. They demonstrate a vigorous sympathetic vasopressor response during the Valsalva maneuver with a blood pressure overshoot in late phase II and an exaggerated phase IV blood pressure overshoot. It is not clear if mast cell activation, releasing vasoactive mediators, represents the primary event in these patients or if sympathetic activation, through release of norepinephrine, neuropeptide Y and ATP, is the cause of mast cell activation [<xref ref-type="bibr" rid="R34">34</xref>].</p><p>In these patients, beta-adrenergic antagonists can actually trigger an episode and worsen symptoms. Centrally acting agents to decrease the sympathetic nervous system discharge (e.g. methyldopa or clonidine) may prove effective. Alternatively, treatment could target mast cell mediators with a combination of antihistamines (H1- and H2-antagonists) and with the cautious use of non-steroidal agents (high dose aspirin) in refractory cases.</p></sec></sec><sec sec-type="" id="s7"><title>Non-Pharmacological Treatment of POTS</title><p>No therapy is successful for all patients with POTS. Initial efforts should focus on identifying and treating any reversible causes. Potentially contributory medications (especially vasodilators, diuretics, and drugs that inhibit NET) should be withdrawn. If a patient has been through prolonged bedrest, their symptoms will gradually improve as they recondition themselves to upright posture. Treatment should be optimized for any chronic disease that is present. If there is clear evidence of a re-entrant supraventricular arrhythmia, then this should be treated, including with radiofrequency ablation as appropriate. However, radiofrequency sinus node modification for the sinus tachycardia of POTS is not recommended. This often makes the patient’s symptoms worse (and occasionally the patient becomes pacemaker dependent). Specific therapies are summarized in <xref ref-type="table" rid="T2">Table 2</xref>.</p><p>It is important to educate the patient about the nature of the disorder. The patient should avoid aggravating factors such as dehydration, and extreme heat. In order to ensure adequate hydration, we ask our patients to consume 8-10 cups of water daily and to rapidly drink 16 fl oz of water to lower their heart rates [<xref ref-type="bibr" rid="R35">35</xref>]. In addition, they are asked to aggressively increase their sodium intake up to 200 mEq/day. This is often hard to achieve without NaCl tablets 1 gm/tablet TID with meals. Elastic support hose can help to minimize the degree of peripheral venous pooling and enhance venous return. We recommend 30-40 mmHg of counter-pressure and they should come up to the waist. If the stockings are only knee-high, a line of edema can form just above the stockings. Their use can be limited by their tolerability as the stockings can be hot, itchy and uncomfortable. Exercise (both aerobic and resistance training) is also encouraged and has been shown to be beneficial [<xref ref-type="bibr" rid="R36">36</xref>]. In addition to reversing any “deconditioning”, this intervention can also increase blood volume. Vigorous exercise may acutely worsen symptoms and may even result in prolonged fatigue. It is important that patients start slowly and remain within range of their “target heart rate” in the early stages to avoid symptoms that might discourage further exercise.</p><p>Acute blood volume expansion is effective at controlling the heart rate and acutely improving symptoms. Jacob et al. [<xref ref-type="bibr" rid="R37">37</xref>] found that 1 liter of physiological saline infused intravenously over 1 hour decreased the orthostatic tachycardia from 33±5 bpm before the infusion to 15±3 bpm immediately following the infusion. The physiological saline was more effective at heart rate control than were treatments with either an alpha-1 agonist or an alpha-2 agonist. This treatment is not practical on a day to day basis as a medical setting is required to insert the intravenous catheter and infuse the saline. Recently, there have been reports of patients having regular saline infusions, typically 1 liter of normal saline every other day or every day. Many report an improvement in symptoms. However, there are not yet objective data to substantiate such benefit. Further, there is a risk of vascular access complications or infection. At this time, such therapy for patients with POTS should be considered cautiously.</p></sec><sec sec-type="" id="s8"><title>Pharmacological Treatment of POTS</title><p>No medicines are approved by the United States Food and Drug Administration for the treatment of POTS. Thus all agents are used for this disorder are “off label”. Furthermore, there are no pharmacological agents that have been tested in a long-term properly powered randomized clinical trial.</p><p>In patients in whom the presence of hypovolemia is either known or strongly suspected, fludrocortisone (an aldosterone analogue) is often used. Through enhanced sodium retention, it should expand the plasma volume, although there is a paucity of data regarding the exact mechanisms of action. Although fairly well tolerated, side effects can include hypokalemia, hypomagnesemia, worsening headaches, acne, and fluid retention with edema. Another volume expanding agent that may be helpful for short-term use is oral vasopressin (DDAVP). This agent causes the kidney to retain free water, but not sodium. Potential side effects include hyponatremia, edema and headache. Erythropoietin has occasionally proven useful in patients with POTS who are refractory to other forms of therapy. While the primary mode of action is likely an increase in intravascular volume via its increase in red cell mass, erythropoietin also appears to have a direct vasoconstrictive effect, possibly through enhanced red cell mediated nitric oxide scavenging [<xref ref-type="bibr" rid="R38">38</xref>]. Treatment with erythropoietin has many drawbacks including the significant expense and the need for subcutaneous administration.</p><p>Central sympatholytic medications are often useful and well tolerated in patients with the central hyperadrenergic form of POTS, but may not be as well tolerated in neuropathic POTS. Clonidine is an alpha 2 agonist that acts centrally to decrease sympathetic nervous system tone. Clonidine, at doses of 0.05 mg to 0.2 mg PO BID, can stabilize heart rate and blood pressure in patients with a large amount of postganglionic sympathetic involvement. Unfortunately, it can also cause drowsiness, fatigue and worsen the mental clouding of some patients. Methyldopa, a false neurotransmitter, is sometimes more successful in controlling symptoms in these patients at doses of 125 mg to 250 mg PO TID [<xref ref-type="bibr" rid="R39">39</xref>].</p><p>When used in low doses, beta-adrenergic antagonists can be useful. We typical use propranolol 10-20 mg PO BID-QID. While this dose range is small, such doses can often have a significant impact on heart rate control, and higher doses are often not tolerated due to hypotension and fatigue.</p><p>Since a failure of vascular resistance may be an integral part of neuropathic POTS, vasoconstrictors such as midodrine (an alpha-1 agonist) can be employed [<xref ref-type="bibr" rid="R40">40</xref>]. Some patients cannot tolerate this agent due to the unpleasant sensation of scalp tingling or goosebumps. Midodrine can also cause hypertension.</p><p>We recently reported that an unselected group of patients seen in our inpatient research unit were given a trial of the acetylcholinesterase inhibitor pyridostigmine. By increasing the levels of synaptic acetylcholine at both the autonomic ganglia and the peripheral muscarinic parasympathetic receptors, pyridostigmine significantly restrained the heart rate in response to standing in our patients with POTS. We prescribe pyridostigmine 30mg to 60 mg PO TID alone or in combination with low dose propranolol. Pyridostigmine can enhance bowel motility, so it is not always well tolerated in patients with diarrhea-predominant irritable bowel syndrome symptoms.</p><p>While most of the treatments discussed above have focused on the control of heart rate, many patients are also greatly troubled by mental clouding. Modafinil, a stimulant whose mechanism is not yet clear, has been used in some patients with resulting improvement in alertness. However, caution is advised as it may aggravate the orthostatic tachycardia [<xref ref-type="bibr" rid="R41">41</xref>].</p></sec><sec sec-type="" id="s9"><title>Conclusions</title><p>POTS is a disorder of the autonomic nervous system in which many symptoms can be treated. The cardinal manifestation is symptomatic orthostatic tachycardia. The disorder can produce substantial disability among otherwise healthy people. Patients with POTS demonstrate a heart rate increase of ≥30 bpm with prolonged standing (5-30 minutes), often have high levels of upright plasma norepinephrine, and many patients have a low blood volume. Therapies aimed at correcting the hypovolemia and the autonomic imbalance may help relieve the severity of the symptoms. Continued research is vital to better understand this disorder and to differentiate its various subtypes.</p></sec>
An Approach to Catheter Ablation of Cavotricuspid Isthmus Dependent Atrial Flutter	<p>Much of our understanding of the mechanisms of macro re-entrant atrial tachycardia comes from study of cavotricuspid isthmus (CTI) dependent atrial flutter. In the majority of cases, the diagnosis can be made from simple analysis of the surface ECG. Endocardial mapping during tachycardia allows confirmation of the macro re-entrant circuit within the right atrium while, at the same time, permitting curative catheter ablation targeting the critical isthmus of tissue located between the tricuspid annulus and the inferior vena cava. The procedure is short, safe and by demonstration of an electrophysiological endpoint - bidirectional conduction block across the CTI - is associated with an excellent outcome following ablation. It is now fair to say that catheter ablation should be considered as a first line therapy for patients with documented CTI-dependent atrial flutter.</p>	<contrib contrib-type="author"><name><surname>O’Neill</surname><given-names>Mark D</given-names></name><degrees>MB BCH DPhil</degrees></contrib><contrib contrib-type="author"><name><surname>Jais</surname><given-names>Pierre</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Jönsson</surname><given-names>Anders</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Takahashi</surname><given-names>Yoshihide</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Sacher</surname><given-names>Frédéric</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Hocini</surname><given-names>Mélèze</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Sanders</surname><given-names>Prashanthan</given-names></name><degrees>MB BS PhD</degrees></contrib><contrib contrib-type="author"><name><surname>Rostock</surname><given-names>Thomas</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Rotter</surname><given-names>Martin</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Clémenty</surname><given-names>Jacques</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Haïssaguerre</surname><given-names>Michel</given-names></name><degrees>MD</degrees></contrib><aff>Hôpital Cardiologique du Haut Lévêque & Université Victor Segalen Bordeaux II, 33604 Bordeaux, Pessac, France.</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Since the first description of atrial flutter in humans almost a century ago, electrophysiological techniques have evolved to facilitate a comprehensive understanding of the arrhythmia mechanism which in turn, has led to a highly effective therapeutic strategy. This review is confined to cavotricuspid isthmus-dependent flutter and details the development of the understanding, diagnosis and therapy of this common arrhythmia. Specifically, the approach used for the catheter ablation of atrial flutter at our institution is described in detail.</p></sec><sec sec-type="" id="s2"><title>Insights from electrocardiography</title><p>Early recognition of the difference between atrial flutter (AFL) and atrial fibrillation (AF) was made by Jolly and Ritchie with the demonstration of regular atrial activity in the inferior leads of the surface electrocardiogram. Lewis and colleagues demonstrated in animal studies that constant and ordered activation of at least part of the atrium was responsible for the flutter waves on the surface ECG and remarkably deduced that atrial flutter was likely due to intra-atrial circus movement around the venae cavae (for historical review, see Lee [<xref ref-type="bibr" rid="R1">1</xref>]).</p><p>Cavotricuspid isthmus-dependent counterclockwise flutter is the archetypal atrial macro-re-entrant arrhythmia, the anatomical circuit of which has now been well described with both conventional and newer mapping modalities [<xref ref-type="bibr" rid="R2">2</xref>-<xref ref-type="bibr" rid="R5">5</xref>]. The route of the re-entrant circuit is bordered anteriorly by the tricuspid annulus and posteriorly by the venae cavae orifices, the crista terminalis and the Eustachian ridge [<xref ref-type="bibr" rid="R6">6</xref>]. The so-called cavotricuspid isthmus (CTI) serves as the critical zone of slowed conduction which facilitates perpetuation of the re-entrant circuit and also serves as an ideal target for its physical interruption by radiofrequency ablation [<xref ref-type="bibr" rid="R4">4</xref>-<xref ref-type="bibr" rid="R7">7</xref>].</p><p>Flutter waves are atrial complexes of constant morphology, polarity and cycle length in a rate range from 240 to 340 per minute [<xref ref-type="bibr" rid="R6">6</xref>] (<xref ref-type="fig" rid="F1">Figure 1</xref>). The presence of flutter waves in the inferior limb leads and in V1 is specific for cavotricuspid isthmus-dependent flutter [<xref ref-type="bibr" rid="R6">6</xref>]: negative flutter waves in the inferior leads and a positive wave in V1 with transition of morphology across the anterior leads is consistent with a counterclockwise atrial circuit with lateral-to-medial activation of the cavotricuspid isthmus (CTI) while activation in the reverse direction (clockwise) is suggested by positive flutter waves in the inferior leads and a negative deflection in V1, which accounts for 5-10% of all CTI-dependent flutter. There are rare forms of tachycardia with similar P wave morphology, for example atrial tachycardia originating from the ostium of the coronary sinus. Endocardial mapping in this case will demonstrate caudo-cranial activation of both the RA septum and lateral wall, which is not in keeping with counter-clockwise rotation around the previously described flutter circuit.</p><p>An important subgroup of patients to consider are those presenting with atrial flutter following catheter ablation of the left atrium for treatment of atrial fibrillation. It has recently been shown that over 50% of patients with electrophysiologically confirmed counterclockwise CTI-dependent flutter as the arrhythmia of recurrence following LA ablation for AF show upright flutter waves in the inferior leads, a finding which is most likely due to the left atrial debulking and resultant voltage abatement among this group [<xref ref-type="bibr" rid="R8">8</xref>]. Therefore, while the presence of negative flutter waves in the inferior leads is highly suggestive of CTI dependent flutter, their absence does not exclude the diagnosis, most notably in the context of structural heart disease or prior left atrial ablation.</p></sec><sec sec-type="" id="s3"><title>Electrophysiological evaluation of CTI-dependent flutter</title><sec id="s3a"><title>Coronary sinus activation sequence</title><p>Following a thorough evaluation of the surface ECG, a quadripolar deflectable diagnostic catheter (Xtrem, ELA Medical) and mapping/ablation catheter are advanced from the right femoral vein to the coronary sinus (CS) and right atrium respectively. The distal pole of the CS catheter is positioned at 4 to 6 o’clock in the LAO position. If the patient is in AFL at the time of the study, the coronary sinus activation sequence should be from proximal to distal (<xref ref-type="fig" rid="F2">Figure 2</xref>), although equally this could be the case for any tachycardia mechanism originating in the right atrium, and may also be seen in some left atrial flutters, for example counterclockwise perimitral flutter. Conversely, flutter localized to the superior right atrium may activate the distal CS earlier than the proximal CS, via Bachmann’s bundle. Together with the typical ECG findings, a proximal-to-distal CS activation sequence is not diagnostic for, but rather is highly suggestive of CTI-dependent flutter as the underlying tachycardia mechanism.</p></sec><sec id="s3b"><title>Endocardial mapping</title><p>Once the diagnosis of CTI-dependent flutter is suspected from the surface ECG and the CS activation sequence, more detailed intra-atrial mapping is performed. Using the CS as a stable reference, up to eight points are sequentially mapped within the right atrium, including approximately four points around the tricuspid annulus, and their timing relative to flutter wave onset and the clearest CS signal determined, to evaluate the direction of impulse propagation during tachycardia. For the majority of patients with CTI-dependent counterclockwise flutter, the atrial electrogram recorded on the mid-isthmus lies within the plateau phase of the flutter wave, between 50 and 90ms prior to the onset of the pronounced negative deflection associated with the septal ascent of the depolarisation wavefront. Demonstration of a right atrial activation time which is significantly shorter than the tachycardia cycle length or the identification of colliding activation fronts in the right atrium are features which are highly suggestive of a left atrial flutter origin [<xref ref-type="bibr" rid="R5">5</xref>].</p></sec><sec id="s3c"><title>Entrainment manoeuvres and post pacing interval</title><p>In addition to this activation mapping, entrainment manoeuvres can be employed in the right atrium and coronary sinus, to confirm and exclude the right and left atria as the chamber of arrhythmia origin respectively. Entrainment involves pacing from multiple, separate sites within the right atrium at cycle lengths of 10-20ms faster than the tachycardia cycle length, observing its effect on flutter wave morphology [<xref ref-type="bibr" rid="R9">9</xref>] and estimating proximity of pacing site to tachycardia circuit by analysis of the post pacing interval. The pacing site is considered to lie within the tachycardia circuit when the post pacing interval is within 30ms of the tachycardia cycle length. Entrainment from sites which are outside the flutter circuit will demonstrate manifest fusion on the surface ECG and the PPI will exceed the flutter cycle length by more than 30ms. Pacing performed from within the circuit but outside a protected isthmus demonstrates manifest entrainment, characterized by surface ECG fusion, progressive fusion at faster pacing rates and all sites which are orthodromically activated having a PPI equal to the flutter cycle length. Transient entrainment performed from within the protected CTI isthmus demonstrates a long stimulus-to-p wave interval, in keeping with its role as an area of slow conduction in which unidirectional block of the antidromic wave front probably occurs. Mapping of the septal aspect of the CTI demonstrates the earliest atrial electrogram relative to the onset of the flutter wave on the surface ECG, and entrainment from this position demonstrates a short stimulus-to-p wave interval with concealed fusion, confirming this as the likely exit site from the zone of slow conduction critical to maintenance of the tachycardia. This manoeuvre is repeated for a minimum of two right atrial sites, usually incorporating the RA septum and lateral wall at the very least.</p><p>Many patients with a history of atrial flutter who been referred for ablation therapy are in sinus rhythm at the time of their procedure. If the 12 lead electrocardiogram recorded during the clinical arrhythmia is consistent with CTI-dependent flutter, and the medical history gives no cause to suspect a left atrial origin, it is our practice to proceed directly to radiofrequency ablation of the CTI rather than to attempt induction.</p></sec></sec><sec sec-type="" id="s4"><title>Catheter ablation of CTI-dependent flutter</title><p>Electrophysiological study and ablation is performed in the post absorptive state and under light sedation (midazolam 2-5mg and/or analgesia (nalbuphine 10 to 20mg). All antiarrhythmic medication, with the exception of amiodarone, is discontinued at least five half lives prior to the procedure. For patients already receiving oral anticoagulation, this is discontinued 48 hours prior to the procedure and continued for one month after the procedure. Otherwise, patients do not routinely receive anticoagulation medication before, during or after an elective CTI ablation.</p><p>All surface ECG leads are recorded continuously throughout the procedure to facilitate rapid review however only leads I, II, III and V1 are displayed in real time. Bipolar intracardiac electrograms are recorded in the coronary sinus using a quadripolar deflectable catheter and from the atrial endocardium using a conventional quadripolar thermocouple D curve 8mm tip catheter (Biosense Webster). All electrograms are sampled at 1KHz, filtered from 30 to 500Hz, amplified at 0.1mV/cm and displayed at a sweep speed of 100ms using a Windows-based digital amplifier/recording system for offline analysis (Bard Electrophysiology).</p><p>Ablation of the CTI is performed in the patient’s presenting rhythm, be it flutter or sinus rhythm. In the case of the latter, pacing is performed from the proximal pole of the CS (cycle length 600ms) as this location is more stable than a pacing catheter positioned in the inferolateral right atrium.</p><sec id="s4a"><title>The anatomical target</title><p>It is now well established that the anatomical target for interruption of the macro re-entrant circuit comprising typical atrial flutter is the cavo-tricuspid isthmus [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R5">5</xref>]. Using an 8mm tip ablation catheter, the linear lesion is made by a continuous application of RF energy beginning at the ventricular end of the CTI (A:V electrogram amplitude ratio of ~1:2) and dragging the catheter towards the atrial aspect of the CTI with a 60-90s delay at successive ablation points before moving the catheter. Radiofrequency energy (550Hz unmodulated sine-wave output) is delivered through a Cordis Stockert generator in temperature control mode with the temperature limited to 60°C and the power to 60W. Ablation is not routinely performed at the septal aspect of the CTI or inside the ostium of the CS in order to minimize the risk of circumflex artery injury, atrioventricular block or cardiac perforation. Nevertheless, lateral or septal positions are occasionally used when linear block cannot be achieved by ablation in the medial isthmus.</p><p>At our institution, an irrigated 3.5 mm tip catheter is used in all patients undergoing pulmonary vein isolation and left atrial ablation for atrial fibrillation. Unless there is evidence of persisting bidirectional conduction block across the CTI from a previous catheter intervention, it is our practice to ablate the CTI routinely in these patients following completion of the left atrial procedure. In those patients admitted electively for conventional CTI ablation for typical flutter, a non-irrigated 8mm tip catheter is used (Cordis Webster D curve).</p><p>The length, depth and anatomical complexity of the cavotricuspid isthmus can vary considerably between patients with a resultant impact on the success of radiofrequency ablation [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R11">11</xref>]. To facilitate effective ablation at the ventricular aspect of a “long” CTI, a long sheath can be used. Similarly, where there is difficulty in completing the IVC end of the line because of a prominent Eustachian ridge preventing a smooth drag back across the isthmus, the catheter may be looped and apposed to the isthmus from the atrial side and advanced from the IVC towards the Eustachian ridge. In patients where conduction block across the isthmus cannot be achieved with a conventional catheter, the use of an irrigated tip catheter has been shown to facilitate achievement of bidirectional CTI conduction block [<xref ref-type="bibr" rid="R12">12</xref>]. Interestingly, there is a significant correlation between the surface atrial ECG amplitude and the amount of RF energy required to achieve complete isthmus block, a finding which may be of use in the preprocedural choice of ablation catheter for typical atrial flutter [<xref ref-type="bibr" rid="R13">13</xref>]. Paradoxically, higher voltages recorded at the CTI using an electroanatomic approach do not correlate with reduced ability to achieve bidirectional conduction block when ablation is performed with a conventional 8mm tip RF catheter [<xref ref-type="bibr" rid="R14">14</xref>].</p></sec><sec id="s4b"><title>The electrophysiological target</title><p>The flutter wave morphology in counterclockwise CTI dependent flutter is characteristic, consisting of a downsloping “plateau” phase, followed by a short, sharp negative deflection, then a sharp positive deflection with a positive ‡overshoot’ which in turn leads to the next plateau. The anatomical location of the medial part of the CTI has been shown to correlate well with the timing of the atrial electrogram relative to the plateau of the flutter wave and ablation targets the atrial electrogram occurring simultaneously with the middle of the plateau phase of the flutter wave in lead II (<xref ref-type="fig" rid="F2">Figure 2</xref> & <xref ref-type="fig" rid="F3">3</xref>). When ablation is performed during flutter, careful attention to this relationship ensures radiofrequency energy is delivered perpendicular to the advancing wavefront and allows immediate recognition of displacement of the catheter medially or laterally, based on delay or advancement of the atrial electrogram respectively, relative to the plateau [<xref ref-type="bibr" rid="R15">15</xref>]. When performed during proximal coronary sinus pacing, RF energy is sequentially delivered at electrogram sites in the isthmus region with a constant stimulus-to-atrial electrogram time across the length of the isthmus.</p></sec><sec id="s4c"><title>The end point</title><p>Completion of the linear lesion by radiofrequency energy requires subsequent quantitative demonstration of bidirectional conduction block across the CTI in order to achieve the best outcomes for catheter ablation of atrial flutter. Complete bidirectional conduction block across the CTI may be defined as: a descending wave front on the lateral atrial wall up to the line of block during proximal CS pacing representative of a reversed atrial depolarization sequence; a greater activation delay measured to the second double potential recorded on the line of block when compared with the delay measured to the atrial electrogram further lateral to the line during proximal CS pacing (<xref ref-type="fig" rid="F4">Figure 4</xref>); a decreasing activation delay measured medial to the line when pacing from immediately lateral to the line and from further lateral to the line [<xref ref-type="bibr" rid="R16">16</xref>]. Other electrophysiological features consistent with conduction block across the CTI include a reversal in bipolar electrogram polarity immediately lateral to the line during proximal CS pacing, signifying reversal of the vector of activation [<xref ref-type="bibr" rid="R17">17</xref>] (<xref ref-type="fig" rid="F4">Figure 4</xref>); change in morphology of the unipolar electrogram (from QS, rS or RS to R or Rs) recorded immediately lateral to the line of block during proximal CS pacing [<xref ref-type="bibr" rid="R18">18</xref>]; mapping of a corridor of double potentials with an intervening isoelectric interval along the line of ablation [<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R20">20</xref>]; demonstration of a reversal in the direction of atrial depolarization during proximal CS pacing either by sequential right atrial mapping with the ablation catheter or the use of a multipolar diagnostic catheter positioned along the lateral aspect of the tricuspid annulus. In practice, a craniocaudal activation of the septum (PCS activated after His bundle region) during pacing of the low lateral RA indicates an counterclockwise block in the isthmus, whereas a craniocaudal activation (low after high RA) of the lateral RA during pacing of the PCS indicates a clockwise block, and is reflected by increased positivity of the terminal component of the P wave in lead II (<xref ref-type="fig" rid="F5">Figure 5</xref>). When taken together with an abrupt increase in the degree of separation of atrial bipolar electrograms recorded on the line of block, and the highly unlikely situation of ablation resulting in unidirectional clockwise conduction block [<xref ref-type="bibr" rid="R21">21</xref>], it is usually sufficient to document the atrial electrogram separation and confirm maximum conduction delay immediately medial to the line during low lateral right atrial pacing.</p></sec></sec><sec sec-type="" id="s5"><title>Complications</title><p>Although uncommon, complications have been reported for ablation confined to the CTI. Injury to the right coronary artery has been described [<xref ref-type="bibr" rid="R22">22</xref>]. The atrioventricular (AV) node may be adversely affected by modification of local vagal afferent input related to RF stimulation or by a global vagal response to pain during energy delivery [<xref ref-type="bibr" rid="R23">23</xref>]. Of course, direct thermal injury to the node or its critical inputs may also result in transient or even permanent AV block.</p></sec><sec sec-type="" id="s6"><title>Conclusion</title><p>Radiofrequency ablation has been shown to be more effective than anti-arrhythmic medication in preventing recurrences of and reducing hospitalizations for CTI-dependent atrial flutter [<xref ref-type="bibr" rid="R24">24</xref>]. Catheter ablation is safe, cost-effective and should be considered as the first line treatment of choice for the majority of patients [<xref ref-type="bibr" rid="R25">25</xref>].</p></sec>
Gene Therapy in Cardiac Arrhythmias	<p>Gene therapy has progressed from a dream to a bedside reality in quite a few human diseases. From its first application in adenosine deaminase deficiency, through the years, its application has evolved to vascular angiogenesis and cardiac arrhythmias. Gene based biological pacemakers using viral vectors or mesenchymal cells tested in animal models hold much promise. Induction of pacemaker activity within the left bundle branch can provide stable heart rates. Genetic modification of the AV node mimicking beta blockade can be therapeutic in the management of atrial fibrillation. G protein overexpression to modify the AV node also is experimental. Modification and expression of potassium channel genes altering the delayed rectifier potassium currents may permit better management of congenital long QT syndromes. Arrhythmias in a failing heart are due to abnormal calcium cycling. Potential targets for genetic modulation include the sarcoplasmic reticulum calcium pump, calsequestrin and sodium calcium exchanger. Lastly the ethical concerns need to be addressed.</p>	<contrib contrib-type="author"><name><surname>V</surname><given-names>Praveen S</given-names></name><degrees>MBBS, MD, DNB</degrees></contrib><contrib contrib-type="author"><name><surname>Francis</surname><given-names>Johnson</given-names></name><degrees>MBBS, MD, DM, FCSI</degrees></contrib><contrib contrib-type="author"><name><surname>K</surname><given-names>Venugopal</given-names></name><degrees>MBBS, MD, DM, FCSI</degrees></contrib><aff> Department of Cardiology, Medical College, Calicut, Kerala, India</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Gene therapy is defined as the transfer of nucleic acids to somatic cells as therapeutically useful molecules. Genetic defects can be corrected or gene products be expressed by gene therapy. This approach has many potential applications, the most obvious being the treatment of inherited monogenic disorders like cystic fibrosis. Human genome has approximately 30,000 genes. The genetic diversity is amplified by alternate splicing of mRNA and post translational modification of proteins. The possible gene targets for arrhythmias is very large. Anti arrhythmic agents act by blocking the ion channels. These antiarrhythmics have poor channel specificity and may cross react with other ion channels. In addition they have poor affinity for the channels and have a narrow therapeutic window. The molecular targets of arrhythmia management are the ion channels and the modulators of ion channels like G proteins [<xref ref-type="bibr" rid="R1">1</xref>].</p></sec><sec sec-type="" id="s2"><title>Vectors for gene therapy</title><p>A vector is the vehicle commonly used to introduce the gene to the target cell. Vectors may be RNA or DNA viruses or non viral in nature. Viruses which have the capacity to incorporate themselves in the host genome are used as vectors for gene therapy. The commonly used viral vectors are genetically modified retroviruses, adenoviruses, adeno associated viruses and lentiviruses. These viral vectors are made replication deficient to ensure safety, but requires large amounts of vector particles for efficacy. Non viral vectors based on plasmids, DNA- lipid complexes and naked DNA are also used since they lack foreign proteins and avoid immunological problems. None of the currently available vectors satisfy the criteria of an ideal gene therapeutic system. The feasibility of gene transfer has been demonstrated in both animals and humans.</p><p>In contradistinction to the experimental laboratory, the extent of gene transfer and expression is low in clinical settings. The period during which a newly introduced gene is expressed is variable and differs with the tissue, but is often short. For example, early-generation non-viral vectors express the gene at maximum levels only for a few days [<xref ref-type="bibr" rid="R2">2</xref>]. Many adenoviral vectors express the gene for 2-3 weeks [<xref ref-type="bibr" rid="R3">3</xref>]. Non viral vectors also have short duration of gene expression. Short duration of gene expression necessitates repeat dosing, although less efficacious. By contrast, expression from adeno-associated viral vectors may not peak for several weeks, but then remain constant in some tissues for several months [<xref ref-type="bibr" rid="R4">4</xref>]. Retroviruses produce a long lasting effect by integration of the transfected gene into the host genome [<xref ref-type="bibr" rid="R5">5</xref>]</p></sec><sec sec-type="" id="s3"><title>Methods of gene delivery</title><p>The classical methods of vector delivery are direct injection into the myocardium, infusion through the coronary arteries or administration to the epicardium.Various novel methods of transfection have been tried in animal models, including DNA polymer coating on inert materials and subsequent transfer to the atrial myocardium, with sustained gene activity [<xref ref-type="bibr" rid="R6">6</xref>]. Intracoronary perfusion is another modality of gene transduction with near complete expression under optimal conditions [<xref ref-type="bibr" rid="R7">7</xref>]. The gene transfer efficiency depends on the coronary flow rate, virus concentration, virus exposure time and microvascular permeability. Agents which increase the microvascular permeability have been used to enhance the delivery. Each disease has its own target tissue and the amount of gene product required for treatment. So only few generalizations can be made about the vector selection and the method of gene delivery.</p></sec><sec sec-type="" id="s4"><title>The need for gene therapy in cardiac arrhythmias</title><p>Antiarrhythmic medications suppress arrhythmias, but their systemic effects are often poorly tolerated and their proarrhythmic effects increase mortality. Radiofrequency ablation can cure only a limited number of arrhythmias. Implantable devices can be curative for bradyarrhythmias and lifesaving for tachyarrhythmias, but require a lifetime commitment to repeated procedures, have a significant expense, and may lead to severe complications. The need for new treatment strategies for cardiac arrhythmias has motivated the continuing development of gene therapeutic options Gene therapy may prove to be a less intrusive, long-term solution to arrythmias than pacemakers or antiarrhythmic agents. Gene therapy for arrhythmias is a field still in its infancy.</p><p>The ongoing research into gene therapy for cardiac arrhythmias can be briefly discussed under: <list list-type="order"><list-item><p>Biological pacemakers</p></list-item><list-item><p>AV nodal modification</p></list-item><list-item><p>Long QT syndromes</p></list-item><list-item><p>Arrhythmias of cardiac failure</p></list-item><list-item><p>Ventricular arrhythmias</p></list-item></list></p></sec><sec sec-type="" id="s5"><title>Biological pacemakers</title><p>Conventional electronic pacemakers although highly efficacious, have a finite battery life and problems related to lead and circuitry. Biological pacemakers aim a near physiological pacing by molecular or cellular targeting.Biological pacemakers can be cell therapy based or gene therapy based. Strategies have included gene therapy using naked plasmids or viral vectors and cell therapy with both adult human mesenchymal stem cells (hMSCs) and human embryonic stem cells [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R9">9</xref>]. Over the past few years, gene therapy has been used to explore the overexpression of β2-adrenergic receptors, the down-regulation of inward rectifier current, and the overexpression of pacemaker current as potential sources of biological pacemakers [<xref ref-type="bibr" rid="R10">10</xref>]. The protein product of HCN2 gene is a candidate for the pacemaker current. Rosen et al from Columbia university have recreated the biological pacemaker by administering the pacemaker gene HCN2 via viral vector or in an hMSC platform to produce pacemaker function in the intact canine heart. Adenovirus containing the HCN2 gene [ad HCN2] on introduction in canine models produced spontaneous rhythms.</p><p>Another potential approach to a biological pacemaker is the induction of pacemaker activity within the left bundle branch, thereby providing a ventricular escape rhythm with physiologically acceptable rates. The long-term stability and feasibility of this approach remain to be tested [<xref ref-type="bibr" rid="R11">11</xref>]. Adeno viral vectors have been tried in animal models to create pacemaker clones from ventricular myocytes. The inward rectifier current [I<sub>K1</sub>] is responsible for the maintenance of the resting membrane potential. The main subunit of this is Kir 2.1. The Kir 2.1 gene was mutated to make it a dysfunctional channel (a dominant-negative), inserted into an adenoviral vector and delivered to the hearts of guinea pigs. The enhanced automaticity of the ventricular myocytes subsequent to the gene transfer conferred pacemaker like properties [<xref ref-type="bibr" rid="R12">12</xref>].</p></sec><sec sec-type="" id="s6"><title>AV nodal modification</title><p>As in case of biological pacemakers, the field is limited to animal studies. The major area of research in gene therapy for tachyarrhythmias now is on AV nodal modification to produce AV nodal block mimicking beta blockade.</p><p>Genetic modification of AV node in an intracoronary perfusion model of porcine heart was reported by Donahue et al from John Hopkins University [<xref ref-type="bibr" rid="R13">13</xref>]. They infected porcine hearts with Adbetagal (recombinant adenovirus expressing Escherichia coli beta-galactosidase) or with AdGi [adenovirus encoding the Galphai2 subunit]. Galphai2 overexpression suppressed baseline atrioventricular conduction and slowed the heart rate during atrial fibrillation without producing complete heart block. In contrast, expression of the reporter gene beta-galactosidase had no electrophysiological effects. Reporter genes are nucleic acid sequences encoding easily assayed proteins.</p><p>AV nodal gene transfer can decrease heart rates in animal models of atrial fibrillation. Inhibitory G protein overexpression can prolong the AV nodal refractory period with slowing of conduction resulting in reduced ventricular rates in atrial fibrillation [<xref ref-type="bibr" rid="R14">14</xref>]. In mouse models, cardiac overexpression of adenosine receptor [A(3)AR] resulted in gene dose-dependent AV block and pronounced sinus nodal dysfunction in vivo. These may have future therapeutic implications for SA and AV nodal modification [<xref ref-type="bibr" rid="R15">15</xref>]. Focal gene transfer to the AV node to produce a genetic calcium channel blocker has been successful in experimental settings. Over expression of the ras related small G- protein [Gem] in AV node slowed AV nodal conduction controlling the heart rate in atrial fibrillation [<xref ref-type="bibr" rid="R16">16</xref>].</p></sec><sec sec-type="" id="s7"><title>Long QT syndromes ( LQTS)</title><p>Sympathetic imbalance was previously thought to be responsible for this disease. Since 1991, 7 LQTS genes have been discovered and more than 300 mutations have been identified to account for the disease. Acquired LQTS, is presumed to be due to the blockade of the rapid component of the delayed rectifier K+ current (I<sub>Kr</sub>). Blockage of the I<sub>Kr</sub> prolongs the QT interval and increases the dispersion of repolarization, predisposing to torsades de pointes. Molecular genetic analysis could be useful to unravel subclinical mutations or polymorphisms.</p><p>Individuals with cardiac potassium channel missense mutation, Q9E-hMiRP1 are predisposed to develop QT prolongation after clarithromycin administration. Experimental studies have demonstrated that cells transfected with plasmid DNA containing Q9E-hMiRP1 have reduced potassium currents on exposure to clarithromycin.</p><p>Site specific gene therapy for arrhythmias by transfecting cell clones with the K+ channel genes is a feasible approach to the management of LQTS [<xref ref-type="bibr" rid="R17">17</xref>]. Mutated K+ channels resulting in loss of function have been implicated in LQT 1 and 2. The potassium channel alpha subunit genes KCNH2 [HERG] and KCNQ1 [KvLQT1] responsible for I<sub>Kr</sub> and I<sub>Ks</sub> respectively are mutated in LQTS. In normal epithelia, KCNE3 [E3] interacts with the KVQT1 [Q1] thereby augmenting the potassium currents. E3 subunit can be genetically expressed in cardiac tissues [where it is normally scarce] to abbreviate the action potential duration and enhance the potassium current.This potentially prevents arrhythmias in LQTS. Adenovirus encoded E3 introduced into guinea pig ventricles shortened QT interval on homogenous transduction, but could be potentially arrhythmogenic if transduction is heterogenous [<xref ref-type="bibr" rid="R18">18</xref>].</p></sec><sec sec-type="" id="s8"><title>Arrhythmias of cardiac failure</title><p>Abnormal calcium cycling plays an important role in the genesis of contractile dysfunction and arrhythmias in the setting of heart failure. Genetic modulation of the sarcoplasmic reticulum calcium ATP ase pump [SERCA] can have an ameliorating effect on the arrhythmias of cardiac failure.</p><p>Delayed repolarization predisposes the failing heart to ventricular arrhythmias and this represents a logical target for gene therapy. The calcium ATPase SERCA1 was coexpressed with the potassium channel Kir2.1 in guinea pig hearts. Such myocytes had bigger calcium transients and shorter action potentials. In vivo, repolarization was abbreviated, but contractile function remained unimpaired [<xref ref-type="bibr" rid="R19">19</xref>]. This shortening of repolarization prevents arrhythmias. Coexpression of SERCA1 prevented the decrease in contractility due to shortening of action potential. This interesting observation is useful in prevention of arrhythmias in the setting of heart failure.</p><p>Calsequestrin, the high capacity calcium binding protein expressed in sarcoplasmic reticulum, also positively controls the rate of calcium release during excitation-contraction coupling. Mutations in the calsequestrin gene have been linked to arrhythmias and sudden death. For example, the recessive form of catecholaminergic polymorphic ventricular tachycardia is associated with calsequestrin mutations [<xref ref-type="bibr" rid="R20">20</xref>]. Modulation of the gene for calsequestrin could be one of the numerous potential targets for therapy.</p><p>The main pathway for calcium efflux from the cell is the Na - Ca exchanger (NCX), a membrane antiporter and a determinant of both the electrical and contractile state of the heart. NCX causes efflux of one Ca(2+) for three Na+ transported into the cell. Enhanced expression of NCX has recently been recognised as one of the molecular mechanisms that contributes to reduced Ca(2+) release, impaired contractility and an increased risk of arrhythmias during the development of cardiac hypertrophy and failure. The NCX also plays a crucial role in the pathogenesis of arrhythmias and cellular injury associated with ischaemia and reperfusion. Hence, NCX blockade represents a potential therapeutic strategy for treating cardiac disease.However, its reversibility and electrogenic properties must be taken into consideration when predicting the outcome. NCX inhibition has been demonstrated to be protective against ischaemic injury and to have a positive inotropic and antiarrhythmic effect in failing heart cells [<xref ref-type="bibr" rid="R21">21</xref>].</p><p>Failing hearts show a delay of repolarization and prolongation of action potential due to diminished potassium currents, which is proarrhythmic. Adenovirus over- expressing the potassium channels [AdShK] can reverse the prolongation of action potential duration, thereby potentially preventing arrhythmias [<xref ref-type="bibr" rid="R22">22</xref>].</p></sec><sec sec-type="" id="s9"><title>Ventricular arrhythmias</title><p>Electrical alternans has been linked to the development of ventricular arrhythmias.Increasing the rapid component of the delayed rectifier current (I<sub>Kr</sub>) may suppress electrical alternans and may be antiarrhythmic. I<sub>Kr</sub> in isolated canine ventricular myocytes was increased by infection with an adenovirus containing the gene for the pore-forming domain of I<sub>Kr</sub> [human ether-a-go-go gene (HERG)]. The voltage at which peak I<sub>Kr</sub> occurred was significantly less negative in HERG-infected myocytes, thereby shifting the steady-state voltage-dependent activation and inactivation curves to less negative potentials [<xref ref-type="bibr" rid="R23">23</xref>]. This observation lend support to the idea that increasing I<sub>Kr</sub> may be a viable approach to suppressing electrical alternans thereby suppressing arrhythmias.</p></sec><sec sec-type="" id="s11"><title>Current problems with gene therapy</title><p>As alluded to earlier, gene medicine is still in its infancy. Except for a few human trials, majority of trials to date have been in experimental animals. The expected result from gene therapy is a permanent cure of arrhythmias with a single stage treatment with minimal or no adverse effects. Obviously we are far from the ideal. In the field of arrhythmias, an expectant waiting for the scenario to unfold in full is needed. The available vectors to date are far from ideal. Problems with vectors include variability in transfection capabilities,inefficient delivery at site,limited period of gene expression, and immunogenicity. The tissue expression of many genes are transient. The level and efficiency of expression of many trans genes are suboptimal. Many viral vectors are potentially immunogenic and carcinogenic. The interaction between vector and host genome can result in the vector being rendered replicant and lose the therapeutic gene. Another area of concern is that the currently available vectors have less ability to transduce vascular cells than nonvascular cells. This could hamper efforts at cardiovascular gene transfer. Traditional vectors need to be engineered to increase their affinity for the target tissue or cell and prevent transduction to other cells [<xref ref-type="bibr" rid="R24">24</xref>].</p><p>Succesful transfer of the therapeutic gene to all the myocytes at the target site is not fully achieved in the experimental settings. The receptors for many viral vectors are present in many tissues therby limiting the specificty of gene delivery.</p><p>Many an arrhythmia with a diffuse substrate like atrial fibrillation needs the gene to be delivered to a wide area [the whole of the atrium]. The transfer methods like direct injection into myocardium fails to deliver the gene a short distance from the injecton site. A special area of concern in arrhythmia gene therapy is the potential for the remedy itself being arrhythmogenic. Incomplete restoration may in also be arrhythmogenic. In a non linear system like biological organisms, making an isolated change in a specific aberration will result in restoration of normal function only if the defect is truly isolated and is the direct cause of the phenotypic response. The long term response of a genetic modification in the myocardium is unknown at present. Only continued research and time can answer these problems with certainty.</p></sec><sec sec-type="" id="s12"><title>The Future</title><p>Newer refinements in vector development and design are needed to have better transduction in cardiovascular tissue. Cell specific regulatory elements and promoters to selectively target the cardiac tissue is a potential area of interest [<xref ref-type="bibr" rid="R25">25</xref>]. Application at bed side awaits further refinement in gene delivery. Bactofection (bacterial gene delivery) as an alternative to viral vectors has been proposed [<xref ref-type="bibr" rid="R26">26</xref>]. Hybrid vectors, gutted vectors and new generation non viral vectors may hold the key to future research.</p></sec><sec sec-type="" id="s13"><title>Conclusion</title><p>Given a wide plethora of potential targets for gene therapeutic strategies, the possible applications are unlimited. We have still a long way to go from animal models to the level of safe and efficacious application at the bedside. This awaits more refinements in gene delivery methods and vector designs. Not to be forgotten are the increasing concerns about safety. A regulated and sustained target tissue expression of the transduced gene with a wide index of safety should be the ultimate goal of any genetic intervention.</p></sec>
Electrophysiological Mechanisms of Atrial Flutter	<p>Atrial flutter (AFL) is a common arrhythmia in clinical practice. Several experimental models such as tricuspid regurgitation model, tricuspid ring model, sterile pericarditis model and atrial crush injury model have provided important information about reentrant circuit and can test the effect of antiarrhythmic drugs. Human atrial flutter has typical and atypical forms. Typical atrial flutter rotates around tricuspid annulus and uses the crista terminalis and sometimes sinus venosa as the boundary. The IVC-tricuspid isthmus is a slow conduction zone and the target of radiofrequency ablation. Atypical atrial flutter may arise from the right or left atrium. Right atrial flutter includes upper loop reentry, free wall reentry and figure of eight reentry. Left atrial flutter includes mitral annular atrial flutter, pulmonary vein-related atrial flutter and left septal atrial flutter. Radiofrequency ablation of the isthmus between the boundaries can eliminate these arrhythmias.</p>	<contrib contrib-type="author"><name><surname>Tai</surname><given-names>Ching- Tai</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Chen</surname><given-names>Shin-Ann</given-names></name><degrees>MD</degrees></contrib><aff>Division of Cardiology, Department of Medicine, National Yang-Ming University School of Medicine, Taipei Veterans General Hospital, Taiwan, R.O.C.</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Atrial flutter (AFL) is a frequent arrhythmia second only to atrial fibrillation in clinical practice. Since Jolly and Ritchie first recorded AFL in 1910, over the next several decades there was surprisingly little progress in understanding its mechanisms [<xref ref-type="bibr" rid="R1">1</xref>]. In 1921, Sir Thomas Lewis and his colleagues were the first to investigate the mechanism of this arrhythmia [<xref ref-type="bibr" rid="R2">2</xref>]. Using a combination of epicardial maps and ECG recordings from a canine model of AFL induced by rapid atrial pacing, they showed that the activation circulated in either a cranial-caudo or a caudo-cranial direction in the right atrium. They concluded that AFL was due to intraatrial circus movement around the vena cava. Subsequent works that supported the notion that AFL was due to reentry included those of Rosenbleuth and Garcia-Ramos who created a crush injury model of this arrhythmia by making a lesion between the vena cava in 1947 [<xref ref-type="bibr" rid="R3">3</xref>]. Based on the epicardial maps, the authors inferred that the reentry loop circled around the atrial crush lesion. In 1947 and 1948, Scherf et al injected aconitine into the atrial subepicardium and found a uniformly discharging ectopic focus to provide an evidence that AFL was due to a focal activation mechanism [<xref ref-type="bibr" rid="R4">4</xref>]. However, it was not until the last three decades that numerous studies in animal models and human AFL developed and advanced in understanding the mechanism of AFL. The purposes of this review article are to review the recent progress in the experimental models of AFL and clinical studies of human AFL and to improve the ablative therapy of AFL.</p></sec><sec sec-type="" id="s2"><title>Experimental Models of AFL</title><sec id="s2a"><title>Tricuspid Regurgitation Model</title><p>Boyden et al cut the chorda tendineae of the anterior and septal leaflets of the tricuspid valve using a knife and produced some degree of tricuspid insufficiency and volume overload induced enlargement of the right atrium [<xref ref-type="bibr" rid="R5">5</xref>]. From their endocardial mapping, they found a progressive delay and inhomogeneity in conduction with successive stimuli. After a critical number of stimuli, a fixed area of functional block occurred at one site. With subsequent stimuli, the line of block was maintained by continuous collision of a wavefront from the previous beat with the current stimulated beat. When the stimulation is terminated, the paced wave front from the last paced beat begins to propagate across the right atrial free wall and produces reentry by circling around the line of functional block. In all episodes of sustained AFL, the rhythm was due to reentrant excitation in tissues of the right atrium. Impulse propagation was either clockwise or counterclockwise and functional block provided an important boundary of AFL.</p></sec><sec id="s2b"><title>Tricuspid Ring Model </title><p>Frame et al made an intercaval incision connected with a second incision in the right atrial free wall to create a Y-shaped lesion [<xref ref-type="bibr" rid="R6">6</xref>]. The atrial flutter can be easily and reliably induced by programmed electrical stimulation. The range of cycle length was 140 to 170 ms. The duration of the excitable gap was 60 to 80 ms , which represented 40% to 50% of the atrial flutter cycle length. High density mapping using a computer multiplexing system demonstrated that the reentrant impulse circulated around the tricuspid annulus in a clockwise or counterclockwise direction. In this model, reentry occurred entirely in normal fast response tissue, with no single area of markedly slower conduction. Sodium channel blocking drugs slow conduction in all parts of the reentrant pathway. The Y-shaped lesion and tricuspid annulus were two fixed barriers of this atrial flutter.</p></sec><sec id="s2c"><title>Sterile Pericarditis Model</title><p>After pericardiotomy, the atrial surfaces were then generously dusted with sterile talcum powder, a single layer of gauze is then put on the right and left atrial free walls, and the pericardiotomy is repaired [<xref ref-type="bibr" rid="R7">7</xref>]. The atrial flutter could be induced by rapid atrial pacing in the first 4 postoperative days. During the onset of atrial flutter, there is a transional rhythm like atrial fibrillation. A period of atrial fibrillation activated the right atrium through wave fronts which produced a relatively large localized area of slow conduction. Then, unidirectional conduction block of the wave front occurred for one beat in the area of slow conduction and this permitted the unblocked wave front to turn around an area of functional block, thereby initiating the reentry. Sequential site atrial mapping using a hand-held probe during atrial flutter in the open-chest state demonstrated either clockwise or counterclockwise reentrant excitation in the right atrial free wall. The mean sustained atrial flutter cycle length was 131 ± 20 ms, with a range of 100 to 170 ms. Double potentials were recorded in the center of the reentrant circuit during atrial flutter and denoted a line of functional conduction block with each deflection representing activation on either side of the area of functional block. Fractionated electrograms were recorded from areas of slow conduction, principally the pivot points of the reentrant wave front.</p></sec><sec id="s2d"><title>Atrial Crush Injury Model</title><p>Atrial crush injury was created with a surgical clamp placed on the right atrial free wall, producing a lesion parallel to and 1.5 cm above the atrioventricular ring, extending from the base of the right atrial appendage 1.5 to 2.5 cm posteriorly toward the intercaval zone and 3 to 4 cm wide [<xref ref-type="bibr" rid="R8">8</xref>]. Atrial flutter was induced by programmed atrial stimulation or rapid atrial pacing. The atrial flutter cycle length was 140 to 150 ms. During atrial flutter the earliest atrial activation relative to F wave onset was noted in the right atrium and the reentrant wave front revolved around the crush injury. Inverted F wave atrial flutter was typically associated with a counterclockwise activation pattern around the crush injury, with initial activation of the left atrium posteriorly. In contrast, upright F wave atrial flutter was typically associated with a clockwise activation pattern around the crush injury, with initial activation of the left atrium anteriorly. Direct induction of atrial flutter was associated with the development of progressive conduction delay in the isthmus between the crush injury and the tricuspid annulus, eventually culminating in unidirectional block and initiation of reentry. In many instances, however, the onset of atrial flutter followed a brief period of atrial fibrillation. The conduction velocity was generally slower in the isthmus between the crush injury and tricuspid annulus.</p></sec></sec><sec sec-type="" id="s3"><title>Human AFL</title><sec id="s3a"><title>Typical AFL</title><p>Human AFL is defined by the undulating P wave in the ECG with saw-tooth appearance. Typical AFL has positive P waves in lead V1, negative P waves in lead V6, and negative P waves in lead II, III, and aVF. Activation mapping using the Halo catheter and 3-D mapping system showed the activation wave front goes downward in the free wall , travels through the cavotricuspid isthmus, spread upward in the septal wall, and crosses the crista terminalis to complete the reentrant circuit. Reverse typical FL has negative P waves in lead V1, positive P wave in lead V6, and positive P waves in lead II, III, and aVF. The action sequence was the reverse of typical AFL.</p><sec id="s3a1"><title>Slow Conduction Zone of the Typical AFL Circuit</title><p>In this laboratory, we have studied the electrophysiologic properties of typical AFL circuit. It was consistent with previous findings that the low right atrial isthmus, defined as a path bounded by the orifice of inferior vena cava, eustachian valve/ridge, coronary sinus ostium, and tricuspid annulus, is a zone of slow conduction during AFL [<xref ref-type="bibr" rid="R9">9</xref>-<xref ref-type="bibr" rid="R11">11</xref>]. Furthermore, this laboratory demonstrated that during sinus rhythm incremental pacing from the low lateral right atrium and coronary sinus ostium could produce rate-dependent conduction delays [<xref ref-type="bibr" rid="R10">10</xref>], culminating in unidirectional block in the low right atrial isthmus, and induction of counterclockwise or clockwise AFL in patients with or without clinical AFL (<xref ref-type="fig" rid="F1">Figure 1</xref>). These findings were confirmed by Feld et al and suggested that slow conduction in the low right atrial isthmus may be mechanistically important for the development of human typical AFL [<xref ref-type="bibr" rid="R11">11</xref>]. In contrast, decremental conduction properties or rate-dependent conduction delays were not found in the right atrial free wall. The mechanism of slow conduction in the isthmus was not clear. Spach et al. have shown that conduction velocity of atrial impulses is faster parallel to the long axis of myocyte fibers and slower along the plane transverse to myocyte fiber orientation [<xref ref-type="bibr" rid="R12">12</xref>]. This phenomenon was explained by higher axial resistance due to scant cell-to-cell coupling encountered when impulses propagated perpendicular to the long axis of muscle fibers [<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R13">13</xref>]. With aging or atrial dilatation, intercellular fibrosis can change the density of gap junctions and produce nonuniform anisotropic conduction through the trabeculations of the low right atrial isthmus [<xref ref-type="bibr" rid="R13">13</xref>]. This hypothesis is supported by a recent anatomic study of the low right atrial isthmus in humans [<xref ref-type="bibr" rid="R14">14</xref>]. Furthermore, observations in dogs with natural and evoked atrial flutter suggest that thinning of atrial myocardium with intervening spaces may predispose to both slow and nonuniform conduction [<xref ref-type="bibr" rid="R15">15</xref>].</p></sec><sec id="s3a2"><title>Conduction Barriers</title><p>Using activation and entrainment mapping from closely spaced sites around the tricuspid annulus during typical AFL, Kalman et al confirmed that all sites around the circumference of the tricuspid annulus were a part of the flutter reentrant circuit, since the postpacing interval was equal to the flutter cycle length[<xref ref-type="bibr" rid="R16">16</xref>]. Thus, the tricuspid annulus is the anterior and fixed barrier in typical AFL. Using intracardiac echocardiography to place a multipolar catheter along the length of the crista terminalis and eustachian ridge, split potentials could be recorded along these structures with disparate activation sequences of each component by Olgin et al[<xref ref-type="bibr" rid="R17">17</xref>]. Moreover, entrainment could be used to demonstrate that one component of the split potential is within the reentrant circuit while the other is not. These findings are strong evidence of these structures forming the posterior barrier in typical AFL. In this laboratory, we have studied the conduction properties of the crista terminalis in patients with and without clinical AFL[<xref ref-type="bibr" rid="R18">18</xref>]. We found that split potentials could be recorded along the length of the crista terminalis during pacing from the low posterior right atrium at a long cycle length in patients with clinical AFL (<xref ref-type="fig" rid="F2">Figure 2</xref>), suggesting that poor transverse conduction property in the crista terminalis may be the requisite substrate for clinical occurrence of typical AFL[<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>]. However, Friedman et al found that a functional line of block was present at the posteromedial (sinus venosa region) right atrium during counterclockwise and clockwise AFL, suggesting that crista terminalis block was not required for the maintenance of typical AFL[<xref ref-type="bibr" rid="R20">20</xref>]. These different results may be due to heterogeneity in the right atrial activation outside of the low right atrial isthmus in patients with typical AFL[<xref ref-type="bibr" rid="R21">21</xref>].</p></sec><sec id="s3a3"><title>Excitable Gaps</title><p>A flat resetting response was observed in most cases of typical AFL, signifying a fully excitable gap [<xref ref-type="bibr" rid="R22">22</xref>,<xref ref-type="bibr" rid="R23">23</xref>]. The total duration of excitable gap is relatively wide and occupies about 13 to 20 % of the flutter cycle length depending on the pacing site.</p></sec><sec id="s3a4"><title>Variant Circuit</title><p>Using the noncontact mapping system, we could demonstrate that some patients had a single incomplete line of block in the crista terminalis during typical atrial flutter. This resulted in double loop reentry during typical atrial flutter, one circulating around the tricuspid annulus, and the other rotating around a part of crista terminalis through the conduction gap (<xref ref-type="fig" rid="F3">Figure 3</xref>). RF ablation of the cavotricuspid isthmus and crista gap could eliminate this atrial flutter.</p></sec></sec><sec id="s3b"><title>Atypical AFL</title><p>Atypical AFL may arise from the right or left atrium. There are no consistent ECG characteristics. However, using three criteria (positive P waves in lead V6, negative P wave in lead aVL, and low amplitude of the P waves in inferior leads) can differentiate left from right AFL [<xref ref-type="bibr" rid="R24">24</xref>].</p><sec id="s3b1"><title>Right Atrial Upper Loop Reentry</title><p>Using a noncontact, 3D mapping technique, we have demonstrated a macroreentrant circuit localized to the upper portion of the right atrium [<xref ref-type="bibr" rid="R25">25</xref>]. The wave front had counterclockwise activation (descending activation sequence in the free wall anterior to the crista) or clockwise activation (ascending activation sequence in the free wall anterior to the crista) around the central obstacle, which was composed of the crista terminalis, the area of functional block and superior vena cava (<xref ref-type="fig" rid="F4">Figure 4</xref>). The lower turn-around points were located at the conduction gap in the crista terminalis. RF linear ablation of the conduction gap in the crista terminalis eliminated atrial flutter.</p></sec><sec id="s3b2"><title>Right Atrial Free wall Reentry (<xref ref-type="fig" rid="F5">Figure 5</xref>)</title><p>Usually there is a low voltage zone in the anterior free wall, which may be due to spontaneous scar formation. The activation wave front circulates around this low voltage zone and the electrograms at this zone show double potentials [<xref ref-type="bibr" rid="R26">26</xref>]. RF ablation of the channel between the Inferior vena cava or tricuspid annulus and the central obstacle can eliminate this atrial flutter.</p></sec><sec id="s3b3"><title>Right Atrial Figure of Eight Reentry</title><p>The type I figure-of-eight reentry (n = 4) demonstrated simultaneous upper and lower loop reentry sharing a common pathway through conduction gap in the crista terminalis [<xref ref-type="bibr" rid="R26">26</xref>]. The two separate central obstacles were the superior vena cava (SVC) combined with upper crista and the inferior vena cava combined with lower crista (<xref ref-type="fig" rid="F6">Figure 6</xref>). The type II figure-of-eight reentry (n = 8) demonstrated simultaneous upper loop reentry and free wall reentry [<xref ref-type="bibr" rid="R26">26</xref>]. The channel between the crista terminalis and the low voltage zone was a common pathway. The two separate central obstacles were the SVC with upper crista and a part of the low voltage zone. RF ablation of the conduction gap in the crista terminalis (for type I reentry) and the channel between the crista terminalis and low voltage zone (for type II reentry) was effective in eliminating atrial flutter.</p></sec><sec id="s3b4"><title>Mitral Annular Atrial Flutter (<xref ref-type="fig" rid="F7">Figure 7</xref>)</title><p>This macroreentrant circuit rotates around the mitral annulus, either counterclockwise or clockwise [<xref ref-type="bibr" rid="R27">27</xref>]. The boundaries of the critical isthmus include the mitral annulus anteriorly, and low voltage zone or scars in the posterior wall of the left atrium posteriorly. RF ablation of the isthmus between the left inferior pulmonary vein and the mitral annulus can eliminate this atrial flutter.</p></sec><sec id="s3b5"><title> Pulmonary Vein-Related Atrial Flutter (<xref ref-type="fig" rid="F8">Figure 8</xref>)</title><p>Macroreentrant circuits can rotate around one or more pulmonary veins and a scar in the posterior wall or roof of the left atrium [<xref ref-type="bibr" rid="R28">28</xref>]. These circuits may have multiple loops. The peri-pulmonary vein circuits can be cured with ablation by creating a lesion from a pulmonary vein to the mitral annulus or to the contralateral pulmonary vein.</p></sec><sec id="s3b6"><title>Left Septal Atrial Flutter</title><p>The macroreentrant circuit rotates around the left septum primum, either counterclockwise or clockwise [<xref ref-type="bibr" rid="R29">29</xref>,<xref ref-type="bibr" rid="R30">30</xref>]. The characteristics of the ECG showed dominant positive P waves in lead V1 and low amplitude waves in the other leads. The critical isthmus is located between the septum primum and the pulmonary veins or between the septum primum and the mitral annulus ring. RF ablation of this isthmus can eliminate this atrial flutter.</p></sec></sec></sec><sec sec-type="" id="s4"><title>Conclusion</title><p>AFL is a reentrant arrhythmia and needs anatomic or functional barriers to maintain its activation. Typical AFL rotates around the tricuspid annulus with the crista terminalis and tricuspid annulus as barriers. Atypical AFL may originate from the right or left atrium without involving the cavotricuspid isthmus. The barriers may be scars, crista terminalis, mitral annulus, pulmonary veins, or septum primum. RF ablation of the isthmus between the boundaries can cure this arrhythmia.</p></sec>
Exercise-induced left septal fascicular block: an expression of severe myocardial ischemia	Could not extract abstract	<contrib contrib-type="author"><name><surname>Uchida</surname><given-names>Augusto Hiroshi</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Moffa</surname><given-names>Paulo Jorge</given-names></name><degrees>PhD</degrees><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Riera</surname><given-names>Andrés Ricardo Pérez</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Ferreira</surname><given-names>Beatriz Moreira Ayub</given-names></name><xref ref-type="aff" rid="aff1"></xref></contrib><aff id="aff1"><label></label>Heart Institute of the University of Sao Paulo, Sao Paulo - Brazil</aff><aff id="aff2"><label>†</label>ABC Medicine Faculty - ABC Foundation - Santo André - Sao Paulo - Brazil</aff>	Indian Pacing and Electrophysiology Journal	<p>The electrocardiogram (ECG) criteria for the left septal fascicular block (LSFB) are not universally accepted and many other denominations can be seen in literature: focal septal block, septal focal block, left septal fascicular block, left anterior septal block, septal fascicular conduction disorder of the left branch, left septal Purkinje network block, left septal subdivision block of the left bundle branch, anterior conduction delay, left median hemiblock, left medial subdivision block of the left bundle branch, middle fascicle block, block of the anteromedial division of the left bundle branch of His, and anteromedial divisional block. During exercise stress test, fascicular blocks (left anterior and posterior) seem to indicate severe coronary artery narrowing of left main coronary or proximal left anterior descending artery disease [<xref ref-type="bibr" rid="R1">1</xref>] and transient exercise-induced left septal fascicular block has been reported a few times [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R3">3</xref>].</p><p>54-year-old male, with a history of essential arterial systemic hypertension, primary hyperlipidemia and six-month typical chest pain during exercise (Class II - Canadian Cardiovascular Society) underwent an exercise stress test. During the exercise stress test, ECG demonstrated abrupt prominent anterior forces, an increase in R wave amplitude from V1 to V4, extreme left axis deviation and minor ST segment depression in DII, DIII and aVF (<xref ref-type="fig" rid="F1">Figure 1</xref>). The post-exercise period showed progressive return of the QRS axis in both frontal and horizontal planes and the ST depression worsened by 1 mm. Coronary angiogram (<xref ref-type="fig" rid="F2A">Figure 2A</xref>) showed a critical proximal left anterior descending artery lesion. An exercise stress test done three months after coronary artery bypass surgery grafting was normal (<xref ref-type="fig" rid="F2B">Figure 2B</xref>).</p><sec sec-type="" id="s1"><title>Discussion</title><p>Intraventricular conduction abnormalities consistent with LSFB are exceptionally recognized. Its diagnosis requires particular attention to changes in the direction of the initial and middle forces of ventricular activation in right and/or middle precordial leads. We divided ventricular activation of LSFB into four successive moments: <list list-type="order"><list-item><p>Initial vector of 20 ms (named 1PI vector) as a consequence of left anterior fascicular block (LAFB) and LSFB. The onset of ventricular activation will depend on the division that is not blocked; i.e. the left posterior-inferior fascicle that ends in the base of the postero-medial papillary muscle of the mitral valve in the postero-inferior region of the left ventricle. The activation of this area originates the so-called 1PI vector (the initial 20 ms vector) directed to left and often to the back, frequently manifesting initial embryonic q waves from V1 to V3.</p></list-item><list-item><p>Vector from 20 ms to 40 ms: it represents the complete activation from right to left of the postero-inferior region of the left ventricle; originating the so-called 2PI vector, heading backward, downward and to the left;</p></list-item><list-item><p>Vectors from 40 ms to 60 ms: from 40 ms the blocked antero-superior and antero-medial regions activate. This activation is processed through septal and free-wall Purkinje arborizations, which interconnect the three areas. This phase is represented by us with the modified names as 2AS and 2AM vectors. The first one causes the classical pattern of left anterior fascicular block in the frontal plane and the second one prominent anterior forces in the horizontal plane, translated by R waves of increased voltage "in crescendo" from V1 to V3 and decreasing from V5 to V6.</p></list-item><list-item><p>Basal vector beyond 60 ms: it corresponds to activation of the basal region of both ventricles, being called basal vector.</p></list-item></list></p><p>The peak exercise tracing (<xref ref-type="fig" rid="F1">Figure 1</xref>) bears similarity to "standard masquerading" bundle branch block, in which the limb leads look like left bundle branch block (LBBB) and precordial leads like right bundle branch block (RBBB). The QRS duration increased from about 80 ms to 120 ms, and time to peak of R wave in V1 from 20 ms to 70 ms. But, if we observe in detail the left limb leads, DI and aVL, there is no final S wave in them (the QRS complex is a pure R), which leads us to think that these prominent anterior forces are not due to a lesser degree of RBBB, but to LSFB. An isolated RBBB (without LAFB) always present S waves in the left sided leads. On the other hand, in this case there are no final S wave in the left leads (<xref ref-type="fig" rid="F1">Figure 1</xref> recovery - post-exercise 04:23 minutes).</p><p>Differential diagnosis of prominent anterior forces (tall R waves in right and/or middle precordial leads) must include3-6: normal variant, misplaced precordial leads, right bundle branch block, left ventricular ectopy, right ventricular enlargement, Wolff-Parkinson-White with posterior accessory pathway, strictly posterior and posterolateral myocardial infarction, obstructive and non-obstructive forms of hypertrophic cardiomyopathy, progressive muscular dystrophy, dextrocardia, and associations of the previous ones. In this case, the basal ECG did not show a tall R wave from V1 to V4. The exercise-induced transient prominent anterior forces that we interpreted as being secondary to LSFB showed as an abrupt transient tall R wave from V1 to V4, without concomitant significant ST depression. After myocardial revascularization this phenomenon was not found on exercise stress test reflecting the ischaemic nature of this abnormality manifested mainly as dromotropic disorders.</p></sec>
Concealed malfunction of the temporary pacemaker	Could not extract abstract	<contrib contrib-type="author"><name><surname>Alasti</surname><given-names>Mohammad</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Haghjoo</surname><given-names>Majid</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Alizadeh</surname><given-names>Abolfath</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Sadr-Ameli</surname><given-names>Mohammad Ali</given-names></name><degrees>MD</degrees></contrib><aff> Department of Pacemaker and Electrophysiology, Rajaie Cardiovascular Medical and Research Center, School of Medicine, Iran University of Medical Sciences, Tehran, Iran.</aff>	Indian Pacing and Electrophysiology Journal	<p>The 12-lead ECG shows sequential atrial and ventricular pacing (<xref ref-type="fig" rid="F1">Figure 1A</xref>). A tracing, obtained simultaneously during pacemaker interrogation, disclosed pacemaker functioning as VDD mode (<xref ref-type="fig" rid="F1">Figure 1B</xref>). The careful examination of this pacemaker tracing showed that there is a pacing stimulus before each P wave (compatible with DDD mode). This paradox can only be explained by displacement of the temporary pacing lead to right atrium and right atrial stimulation by temporary pacemaker. In this setting, each temporary pacemaker-induced atrial depolarization is tracked by the right atrial lead of the permanent pacemaker as intrinsic P wave. Fluoroscopic study confirmed this explanation (<xref ref-type="fig" rid="F2">Figure 2</xref>). The displaced temporary pacing lead was seen near the lateral right atrial wall. Temporary pacemaker lead had been inserted before replacement of permanent pacemaker.</p>
Endocardial Pacemaker Implantation in Neonates and Infants	<p>Transvenous pacemaker lead implantation is the preferred method of pacing in adult patients. Lead performance and longevity are superior and the implantation approach can be performed under local anaesthetic with a very low morbidity. In children, and especially in neonates and infants, the epicardial route was traditionally chosen until the advent of smaller generators and lead implantation techniques that allowed growth of the child without lead displacement. Endocardial implantation is not universally accepted, however, as there is an incidence of venous occlusion of the smaller veins of neonates and infants with concerns for loss of venous access in the future. Growing experience with lower profile leads, however, reveals that endocardial pacing too can be performed with low morbidity and good long-term results in neonates and infants.</p>	<contrib contrib-type="author"><name><surname>Ayabakan</surname><given-names>Canan</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Rosenthal</surname><given-names>Eric</given-names></name><degrees>MD FRCP</degrees></contrib><aff>Evelina Children’s Hospital, Guy’s & St Thomas’ Hospital Trust, Lambeth Palace Road, London SE1 7EH, UK</aff>	Indian Pacing and Electrophysiology Journal	<p>Cardiac pacing in infants and children has evolved considerably since the initial implants over half a century ago. Pacemaker generators have become smaller and are now easily accommodated subcutaneously or submuscularly either in the chest or the abdominal wall in the smallest of infants. Improvements in lead technology have enabled lower chronic thresholds with greater mechanical and electrical integrity and longevity of the leads and the pacemaker system. While endocardial leads have consistently proven to be superior to epicardial leads both mechanically and electrically, the development of steroid eluting epicardial leads has improved the effectiveness of epicardial systems [<xref ref-type="bibr" rid="R1">1</xref>]. The child who receives a pacemaker will undergo several additional interventions to replace the generator with or without the lead itself during his/her lifetime. Special consideration should therefore be given when selecting the most appropriate pacing system in the very small.</p><p>The choice of the method of permanent pacing is determined by the patient’s size, and presence of structural heart disease. While a large body of opinion holds that neonates and infants should receive epicardial pacing systems [<xref ref-type="bibr" rid="R2">2</xref>-<xref ref-type="bibr" rid="R4">4</xref>], experience with endocardial leads in the smallest of infants is building up (<xref ref-type="fig" rid="F1">Figure 1</xref>). The favourable outcome of endocardial leads in small infants is not limited to anecdotal case reports [<xref ref-type="bibr" rid="R5">5</xref>-<xref ref-type="bibr" rid="R10">10</xref>]. The largest experience to date from two centres (Guy’s Hospital and Wilhelmina Children’s Hospital) in 39 infants ≤ 10kg was published recently by Kammeraad et al [<xref ref-type="bibr" rid="R11">11</xref>]. After a relatively long follow-up period of a median 4.3 years (up to 15.3 years), 31 of 36 patients (86%) continued with an endocardial system and 27 of them had the original endocardial lead. Two endocardial systems had to be replaced with epicardial systems and pacing therapy was stopped in three patients because it was no longer indicated. Lead survival was excellent with 87% of those with their first or second generator continuing with the original lead - comparable to the results of epicardial leads in older children [<xref ref-type="bibr" rid="R12">12</xref>]. Indeed in a similarly aged cohort, a significant number of epicardial lead failures were reported in neonates within the first year of implantation due to acute exit block and lead fractures [<xref ref-type="bibr" rid="R13">13</xref>].</p><sec sec-type="" id="s1"><title>Venous obstruction</title><p>Endocardial leads are increasingly preferred for small infants in other centres as well [<xref ref-type="bibr" rid="R14">14</xref>-<xref ref-type="bibr" rid="R16">16</xref>]. The reasons for this preference are lower acute and chronic threshold values and decreased surgical morbidity from the procedure. The risk of venous occlusion, however, is the major drawback of endocardial lead placement in neonates and infants. If an adequate amount of slack is not left in the heart to cater for growth and the lead cannot be advanced, then replacement will be needed. If the lead cannot be replaced due to venous occlusion then the contralateral subclavian vein will need to be used and this may reduce the options for venous access in the future. Although transfemoral and transhepatic pacing can be performed with good results, few centres advocate these routinely [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R18">18</xref>]. Accordingly some centres still prefer to implant epicardial leads in children less than 10 kg [<xref ref-type="bibr" rid="R2">2</xref>-<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R13">13</xref>]. In some centres the policy is one of replacing epicardial systems electively with an endocardial system at the end of life of the first generator due to an increase in lead fractures with time [<xref ref-type="bibr" rid="R13">13</xref>].</p><p>Studies specifically comparing pacing via the endocardial route versus the epicardial route in small infants are lacking. The incidence of venous obstruction and occlusion has rarely been addressed. Venous obstruction due to endocardial leads in a small group of newborns and infants was assessed by Stojanov et al [<xref ref-type="bibr" rid="R16">16</xref>]. In 12 children younger than 12 months, who had endocardial leads implanted via the cephalic vein, only two had ultrasonographic evidence of partial (up to 20%) venous occlusion at the end of a mean follow-up of 85.2 pacing months. Only four patients were reported to have asymptomatic venous occlusion during lead extraction among the 36 neonates and infants reported by Kammeraad et al [<xref ref-type="bibr" rid="R11">11</xref>]. Asymptomatic venous occlusion was not systematically looked for however, and the incidence may therefore be higher in this cohort who had their leads implanted via a subclavian puncture. Even in older children, the incidence of venous occlusion may be as high as 21% in endocardial leads depending on the lead diameter, body surface area and the introduction site (subclavian vein versus cephalic vein) of the lead [<xref ref-type="bibr" rid="R19">19</xref>].</p><p>Stojanov et al promote a cephalic vein cut down rather than a subclavian puncture approach in the very small [<xref ref-type="bibr" rid="R15">15</xref>]. It may be that avoidance of the relatively large sheath needed for a subclavian puncture to introduce the lead tip, which is larger than the lead body, has an advantage. Whether late venous obstruction depends more on the size of the subclavian introducer and acute trauma produced at the time of vein puncture or the diameter of the lead left in the vein has not been systematically investigated.</p></sec><sec sec-type="" id="s2"><title>Lead Extraction</title><p>Endocardial pacing in neonates is not for the faint-hearted as lead replacement is likely to be needed in at least some of the patients. While extracting 11 leads uneventfully in nine patients (including atrial leads placed at the second or third system), Kammeraad et al [<xref ref-type="bibr" rid="R11">11</xref>] had to abandon a ventricular lead, which was adherent to the atrial myocardium. Extraction techniques are continuing to improve and lead replacement through an occluded subclavian vein is now possible using countertraction sheaths. A technique for femoral extraction that allows preservation of guidewire access across an occluded subclavian vein (<xref ref-type="fig" rid="F2">Figure 2</xref>) is described by Kammeraad et al [<xref ref-type="bibr" rid="R11">11</xref>]. Downsizing extraction countertraction sheaths for smaller leads may allow their safe use even in small children. In older children, standard diathermy and laser extraction sheaths can be used safely.</p></sec><sec sec-type="" id="s3"><title>Concomittant Cardiac Surgery</title><p>Although epicardial leads do not carry a risk of venous thrombosis, there is an increased perioperative morbidity [<xref ref-type="bibr" rid="R13">13</xref>]. Patients who are already a candidate for a sternotomy or thoracotomy for the correction of heart defects may benefit from epicardial leads especially when dual chamber pacing is required for atrioventricular synchrony to improve the postoperative hemodynamics. However postoperative infections and pericarditis may preclude epicardial lead placement, in which case an endocardial pacing system is the best choice [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R13">13</xref>]. While some surgical heart block patients require dual chamber pacing, the majority of infants and neonates who require pacing, are well served by a single chamber system that can be implanted endocardially. Transvenous dual chamber pacing is possible but generally avoided due to the increased risks of venous occlusion [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R14">14</xref>]. Early pacing with a dual chamber system may lead to a cardiomyopathy due to pacing induced ventricular dysynchrony which is aggravated by the higher rates found with atrial tracking in the very young [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R21">21</xref>]</p></sec><sec sec-type="" id="s4"><title>Conclusions</title><p>Neither the epicardial nor the endocardial approach in permanent pacing is free from complications. The experience reported with endocardial pacing in neonates and infants while encouraging is limited and no directly comparative data with epicardial pacing are available. Although the clear advantage of endocardial pacemaker implantation over the epicardial approach has yet to be demonstrated, the growing experience indicates that endocardial pacing is feasible and effective even in neonates and small infants - and it is an acceptable alternative to epicardial systems. The current disadvantages of endocardial leads may be overcome in the future by downsizing of the leads and extraction systems. The ultimate approach in very small children currently will depend on the facilities and experience of the surgeons and cardiologists in each centre.</p></sec>
Pacemaker Prevention Therapy in Drug-refractory Paroxysmal Atrial Fibrillation: Reliability of Diagnostics and Effectiveness of Prevention Pacing Therapy in Vitatron<sup>™</sup> Selection<sup>®</sup> Device	Could not extract abstract	<contrib contrib-type="author"><name><surname>Terranova</surname><given-names>Paolo</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref></contrib><contrib contrib-type="author"><name><surname>Valli</surname><given-names>Paolo</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">2</xref></contrib><contrib contrib-type="author"><name><surname>Terranova</surname><given-names>Peppino</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1">1</xref></contrib><contrib contrib-type="author"><name><surname>Dell’Orto</surname><given-names>Simonetta</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">2</xref></contrib><contrib contrib-type="author"><name><surname>Greco</surname><given-names>Enrico Maria</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">2</xref></contrib><aff id="aff1"><label>1</label>Divisione di Cardiologia, Azienda Ospedaliera “Luigi Sacco” – Polo Universitario, Istituto di Scienze Biomediche LITA; University of Milan, Italy</aff><aff id="aff2"><label>2</label>Unità Operativa di Cardiologia, Presidio Ospedaliero “Causa Pia Ospedaliera Uboldo”, Cernusco sul Naviglio, Azienda Ospedaliera di Melegnano, Milano</aff>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>Atrial fibrillation (AF) represents one of the major clinical, social and economical medical challenge [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>]. Progressive ageing is associated with an inevitable rising in incidence and prevalence of this rhythm disorder [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R4">4</xref>], which causes not only impairments in functional capability being associated to a greater incidence of cerebro-vascular events, but also a rising request for emergency room visits and hospital admission, mainly in patients affected by left ventricular dysfunction or by other significant cardiovascular diseases.</p><p>Different therapeutic strategies, both pharmacological and non pharmacological, are available to restore sinus rhythm, prevent arrhythmia recurrences and cardio-embolic events and control the mean ventricular rate [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R5">5</xref>]. What is still a challenge is how to tailor appropriate strategy and therapy for each patient.</p><p>It is an affirming concept that AF should not be considered as a simple cardiac rhythm disturbance, but rather as a syndrome with a variety of clinical presentations and two frequently negative hemodynamic feature: (1) the lack of atrial contribution to ventricular filling and (2) the irregularity and/or the fast shortening of cardiac cycle lenght [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R6">6</xref>]. The abnormal automaticity of pulmonary veins’ foci [<xref ref-type="bibr" rid="R7">7</xref>] represents a critical determinant for AF onset. It’s well known, nowadays, that a (1) critical atrial mass is necessary to maintain the re-entry atrial circuits [<xref ref-type="bibr" rid="R8">8</xref>] and that (2) the adrenergic activation determines a pro-arrhythmic effect on bursts originating from the pulmonary veins [<xref ref-type="bibr" rid="R9">9</xref>] or on arrhythmia recurrence after a cardioversion [<xref ref-type="bibr" rid="R10">10</xref>]. Notwithstanding, it’s almost impossible to figure out a temporal prevision of the evolution of this “AF syndrome” in each patient.</p><p>In 1995, Wijffels and coll. [<xref ref-type="bibr" rid="R11">11</xref>] showed, for the first time, that AF causes several changes in atrial electrophysiological properties that could be responsible for the clinical progression and maintenance of this tachyarrhythmia. The persistence of electrical disorders was found to be proportional to the duration of atrial high-rate pacing [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R12">12</xref>].</p><p>There are three main electrophysiological assumptions [<xref ref-type="bibr" rid="R12">12</xref>-<xref ref-type="bibr" rid="R24">24</xref>] for pacing prevention of AF widely accepted: <list list-type="order"><list-item><p>to reduce the arrhythmogenetic effects of bradycardia and, in particular, the atrial refractory periods dispersion;</p></list-item><list-item><p>to override atrial premature contractions (PACs), as a possible cause of AF onset;</p></list-item><list-item><p>to reduce the compensatory pauses after an atrial or a ventricular premature contraction, thus reducing the dispersion of atrial refractoriness induced by the “short cycle – long cycle” mechanism.</p></list-item></list></p><p>A prospective randomised study from Andersen and coll. [<xref ref-type="bibr" rid="R15">15</xref>] on 225 patients with sick sinus syndrome, showed that atrial pacing (AAI) not only significantly reduced the incidence and burden of AF and thromboembolisms when compared with ventricular one, but also that there was a significant reduction in total and cardiovascular mortality.</p><p>To improve the capability to prevent the onset of AF, other several approaches have been tested. In patients with marked intraatrial conduction period delays, Saksena [<xref ref-type="bibr" rid="R16">16</xref>] and Daubert [<xref ref-type="bibr" rid="R17">17</xref>] suggested to resynchronize both atria with dual-site atrial pacing, in different stimulation sites, in order to reduce the consequent atrial electrical abnormalities. Nowadays, this approach has been less considered, because of a low efficacy on AF incidence and burden, with frequent episodes of cross-talk and interference due to the dipole wide range, and because of the higher threshold needed for left atrial pacing. As alternative solutions, Padeletti and coll [<xref ref-type="bibr" rid="R18">18</xref>]. suggested the use of a single atrial lead on atrial septum. Several studies are still ongoing, with contrasting preliminary results.</p><p>It’s also well known that variability in the origin of PACs should increase the probability of a re-entry mechanism, mainly in presence of an anomalous substrate. On this account, it has been proposed to reduce atrial rate variability using an overdrive atrial steady stimulation alone, although this approach did not obtain significant results. Alternatively, other pacing techniques such as bursts and drives algorithms have been proposed. Several trials [<xref ref-type="bibr" rid="R19">19</xref>-<xref ref-type="bibr" rid="R24">24</xref>] have been done to evaluate the safety and efficacy of these devices in terminating spontaneous atrial tachyarrhythmias, but the results were quite disomogeneous and so this approach is far from being definitively accepted.</p><p>In particular, the Atrial Dynamic Overdrive Pacing Trial (ADOPT) [<xref ref-type="bibr" rid="R25">25</xref>] evaluated the effects of atrial overdrive pacing algorithms in 319 patients (mean age 71±10 years) with paroxysmal atrial fibrillation (3-3.5 documented atrial fibrillation episodes in the 12 weeks prior to implant). Atrial pacing reduced symptomatic atrial fibrillation burden (On: 1.9% vs. Off: 2.5%). Symptomatic atrial fibrillation burden decreased over time for both groups. No significant difference was noted in the number of atrial fibrillation episodes (On: 3.2 ± 8.6 vs. Off: 4.3 ± 11.5) or total hospitalizations (On: 9% vs. Off: 13%) between the two groups. Quality-of- life improved in both groups.</p><p>Moreover, in the ATTEST study [<xref ref-type="bibr" rid="R26">26</xref>] was evaluated the effectiveness of an atrial therapy device utilizing preventative and antitacycardia pacing in patients with symptomatic atrial fibrillation. The implanted device was an AT500 (Medtronic Inc, Minneapolis, MN) DDDR pacemaker, with a mode switching algorithm (DDIR) during atrial tachyarrhythmias. Atrial antitachycardia pacing was available in rate adaptive burst or ramp modes, and the device also could employ 3 preventative atrial pacing algorithms to attempt to prevent AF/AT recurrences. After a one-month run in period, patients were randomized to either atrial prevention or termination on or off, and followed for three months. At the end of the trial, there was observed no significant difference between the two groups in the run in period or in follow up in either frequency, burden or symptomatic frequency of AT/AF. The efficacy of device ATP was 54% in converting atrial arrhythmias to sinus rhythm as evaluated by the device and the accuracy of the detection algorithm was 99.9%.</p><p>Based on the hypothesis of the effectiveness of a non-pharmacological antiarrhythmic electrical therapy, our study was aimed at evaluate the reliability of the four specific pacing algorithms of the DDDRP pacemaker, the Selection<sup>®</sup> 900E/AF2.0 Vitatron<sup>™</sup>, and to reduce the trigger mechanisms, possibly responsible for AF onset.</p></sec><sec sec-type="materials\|methods" id="s2"><title>Materials and Methods</title><p>We initially evaluated the atrial arrhythmia recording reliability in 15 patients, implanted with a Selection™ 900E/AF2.0 Vitatron® device, for a “brady-tachy” syndrome, with advanced atrioventricular conduction disturbances. Patients were enrolled in a period of 10 months, starting from January 2000 and ending in October 2000. Patients were 9 males and 6 females, with a mean age of 71±5 years, without any previous history of myocardial infarction, angina, diabetes and other known risk factors and in I NYHA functional Class. All patients were initially under oral anticoagulant therapy (INR 2.0-3.0) and IC antiarrhythmic drug therapy (Acetate Flecainide 200 mg per day) for at least 6 months prior to implant. Therapies were not discontinued during the follow-up.</p><p>In order to evaluate the Selection™ 900E/AF2.0 Vitatron® atrial arrhythmia recording reliability, we compared the number of atrial arrhythmic episodes and their onset and duration,properly stored in the pacemaker, with a contemporaneous 24h standard Holter cassette recording (Ela Medical™ Synetec® System, version 1.20). A statistical correlation analysis was performed. The pacemaker parameters were settled in order to record atrial arrhythmic episodes lasting at least 6 ventricular cycles for their onset and 10 ventricular cycle for their ending. The atrial cut-off detection rate was of 180 atrial waves per minute. Clinical and follow-up features of our population are summarized in <xref ref-type="table" rid="T1">Table 1</xref>. Data are expressed as mean ± SEM. Paired t-Student test was performed.</p><p>The second aim was a middle-term efficacy evaluation (mean follow up 24±8 months, i.e. from 20 to 32 months) of all the four Selection™ 900E/AF2.0 Vitatron® available pacing algorithms, in order to prevent the trigger mechanisms, possibly responsible for AF onset. Each pacing algorithm was tailored according to the clinical and electrophysiological features of each patient, according to the previously recorded atrial arrhythmic events.</p><p>The four used atrial pacing algorithms were the following four: <list list-type="order"><list-item><p><bold><italic>Pace Conditioning<sup>™</sup></italic></bold> this algorithm consist on a permanent overdrive atrial pacing with an atrial pacing rate of about 15 b.p.m. higher than the beneath intrinsic effective atrial rate.</p></list-item><list-item><p><bold><italic>PAC Suppression<sup>™</sup></italic></bold> this algorithm is designed to reduce the incidence of atrial tachyarrhythmia by a temporary stable atrial overdrive pacing following a PAC. This algorithm provides a temporary atrial overdrive pacing lasting for 600 ventricular cycles after a sensed PAC. At the end of this 600 ventricular cycles, the pacing rate progressively reduces till the lower rate limit or till the emergence of a stable sinus rhythm.</p></list-item><list-item><p><bold><italic>Post-PAC Response<sup>™</sup></italic></bold> this algorithm is designed to reduce the post-extrasystolic pauses by controlling the atrial rate in the 2 beats after a PAC. The first atrial paced beat’s rate is determined by an averaging between the previous physiologic RR interval and the PAC pairing rate. From the second beat onwards, the atrial fleeing rate returns to the atrial physiologic rate.</p></list-item><list-item><p><bold><italic>Post-Exercise Rate Control<sup>™</sup></italic></bold> the post-exercise rate control has been specifically designed to prevent a too fast lowering of heart rate after a physical activity. The post-exercise rate increases proportionally to the difference between the physiological heart rate and the target heart rate, i.e. 90% of the physiological heart rate.</p></list-item></list></p></sec><sec id="s3"><title>Results</title><p>During the first part of this trial, i.e. that comparing Holter recordings and pacemaker storage data, we observed 59 episodes of paroxysmal AF, lasting from some seconds to some minutes, with a highly significant correlation coefficient between Holter and pacemaker recordings (r=0.96) (<xref ref-type="fig" rid="F1">Figure 1</xref>). The AF mean duration was of 70,9±46,8 s. in the Holter group and of 69,6±47,4 in pacemaker recordings (p = 0.881; 95% confidence interval: -18.48 to 15.88). Differences between timing and duration of the two recordings is likely to be due to the cassette tracking speed features.</p><p>As shown in <xref ref-type="fig" rid="F2">Figure 2</xref>, during the second half of our trial, the antiarrhythmic algorithms produced a significant and progressive reduction in AF episodes (from 46±7 episodes/day to 0,12±0,03 episodes/day; p<0.001; 95% confidence interval: 41.69 to 50.07) and in AF burden (from 93%±6% to 0,3%±0,06%; p<0.001; 95% confidence interval: 89.53 to 95.87), with a concomitant increase in atrial pacing percentage (from 3%±0.5% to 97%±3%; p<0.001; 95% confidence interval: -95.61 to -92.39). AF burden was evaluated and measured during the follow up period using the pacemaker storage data, and no comparison were made with AF burden prior to implant, because of the lack of a real value of this data in our patients, that were sometimes asymptomatic ones. We observed also a decrease in premature atrial contractions from 83±12 PACs/day to 2,3±0,8 (p<0.001; 95% confidence interval: -87.06 to -74.34) (<xref ref-type="fig" rid="F2">Figure 2</xref>). The same method was used to record APCs in the pre- and post- implant period, i.e. the Holter monitoring, and also during all the follow up period, i.e. a retrospective analysis of pacemaker data storage.</p><p>However, it was unfeasible to perform a reliable multivariate analysis because of the reduced number of patients, because of the lack of multiple risk factors, because of the contemporaneous usage in the same patients of acetate flecainide and of pacing prevention algorithms.</p></sec><sec id="s4"><title>Discussion and Conclusion</title><p>Presently, very few clinical randomised trials. [<xref ref-type="bibr" rid="R29">29</xref>-<xref ref-type="bibr" rid="R32">32</xref>] have compared the overlapping between pacemaker stored data and Holter monitoring ones, so that there is not a significant evidence of data about the reliability of these devices in properly recognising and monitoring supraventricular tachyarrhythmias. It’s also well known that bipolar leads often are not enough able to discriminate ventricular far field from normal P-wave amplitude or from wavelets of AF. An Atrial Blanking feature, long enough, may completely mask this phenomenon, but it also represents a blind interval for the atrial channel, sometimes subsequently resulting in a possible negative effect for an optimal arrhythmia detection. Recently, Nicotra [<xref ref-type="bibr" rid="R33">33</xref>] carried out a method for atria l sensing and blanking programming in order of guarantee reliable diagnostics in patients with paroxysmal AF. They proposed a decisional flow-chart based on the scanning of ventricular far field timing and amplitude that could focus and quantify this phenomenon for every kind of implanted device (<xref ref-type="fig" rid="F3">Figure 3</xref>). This approach was also used in the present study during the follow-up.</p><p>Our data confirm the significant reliability of Selection<sup>™</sup> 900E/AF2.0 Vitatron<sup>®</sup> algorithms for detecting and monitoring AF. Moreover, we recorded a significant number of supraventricular tachyarrhythmia’s episodes, thus making stronger the statistical reliability of these data. Subsequently, we detected a progressive and significant reduction in AF burden and episodes. This decreasing in AF episodes and burden may be mainly related to the increase of atrial pacing, obtained by using a newly designed pacing overdrive algorithm responsible of atrial ectopic beats’ suppression in keeping with several trials [<xref ref-type="bibr" rid="R11">11</xref>-<xref ref-type="bibr" rid="R28">28</xref>].</p><p>Most of the discussed trials showed a lower reduction in AF episodes and burden if compared with our study (AF burden decreasing from 85% to 35% vs. a reduction from 93% to 0.3%; p<0.001; 95% confidence interval: -9.253 to -6.747). As showed in our recordings, the exceeding share of AF episodes and AF burden reduction is more likely to be entrusted to the newly designed four preventive pacing algorithms stored in the tested device (Selection<sup>®</sup> 900E/AF2.0<sup>®</sup> Vitatron<sup>™</sup> pacemaker). Alternatively, as the data shown in <xref ref-type="fig" rid="F2">Figure 2</xref> seems to underline, i.e. that the efficacy of pacing for reducing AF episodes/day was evident after only six months of treatment, may suggest that also atrial remodelling could play a role. Electrical, mechanical and anatomical remodelling indicate structural alterations that, once established, may vanish any attempt to restore sinus rhythm. Atrial fibrosis is probably the most critical factor of the remodelling process and appears to be largely media ted by several mechanisms. Our clinical data indicate that these non pharmacological interventions may reduce, in a roundabout way, AF burden and episodes, probably interfering also with someone of those electrical and structural remodelling processes. It is possible, however, that having a very few patients, although followed for a median time of 24 months, we have overestimated the general possible reduction in AF episodes and burden because of a strong selection of our patients. In particular, we selected our patients on the basis of their atrial tachyarrhythmias’ onset mechanisms, and we tailored the available preventive pacing algorithms in each patient on the basis of each onset. This tailoring of pacing preventive algorithms on the basis of the different onsets in each patient, associated with the well known anti-remodelling effect of atrial overdrive pacing on atrial refractoriness dispersion, was the more important goal of our study and the likely reason of the observed marked reduction in AF episodes and burden. Therefore, these examined pacing algorithms may represent an effective therapeutical options to contrast the nearly inevitable progression of this arrhythmia towards its permanent form.</p><p>The AF antiarrhythmic drugs therapy represents the first and more effective therapy, although it is well known that it is not always effective and free from side effects. In our patients, implanted for a sick sinus syndrome and at least partially resistant to AF pharmacological treatment, new interventional non-pharmacological solutions, such as tailored antiarrhythmic pacing algorithms, may represent a further effective therapeutical option. Recent analysis of AFFIRM [<xref ref-type="bibr" rid="R33">33</xref>] and RACE trial [<xref ref-type="bibr" rid="R34">34</xref>] showed that, in patients older than 65 years, with well defined risk factors, a therapeutical strategy based on ventricular rate control is not inferior to a strategy of maintaining sinus rhythm (rhythm control). However, this issue is still debated. Indeed, other clinical trials and sub-analysis [<xref ref-type="bibr" rid="R35">35</xref>-<xref ref-type="bibr" rid="R40">40</xref>] showed that sinus rhythm restoration has to be preferred compared to rhythm control.</p><p>In conclusion, our experience suggests the reliability of well tailored pacing algorithms on AF control in selected patients with brady-tachy syndrome, refractory to pharmacological rhythm control. However, we should say that this study is partially limited by the small and highly selected sample used, that reduces the possibility of translating theses results to general population. Moreover, this was not a truly randomized, controlled study, and so interpretation of the clinical significance of the reduction of AF episodes, as also shown however by several other trials in International Literature, is uncertain, thus not allowing us for making really definitive conclusions. The main information, however, we can derive from this trial is that we should better evaluate, before every pacemaker implantation, the different AF onsets of each patient eligible for such a procedure, because only a really profound knowledge of the different onset mechanisms of atrial tachyarrhythmias in each subject may give us the possibility of choosing a really effective implantable device for these sometimes very different patients, thus obtaining the maximum effective clinical results from these pacing algorithms in terms of reduction of AF episodes and burden.</p></sec>
Complications and Mortality of Single Versus Dual Chamber Implantable Cardioverter Defibrillators	Could not extract abstract	<contrib contrib-type="author"><name><surname>Bagherzadeh</surname><given-names>Ataallah</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Emkanjoo</surname><given-names>Zahra</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Haghjoo</surname><given-names>Majid</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Farahani</surname><given-names>Maryam Moshkani</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Alizadeh</surname><given-names>Abolfath</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Sadr-Ameli</surname><given-names>Mohammad Ali</given-names></name><degrees>MD</degrees></contrib><aff> Department of Pacemaker and Electrophysiology, Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Sciences, Mellat Park, Vali-e-Asr Avenue, Tehran, Iran</aff>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>The implantable cardioverter defibrillators (ICDs) are increasingly being used as a treatment modality for high risk survivors of life-threatening tachyarrhythmia [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>]. Major technological advances like the introduction of transvenous defibrillation leads and "active can" concept have simplified the implantation of ICDs and have reduced implantation-related complications but still there are considerable rate of complications in most of the electrophysiologic centers [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R4">4</xref>].</p><p>The purpose of the current study was to identify and characterize prospectively the frequency of lead and implant-related complications and also mortality in patients with single-chamber versus dual-chamber ICD implantation.</p></sec><sec sec-type="methods" id="s2"><title>Methods</title><sec id="s2a"><title>Patients</title><p>Between January, 2000 and December, 2004, 234 patients received ICDs in our center. All of them included in the study and were followed for 33 ±23 months. The primary end point of the study was actuarial survival. The secondary end point was nonfatal complications, including inappropriate ICD therapy, lead-related and implant-related complications. The study was approved by local Ethics Committee and written informed consent was obtained from all the patients.</p></sec><sec id="s2b"><title>Implanted ICDs and programming</title><p>ICD system manufacturers included Medtronic and St Jude systems in 160 and 75 patients respectively. Atrial leads included 1688 T or 1642T for St Jude and 5076 for Medtronic systems. Ventricular leads included 1570,1572,1571,1580 and 1581 for St Jude and 6944, 6947 and 6948 for Medtronic systems. All of the ventricular leads were bipolar. In implanted devices all the detection and discrimination criteria were activated with the nominal values. In all the devices we defined ventricular fibrillation zone (300ms) plus one VT zone (400 ms). If the patient had an episode of spontaneous or induced sustained monomorphic VT slower than 370 ms we extended the VT zone to VT cycle length plus 40ms. In the VT detection zone the first therapy was three antitachycardia burst pacing. We used the nominal values of the ICDs for the duration and tachyarrhythmia detection criteria. Antibradycardia pacing was programmed in VVI mode with pacing rate of 40 beats per minute in all VR and DR ICDs, unless the patients needed dual chamber pacing, those with long QT syndrome and hypertrophic obstructive cardiomyopathy.</p></sec><sec id="s2c"><title>Implantation techniques</title><p>The implantation routes were either subclavian or Axillary veins percutaneously in electrophysiology laboratory by an electrophysiologist.</p></sec><sec id="s2d"><title>Follow-up protocol</title><p>The patients were followed after 1 month postoperatively and every 3 months thereafter and upon receiving high voltage therapy in our outpatient ICD clinic. ICD evaluations during follow-up visits included routine clinically appropriate measurements including interrogation of the device for tachyarrhythmia episodes, evaluation of sensing and pacing thresholds and lead impedance. Floppy diskettes were used to retrieve all episodes of ICD therapy. All the episodes resulted in ICD therapy, studied independently by two electrophysiologists to define the diagnosis. All the episodes were categorized as appropriate or inappropriate. ICD-related complications were documented, as were mortality and cause of death. Chest radiography (CXR) was obtained before hospital discharge and every 6 months to confirm radiographic lead integrity. Clinical evaluation of patients included history and physical examination of the implantation site. Complications were defined as unexpected adverse events that were felt to be related to the implant procedure or to the ICD lead and generator system.</p></sec><sec id="s2e"><title>Statistical analysis</title><p>Baseline characteristics and complications were summarized as the mean ± SD for continuous variables and otherwise as percentage. SPSS 13 software (SPSS Inc., Chicago, IL, USA) was used for data storage and analysis. Qualitative characteristics were compared using the Pearson’s chi square and Fischer exact tests. Kaplan-Meyer survival curves were calculated. Log rank statistics were used to compare the distribution of time to complications and mortality between different ICD groups. Two-tailed P < 0.05 were considered significant. Regression analysis was performed to evaluate the effect of variables on mortality and complications.</p></sec></sec><sec id="s3"><title>Results</title><sec id="s3a"><title>Baseline characteristics</title><p>Baseline characteristics of the patients are summarized in <xref ref-type="table" rid="T1">Table 1</xref>. <xref ref-type="table" rid="T2">Table 2</xref> compares the baseline characteristics of the patients in VR and DR ICD groups.</p></sec><sec id="s3b"><title>Mortality</title><p>Three patients (1.2%) died during the first month after ICD implantation because of multi-organ failure, massive pulmonary emboli and acute coronary syndrome. The 1-year survival rate was 94.9% and 94.1% for VR and DR ICDs respectively. The overall survival rate during follow-up was 90.6% and 88.9% for VR and DR ICDs respectively. There was no significant difference in mortality between different sex groups (P=0.37), different indications for ICD implantation (P=0.43) or between VR and DR ICD groups (P= 0.55) (<xref ref-type="fig" rid="F1">Figure 1</xref>). Univariate regression analysis showed predictors of mortality as NYHA class III or more (P<0.001), age >65 years (P=0.011) and LVEF<30% (P<0.001). The mortality in patients with CAD and DCM were significantly higher than those with other structural heart diseases (P=0.001). During the follow-up, 59 patients (25.2%) had at least one inappropriate ICD therapy. The mortality of this group was significantly higher than those who had not experienced inappropriate therapy (P=0.03).</p></sec><sec id="s3c"><title>Complications</title><p>Overall, 22 patients (9.4%) suffered from ICD-related complications during follow-up period. There was no significant difference in overall complication rate between VR and DR ICD groups in the follow-up period (P= 0.11) (<xref ref-type="fig" rid="F2">Figure 2</xref>). The risk of complications was not associated with any patient-specific factor including age (P=0.32) and gender (P=0.19) (<xref ref-type="table" rid="T1">Table 1</xref>). The risk of complications did not have any statistically significant difference in secondary versus primary prevention groups (P=0.06). The complications were not associated with the severity of left ventricular systolic dysfunction (P=0.16). Also, no statistically significant difference was observed between different NYHA classes (P=0.45) and underlying heart diseases (P=0.06). During the follow-up, 59 patients (25.2%) had at least one inappropriate ICD therapy. The risk of complications did not have any statistically significant difference in this group with those who had not experienced inappropriate therapy (P=0.16). The comparison of complications in different models of implanted leads was not possible due to variety of the models and small numbers of each model.</p></sec><sec id="s3d"><title>Implant-related complications</title><p>Implant-related complications occurred in 13 patients (5.5%). Hematoma or bleeding was the most frequent complication in this group, found in 7 patients (3%). Pneumothorax and hemothorax were found in 3(1.3%) and 2(0.9%) patients, all of them managed with chest tube insertion. No significant pocket infection was found but only one patient suffered from stitch abscess, treated conservatively with antibiotics.</p></sec><sec id="s3e"><title>Lead-related complications</title><p>Lead-related complications occurred in 9 patients (3.8%). Lead dislodgement was the most frequent lead-related complication, found in 5 patients (2.1%). Dislodgement of ventricular and atrial leads was seen in 3 and 2 patients respectively. Lead fracture was found in 2 patients (0.9%). Exit block and insulation break was rare, each of them occurred only in 1 patient (0.4%). 4 of 5 lead dislodgements and all of hematomas occurred in the 2 month of implantation. No cardiac perforation was found. The frequency of lead-related complications was significantly higher in dual chamber ICDs in comparison with single chamber ICDs (P=0.02).</p></sec></sec><sec id="s4"><title>Discussion</title><p>This prospective study characterized the incidence of lead and implant-related complications and mortality of ICD implantations in our center. The baseline demographic data of the patients in this study were comparable to large studies (<xref ref-type="table" rid="T1">Table 1</xref>) [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R8">8</xref>]. With the downsizing of ICD pulse generators, pectoral implantation has become standard for ICD placement [<xref ref-type="bibr" rid="R9">9</xref>]. Older generation pacemakers of comparable size to ICD pulse generators had a significant risk of erosion and other pocket complications [<xref ref-type="bibr" rid="R10">10</xref>]. During this study, 234 consecutive ICD implantations over a 33 ±23 months follow-up showed low (9.4%) incidence of complications. The most common implant and lead related complications were hematoma and lead dislodgement which was found in 3% and 2.1% of patients. These findings are consistent with or even rarer than prior report [<xref ref-type="bibr" rid="R5">5</xref>-<xref ref-type="bibr" rid="R18">18</xref>]. In the majority of previous studies the ICD-related complications did not differ with increasing age which was consistent with this study [<xref ref-type="bibr" rid="R15">15</xref>-<xref ref-type="bibr" rid="R18">18</xref>]. The infection rate in this case series was very low (0.4%). In large studies, infection rates of 1.3-2.7% were reported. Because most of the infections that were seen in previous reports occurred with abdominal pocket, it seems reasonable to expect a lower infection rate with current generation of ICDs implanted transvenously in the pectoral region [<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R20">20</xref>-<xref ref-type="bibr" rid="R22">22</xref>]. Intensive use of prophylactic antibiotics before and 48 hours after the procedure and avoidance of cauterization for hemostasis may help us reduce the infection rate.</p><p>Some authors encourage the usage of cephalic vein cut downs as the preferred approach for ICD lead venous approach for prevention of vascular complications of blind, percutaneous subclavian venipuncture such as pneumothorax and hemothorax [<xref ref-type="bibr" rid="R23">23</xref>,<xref ref-type="bibr" rid="R24">24</xref>]. The time course of specific complications reinforces the need for continued lead surveillance in ICD patients. Although bleeding, pneumothorax, lead dislodgement and infection tends to occur soon after implant, lead-related problems such as lead fracture and insulation breaks tend to occur throughout follow-up [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R25">25</xref>,<xref ref-type="bibr" rid="R26">26</xref>].</p></sec><sec id="s5"><title>Limitations</title><p>The present study must be interpreted in light of certain methodological limitations. The ICDs were implanted with multiple operators of variable training and experience. Therefore, the effects of differences in experience or surgical technique on complication rates were not controlled. However, the inclusion of the large number of cases and relatively long follow-up periods makes it more likely that these results are reliable. The second limitation was lack of cooperation of some patients for regular follow-up and routine CXRs. Therefore, the incidence of lead fractures and insulation breaks may have been underestimated. Also, small hematomas usually managed by sand bag and under- reported.</p></sec><sec id="s6"><title>Conclusion</title><p>This study showed that the pectoral implantation of ICDs by electrophysiologists in our center has a low rate of complications, comparable with most of the qualified centers. Since most of implant-related and some of lead-related complications occur soon after the implant, close early monitoring of patients during first 2 month after implantation is recommended. The continued occurence of lead fractures and the need for premature system revision over time supports the practice of close, routine ICD surveillance. Vascular complications of blind percutaneous subclavian venipuncture could be avoided if cephalic vein cut down is used for ICD lead venous access.</p></sec>
The Potential Benefit of 5% Sulfamylon Solution in the Treatment of <italic>Acinetobacter baumannii</italic>–contaminated Traumatic War Wounds	<p><bold>Background:</bold> The recent report of high numbers of <italic>Acinetobacter baumannii</italic> bloodstream infections among service members injured in Iraq and Afghanistan during the period January 2002 through August 2004 has prompted an investigation into their etiology. A review of the current guidelines for open combat casualty wounds as part of this broad investigation was not mentioned in the report. <bold>Objective:</bold> The objective of this study was 2-fold: to ascertain the susceptibility of <italic>A baumannii</italic> to currently available topical antibacterial agents and (2) to propose an alternative, effective treatment protocol for contaminated combat-related wounds so as to reduce or eliminate the likelihood of the wound serving as the source of <italic>A baumannii</italic> infection or septicemia. <bold>Methods:</bold> A standardized antimicrobial susceptibility study of 43 strains of <italic>A baumannii</italic> collected from a tertiary care burn center was conducted using 2 commonly used topical antibacterial agents, 1% silver sulfadiazine cream (Silvadene) and 5% mafenide acetate solution (5% Sulfamylon Solution). <bold>Results:</bold> Both were effective, but 5% Sulfamylon Solution demonstrated significantly greater antibacterial activity. <bold>Conclusion:</bold> Five percent Sulfamylon Solution, initially developed for wartime use, and currently limited by the Food and Drug Administration to soaks following meshed split-thickness autografts following excision of second-degree and third-degree burns, has a broad spectrum of antibacterial activity and extensive off-label applicability. It is an ideal agent for use in the treatment of war wounds, and should be considered as a superior replacement for normal saline in the current guidelines for open combat casualty wounds</p>	<contrib contrib-type="author"><name><surname>Kucan</surname><given-names>John O.</given-names></name><degrees>MD, FACS</degrees><xref rid="aff1" ref-type="aff">a</xref><xref ref-type="corresp" rid="cor1"/></contrib><contrib contrib-type="author"><name><surname>Heggers</surname><given-names>John P.</given-names></name><degrees>PhD</degrees><xref rid="aff2" ref-type="aff">b</xref></contrib>	Journal of Burns and Wounds	<p>The recent report of an increasing number of <italic>Acinetobacter baumannii</italic> bloodstream infections among service members injured in Iraq/Afghanistan operations prompted this investigation.<xref ref-type="bibr" rid="B1">1</xref> The number of such infections and the resistance of <italic>A baumannii</italic> infections to multiple antibiotics suggested the need to consider and address the wounds and their treatment as a possible cause of septicemia and to suggest an alternative method to decrease or eliminate colonized or contaminated wounds as a potential source of <italic>A baumannii</italic> sepsis.</p><p>The present guidelines for care of open combat casualty wounds are as follows:</p><p>The standard care for the type of wounds we are seeing from Iraq include vigorous and complete early irrigation and debridement (Figs <xref ref-type="fig" rid="F1">1</xref>–<xref ref-type="fig" rid="F3">3</xref>). The wounds are packed with saline-soaked Kling or fine mesh gauze and left alone for 4 days (Fig <xref ref-type="fig" rid="F4">4</xref>). The only reason to take down the dressing and inspect the wound is foul odor, discharge, bleeding, or fever that cannot be explained without inspecting the wound. At 4 days, the patient is returned to the operating room for a dressing change. If the wound has that “sticky” appearance without any areas of necrosis and need for further extensive debridement, a Delayed Primary Closure (DPC) is done. If the wound does not appear clean, it is debrided, irrigated, and packed again (sometimes with Dakin's solution or whatever solution the surgeon chooses) and the 4-day clock restarted (judgement call). Those wounds that fail DPC or are so large that a DPC is not in the plan are treated by wet-to-dry dressing changes. As soon as the wounds appear clean, Dakin's solution (or whatever solution is used to clean the wound) is switched <italic>to normal saline</italic>.<xref ref-type="bibr" rid="B2">2</xref></p><p>Normal saline solution possesses no inherent antibacterial activity, and therefore cannot be expected to decrease or control bacterial growth in open wounds. Likewise, simple inspection of such wounds is unreliable in predicting successful DPC.<xref ref-type="bibr" rid="B3">3</xref> Because systemic antibiotics are ineffective in reducing bacterial counts in granulating wounds, the use of a topical antibacterial agent that is active in controlling or reducing bacterial proliferation in open wounds may substantially decrease wound sepsis as a source of morbidity and may have a beneficial effect on overall management.<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref> Furthermore, the addition of microbiological quantification within the wounds (quantitative bacteriology or swab techniques) may provide valuable information regarding efficacy of treatment and the likelihood of successful DPC.<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B7">7</xref></p><p>The efficacy of 5% mafenide acetate solution (5% Sulfamylon Solution [Bertek Pharmaceutical Inc, Research Triangle Park, NC]) and 1% silver sulfadiazine cream (Silvadene [Kendall Company, Mansfield, Mass]) was tested against <italic>A baumannii</italic> derived from burn isolates at the Shriners' Burn Institute, Galveston, Tex, in order to determine whether these topical antibacterial agents, successfully employed in burn care, might be effective adjuncts in controlling <italic>A baumannii</italic> wound infections in traumatic war wounds.</p><sec sec-type="materials\|methods" id="S1"><title>MATERIALS AND METHODS</title><p>All <italic>A baumannii</italic> isolates from the burn population were routinely tested for their susceptibility to 5% Sulfamylon Solution and Silvadene by a modified Nathan's agar well assay.</p><p>Mueller-Hinton 150 mm agar plates received 6 mm agar wells using a sterile 6 mm biopsy punch. Once the wells were made, 10<xref ref-type="bibr" rid="B5">5</xref> colony-forming units of the organism were inoculated to each plate. The plates were allowed to dry, and then 0.1 mL of each topical antimicrobial was added to each well and numerically identified. A total of 43 <italic>Acinetobacter</italic> isolates were tested.</p><p>The assay plates were incubated at 37°C for 24 hours. Zones of inhibition were measured and compared for susceptibility.<xref ref-type="bibr" rid="B8">8</xref></p></sec><sec sec-type="results" id="S2"><title>RESULTS</title><p>Table <xref ref-type="table" rid="T1">1</xref> gives the average zones of inhibition for Silvadene and Sulfamylon against <italic>A baumannii.</italic> Sulfamylon treatment had a significantly greater zone of inhibition than did Silvadene treatment (<italic>P</italic> = .05).</p><sec id="S2-1"><title>Statistical assessment</title><p>Zones of inhibition for the 2 topical antimicrobials were compared using a 1-way analysis of variance followed by a Tukey's test. Statistical significance was set at <italic>P</italic> < .05.</p></sec></sec><sec sec-type="discussion" id="S3"><title>DISCUSSION</title><p>The emergence of an increased number of bloodstream infections due to antibiotic-resistant <italic>Abaumannii</italic> among service members wounded during Operation Enduring Freedom (Afghanistan) and Operation Iraqi Freedom (Iraq/Kuwait) from the period January 2002 through August 2004 has focused attention on the issues of infection control in combat settings and healthcare facilities (field hospitals, combat theater medical facilities, military medical centers) and the need for new, effective antibiotics to treat these infections. Using the criteria established by the Centers for Disease Control and Prevention's National Nosocomial Infection Surveillance system, data compiled during this period identified 102 patients with blood cultures positive for <italic>A baumannii.</italic> Of the 102 service members, 85 (83%) sustained wounds during either military operation.<xref ref-type="bibr" rid="B1">1</xref></p><p><italic>A baumannii</italic> is ubiquitous and has recently become one of the most important healthcare-associated hospital pathogen, resulting in significant morbidity and mortality. Over the last 40 years, nosocomial infections caused by this agent have become increasingly problematic, posing significant therapeutic challenges due to multiple antibiotic resistance, persistent colonization, and prolonged environmental survivability. <italic>A baumannii</italic> has been the cause of infections in patients suffering from traumatic injuries and was the most common gram-negative bacillus isolated from traumatic extremity injuries during the Vietnam War.<xref ref-type="bibr" rid="B9">9</xref>–<xref ref-type="bibr" rid="B14">14</xref> Therefore, environmental contamination of wounds with <italic>A baumannii</italic> must be considered as a possible source of infection. Presently, it is not known when or where military personnel who served in Operation Enduring Freedom or Operation Iraqi Freedom acquired the infections: in the field, during evacuation, or during treatment.</p><p>The information compiled on these patients is currently under review. Multiple factors, including antimicrobial exposure, mechanical ventilation, vascular access, and evacuation of patients through treatment facilities, are implicated. Microbiological and molecular analyses of patient and environmental isolates are being conducted to determine the likely sources of infection. However, neither the current standard guidelines for care of open combat casualty wounds have been reviewed nor has any consideration been given to assess more effective ways to ensure bacteriologic control of the wounds to reduce the possibility of the wound serving as the septic source.<xref ref-type="bibr" rid="B1">1</xref></p><p>The present guidelines for the care of open combat casualty wounds require vigorous irrigation, debridement, saline-soaked dressings, and reinspection at 96 hours, unless otherwise required. DPC is performed on wounds with that sticky appearance. Wounds not amenable to DPC undergo additional wound care and dressing changes employing wet-to-dry dressings, Dakin's solution or “whatever solution the surgeon chooses,” and normal saline.<xref ref-type="bibr" rid="B2">2</xref></p><p>The purpose of this study was 2-fold: (1) to ascertain the susceptibility of <italic>A baumannii</italic> to currently available topical antibacterial agents and (2) to propose an alternative, effective treatment protocol for contaminated combat-related wounds so as to reduce or eliminate the likelihood of the wound serving as the source of <italic>A baumannii</italic> infection or septicemia.</p><p>This study showed the efficacy of 2 commonly used topical antibacterial agents, 1% silver sulfadiazine cream (Silvadene) and 5% mafenide acetate solution (5% Sulfamylon Solution), against a large number of strains of <italic>A baumannii.</italic> The latter demonstrated significantly greater activity against <italic>A baumannii</italic> than did 1% silver sulfadiazine cream.</p><p>Topical antibacterial agents are routinely employed in burn care to prevent burn wound sepsis, but their use in nonburn wounds has been generally limited and underappreciated. Nevertheless, these agents, and 5% mafenide acetate solution in particular, have been effectively utilized in the management of nonburn wounds and should be considered as important adjuncts in the management of acute and chronic open wounds. The product indications for 5% Sulfamylon Solution have been severely limited by the Food and Drug Administration to use over meshed split-thickness skin grafts following excision of second-degree and third-degree burn wounds. However, the overall utility of this agent in off-label indications, including contaminated and infected wounds, has been clearly demonstrated since World War II and is unsurpassed.<xref ref-type="bibr" rid="B15">15</xref>–<xref ref-type="bibr" rid="B17">17</xref></p><p>The demonstration in this study of the antibacterial efficacy of 5% mafenide acetate solution against <italic>A baumannii</italic> is neither an accident nor serendipity. Mafenide, either as hydrochloride or as acetate, has been shown to be an effective topical antibacterial agent since its adoption by the German Army in World War II. Numerous studies conducted by Dr Janice Mendelson and by Drs John Moncrief and Robert Lindberg at the US Army Institute of Surgical Research, Brooke Army Medical Center, Fort Sam Houston, Tex, confirmed its effectiveness in the research laboratory and in the clinical setting.<xref ref-type="bibr" rid="B17">17</xref>–<xref ref-type="bibr" rid="B21">21</xref></p><p>Five percent mafenide acetate solution (5% Sulfamylon Solution) is especially effective in the treatment of burns, blast injuries, open fractures, synergistic gangrene, and necrotizing fasciitis. It retains its activity in the presence of blood and pus, and can penetrate beneath the surface of the wound to exert its antibacterial effects on both viable and nonviable tissue. It is not cytotoxic, may stimulate angiogenesis, and exerts its activity directly at the wound site. It is highly effective against both gram-positive and gram-negative bacteria, including anaerobes. It has no antifungal activity. Studies of the agent conducted against numerous strains of gram-positive and gram-negative organisms, including methicillin-resistant <italic>Staphylococcus aureus</italic>, vancomycin-resistant enterococcus, <italic>Pseudomonas aeruginosa</italic>, and <italic>Acinetobacer</italic> species, isolated from a broad array of clinical situations showed complete activity. In another study of more than 11,000 strains of <italic>P aeruginosa</italic>, collected over 25 years at the US Army Institute of Surgical Research, Brooke Army Medical Center, Fort Sam Houston, Tex, there was no evidence of resistance to 5% Sulfamylon Solution.<xref ref-type="bibr" rid="B15">15</xref>–<xref ref-type="bibr" rid="B28">28</xref></p><p>Mafenide acetate powder is stable and autoclavable and requires no refrigeration. It is supplied in drums and is then dispensed in 50 g premeasured packets of powder. Fifty grams of powder is mixed with 1 L of sterile water to yield a 5% solution of the agent. The resultant solution is clear and colorless and can be applied to wounds in the same fashion as is normal saline. Rather then relying on a single application of normal saline, which has no antibacterial activity and may predispose the wound to bacterial proliferation, 5% Sulfamylon Solution exerts a powerful, continuous antibacterial effect on a broad range of bacteria, including <italic>A baumannii</italic>. The dressings need to be remoistened with Sulfamylon Solution at regular intervals (6–8 hours) to prevent wound dessication and to maintain antibacterial potency.<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B27">27</xref></p><p>This agent was developed for wartime requirements and is the ideal topical agent for wound care in difficult environments. It is highly effective against the pathogens most commonly encountered in massive traumatic injuries, blasts, gunshot wounds, burns, and open fractures. Because of its aqueous nature it can be easily incorporated into dressings as a soak and it permits easy inspection of the wound. It can be applied with spray bottles or injected into rubber catheters that have been incorporated into the dressings in order to maintain antibacterial control.</p><p>The inclusion of bacteriological monitoring of the wound (quantitative bacteriology or swab techniques) into the overall wound management protocol may be helpful in gauging the efficacy of wound treatment and the suitability of the wound for DPC.<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B7">7</xref></p></sec><sec sec-type="conclusions" id="S4"><title>CONCLUSION</title><p>The increased incidence of <italic>A baumannii</italic> septicemia involving service members wounded during Operation Iraqi Freedom and Operation Enduring Freedom has prompted a thorough investigation to determine the source of the infections and to decrease their likelihood in the future. The investigation is focusing on many variables and factors that may play a causative role. Current treatment guidelines for open combat casualty wounds were not mentioned as part of the review.</p><p>The standard protocol currently in use relies on the fundamental concepts of adequate debridement and copious wound irrigation but does not provide for ongoing antibacterial activity within the wound prior to DPC.</p><p>Five percent Sulfamylon Solution was tested against numerous <italic>A baumannii</italic> isolates and exerted a strong antibacterial effect as demonstrated by large zones of inhibition in a standardized microbial susceptibility assay. The historical and current clinical and laboratory data clearly support the efficacy and utility of this topical agent in a wide variety of soft tissue wounds and infections. This agent was initially developed for wartime use and should be considered as part of the battlefield casualty antibacterial armamentarium. By substituting this agent for normal saline in the treatment protocol of open combat casualty wounds, reliable and effective bacteriologic control of the wound from the entire spectrum of battlefield bacterial contaminants (gram-positive and gram-negative) can be achieved.</p></sec>
Immunofluorescence Deconvolution Microscopy and Image Reconstruction of Human Defensins in Normal and Burned Skin	<p><bold>Objective:</bold> The aim of this study was visualization and localization of the human antimicrobials human beta defensins 1, 2, and 3, neutrophil defensin alpha (human neutrophil peptide), and the cathelicidin LL-37 in normal and burned skin, and determination of the cell types in which these antimicrobials were localized. <bold>Methods:</bold> Tissue sections were probed with antimicrobial antibodies, tagged with fluorescently labeled secondary antibodies, and subjected to fluorescence deconvolution microscopy and image reconstruction. Images were generated by stacking multiple-section scans, which were then volume rendered by rotating stacks 360° about an axis, or modeled in 3 dimensions. <bold>Results:</bold> This technique yields a definitive image, providing a rapid basis for further quantification and manipulation from a full 3-dimensional aspect. In normal skin, human beta defensin-1 was localized to the perinuclear region of keratinocytes; human beta defensin-2 was primarily localized to the stratum germinativum; human beta defensin-3 was found in dendritic cells of the stratum spinosum; human neutrophil peptide was randomly distributed in the papillary dermis; and LL-37 was concentrated in the stratum corneum and along ducts. In burned skin, in which keratinocytes are lost or destroyed, human beta defensin-1 was present in dermal glandular structures including hair shafts; human beta defensin-2 and human beta defensin-3 were found in the remaining keratin layers and glands of the lower dermis; human neutrophil peptide was primarily localized to hair shafts, though visible in residual keratin layers; and LL-37 was evident in very high concentrations in the epithelium of sweat ducts. <bold>Conclusion:</bold> We conclude via this technique that cells in the lower dermal and subdermal regions of burned skin synthesize antimicrobials after burn injury, and maintain something of a barrier against infection. This methodology is discussed and explained in this article.</p>	<contrib contrib-type="author"><name><surname>Poindexter</surname><given-names>Brian J.</given-names></name><degrees>MS</degrees><xref rid="aff1" ref-type="aff"/><xref ref-type="corresp" rid="cor1"/></contrib>	Journal of Burns and Wounds	<p>The skin has many natural defenses against infection. Tight junctions between keratinocytes prevent the invasion of microbes, and the fatty acid–rich and lipid-rich environment in the epidermis is toxic not only to bacteria but also to fungi and viruses.<xref ref-type="bibr" rid="B1">1</xref> <italic>Defensins</italic> are natural antimicrobial peptides, produced by various cells in human skin,<xref ref-type="bibr" rid="B2">2</xref>,<xref ref-type="bibr" rid="B3">3</xref> that also offer protection against invasion, especially when the skin barrier has been compromised by injury.<xref ref-type="bibr" rid="B1">1</xref> Keratinocytes of the epidermis synthesize cathelicidins,<xref ref-type="bibr" rid="B4">4</xref> eccrine glands produce sweat, which contains LL-37,<xref ref-type="bibr" rid="B5">5</xref> mast cells produce LL-37,<xref ref-type="bibr" rid="B4">4</xref> and neutrophils contain both human neutrophil peptides (HNPs) and LL-37.<xref ref-type="bibr" rid="B6">6</xref> Therefore, the loss or destruction of skin removes many of our natural defense mechanisms.</p><p>We previously determined that natural antimicrobial peptides were present in burned skin,<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B8">8</xref> despite the loss of the epidermis and even the upper dermis, and localized these peptides to specific cell types and particular layers of the remaining skin.<xref ref-type="bibr" rid="B9">9</xref> Fluorescence deconvolution microscopy yields definitive images that allow us to localize peptides and proteins to specific cell types and structures, and directs future studies to the upregulation of many of these microbials and the culture of multiple cell types for formulating wound cover matrices. Fluorescence deconvolution microscopy provides another research tool directed at the treatment of wounds and cellular regeneration.</p><sec sec-type="materials\|methods" id="S1"><title>MATERIALS AND METHODS</title><p>All chemicals were purchased from Sigma Chemical Corp (St. Louis, Mo), except where stated, and were of the highest grade available.</p><sec id="S1-1"><title>Tissue Preparation</title><p>Skin samples were obtained for frozen sectioning from patients admitted to the Regional Burn Center in Springfield, Ill, with partial- and full-thickness burns, ranging from 10% to 35% of total body surface area. Representative tissue specimens were harvested on the second or third day after injury during excision and grafting. Normal skin samples were taken from remnants of split-thickness autografts (0.30 mm). Samples were embedded in sucrose-based O.C.T. compound (Tissue-Tek, Torrence, Calif) and frozen on dry ice. Sections were cut at a thickness of 12 ± 3 μm with a Microm HM 505 E cryotome (Microm Laboratories, Walldorf, Germany) and placed on 18 mm glass cover slips (Fisher, Pittsburgh, Pa), which had been acid cleaned and coated with poly-<sc>l</sc>-lysine. Sections were fixed in 3.7% paraformaldehyde (Tousimis Research, Rockville, Md) for 5 minutes at room temperature, rinsed 5 times with phosphate buffered saline at room temperature, and cover slips were inverted and floated on 10% goat serum for 1 hour at 37°C to reduce nonspecific antibody binding. Antibodies for human defensins human beta defensin-1 (HBD-1), human beta defensin-2 (HBD-2), human neutrophil peptide-1 (HNP-1) (Alpha Diagnostic, San Antonio, Tex), human beta defensin-3 (HBD-3) (Novus Biologicals, Littleton, Colo), and LL-37 (Hycult Biotechnology b.v., Uden, The Netherlands) were diluted 1:100 in 10% goat serum and incubated with the sections for 45 minutes at 37°C. After rinsing the cover slips in 0.05% Tween-20 to remove unbound antibody, fluorescently tagged secondary antibodies (Molecular Probes, Eugene, Ore) were added and the sections were incubated for 30 minutes at 37°C. Finally, F-Actin and the nuclei were simultaneously fluorescently stained with phallicidin (Molecular Probes, Eugene, Ore) and 4′,6-diamidino-2-phenylindole (DAPI) (Molecular Probes) for 15 minutes at room temperature, and cover slips were mounted onto glass slides with Elvanol (DuPont, Willmington, Del) as the mounting media and attached with nail polish.</p></sec><sec id="S1-2"><title>Reconstructive Microscopy (Deconvolution)</title><p>Specimens were scanned with an Applied Precision DeltaVision (Issaquah, Wash) system fitted with an Olympus IX 70 inverted microscope employing a 100-W mercury arc lamp for illumination (Olympus America, Melville, NY) and excitation/emission filter sets (Chroma Technology Corp, Brattleboro, Vt) specific for each of the fluorescent antibodies. The filter set combination for DAPI (nucleus) was a 340 nm excitation filter with a band-pass of 20 nm and a 390 nm emission filter with a band-pass of 20 nm. Phallicidin (F-Actin) fluorescence was acquired with an excitation filter of 488 nm (band-pass 10 nm) and an emission filter of 520 nm (band-pass 25 nm). Defensin antibodies were visualized with an excitation filter of 585 nm (band-pass 10 nm) and an emission filter of 640 nm (band-pass 40 nm). Image scans for each probe were acquired in series at a step-size of 0.2 μm with a Sony Interline CCD camera. At least 30 sections were scanned per sample for each probe (ie, 90 total images for the 3 probes used). Specimen magnification was 400× unless otherwise noted.</p><p>Deconvolution and image analysis were performed by transferring the data sets to a Linux/RedHat workstation employing SoftWoRx software (Applied Precision) that uses an algorithm experimentally produced on the system from the convolution of a point spread function (PSF) to differentiate and reduce extraneous light or scattered light captured by the camera. A PSF describes the imaging and resolution characteristics of light collected by the optics of the microscope and was derived by scanning a 0.1 μm fluorescent bead (Molecular Probes, Eugene, Ore) 4 μm above and below the plane of focus. The resulting PSF was Fourier transformed into an optical transfer function that manipulated the data to produce images with a higher signal-to-noise resolution of the probe emission patterns. All data sets were subjected to 10 deconvolution iterations and then used for image analysis, image reconstructions, volume rendering, and modeling. Subtraction of background fluorescence and change of intensity gain were optimally set for each emission.</p><p>An image projection was produced by stacking each of the individual <italic>z</italic> sections of all 3 fluorescent probes into one image, resulting in a three-dimensional (3D) end product and an overlay of all colors. Volume rendering used the stack of <italic>z</italic> sections and rotated them about the <italic>x</italic> or <italic>y</italic> plane. Each of the volume rotation movies was produced with a 6° view angle. A 3D computer-generated virtual model of fluorescence emission patterns was produced to view the relative positions of each emission pattern and elucidate localization of the defensins to specific cells types. Briefly, positive and background intensities were measured and thresholds set to produce polygons based on the increased signal-to-noise ratio of the positive signal intensities. The software used these polygons to generate 3D objects and a virtual model that represented the emission patterns of each fluorescent probe. The 3D model was fully interactive and could be rotated in any plane for viewing at any angle or perspective.</p><p>The Institutional Review Boards of both Southern Illinois University and the University of Texas Health Science Center at Houston approved the study and appropriate consent was received.</p></sec></sec><sec sec-type="results" id="S2"><title>RESULTS</title><p>Figure <xref ref-type="fig" rid="F1">1</xref> shows stacked image acquisitions of the 5 antimicrobial peptides examined in sections of normal skin. Each image has been deconvoluted, the sections stacked, and the 3 colors overlaid (F-actin is green, nuclei are blue, and the defensins are red). Panel A shows HBD-1 localized to the keratinocytes, primarily in the peri-nuclear area; Panel B shows HBD-2 concentrated in the stratum basale, and distributed throughout the stratum spinosum and corneum; Panel C shows HBD-3 localized to dendritic cells; and Panel D is HNP, which was not found in the epidermis but was located in the upper levels of the dermis (determined by the presence of large cytoskeletal fibers). Finally, Panel E shows LL-37 associated with glands and ducts, absent from the spinosum but present in the outer corneum, indicating its deposition from secreted sweat.</p><p>Figure <xref ref-type="fig" rid="F2">2</xref> is a compilation of selected images acquired from samples of burned skin taken from areas where there were portions of epidermis remaining. Each of the 5 panels demonstrates a loss of viable keratinocytes, a reduced definition of the layers, and a clumped, denatured appearance of the cytoskeletal elements. In panel a, because of loss of keratinocytes, there is no distinct nuclear pattern of HBD-1, though there are clumped masses in the upper papillary dermis and a cluster in the region of the reticular dermis. Panel b demonstrates localization of HBD-2 to eccrine glands, and in Panel c, HBD-3 is shown associated with vascular elements in some sections. Panel d is HNP and again demonstrates localization with a duct and some located diffusely throughout the upper dermis, while LL-37 (panel e) has outer corneal deposition and a high colocalization to duct epithelium. Ducts are distinguished from hair shafts by the lack of F-actin and connective tissue in the former.</p><p>Figure <xref ref-type="fig" rid="F3">3</xref> is an example of volume rendering and volume rotation. This image of normal skin probed for HBD-1 rotates 360° about the <italic>y</italic> axis at a rotational angle of 6°. The rotation is of Panel A, in Figure <xref ref-type="fig" rid="F1">1</xref>, and demonstrates the specificity of HBD-1 to the nuclear region of keratinocytes. It also reveals the overall abundance and widespread distribution of HBD-1 in the epidermis, effecting protection against infection.</p><p>Figure <xref ref-type="fig" rid="F4">4</xref> is another volume rotation of normal skin highlighting HBD-3. In normal skin, HBD-3 is predominantly found in the dendritic cells of the epidermis. The 2-dimensional image in Figure <xref ref-type="fig" rid="F1">1</xref> (Panel C) does not fully display either the specificity of the peptide to the dendritic cells, or the 3-dimensional morphology of these dendritic cells. Therefore, Figure <xref ref-type="fig" rid="F4">4</xref> shows the area of the image that is cut out, been volume rendered and set to rotate 360° about the <italic>y</italic> axis to better visualize the specificity of HBD-3.</p><p>Figure <xref ref-type="fig" rid="F5">5</xref> is a volume rotation of burned skin probed for HNP, taken from Panel d in Figure <xref ref-type="fig" rid="F2">2</xref>. Rotation is 360° about the <italic>x</italic> axis. This image rotation clearly demonstrates a loss of viable keratinocytes but some HNP localization to the upper dermis. It also reveals some specificity of HNP to a sweat duct coursing through the upper part of the dermis. These volume rotations are not only useful for localizing positive emission patterns but also valuable for visualizing the true 3D nature of the specimens.</p><p>Figure <xref ref-type="fig" rid="F6">6</xref> shows images of a computer-generated model, with the colors consistent with previous images and figures. This method of volume rendering helps highlight the specific 3D localization pattern of HBD-3, and further demonstrates localization of HBD-3 to dendritic cells in the epidermis. Each of the images in the figure is produced by sequential rotation of the acquisition along the 3 axes.</p></sec><sec sec-type="discussion" id="S3"><title>DISCUSSION</title><p>The presence of natural antimicrobials in skin plays an important role in the body's natural defenses, and the loss of these compounds reduces our ability to combat infection and sepsis.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B10">10</xref> This is particularly true in cases of thermal injury, in which the resulting sepsis is often fatal.<xref ref-type="bibr" rid="B10">10</xref> Thermal injury results in the removal or destruction of tight junctions and intercellular phospholipid-lipid milieu defenses. It would therefore be beneficial to determine which antimicrobials, if any, remain after burn injury has removed or destroyed the epidermal keratinocytes. In addition, determining which specific cell types continue to synthesize antimicrobials after burn injury would allow us to further cell culture research to identify a better matrix for application to wounds. Finally, it would also be beneficial to ascertain whether antimicrobial upregulation in remaining cells and structures after thermal damage is something that should be vigorously pursued.</p><p>Our earlier work<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B8">8</xref> demonstrated the loss of epidermal HBD-2 after thermal injury, even in small patches of the remaining epidermis. Also, defensin has been found to be absent from burn blister fluid.<xref ref-type="bibr" rid="B11">11</xref> However, we reported that HBD-2 was located in deeper, dermal cells of burned skin, and this fact directed us to hypothesize that eccrine gland epithelia, or hair root–associated melanocytes, might be potential factories for antimicrobial production and that specific cell types could be cultured for production and harvesting of defensins to produce a topical agent, or wound cover matrix. Furthermore, upregulation of antimicrobial production by specific cells of the dermis would be a beneficial adjunct in the treatment of wounds and for combating infection.</p><p>In our recent studies, we imaged skin samples to localize natural antimicrobials before and after thermal injury to ascertain how much of this natural defense barrier against infection had been destroyed and to determine whether other cells and structures deeper to the skin should be considered for future cell culture and antimicrobial production experimentation. We employed fluorescence deconvolution microscopy to produce 3D, volume rendered images (computer-generated virtual models),<xref ref-type="bibr" rid="B12">12</xref> which allowed us to determine in which specific cell type(s) each antimicrobial may be found. Imaging showed that much of this antimicrobial defense mechanism had been destroyed after thermal injury, with HBD-1 all but missing from the skin, except in clumped conglomerations near areas where Meissner's corpuscles might be found. HBD-2 gave a positive fluorescence in the reticular dermal layer, clustered around eccrine cells. HBD-3 was present in a clumped pattern, and seemed not to be associated with any particular cell type or dermal structure. HNP was located near larger vessels and along hair shafts, and LL-37 showed high intensity colocalizations with sweat ducts.</p><p>We therefore demonstrated that natural antimicrobials are synthesized in the deeper portions of burned skin and that cells other than keratinocytes that make these compounds are eccrine epithelia, duct epithelia, and cells found in the base of hair roots and hair bulbs. Thus, we may ask ourselves that when cells are cultured to produce a suitable wound covering matrix, should the coculture, and/or admixing, of other cell types with keratinocytes give us some basis toward a topical treatment in which multiple antimicrobials are produced, or are there drugs or chemicals that will allow us to target specific cell types to induce increased synthesis of these defensins in the remaining skin elements following burn injury?</p><p>This work gives us more insights into the role of antimicrobial peptides in burned skin and its protective mechanisms. This report also demonstrates the use of fluorescence microscopy not only in the research arena but also as a valuable and powerful tool in pathological studies. Questions remain regarding future studies into therapies that might be developed to combat burn sepsis and initiate a rapid healing response. Coupled with the ever-improving specificity of fluorescent probes and antibodies, and the continuing evolution of better microscopes, we should seriously consider more, everyday uses for fluorescence targeting.</p></sec>
Comparison of Dermal Substitutes in Wound Healing Utilizing a Nude Mouse Model	<p><bold>Background:</bold> Dermal skin substitutes have become a standard of care in burn treatment. <bold>Objective:</bold> To compare and assess wound contracture reduction and histologic incorporation into the wound, dermal substitutes were implanted into full-thickness skin wounds in nude mice. <bold>Materials and Methods:</bold> Thirty-seven mice received a full-thickness 2 × 2 cm dorsal skin wound, and were either implanted with an acellular dermal matrix, Alloderm, Dermagraft-TC, Dermalogen, or Integra or assigned to the control group (with no dermal substitute). At 28 days postsurgery, the wounds were assessed for contraction, epithelialization, and other histological characteristics. <bold>Results:</bold> Each dermal substitute decreased wound contracture, but Alloderm and the acellular dermal matrix did so significantly compared to the control (<italic>P</italic> < .01 and <italic>P</italic> < .03, respectively). Within-group and control comparisons showed no significant differences with respect to the presence of dystrophic calcification, squamous hyperplasia, infiltration of neutrophils, fibroblasts, and macrophages, epidermal keratinocyte stratification, or collagen fiber configuration. <bold>Conclusions:</bold> Integra elicited the greatest foreign body response. Although the Dermalogen group had the thickest elastin fiber fragments, Dermagraft may have initiated the earliest elastin fiber formation in the wounds. While all dermal substitutes were incorporated into the wound bed and wound contracture was decreased, acellular dermal matrix and Alloderm, both human skin–derived products, produced less contraction and the thickest new “dermis” in the healed wounds compared to the control or synthetic dermal substitutes.</p>	<contrib contrib-type="author"><name><surname>Truong</surname><given-names>Anh-Tuan N.</given-names></name><degrees>MD</degrees><xref rid="aff1" ref-type="aff">a</xref></contrib><contrib contrib-type="author"><name><surname>Kowal-Vern</surname><given-names>Areta</given-names></name><degrees>MD</degrees><xref rid="aff2" ref-type="aff">b</xref><xref ref-type="corresp" rid="cor1"/></contrib><contrib contrib-type="author"><name><surname>Latenser</surname><given-names>Barbara A.</given-names></name><degrees>MD</degrees><xref rid="aff3" ref-type="aff">c</xref></contrib><contrib contrib-type="author"><name><surname>Wiley</surname><given-names>Dorion E.</given-names></name><degrees>MD</degrees><xref rid="aff2" ref-type="aff">b</xref></contrib><contrib contrib-type="author"><name><surname>Walter</surname><given-names>Robert J.</given-names></name><degrees>PhD</degrees><xref rid="aff2" ref-type="aff">b</xref></contrib>	Journal of Burns and Wounds	<p>Early excision of the burn eschar has greatly improved burn patients' survival. In some cases, the patient is left with extensive regions devoid of dermis and poor cosmesis. Currently, a variety of skin substitutes or artificial dermal replacements are used not only to decrease morbidity and wound contracture in severely burned patients but also to enhance cosmesis of partial- and full-thickness burn wounds.<xref ref-type="bibr" rid="B1">1</xref>–<xref ref-type="bibr" rid="B9">9</xref> Dermal substitutes serve as a scaffold into which cells can migrate and repair the injury. Although dermal substitutes and their histology have been described in the literature and several have been compared to a split-thickness skin graft as the “criterion standard,” there have not been any studies comparing the efficacy of these dermal substitutes as a group in respect to wound contracture and histologic features. The purpose of this study was to compare the effectiveness of several dermal substitutes in an animal model of wound healing. Human cadaver skin was not used because it is viable rather than processed tissue. For this reason, acellular dermal matrix (ADM) was a more compatible dermal substitute for this study because it was a processed nonviable dermal substitute derived from human cadaver skin. We hypothesized that human skin–derived dermal substitutes (ADM, Alloderm, and Dermalogen) would generate a thicker dermis with less wound contraction than would control (Fibrin Sealant Tisseel VH) or synthetic matrices (eg, Dermagraft-TC, Integra). The study compared the wound healing attributes of various dermal substitutes grossly and histologically.</p><sec sec-type="materials\|methods" id="S1"><title>MATERIALS AND METHODS</title><sec id="S1-1"><title>Dermal matrices and fibrin glue</title><p><italic>ADM</italic> was prepared as described below and in the literature by Takami et al.<xref ref-type="bibr" rid="B10">10</xref> It is a dermal collagen matrix derived from banked human skin that is treated to remove all cellular components.<xref ref-type="bibr" rid="B11">11</xref></p><p><italic>Alloderm</italic> (LifeCell Corporation, Branchburg, NJ) is a dermal collagen matrix derived from banked human skin that is treated to remove most cellular components. It is freeze-dried for shipping and storage.</p><p><italic>Dermagraft-TC</italic> (Advanced Tissue Sciences, La Jolla, Calif) is composed of a woven bioabsorbable polymer (polyglycolic and polylactic acids) membrane within which human dermal fibroblasts are grown and then devitalized. It is used for reepithelialization of middermal and mid-to-deep–dermal indeterminate burn wounds and contains Type I collagen fibers, glycosaminoglycans, and growth factors such as TGF-beta and decorin. This product is not used as a dermal substitute in humans, although it was used as such in this study.</p><p><italic>Dermalogen</italic> (Collagenesis, Beverly, Mass) is a powdered human dermal collagen matrix that is treated to remove some cellular components, is freeze-dried, and is then milled into a fine powder. The collagen concentration was mixed with RPMI (MP Biomedicals, Inc, Aurora, Ohio) to form a 15 mg/mL solution. It is used primarily for aesthetic plastic surgery, as a “filler.”</p><p><italic>Integra</italic> (Integra Life Sciences Corporation, Plainsboro, NJ) is a bilayer artificial skin replacement with a “dermal” layer composed of bovine collagen gel cross-linked with shark chondroitin-6-sulfate. The synthetic “epidermal” layer is composed of a polysiloxane polymer that was removed before use in this study.</p><p><italic>Fibrin Sealant Tisseel VH</italic> (Baxter Health, Deerfield, Ill) is a 2-component fibrin glue mixture: fibrinogen + calcium and thrombin + aprotinin (protease inhibitor) were combined quickly and dispensed onto a wound, forming a fibrin clot.</p></sec><sec id="S1-2"><title>Preparation of ADM</title><p>Cryopreserved normal human skin (U.S. Tissue & Cell, Cincinnati, Ohio) obtained from cadavers, using a dermatome set at 0.012 in thick, was thawed rapidly at 37°C. It was then treated with 2.5 units/mL Dispase II (Boehringer Mannheim, Indianapolis, Ind) in phosphate-buffered saline containing 0.2 mM CaCl<sub>2</sub> at 4°C for 24 hours to remove epidermis and other cellular components from the dermal matrix. Subsequently, the dermal matrix was incubated in buffered 0.5% Triton X-100 (United States Biochemical Corp, Cleveland, Ohio) for 24 hours at room temperature with continuous shaking. ADM was extensively washed with phosphate-buffered saline and stored in phosphate-buffered saline at 4°C until use. All solutions used for ADM preparation were filter-sterilized, and all procedures were performed aseptically. Sodium azide (0.02% w/v) was present at all times in the extraction solutions to prevent microbial growth, and was thoroughly washed out before implanting the ADM.</p></sec><sec id="S1-3"><title>Animals and surgery</title><p>Thirty-seven male NIH Swiss nude mice (NSWNU, homozygous, outbred, 4–5 weeks, 20–25 g) were used. The study was reviewed and approved by the Institutional Animal Care and Use Committee. Animals were housed in sterilized cages with sterile feed, water, and bedding. A single dose of antibiotic (kanamycin; 25 units/kg) was administered intramuscularly to all animals approximately 1 hour prior to surgery. All survival surgery was performed using aseptic technique in the Animal Facility operating rooms. Anesthesia consisted of ketamine/xylazine (150/8 mg/kg) administered intraperitoneally with additional ketamine (30 mg/kg) given intraperitoneally as needed to maintain deep anesthesia. Standardized full-thickness 20 mm × 20 mm skin wounds (removing the panniculus carnosus muscle layer) were excised from the dorsum of each anesthetized mouse. On Day 0, the mice weighed (mean ± SD) 27 ± 3 g and on Day 28, they weighed 31 ± 2 g, showing an overall weight gain of 4.6 ± 2 g. Mice with the Alloderm implant gained the most weight (6.5 ± 2 g) compared to the other groups. The weight gain included the weight of the implanted dermal substitute, which was not weighed at the time of implantation.</p></sec><sec id="S1-4"><title>Implantation of dermal substitutes</title><p>There were 6 study groups: ADM, Alloderm, Dermagraft-TC, Integra, Dermalogen, and control. The control group had the same type of wound and wound treatment as the other groups but did not receive a dermal substitute. Dermalogen solution (0.75 cc) was placed into the wound; skin at the wound corners was sutured to the underlying muscle. Each of the dermal matrices was cut to the size of the wound and sutured to the adjoining skin and the underlying muscle at their corners using nylon suture. The Integra silicone layer was removed after the Integra gel was placed into the wound. Fibrin Sealant Tisseel VH was sprayed on the wounds, which were then covered with semipermeable adhesive film (Op-Site, Smith and Nephew, Largo, Fla), Xeroform (Sherwood Medical, St. Louis, Mo), and dry cotton gauze. This dressed wound was then covered with fine stainless steel mesh fixed onto the animal's back with 4 sutures to the skin, each about 1 cm distant from the wound. The mesh was used to prevent the wounds from being disturbed by chewing or scratching. Animals and wounds were inspected daily to ensure that the dressings were intact. After 2 to 3 weeks postsurgery, the wounds were partially healed and the stainless steel mesh was removed. At the end of the 4-week study period, all animals were euthanized and the wounds were harvested, fixed in 10% buffered formalin, cut into five to seven 2-mm sections, and prepared for histological analysis.</p></sec><sec id="S1-5"><title>Gross and histologic measurements and assessments</title><p>Wound characteristics were measured grossly as well as histologically on Day 28 postsurgery. Direct measurements of wound area and the extent (area) of epithelialization and wound contraction were determined from digital photographs (Olympus C3030) using computer planimetry software (UTHSCSA ImageTool for Windows v. 2.00). All excised healed wounds were paraffin-embedded, and sections were stained with Hematoxylin & Eosin, Masson Trichrome, and Elastin Van Gieson. A “blinded” pathologist (A.K.V.) evaluated all samples histologically for depth and length of the healed wound area, epithelial stratification, incorporation of the dermal substitute, degree of neutrophil, macrophage, fibroblast, and foreign body giant (FBG) cell infiltration, and extent of elastin formation. This examination was performed using an Olympus microscope with a millimeter ruler eyepiece adapter. Histologic assessment was performed in low-power fields (100× magnification) on all cross sections of each wound area.</p></sec><sec id="S1-6"><title>Statistical analysis</title><p>Study groups were compared utilizing Statistica (Statsoft, Tulsa, Okla). Summary descriptive statistics such as median, means, standard deviation and error, 1-way analysis of variance, chi-square 2 × 2 summary frequencies (Pearson and Maximum Likelihood), and the Tukey test for unequal numbers were calculated. Nonparametric analyses utilized the Kruskal-Wallis analysis of variance by Ranks and Mann-Whitney <italic>U</italic> test. A <italic>P</italic> value of less than .05 was considered significant.</p></sec></sec><sec sec-type="results" id="S2"><title>RESULTS</title><sec id="S2-1"><title>Wound contracture</title><p>Table <xref ref-type="table" rid="T1">1</xref> shows the dimensions of the healed wounds at 28 days postsurgery for each group.</p><p>Dermal substitute implantations and control decreased wound contraction when assessed by percentage retention of the original wound area: Control by 34% ± 7%; Dermalogen by 39% ± 9%; Integra by 46% ± 12%; Dermagraft-TC by 50% ± 22%. Based on the initial 400 mm<sup>2</sup> wound surface area, Alloderm and ADM underwent the least contracture and retained the greatest mean surface area compared to control and other dermal substitutes (63% ± 14% [<italic>P</italic> < .01] and 57% ± 7% [<italic>P</italic> < .03) total wound surface area, respectively), Figure <xref ref-type="fig" rid="F1">1</xref>.</p></sec><sec id="S2-2"><title>Epithelialization</title><p>Gross and histologic evaluation of surface epithelialization correlated well; wounds were 80% to 99% epithelialized as determined by both methods. Of interest, Dermagraft-TC was fully incorporated as a neodermis in this mouse model even though it is not used in this manner in human burn patients. The length and depth of the healed wound were evaluated histologically (Fig <xref ref-type="fig" rid="F2">2</xref>). ADM produced a significantly thicker (<italic>P</italic> < .04) neodermis compared to the control.</p><p>Mouse keratinocytes migrated inward from the wound margins to regenerate the epidermis. The normal uniform 4-cell layer thick epidermis was replicated in control and Alloderm groups, but Integra, Dermagraft-TC, and Dermalogen groups exhibited a significantly thicker epidermal layer than did the control group, <italic>P</italic> < .04. Integra, Alloderm, and Dermagraft-TC groups showed a highly variable number of cell layers (1–5) in the epidermis (Fig <xref ref-type="fig" rid="F3">3</xref>).</p></sec><sec id="S2-3"><title>Neodermis</title><p>The healing process was characterized by mouse macrophages and fibroblasts initiating the repair and deposition of new fibrous tissue above and below the dermal substitutes enveloping them under an epidermal layer. New blood vessels and capillaries were observed within the dermal substitutes and the new fibrous tissue beneath the dermal substitutes. Histologically, the healed wound area was well delineated due to the lack of hair follicles and adnexa, which remained at and outside the healed wound margins. Because the dermal substitutes extended laterally to a greater extent than visualized grossly, the microscopically recorded healed wound lengths were increased in all groups; ADM and Alloderm groups maintained the original wound dimensions after healing, with the least contracture of the wound surface area (Table <xref ref-type="table" rid="T1">1</xref>).</p></sec><sec id="S2-4"><title>Miscellaneous</title><p>The histologic characteristics of 28-day wounds are shown in Table <xref ref-type="table" rid="T2">2</xref>.</p><p>The Integra matrix was encased in fibrous tissue with a significantly greater number of FBG cells compared to the control group and the other study groups, <italic>P</italic> < .003. Squamous hyperplasia, a reactive florid overgrowth of the epidermal layer, was seen most frequently with Integra, Dermagraft-TC, and Dermalogen. Squamous pearls (a result of keratinocyte degeneration) were prominent in Integra (83%) and control (71%) groups, followed by Alloderm (50%), Dermagraft-TC (40%), and Dermalogen (22%) groups but were absent from the ADM group (<italic>P</italic> < .04). By Day 28, Integra and Dermagraft-TC (dermal substitutes with the least resemblance to human dermis) were not incorporated into the fibrous dermal network and granulation tissue as well as were Alloderm, Dermalogen, and ADM. Histologically, the Integra framework and Dermagraft-TC polymers and membrane were still evident, not resorbed, although the wounds were healed. Within-group and control comparisons showed no significant differences with respect to the presence of dystrophic calcification, squamous hyperplasia, extent of granulation tissue, numbers of infiltrating neutrophils, fibroblasts, and macrophages, epidermal keratinocyte stratification, or collagen fiber configuration.</p></sec><sec id="S2-5"><title>Integra</title><p>In contrast to the other dermal substitutes used here, Integra showed the least propensity for a fibrous dermal restructuring by Day 28; squamous pearls filled the crevices of the Integra “scaffold,” and the hyalinized interstitium was extensively infiltrated by FBG cells. In the healed wound, the Integra matrix was enveloped by fibrous tissue with “islands” of Integra isolated by the ingrowth of fibers into areas that had been cleared by FBG cells. This was a morphology unlike that seen with any of the other dermal substitutes. Horizontal fibers, both superficial and deep, arrayed parallel to the tissue interfaces surrounded and infiltrated the other dermal substitutes in the healed wound areas.</p></sec><sec id="S2-6"><title>Elastin</title><p>The newly formed fibrous tissue in all groups contained few elastin fibers. Some of the dermal substitutes contained intrinsic elastin fibers in varying quantities. There were no elastin fibers seen in the control or Integra group, but fibers were prominent in wounds implanted with Dermalogen as short, stubby thick strands. They were also present to varying degrees in wounds implanted with Dermagraft-TC, Alloderm, and ADM. Slim threads that stained with the Elastin Van Gieson stain were seen in the newly formed fibrous tissue enveloping the Dermagraft-TC–implanted healed wound area. To a lesser degree, ADM and Dermalogen groups showed elastin strands that may have been displaced in the healing process into the newly generated fibrous tissue around the periphery of the implanted dermal substitute.</p></sec></sec><sec sec-type="discussion" id="S3"><title>DISCUSSION</title><sec id="S3-1"><title>Wound contracture</title><p>All full-thickness wounds implanted with/without dermal substitutes produced epithelialized healed wounds without dermal appendages by Day 28 postsurgery. The dermal substitutes were incorporated into the healed wounds to a greater or lesser extent depending on the initial composition of the substitute. Although there was histologic variability in the composition of the healed wounds, all dermal interventions decreased wound contraction and supported the formation of an epidermis. ADM and Alloderm, both human skin–derived materials, showed less contraction in the healed wounds compared to the control or the non-human skin–derived dermal substitutes. Dermalogen, while replete with fragmented elastic fibers, remained amorphous during healing and expanded along the wound margins, resulting in a thinner healed dermis. The control, which had only Fibrin Sealant on the wound surface, also decreased wound contracture by retaining 30% of the original wound area. The use of fibrin glue as a “dermal substitute” had been previously described.<xref ref-type="bibr" rid="B12">12</xref> Although it decreased wound contracture in this study, Fibrin Sealant was not as effective as the other dermal substitutes in creating a substantial neodermis.</p></sec><sec id="S3-2"><title>Epithelialization</title><p>All wounds implanted with dermal substitutes were epithelialized. Integra and ADM require iatrogenic intervention in patients (eg, for skin grafting), but in this model, competent mouse-derived epidermal layers (although thin) formed over the matrix without the need for additional skin grafting. Reports in the literature indicate that adult mouse bone marrow cells can differentiate into all skin components such as epidermal keratinocytes, sebaceous gland cells, and epithelial, dendritic, and endothelial cells.<xref ref-type="bibr" rid="B13">13</xref> It is unknown at this time whether the epidermal layer expanded from the epidermis along the wound margin or formed from adult mouse bone marrow cells that populated the area. Dermalogen and Integra showed a propensity for squamous hyperplasia, an exuberant regeneration of the epidermal layer. Human burn scars may develop pseudoepitheliomatous hyperplasia (disordered progression of squamous hyperplasia), with a proclivity toward aneuploidy, which may predispose toward the development of squamous cell carcinoma.<xref ref-type="bibr" rid="B14">14</xref> The present study showed that while squamous hyperplasia was seen, all wounds healed without evidence of malignant progression. Current literature reporting long-term follow-up of patients who received these dermal substitutes has shown no tendency toward the development of malignancy.</p></sec><sec id="S3-3"><title>Neodermis</title><p>Successful treatment with dermal skin substitutes requires low antigenicity, the capacity for rapid vascularization, and stability as a dermal template.<xref ref-type="bibr" rid="B7">7</xref>–<xref ref-type="bibr" rid="B9">9</xref> Dermal substitutes provide stability to split-thickness skin grafts and to cultured epithelial autografts. In this study, the dermal substitutes may have formed a barrier that prevented hair follicle and skin adnexa reconstitution within the 28-day study period. The most acceptable dermal substitutes were derived from full- or split-thickness allogeneic skin treated to remove epithelial components (keratinocytes, sweat glands, and sebaceous glands) and dermal components (fibroblasts, vascular endothelium, and smooth muscle) compared to those derived from xenogenic materials (pig or bovine skin, shark cartilage, etc). “Scaffolding” in ADM and Alloderm consisted of already well-formed fibrous tissue. In contrast, Integra required a dissolution of its “foreign body scaffolding” before the more permanent fibrous tissue could be laid out in the area it occupied. The human skin–derived dermal substitutes provided a thicker and longer dermis because they had, albeit altered, greater mass to start with in terms of collagen, and other fibers into which the macrophages and fibroblasts could infiltrate, initiating angiogenesis and dermal substitute incorporation.</p></sec><sec id="S3-4"><title>Integra</title><p>It has been previously reported that 14.4% of human patients who received Integra developed the FBG cells and eosinophils.<xref ref-type="bibr" rid="B15">15</xref> The present study showed numerous FBG cells in the Integra-implanted healed wound. This was probably a reaction to the constituents (ie, denatured bovine collagen and shark chondroitin sulfate) of Integra, which were perceived as foreign bodies. It is possible that this process contributed to the thinner “new dermis” observed here with Integra and it deserves further investigation, especially at later postsurgical intervals. This foreign body reaction to Integra components in some patients may be impacting the generation and vascularization of the new dermis as the wound heals.</p></sec><sec id="S3-5"><title>Elastin</title><p>Although it takes years for elastin fibers to become established in human scars, elastin fibers may form as early as 90 days into the wound healing process in mice.<xref ref-type="bibr" rid="B16">16</xref> While elastin has a complex structure, there have been reports of earlier elastin deposition, as early as 40 days after grafting in humans.<xref ref-type="bibr" rid="B16">16</xref>–<xref ref-type="bibr" rid="B19">19</xref> Dermagraft-TC when used as a dermal substitute appeared to promote the formation of elastin fibers earlier than did other dermal substitutes. Further study is needed to determine whether the elastin fibers were newly formed or whether they were transposed or redistributed from the dermal substitute during the progression of the fibrous network repair.</p></sec><sec id="S3-6"><title>Potential areas of investigations</title><p>This study, while showing that the human skin–derived dermal matrices formed the thickest neodermis and produced the least amount of contracture, was a relatively small study that investigated wound healing during a limited time period. Epithelialized wounds remodel continually for more than a year. Several questions were raised in this study that merit continued comparison of dermal substitutes and their long-term effects on wound healing. It is unknown whether the characteristics of each dermal substitute, observed at 28 days, would have been sustained for longer time intervals. Do patients receiving Integra as a dermal substitute retain the same skin pliability several years later? Why do some dermal substitutes elicit a FBG cell reaction and others do not? How effective is Fibrin Sealant alone as a dermal substitute? Do these dermal substitutes perform differently in burn/excised wounds? Investigations are underway to assess these questions, and to determine the long-term sequelae and wound remodeling characteristics of these dermal substitutes.</p></sec></sec>
Wound Healing of Cutaneous Sulfur Mustard Injuries	<p>Sulfur mustard is an alkylating chemical warfare agent that primarily affects the eyes, skin, and airways. Sulfur mustard injuries can take several months to heal, necessitate lengthy hospitalizations, and result in significant cosmetic and/or functional deficits. Historically, blister aspiration and/or deroofing (epidermal removal), physical debridement, irrigation, topical antibiotics, and sterile dressings have been the main courses of action in the medical management of cutaneous sulfur mustard injuries. Current treatment strategy consists of symptomatic management and is designed to relieve symptoms, prevent infections, and promote healing. There are currently no standardized or optimized methods of casualty management that prevent or minimize deficits and provide for speedy wound healing. Several laboratories are actively searching for improved therapies for cutaneous vesicant injury, with the aim of returning damaged skin to optimal appearance and normal function in the shortest time. Improved treatment will result in a better cosmetic and functional outcome for the patient, and will enable the casualty to return to normal activities sooner. This editorial gives brief overviews of sulfur mustard use, its toxicity, concepts for medical countermeasures, current treatments, and strategies for the development of improved therapies.</p>	<contrib contrib-type="author"><name><surname>Graham</surname><given-names>John S.</given-names></name><degrees>PhD</degrees><xref rid="aff1" ref-type="aff">a</xref><xref ref-type="corresp" rid="cor1"/></contrib><contrib contrib-type="author"><name><surname>Chilcott</surname><given-names>Robert P.</given-names></name><degrees>PhD</degrees><xref rid="aff2" ref-type="aff">b</xref></contrib><contrib contrib-type="author"><name><surname>Rice</surname><given-names>Paul</given-names></name><degrees>BM FRCPath</degrees><xref rid="aff2" ref-type="aff">b</xref></contrib><contrib contrib-type="author"><name><surname>Milner</surname><given-names>Stephen M.</given-names></name><degrees>MD</degrees><xref rid="aff3" ref-type="aff">c</xref></contrib><contrib contrib-type="author"><name><surname>Hurst</surname><given-names>Charles G.</given-names></name><degrees>MD</degrees><xref rid="aff1" ref-type="aff">a</xref></contrib><contrib contrib-type="author"><name><surname>Maliner</surname><given-names>Beverly I.</given-names></name><degrees>DO</degrees><xref rid="aff1" ref-type="aff">a</xref></contrib>	Journal of Burns and Wounds	<sec id="S1"><title>HISTORICAL PERSPECTIVE OF SULFUR MUSTARD USE</title><p>Sulfur mustard [bis(2-chloroethyl)sulfide; or HD] is a potent vesicating (blistering) chemical warfare agent that was first used in the First World War by Germany against French troops at Ypres, Belgium (1917). Since then, there has been evidence or allegations of HD use in 11 conflicts, including by Italy against Ethiopia in 1936, by Japan against China in 1937, by Poland against Germany in 1939, by Egypt against Yemen from 1963 to 1967, and by Iraq against Iran in the 1980s.<xref ref-type="bibr" rid="B1">1</xref> In addition, its use was threatened in the early 1990s during the Persian Gulf War. There are no peacetime industrial uses for HD. Aging HD stocks remain hazardous to civilians. These munitions and stockpiles were discarded on land and into the ocean during and after the Second World War. Periodically they surface during agricultural or fishing activities, causing serious injury. Extensive production and stockpiling during the Second World War, along with the known effects of HD on epithelial tissues, have spurred agreements to ban its production and use. These agreements include the 1993 Chemical Weapons Convention (CWC)<xref ref-type="bibr" rid="B2">2</xref> and agent destruction programs. Nonetheless, this chemical warfare agent remains a threat owing to its popularity in some countries that are not signatories to the CWC and to accidental exposures. Dispersed as a vapor, aerosol, or in liquid droplets, HD remains a threat to war fighters and civilians worldwide.</p><p>As a class, vesicants include sulfur mustard, the nitrogen mustards (HN<sub>1</sub>, the chemotherapeutic agent HN<sub>2</sub> or “Mustargen,” and HN<sub>3</sub>), arsenicals such as Lewisite, and halogenated oximes such as phosgene oxime. A plethora of information on vesicants is available in the published literature. This report is a review of some of the literature available on HD.</p></sec><sec id="S2"><title>SULFUR MUSTARD TOXICITY</title><sec id="S2-1"><title>General toxicology</title><p>The known molecular mechanisms of action, chemistry, toxicodynamics, and genotoxicity of HD as well as the pathogenesis and histopathology of HD injuries have been widely described.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B3">3</xref>–<xref ref-type="bibr" rid="B16">16</xref> Because its actions partly resemble those of ionizing radiation, HD is sometimes considered a radiomimetic compound. Papirmeister et al<xref ref-type="bibr" rid="B1">1</xref> wrote an extensive review of HD research, offered several theories of its cytotoxicity, and summarized what is known of its absorption, distribution, biotransformation, and excretion. Somani et al<xref ref-type="bibr" rid="B3">3</xref> have also provided an excellent overview of the toxicodynamics of HD.</p><p>Current research is not only investigating various aspects of Papirmeister's theories, but also actively pursuing the development and use of topical skin protectants, decontaminating agents, prophylactic and therapeutic countermeasures, and improved therapies for established lesions. Most of the research to date has focused on mitigating cutaneous injury. Larger knowledge gaps exist for injury to the eyes and the respiratory system, and for systemic effects.</p></sec><sec id="S2-2"><title>General clinical manifestations</title><p>Sulfur mustard is a bifunctional alkylating agent that causes extensive incapacitating injuries to the skin, respiratory tract, and eyes. It reacts with a large variety of molecules of biologic interest, ranging from compounds of low molecular weight to macromolecules such as DNA, RNA, and proteins.<xref ref-type="bibr" rid="B1">1</xref> Similar to other alkylating agents, it can cross-link complementary strands of DNA and thereby inhibit cell division. Although injury occurs very rapidly after tissue contact with HD, clinical signs of injury are delayed from 2 to 24 hours following exposure, with the length of the delay being inversely proportional to exposure level and other factors.<xref ref-type="bibr" rid="B16">16</xref> Because of this delay and lack of pain on contact, exposures can go unrecognized, and exposed persons may often fail to take immediate decontamination and protective measures, resulting in extensive exposure and severe injury.</p><p>Characteristic lesions are a function of dose and time after exposure.<xref ref-type="bibr" rid="B1">1</xref> At low toxic doses, skin lesions are characterized by an initial asymptomatic period of highly variable duration followed by localized pruritic erythema.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B15">15</xref> Higher doses initially produce small vesicles within or on the periphery of the erythematous areas that may later merge to form large, pendulous blisters.<xref ref-type="bibr" rid="B15">15</xref> Vesication may take several days to complete. Very high doses produce coagulation necrosis, which is often characterized as “doughnut burns” owing to the appearance of a central necrotic region of skin surrounded by a circular periphery of less damaged tissue.<xref ref-type="bibr" rid="B15">15</xref> Secondary bacterial infections of this type of wound pose a serious threat. In the eyes, low vapor exposures are characterized by reddening. As the concentration increases, conjunctivitis becomes progressively more severe, with photophobia, blepharospasm, pain, and corneal damage seen at higher doses.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B15">15</xref> Liquid or very large vapor exposure to the eyes results in severe corneal damage, with potential for blindness.<xref ref-type="bibr" rid="B15">15</xref> Inhalation of HD vapor induces changes in laryngeal and tracheobronchial mucosa, with mild to severe inflammation.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B15">15</xref> Depending on the degree of exposure, symptoms can range from a mild irritation of the upper respiratory tract to severe bronchiolar damage leading to epithelial necrosis, hemorrhage, inflammatory exudate, and pseudomembrane formation in the tracheobronchial tree.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B15">15</xref> While rare, massive pulmonary exposure can result in hemorrhagic pulmonary edema. In the First World War, respiratory injury from vapor exposures resulted in death due to pneumonia secondary to chemical pneumonitis.<xref ref-type="bibr" rid="B15">15</xref> More profound respiratory injury from aerosol exposures, as in the Iran-Iraq war of the late 1980s, have resulted in 2 waves of death.<xref ref-type="bibr" rid="B17">17</xref> The first wave occurred within 3 days of the attack from respiratory failure due to extreme injury to respiratory epithelium and alveoli. The second wave of deaths occurred between 1 and 3 weeks postexposure and resulted from secondary bronchopneumonia and sepsis secondary to marrow failure. High doses may produce a systemic toxicity that includes destruction of bone marrow precursor cells. Leukopenia typically begins 3 to 5 days after exposure<xref ref-type="bibr" rid="B15">15</xref> and may progress to pancytopenia. Very high doses may induce destruction of lymphoid organs and the mucosa of the small intestine and produce effects in the central nervous system, including apathy, depression, hyperexcitability, abnormal muscular movements, and convulsions.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B15">15</xref></p><p>Sulfur mustard has been shown to be carcinogenic and mutagenic in animal studies, and epidemiological studies of factory workers involved in HD production who were chronically exposed to low doses of HD have implicated it in human cancers (primarily respiratory carcinomas).<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B16">16</xref> Skin cancers have been experimentally generated in rats and mice exposed to HD and can potentially arise in HD-induced scar tissue in humans.<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B16">16</xref> There is no evidence that HD is a cause for adverse reproductive events, although epidemiological study of heavily affected civilian populations may alter current understanding.</p><p>Persistent problems of soldiers and factory workers exposed during the First World War include chronic bronchitis, asthma, laryngitis, recurrent pneumonia, and long-term keratitis.<xref ref-type="bibr" rid="B16">16</xref> Recent clinical reports of follow-up examinations of survivors of the 1987 Iraqi attack on Sardasht, including those who were children at the time, indicate chronic dermatological, ophthalmic, and respiratory problems.<xref ref-type="bibr" rid="B17">17</xref>,<xref ref-type="bibr" rid="B18">18</xref> Pulmonary dysfunction was the most common problem reported. It was noted that chronic effects tended to be significantly more pronounced in adults, possibly because mechanisms of cellular repair are more dynamic in children.<xref ref-type="bibr" rid="B18">18</xref></p></sec><sec id="S2-3"><title>Hypotheses of cytotoxicity</title><p>There are several hypotheses of HD cytotoxicity: (1) the poly(ADP-ribose) polymerase hypothesis, (2) the thiol-Ca<sup>2+</sup> hypothesis, and (3) the lipid peroxidation hypothesis.<xref ref-type="bibr" rid="B1">1</xref> None of these hypotheses has been fully accepted, and the mechanisms proposed by these theories may be active at the same time. At present, the exact mechanism of action of HD is not known.</p><p>The first hypothesis, originally proposed by Papirmeister,<xref ref-type="bibr" rid="B1">1</xref> has been coined the poly(ADP-ribose) polymerase (PADPRP) hypothesis. Sulfur mustard injury is initiated by a rapid alkylation of DNA. Alkylated DNA purines undergo spontaneous enzymatic depurination, resulting in numerous apurinic sites that are subsequently cleaved via apurinic endonucleases to form DNA breaks. Accumulation of these breaks activates the chromosomal enzyme PADPRP, which uses NAD<sup>+</sup> as a substrate to ADP-ribosylate several nuclear proteins, resulting in a depletion of cellular NAD<sup>+</sup>. NAD<sup>+</sup> depletion appears to be the result of an accelerated rate of use, as opposed to interference with its synthesis. This depletion results in the inhibition of glycolysis and subsequent stimulation of the NADP<sup>+</sup>-dependent hexose monophosphate shunt (HMS) owing to accumulation of glucose-6-phosphate. Activation of the HMS results in induction and secretion of proteases, which in turn lead to the typical pathology changes noted above. Results from in vivo studies using a human skin graft model and in vitro work using human lymphocytes, human keratinocytes, and human skin organ cultures suggest that NAD<sup>+</sup> levels that decrease below a critical value become irreversible and injury-producing. Nuclear pathology appears to precede cytoplasmic damage. This nuclear damage is characterized by loss of euchromatin and condensation and margination of heterochromatin.<xref ref-type="bibr" rid="B1">1</xref> Severe damage to the nuclear envelope, such as blebbing and breakage, has been demonstrated by electron microscopy, along with the formation of perinuclear vacuoles. The extent of nuclear damage has been shown to be dose- and time-related.<xref ref-type="bibr" rid="B1">1</xref></p><p>Evidence that may contradict the PADPRP hypothesis includes the facts that (1) NAD<sup>+</sup> depletion does not occur in rat keratinocytes until the concentration of HD is sufficient to inhibit DNA repair<xref ref-type="bibr" rid="B19">19</xref>; (2) DNA repair in human keratinocytes may be accomplished within 90 minutes of exposure,<xref ref-type="bibr" rid="B20">20</xref> whereas significant NAD<sup>+</sup> depletion is not observed until 1 to 3 hours postexposure; and (3) elevation or maintenance of NAD<sup>+</sup> in human keratinocytes does not confer protection.<xref ref-type="bibr" rid="B21">21</xref></p><p>The second hypothesis of HD cytotoxicity, the thiol-Ca<sup>2+</sup> theory, is based on a mechanism originally proposed by Orrenius and Nicotera in their studies of rat hepatocytes.<xref ref-type="bibr" rid="B22">22</xref> The initiating event here is a reduction in cellular protein thiol levels, leading to toxic increases in free cytosolic Ca<sup>2+</sup> levels. The thiol levels are depleted as a result of direct reactions of glutathione (GSH) and protein sulfhydryls with oxidants or other electrophilic xenobiotics. Glutathione is an intracellular scavenger of HD, which may be the cause of GSH depletion. Depletion of GSH occurs following reaction with electrophiles, thus exposing protein sulfhydryls to damage by the xenobiotic or by endogenously produced toxic oxygen species that arise as by-products of oxidative metabolism. Alternatively, HD may also react directly with protein thiols to cause the inactivation of enzymes. One group of enzymes affected by modification of its sulfhydryl groups is Ca<sup>2+</sup> translocases, resulting in alteration of intracellular Ca<sup>2+</sup> homeostasis. This leads to an increase in cytosolic free Ca<sup>2+</sup>, which in turn leads to activation of Ca<sup>2+</sup>-dependent catabolic processes, including stimulation of proteases, endonucleases, and phospholipases. Protease activity leads to protein degradation and ultimate perturbation of the cytoskeleton. Changes in the cytoskeleton may also arise from direct alkylation by HD on microfilamentous proteins, although this does not appear to play a critical role. Endonuclease activity leads to DNA breaks, leading to chromatin condensation and energy loss in the cell. Activation of phospholipases leads to phospholipid hydrolysis, with subsequent alterations in membrane fluidity and loss of membrane protein function and integrity. The cytoskeletal, nuclear, and membrane changes all lead to cell death.<xref ref-type="bibr" rid="B1">1</xref> It is not known whether vesicating doses of HD deplete GSH sufficiently to produce these cytotoxic effects. Slow GSH depletion may be less toxic because it allows time for cell adaptation. The lengthy latent period prior to clinical manifestation of HD lesions is inconsistent with the more rapid injury expected from an alkylation-induced GSH depletion, suggesting that damage to a cellular target other than GSH is responsible for initiating the cytotoxicity and that any involvement of the thiol-Ca<sup>2+</sup> cytotoxic pathway results from this initial damage.</p><p>The third hypothesis involves lipid peroxidation distinct from the thiol-Ca<sup>2+</sup> hypothesis, where the principal toxic consequence of GSH depletion is the formation of toxic lipid peroxides. Cell death is thus proposed to be due to an accumulation of endogenous oxidants (eg, H<sub>2</sub>O<sub>2</sub> accumulation resulting in hydroxyl and perferryl ion formation), leading to lipid peroxidation and irreversible membrane damage.<xref ref-type="bibr" rid="B1">1</xref></p><p>More recent theories to account for the mechanism of HD toxicity include a melanocyte free-radical hypothesis<xref ref-type="bibr" rid="B23">23</xref> and one in which HD requires metabolic activation.<xref ref-type="bibr" rid="B24">24</xref> Owing to the multiple molecular targets of HD, multiple pathways to cell death may be initiated. These multiple pathways may have some common steps; hence, all of the above hypotheses may play a role in the cytotoxicity of sulfur mustard. Papirmeister<xref ref-type="bibr" rid="B25">25</xref> also explained how HD could induce apoptosis (programmed cell death) and postulated how it might contribute to or modify the pathogenesis of HD injury. Kan et al<xref ref-type="bibr" rid="B26">26</xref> postulated that HD-induced cell death involves early apoptosis (6–12 hours postexposure) and late necrosis (24 hours), which temporally overlap to produce a single cell-death pathway along an apoptotic-necrotic continuum. Several lines of research are underway to further elucidate the role of apoptosis in HD toxicity.</p></sec><sec id="S2-4"><title>Pathogenesis of blisters</title><p>The cytotoxic effects of HD on skin have been widely described for a number of species.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B10">10</xref>,<xref ref-type="bibr" rid="B27">27</xref>–<xref ref-type="bibr" rid="B35">35</xref> The primary cutaneous cell population targeted by HD is the basal cell of the epidermis. Blisters (humans) and microblisters (animal models) occur on separation of the epidermis from the dermis at the dermal-epidermal junction. This separation is dependent on the loss of integrity of basal cells and anchoring filaments.<xref ref-type="bibr" rid="B4">4</xref> In animal model studies, the development of an apparent initial nuclear pathologic condition of basal cells of the stratum germinativum was followed by progressive cytoplasmic changes, leading to the eventual death of affected basal cells. Petrali et al<xref ref-type="bibr" rid="B4">4</xref> described degenerative subcellular effects of HD on the skin of hairless guinea pigs 24 hours after exposure. Degeneration was also observed in human lymphocytes and keratinocytes in vitro by Petrali et al,<xref ref-type="bibr" rid="B4">4</xref> in a variety of animal models by Papirmeister et al,<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B34">34</xref> and in rabbits and guinea pigs by Vogt et al.<xref ref-type="bibr" rid="B6">6</xref> While degenerative changes are noted in the upper layers of the epidermis as well as in epidermal cells of hair follicles, changes are most prevalent among basal cells, with the earliest and most severe degeneration seen in cells located above the dermal papillae.<xref ref-type="bibr" rid="B1">1</xref> Microblisters are observed to arise from focal areas of epidermal-dermal separation in areas of widespread basal cell pyknosis, 24 to 48 hours after HD exposure, as seen by light microscopy. Progressive changes reported in basal epithelial cells include formation of perinuclear or paranuclear vacuoles, a decrease in nuclear staining intensity, cytoplasmic swelling, relocation of chromatin to the periphery of the nucleus, loss of chromatin, and pyknosis.<xref ref-type="bibr" rid="B33">33</xref> These changes are followed or accompanied by necrosis, vacuolization, or hydropic degeneration of the cytoplasm.<xref ref-type="bibr" rid="B33">33</xref> While the extent of nuclear damage is dose- and time-related, it does not take place simultaneously in all basal cells examined, probably reflecting differing repair efficiencies of cells in various phases of the cell cycle.<xref ref-type="bibr" rid="B1">1</xref> Nuclear pyknosis has also been described following HD exposure in human skin equivalent.<xref ref-type="bibr" rid="B36">36</xref>,<xref ref-type="bibr" rid="B37">37</xref> It was also described in isolated perfused porcine skin flaps exposed to the HD simulant 2-chloroethyl methyl sulfide, a monofunctional analog of HD.<xref ref-type="bibr" rid="B38">38</xref> The time of onset of nuclear or cytoplasmic changes is species-specific. In general, pyknotic nuclei begin to appear in the basal cell layer 3 to 6 hours after HD exposure. By 12 hours, focal areas of pyknosis are seen and become widespread by 24 hours, with the nuclei of many suprabasal keratinocytes becoming involved.</p><p>The pathogenesis of microblisters is not fully understood. Petrali et al<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B39">39</xref> found indications that proteins of extracellular matrices of the basement membrane zone (BMZ) are affected during the development of HD-induced skin pathology in hairless guinea pigs and postulated that they may contribute to the formation of microblisters. Immunohistochemical staining for bullous pemphigoid antigen, a noncollagenous protein shared between basal cell hemidesmosomes and the lamina lucida, revealed a diminishing of bullous pemphigoid antigen reactivity at early times and subsequent loss of antigenicity at later time periods after an 8-minute HD vapor exposure (tissues were harvested at selected postexposure time periods up to 24 hours). Laminin, the major glycoprotein of the lamina lucida, showed scanty immunolocalization at the later time periods, conforming to the structurally altered lamina lucida at microblister lesion sites. The reactivity of Type IV collagen, a ubiquitous protein assigned to the lamina densa of basement membranes, was unaltered to specific antisera throughout prevesication and vesication time periods. The influence of these altered macromolecules on repair mechanisms following HD toxicity is not known. Other proteins of the BMZ yet to be investigated or discovered likely play a role in the pathogenesis of HD-induced microblisters.</p></sec><sec id="S2-5"><title>Chemical properties</title><p>A summary of the chemical, physical, environmental, and biological properties of HD and other vesicating agents (Lewisite, phosgene oxime) can be found in <italic>The Textbook of Military Medicine</italic>.<xref ref-type="bibr" rid="B15">15</xref> Sulfur mustard is a pale yellow to dark brown oily liquid, with an odor of garlic or mustard. It has low solubility in water and is highly lipophilic, so it readily partitions into the skin. Rates of 150 μg cm<sup>−2</sup> min<sup>−1</sup> through human skin in vivo<xref ref-type="bibr" rid="B40">40</xref> and 157 μg cm<sup>−2</sup> min<sup>−1</sup> through heat-separated human skin in vitro<xref ref-type="bibr" rid="B41">41</xref> have been reported. Penetration is enhanced by moisture and heat, and in thin skin. The LD<sub>50</sub> of liquid HD is about 100 mg kg<sup>−1</sup>. This is enough fluid (5–6 mL) to cover about 25% total body surface area (TBSA) in an adult.<xref ref-type="bibr" rid="B15">15</xref> A 10-μg droplet is enough to cause vesication. The threshold for vapor/aerosol to induce damage is 200 to 2,000 mg min m<sup>−3</sup> (concentration time, C<italic>t</italic>), dependent on the anatomical location, environmental conditions (temperature and humidity), sweating, and other factors.<xref ref-type="bibr" rid="B15">15</xref> Vapor injury generally induces superficial or partial-thickness dermal injury. Liquid HD can produce full-thickness injury. Because of its high freezing point (14°C), HD is very persistent in cold and temperate climates. Its persistence is lower in warmer climates, where the agent vaporizes more easily.</p></sec><sec id="S2-6"><title>Clinical dermal effects</title><p>Clinically, erythema is the first symptom noted, typically with a delayed onset of 4 to 8 hours. Pruritis, burning, or stinging may accompany the erythema.<xref ref-type="bibr" rid="B15">15</xref> Slight edema may also be present. If the exposure was small, the erythema will not progress to vesication. If the lesion progresses, very small vesicles will develop within or on the periphery of the erythematous areas beginning at about 2 to 18 hours postexposure.<xref ref-type="bibr" rid="B15">15</xref> These vesicles can later coalesce to form large blisters/bullae. Vesication may take several days to complete. Additional (new) blisters may arise a week later. Extremely high doses may induce a central zone of coagulation necrosis, with blisters forming along the periphery. Sulfur mustard blisters are subepidermal. Smaller ones are quite durable but large bullae are vulnerable to friction. The blister fluid is initially thin and clear, or slightly straw-colored, and later turns yellowish and tends to coagulate. It does not contain unreacted agent and is not a vesicant itself, unlike the fluid in Lewisite-induced blisters.<xref ref-type="bibr" rid="B15">15</xref> High environmental temperatures, hydrated skin, thin or delicate skin, or skin occluded by clothing will generally exhibit more severe lesions and shorter times to onset of symptoms than cooler, less hydrated, thicker, or unoccluded skin under similar levels of exposure. The face, neck, antecubital fossae, axillae, perineum, and external genitalia tend to be very sensitive areas. Differences in individual sensitivity to HD have been noted. Skin color has not been shown to affect HD penetration rates.<xref ref-type="bibr" rid="B42">42</xref> Transient hyperpigmentation was reported in casualties of the 1980s Iran-Iraq War, primarily due to accumulation of melanin derived from dead melanocytes at the base of an epidermis about to desquamate, and from opacification and darkening of nonviable epidermal cells.<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B43">43</xref>,<xref ref-type="bibr" rid="B44">44</xref> Following desquamation, the skin may appear hypopigmented. It is our opinion that long-term hyperpigmentation results from stimulation of melanocytes and hypopigmentation from melanocyte destruction. These pigmentary changes usually reverse in 6 to 12 months, but can be permanent.</p><p>Healing time of these dermal injuries depends on severity, with erythema alone taking several days to subside and severe lesions taking weeks to months depending on the anatomical location, depth of injury, and size. Lesions can be painful and secondary bacterial infection is a common problem (only in humans). While full reepithelialization has sometimes been considered as an endpoint for healing, barrier function must return for the lesion to be considered completely healed. Preliminary data from our laboratory indicate that barrier function, as indicated by transepidermal water loss (TEWL) measurements, is severely disrupted in ventral abdominal pig skin by 3 to 7 days after 30 to 120 minutes of exposure to liquid HD. While full reepithelialization of these deep dermal/full-thickness injuries appears grossly to occur 35 to 56 days after exposure, TEWL values do not return to baseline until 63 to 70 days postexposure (J. S. Graham et al, unpublished data, 1998). The delay in barrier disruption following agent exposure can be explained by the in situ existence of an intact stratum corneum for the first few days following exposure. The stratum corneum remains intact for 2 to 3 days, after which barrier function becomes compromised (ie, TEWL rates greatly increase) owing to loss of sloughing epidermis or deroofing of a blister. Following wound closure (eg, complete coverage of the damaged area by migrating keratinocytes), additional time is needed to form a fully stratified epidermis with a well-formed stratum corneum. Poor barrier function is generally attributed to an imbalance in water content, lack of proper organization, and/or adhesion of the corneocytes that make up the stratum corneum.</p></sec><sec id="S2-7"><title>Long-term cutaneous effects</title><p>Residual health effects of significant HD exposure are usually respiratory, ocular, or cutaneous. The permanent consequences of cutaneous injury can include hypopigmentation, hyperpigmentation, fragile skin that is easily damaged by trauma, and hypertrophic scarring. Skin hypersensitivity and chronic ulceration problems have also been reported.<xref ref-type="bibr" rid="B16">16</xref> Hypertrophic scarring is a result of uncontrolled fibroblastic activity and overgrowth of connective tissue during wound repair.<xref ref-type="bibr" rid="B16">16</xref> Excessive wound contraction has been noted in a weanling pig model following deep dermal/full-thickness HD injury<xref ref-type="bibr" rid="B45">45</xref> and observed in human casualties.<xref ref-type="bibr" rid="B44">44</xref> Excessive scarring and/or wound contraction over joints is disfiguring and impedes dexterity and locomotion. If initial treatment is not sufficiently aggressive and excessive wound contraction and scar tissue formation occur, surgical release may be indicated. As it is generally accepted that skin cancers can arise in scar tissue, HD-induced scars can be considered to have a carcinogenic potential. It has previously been noted that cutaneous cancers resulting from acute HD exposures usually localize in scars.<xref ref-type="bibr" rid="B16">16</xref> While cutaneous malignancies have been reported, they appear to be an uncommon consequence of HD exposure if indeed they were due to HD exposure alone.<xref ref-type="bibr" rid="B46">46</xref> Theories on the mechanism for malignant expression abound.</p></sec></sec><sec id="S3"><title>CONCEPTS FOR MEDICAL COUNTERMEASURES TO SULFUR MUSTARD</title><p>Concepts for medical countermeasures to vesicant agents have been developed and active research programs are underway in a number of North Atlantic Treaty Organization (NATO) countries. Institutes actively involved in this research area include the US Army Medical Research Institute of Chemical Defense, located at Aberdeen Proving Ground, Md, the Biomedical Sciences Department of the Defence Science and Technology Laboratory, Porton Down, Salisbury, Wiltshire, UK, and the Defence Research and Development Canada—Suffield Medicine Hat, Alberta. Developed research concepts include elimination of body contact, improved decontamination, pharmacological intervention, and chemical casualty management.</p><sec id="S3-1"><title>Elimination of contact</title><p>The initial step in protecting a person from the deleterious effects of HD is to eliminate contact with the agent. Respirators/protective masks can be donned to eliminate respiratory and ocular exposures,<xref ref-type="bibr" rid="B47">47</xref> specialized protective clothing (including suit, gloves, and foot protection) can be donned to prevent the agent from reaching the skin,<xref ref-type="bibr" rid="B47">47</xref> and topical skin protectants<xref ref-type="bibr" rid="B48">48</xref>–<xref ref-type="bibr" rid="B62">62</xref> can be used to protect areas of the skin vulnerable to the agent (eg, wrist, ankle, waist, and neck junctions in protective clothing).</p></sec><sec id="S3-2"><title>Decontamination</title><p>Skin injury from HD can be completely avoided through rapid physical removal, ideally within minutes. Removal in less than 2 minutes completely prevents damage, but even later removal lessens injury.<xref ref-type="bibr" rid="B15">15</xref> Physical removal without agent neutralization (eg, wiping with cloth or gauze, scraping with tongue depressor) can remove bulk agent from the skin and help reduce the severity of any resultant lesion. Copious quantities of water with soap appear to be quite effective in physically removing liquid HD from the skin surface.<xref ref-type="bibr" rid="B63">63</xref> Decontamination can also be performed by physical adsorption with or without chemical inactivation. There are a number of decontaminating agents available to remove the agent that are based on reactive powders (eg, Ambergard XE-555 Resin [Rohm and Haas Company, Philadelphia, Pa]<xref ref-type="bibr" rid="B63">63</xref>), neutralizing solutions (eg, 0.5% hypochlorite<xref ref-type="bibr" rid="B63">63</xref>), reactive skin lotions (eg, Reactive Skin Decontamination Lotion [RSDL, O'Dell Engineering Ltd/E-Z-EM, Inc, Lake Success, NY]<xref ref-type="bibr" rid="B64">64</xref>), or absorbent powders (eg, fuller's earth). Equally effective, however, are commonly available household products such as baking flour followed by wet wipes.<xref ref-type="bibr" rid="B65">65</xref> Solvents such as kerosene and surgical spirit may also be effective for removing HD from skin.<xref ref-type="bibr" rid="B66">66</xref></p><p>Wound healing studies using weanling swine have demonstrated that following exposure to undiluted liquid HD for 120 minutes, there is a significant period of off-gassing of unbound agent, as measured by a MINICAMS air monitor (OI Analytical, College Station, Tex; J. S. Graham et al, unpublished data, 2001). Quantification and localization of the HD depot responsible for this lengthy off-gassing in this animal model has not been performed. The existence of a dermal reservoir of HD in humans was first suggested in World War 1 by Smith et al,<xref ref-type="bibr" rid="B66">66</xref> who demonstrated that HD injuries could be prevented by washing contaminated skin with an appropriate solvent up to 45 minutes postexposure.</p><p>Furthermore, Smith et al demonstrated that the skin reservoir of HD could be transferred to a second individual, even after the exposed surface had been decontaminated. However, studies conducted during the Second World War reported the opposite effect in that HD was rapidly “fixed” by skin constituents such as proteins.<xref ref-type="bibr" rid="B40">40</xref> Contemporary in vitro studies have confirmed the original finding of Smith et al that a substantial reservoir of HD is formed in human skin that can account for up to 35% of the applied dose after 24 hours.<xref ref-type="bibr" rid="B41">41</xref> This reservoir has been localized to the stratum corneum and upper epidermis. This substantiates work conducted by Cullumbine<xref ref-type="bibr" rid="B67">67</xref> which demonstrated that the process of vesication could be blocked by the timely application of “peeling” (keratinlytic) agents up to 14 hours postexposure. The existence of an HD depot in human skin for a period of time following exposure has implications for the safety of medical emergency personnel treating HD casualties; for example, use of protective apparel may be warranted. Prior to medical treatment or casualty transport in enclosed vehicles, thoroughness of cleansing should be ensured through multiple washings and/or use of a detector. The existence of an agent depot could also influence the design of decontaminating agents. Any decontaminating agent that is capable of pulling the targeted agent out of the skin as well as neutralizing it on the surface of the skin has the potential to decrease HD-induced pathology, even beyond the 2-min efficacy window<xref ref-type="bibr" rid="B15">15</xref> of conventional decontamination procedures.</p><p>Understanding the kinetics of HD skin absorption, and the amount and persistence of unbound HD in putative agent reservoirs, will aid in choosing the most appropriate decontaminating agent and in determining its application doctrine and its window of effective use. The best agent will be the one that not only decontaminates HD sitting on the surface of the skin, but also is capable of fully penetrating the stratum corneum and neutralizing any unbound agent reservoir located there, unlike previous reactive therapies.<xref ref-type="bibr" rid="B68">68</xref> Research is underway to develop an in vitro model for efficacy testing of advanced decontaminating agents capable of pulling HD out of the skin reservoir and neutralizing HD on the surface of the skin, to identify the most appropriate animal model for extrapolation of animal data to humans, and to conduct in vivo efficacy tests of candidate decontamination systems.</p></sec><sec id="S3-3"><title>Pharmacological intervention</title><p>Because HD is not painful on contact, the exposed person may not be aware of the exposure until symptoms begin to appear after the latent period. Pharmacological approaches are being studied for their efficacy in minimizing or preventing that damage.<xref ref-type="bibr" rid="B69">69</xref>–<xref ref-type="bibr" rid="B71">71</xref> The most successful strategies to date have included the use of anti-inflammatories, protease inhibitors, intracellular scavengers, cell cycle inhibitors, PADPRP inhibitors, and calcium modulators, (Dr William J. Smith, PhD, US Army Medical Research Institute of Chemical Defense, Aberdeen Proving Ground, Md, oral communication, 2004). It is interesting to note that pretreatments (eg, antioxidants) that reside in the extracellular matrix are generally more effective than those designed for intracellular protection against HD (Defence Science and Technology Laboratory, Porton Down, unpublished data, 2004). This may imply that the cytotoxic effects of HD are mediated via interaction with cell surface components rather than intracellular targets. Thus, further investigations of the interaction of cell-surface molecules with HD may provide a new insight into the mechanisms of HD toxicity.</p></sec><sec id="S3-4"><title>Chemical casualty management</title><p>When HD comes in contact with the skin, decontamination is not performed in a timely fashion, and pharmacological intervention is absent or not adequately effective, a chemical casualty will be produced that requires medical attention. Casualty management now comes into play, discussed in depth in the next section. Educational material and training courses are available through the Chemical Casualty Care Division of the US Army Medical Research Institute of Chemical Defense, Aberdeen Proving Ground, Md (on the Web at <ext-link ext-link-type="uri" xlink:href="https://ccc.apgea.army.mil/default.htm">https://ccc.apgea.army.mil/default.htm</ext-link>, by phone at 410-436-2230/3393, or by e-mail at <email xlink:href="email://[email protected]">[email protected]</email>).</p></sec></sec><sec id="S4"><title>CURRENT TREATMENTS FOR CUTANEOUS SULFUR MUSTARD INJURIES</title><p>There are currently no standardized or optimized methods of casualty management and no drugs available to prevent the effects of HD on skin and mucous membranes. Historically, blister aspiration and/or deroofing (epidermal removal), physical debridement, irrigation, topical antibiotics, and sterile dressings have been the main courses of action in the medical management of cutaneous HD injuries.<xref ref-type="bibr" rid="B12">12</xref>,<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B72">72</xref>,<xref ref-type="bibr" rid="B73">73</xref> Current treatment strategy consists of symptomatic management and is designed to relieve symptoms, prevent infections, and promote healing. The general recommendations listed below are fully described in the <italic>Textbook of Military Medicine</italic>,<xref ref-type="bibr" rid="B15">15</xref> the <italic>Field Management of Chemical Casualties Handbook</italic>,<xref ref-type="bibr" rid="B74">74</xref> the <italic>Medical Management of Chemical Casualties Handbook</italic>,<xref ref-type="bibr" rid="B75">75</xref> and the <italic>NATO Handbook on the Medical Aspects of NBC Defensive Operations</italic>.<xref ref-type="bibr" rid="B76">76</xref></p><p>The decision to evacuate and hospitalize an HD casualty is based on the extent and severity of the skin lesions, in consideration with other injuries that may be present (eg, respiratory, ocular). For patients experiencing only cutaneous HD injuries, erythema covering more than 5% TBSA in noncritical areas requires hospitalization. Erythema covering less than 5% TBSA may require hospitalization, depending on the site of the injury (eg, face, inguinal area) and level of impairment (eg, limitation of limb movement due to pain, edema). Total body surface area can be determined using Wallace's Rule of Nines and the Lund and Browder chart for estimating burn severity.<xref ref-type="bibr" rid="B77">77</xref>,<xref ref-type="bibr" rid="B78">78</xref> Multiple or large areas of vesication will also require hospitalization. Since blister formation may initially be slight, the patient should be watched for a progression in the size and number of blisters. Topical antibacterial creams such as silver sulfadiazine or 10% mafenide acetate can be prescribed to patients not requiring close medical monitoring, with instructions to apply a thin layer to the affected area 4 times a day. Following application of the cream, the area should be covered with a petrolatum gauze bandage.</p><p>Not all burn injuries require specialized care in a burn center. The American Burn Association has well-defined criteria for patient transfer to such a center,<xref ref-type="bibr" rid="B78">78</xref> which should serve as additional guidance in deciding where to hospitalize an HD casualty.</p><p>Before commencement of any treatment, clothing should be carefully removed and treated as potentially contaminated, and the patient thoroughly decontaminated. For a general overview of decontamination procedures, the reader is directed to Hurst's chapter on decontamination in the <italic>Textbook of Military Medicine</italic>.<xref ref-type="bibr" rid="B63">63</xref></p><p>Direct comparisons in the literature between HD and thermal burns are scarce. Papirmeister noted that disintegration of the basal cell layer caused by thermal burns has been shown to produce an intraepidermal blister that contains fragments of the basal cell layer attached to the basal lamina, unlike the almost totally denuded basement membrane in HD lesions.<xref ref-type="bibr" rid="B1">1</xref> An argument against the adage “a burn is a burn” is that HD initially targets a specific cell type (epithelial basal cells), unlike a thermal burn that begins damage at the stratum corneum and then works its way downward. Since stratum corneum is the structure largely responsible for barrier function, water loss rates are very high immediately after a thermal burn (140–180 g m<sup>−2</sup> h<sup>−1</sup> in humans).<xref ref-type="bibr" rid="B79">79</xref> The stratum corneum remains intact for 2 to 3 days after a cutaneous HD injury, after which barrier function becomes compromised by loss of sloughing epidermis or deroofing of the blister. The systemic fluid derangement seen in cutaneous HD injury is less than that seen with thermal burns. Fluids and electrolytes should be closely monitored, since fluids may be lost to edematous areas, with resultant dehydration. Medical personnel are cautioned not to overhydrate the patient; hypervolemia and pulmonary edema can be iatrogenically induced in HD casualties.<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B44">44</xref> Fluid requirements in Iranian casualties during the Iran-Iraq war appeared to have been relatively independent of TBSA.<xref ref-type="bibr" rid="B44">44</xref> The recommended infusion rates and formulas used to calculate total volume requirements for thermal burn patients, based on body weight and TBSA,<xref ref-type="bibr" rid="B80">80</xref> should not be routinely applied in HD casualty management. The exact fluid replacement requirements for cutaneous HD injuries should be based on patient status and considered on a case-by-case basis. The fluids used in replacement fluid therapy for non-HD burns, which would likely be appropriate for use in HD injuries if fluid replacement is required, are described by Settle,<xref ref-type="bibr" rid="B80">80</xref> Brisebois,<xref ref-type="bibr" rid="B81">81</xref> and Thomas et al.<xref ref-type="bibr" rid="B82">82</xref></p><p>Sulfur mustard casualties should be kept comfortable and their lesions regularly cleansed to prevent infections. Limbs may need to be immobilized, as movement of joints can aggravate existing lesions. Blisters arising on the trunk require protective dressings to avoid or minimize damage as a result of friction with clothing or bedding. As burning and itching sensations are typically present after the appearance of erythema is noted, topical antipruritics are applied (eg, calamine lotion, 0.25% camphor, menthol, corticosteroidal preparations, and silver sulfadiazine cream). Systemic analgesics and antipruritics may be indicated, depending on the discomfort level of the patient.</p><p>Infection is a significant factor in causing delayed healing of cutaneous HD injuries. These injuries are covered by necrotic debris, which is a nidus for infection. There is no consensus, however, on whether intact blisters should be deroofed. Blister fluid from intact blisters provides a sterile wound covering, but the blisters are fragile and easily ruptured. Once blisters have broken, ragged roofs should be removed and sterile dressings put in place as soon as possible. Wounds should be inspected periodically for signs of infection.</p><p>It is generally recommended in military medical manuals to deroof blisters that are greater than 1 cm in diameter, irrigating the underlying area 2 to 4 times per day with saline, sterile water, clean soapy water, or Dakin's solution. Following cleansing, the area should be liberally covered with a topical antibiotic cream (eg, silver sulfadiazine; mafenide acetate; bacitracin; and triple combination preparations of neomycin sulfate, polymyxin B sulfate, and bacitracin zinc [Neosporin, Pfizer Inc, New York, NY]). A sterile dressing should then be put in place. Blisters that have already broken should have their ragged edges removed and the area irrigated, treated with antibiotic, and dressed with a sterile dressing. Blisters less than 1 cm in diameter should be left intact, with the area surrounding the blister irrigated at least once per day followed by application of a topical antibiotic. A petrolatum gauze bandage can be put in place over these unbroken blisters, if desired. Any such dressings should be changed every 3 to 4 days.</p><p>While these handbooks recommend the use of bacitracin and triple combination preparations following cleansing of deroofed blisters, they do not provide specific application guidelines. We feel that the use of these ointments should be limited to small wounds (less than 1% TBSA) and employed for very brief periods (3–5 days) because of their high capacity to provoke allergic cutaneous reactions. Likewise, the use of 10% mafenide acetate cream should be avoided because of the severe pain that it causes when applied to partial-thickness wounds and the possibility of metabolic derangements. Such problems are not encountered with the use of 5% mafenide acetate solution, which should be used instead of the 10% cream.</p></sec><sec id="S5"><title>STRATEGIES FOR THE DEVELOPMENT OF IMPROVED THERAPIES</title><p>Previous animal studies have shown that surgically aggressive approaches are needed to prevent or minimize significant cosmetic and functional deficits that result from deep HD injury. For the best outcome, deep dermal/full-thickness cutaneous HD injuries require full-thickness debridement followed by autologous split-thickness skin grafting.<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B83">83</xref> These surgically aggressive approaches to deep HD injuries resulted in the return of barrier function, skin color, and mechanical properties (hardness and elasticity) to near-normal levels within 15 days of treatment in a weanling pig model.<xref ref-type="bibr" rid="B83">83</xref> To be successful, the skin grafts must be placed on a hemostatically secure wound bed, devoid of blood clots, debris, or necrotic tissue. The recipient bed must have an adequate blood supply to nourish the skin grafts, and the grafts must be protected from shearing forces, motion, and mechanical disruption. A variety of modalities is available and may be employed in achieving initial graft adherence and subsequent acceptance (“take”). These include sutures, surgical staples, fibrin glue, tie-over bolsters, compression dressings, and a variety of antishear dressing techniques. The choice of fixation and dressing technique is determined by the size and location of the wounds, and the experience and preferences of the surgeon.</p><p>Split-thickness skin grafting was used late in a few cases during the Iran-Iraq conflict where healing was particularly slow<xref ref-type="bibr" rid="B12">12</xref> and for the late treatment of some poorly healing, deep injuries sustained in 1992 by a civilian who came across an unexploded artillery shell from the First World War.<xref ref-type="bibr" rid="B72">72</xref> The <italic>NATO Handbook on the Medical Aspects of NBC Defensive Operations</italic> states that grafting has rarely been required in the past and when it was attempted, graft acceptance has been poor.<xref ref-type="bibr" rid="B76">76</xref> Surgical details of the grafting procedures used are not readily available, and the procedures may not have been optimal. In contrast to this handbook, Graham et al<xref ref-type="bibr" rid="B45">45</xref> noted equally high graft acceptance rates following either full-thickness sharp surgical tangential excision or laser debridement using a deep dermal/full-thickness HD injury model in weanling pigs. In thermal burns management, deep burns are grafted to promote timely wound closure and improve outcome with minimal cosmetic and functional deficits. The decision to graft is based on depth of injury. As with thermal burns, depth of HD injury should be accurately assessed before treatment begins. Reported long-term effects such as fragile skin and scarring likely indicate that injury depth was not accurately diagnosed and treatment was not sufficiently aggressive. As with deep thermal burns, deep HD injuries will require surgically aggressive approaches.</p><p>While past HD wound-healing research in swine has concentrated on deep dermal/full-thickness injuries, superficial (epidermis only) and superficial dermal injuries may have greater clinical relevance on the battlefield. Partial-thickness injuries will likely not require such surgically aggressive approaches (eg, split-thickness skin grafting). Treatment strategies for improved healing of partial-thickness cutaneous HD injury have recently been formulated by a working group of researchers and physicians at government laboratories in the United States and United Kingdom. The strategies are described below. Research is underway to experimentally support these strategies and determine which medical devices, supplies, and pharmaceuticals are most efficacious.</p><p>It is important to recognize that for any therapeutic regimen to be successful, a healthy immunological,<xref ref-type="bibr" rid="B84">84</xref> nutritional/metabolic,<xref ref-type="bibr" rid="B85">85</xref> and psychological<xref ref-type="bibr" rid="B86">86</xref> status needs to be maintained. Infections and perturbations in organ function also need to be closely monitored and addressed as needed.</p><p>The ultimate goal is to determine the most efficacious treatment regimen to be applied in the clinical management of HD casualties. The ideal regimen should return damaged skin to optimal appearance and normal function in the shortest time. Improved treatment will result in a better cosmetic and functional outcome for the patient and will enable the casualty to return to normal activities sooner.</p><sec id="S5-1"><title>The pig as an animal model for efficacy testing of candidate treatment regimens</title><p>Since human testing with vesicating agents such as HD is unethical, candidate treatment regimens need to be tested in an appropriate animal model. The animal model of choice is the pig, due to the similarities between human and porcine skin.<xref ref-type="bibr" rid="B87">87</xref>–<xref ref-type="bibr" rid="B97">97</xref> The comparable histological characteristics of pig and human skin are similarities in epidermal thickness and composition, epidermal enzyme patterns, epidermal tissue turnover time, lipid content, character of keratinous proteins, pelage density and pattern of hair growth, dermal structure, deposition of subdermal fat, and general morphology.<xref ref-type="bibr" rid="B90">90</xref>–<xref ref-type="bibr" rid="B92">92</xref> In addition, pig skin is antigenically closer to human skin than is rodent skin. A number of human antibodies have been shown to cross-react in pig skin.<xref ref-type="bibr" rid="B98">98</xref> US Environmental Protection Agency guidelines for dermal exposure assessment state that the percutaneous absorption of many compounds in the pig is similar to that found in humans.<xref ref-type="bibr" rid="B93">93</xref> Dick and Scott<xref ref-type="bibr" rid="B90">90</xref> found that pig skin permeability to selected lipophilic penetrants was closer to that of human skin than was rat skin. Klain et al<xref ref-type="bibr" rid="B94">94</xref> concluded that pig skin was a good model for human skin metabolic studies. Meyer et al<xref ref-type="bibr" rid="B95">95</xref> concluded that among the domestic species, the pig provides the most suitable experimental model for dermatological research on humans. Results from studies of experimental treatments in porcine models of partial-thickness wound healing have correlated well with results of clinical studies.<xref ref-type="bibr" rid="B91">91</xref> These findings suggest that the pig is a suitable research animal to use for predicting cutaneous effects of xenobiotics in humans. Some differences between human and porcine skin have been noted, however. Pigs lack eccrine sweat glands,<xref ref-type="bibr" rid="B92">92</xref> although hair and sebaceous gland number and distribution are similar.<xref ref-type="bibr" rid="B91">91</xref> Tubular apocrine glands are present and lie adjacent to hair follicles, and are more numerous than in humans.<xref ref-type="bibr" rid="B92">92</xref> The subepidermal vascular network is less dense than it is in humans; however, the pattern of vascularization in the lower region corresponds to that found in humans.<xref ref-type="bibr" rid="B92">92</xref> In addition, the permeability of pig skin to HD may be significantly higher than that of human skin under certain conditions.<xref ref-type="bibr" rid="B96">96</xref></p><p>Pigs have been widely used in vesicant research.<xref ref-type="bibr" rid="B10">10</xref>,<xref ref-type="bibr" rid="B11">11</xref>,<xref ref-type="bibr" rid="B38">38</xref>,<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B83">83</xref>,<xref ref-type="bibr" rid="B96">96</xref>–<xref ref-type="bibr" rid="B109">109</xref> Weanling pigs have often been used for their ease of handling during lengthy wound healing studies. They are small enough to be easily placed under chemical fume hoods during agent exposures and have a large enough body surface to allow multiple, large (3-cm-diameter) lesions to be placed on either the dorsum or the ventral abdominal surface. While they do grow during wound healing studies, functional data (eg, microcutaneous blood flow, skin color) can be normalized to surrounding unaffected skin<xref ref-type="bibr" rid="B83">83</xref> and morphometric data can be normalized to total body surface area<xref ref-type="bibr" rid="B45">45</xref> to take this growth into account. The healing process appears to be optimal in young pigs.<xref ref-type="bibr" rid="B92">92</xref> The healing rate is faster in Yorkshire piglets than in mature Yucatan miniature pigs,<xref ref-type="bibr" rid="B91">91</xref> and young pigs are highly resistant to contamination and infection.<xref ref-type="bibr" rid="B92">92</xref> Furthermore, pigs are amenable to habituation and can be trained to allow noninvasive biophysical skin measurements to be obtained from HD-exposed sites on the dorsum without the need for restraint or anesthesia.<xref ref-type="bibr" rid="B110">110</xref> Finally, the use of pigs in cutaneous ulcer and burn wound research is supported by the US Food and Drug Administration.<xref ref-type="bibr" rid="B111">111</xref> Thus, weanling pigs appear to be a suitable model for examining the efficacy of treatment regimens without significant interference caused by high infection rates or slow healing rates. However, they would not likely be the ideal model for studying wound colonization or infection.</p><p>While there is no common laboratory animal species, including the pig, that generates frank blisters, as do humans, HD has been noted to induce microblisters in pigs.<xref ref-type="bibr" rid="B10">10</xref> This lack of frank blistering is thought to be the result of a diminished superficial dermal vascular plexus, a densely arranged dermis, and lack of loose areolar tissue that precludes intercellular fluid accumulation.<xref ref-type="bibr" rid="B92">92</xref></p></sec><sec id="S5-2"><title>Immediate treatment of cutaneous sulfur mustard casualties</title><p>This section describes the recommendations of the US-UK working group for the immediate treatment of HD casualties. The uses of anti-inflammatory agents, antioxidants, and occlusive/semiocclusive dressings are discussed. The potential need for replacement fluid therapy and management of intact blisters are also addressed.</p><p>For those patients who are beginning to present with erythema or those who are in the latent period and suspect an exposure may have occurred, systemic administration of an anti-inflammatory agent will likely help to decrease the amount of damage ultimately induced. The pro-inflammatory mediators IL-1β, IL-6, IL-8, and TNF-α are released by normal human epidermal keratinocytes in culture on exposure to HD.<xref ref-type="bibr" rid="B112">112</xref> Sulfur mustard has also been shown to provoke an edema response and release of IL-6 in 2 different mouse models.<xref ref-type="bibr" rid="B113">113</xref> Sulfur mustard–induced inflammatory responses themselves likely contribute to the severity of the pathology, and numerous animal studies have shown the benefits of prophylactic or therapeutic use of anti-inflammatory agents.<xref ref-type="bibr" rid="B43">43</xref>,<xref ref-type="bibr" rid="B69">69</xref>–<xref ref-type="bibr" rid="B71">71</xref> There are active programs researching a variety of nonsteroidal anti-inflammatory drugs (NSAIDs) administered topically or systemically, alone and in various combinations. Top candidates that have shown efficacy in a mouse ear model include indomethacin, fluphenazine dihydrochloride, olvanil, retro olvanil, octyl homovanillamide, and other analogs of capsaicin.<xref ref-type="bibr" rid="B69">69</xref>–<xref ref-type="bibr" rid="B71">71</xref> It remains to be determined which NSAID (or combination), route of administration, length of administration, and dosing regimen is the most efficacious in preventing or ameliorating the effects of HD on skin. It is likely that administration for 2 to 5 days will be required for an NSAID. Topically delivered intracellular scavengers such as 4-methyl-2-mercaptopyridine-1-oxide<xref ref-type="bibr" rid="B69">69</xref> and dimercaprol<xref ref-type="bibr" rid="B69">69</xref>,<xref ref-type="bibr" rid="B71">71</xref> have proved effective in animal experiments in reducing the severity of HD-induced cutaneous injuries, and concurrent use of one of these agents with an NSAID may yield the best results (Dr William J. Smith, PhD, US Army Medical Research Institute of Chemical Defense, Aberdeen Proving Ground, Md, oral communication, 2004). (It should be noted that mouse skin is very thin and permeable. Any topical agent showing efficacy in a mouse model should also be tested in another animal model such as the weanling pig.) Corticosteroid anti-inflammatory agents such as hydrocortisone (given systemically or topically for cutaneous HD injuries)<xref ref-type="bibr" rid="B70">70</xref>,<xref ref-type="bibr" rid="B71">71</xref> and dexamethasone (tested in vitro on primary alveolar macrophages and given topically for ocular HD injuries)<xref ref-type="bibr" rid="B114">114</xref>,<xref ref-type="bibr" rid="B115">115</xref> also appear to be promising therapeutic agents. There are other topical steroidal anti-inflammatory agents of much greater potency that would likely be very efficacious if used early in the lesion development stage, such as betamethasone dipropionate, clobetasol propionate, and diflorasone diacetate. Superpotent (Class 1), potent (Class 2), and upper mid-strength (Class 3) topical corticosteroids should be tested for their efficacy in ameliorating HD-induced cutaneous injury.</p><p>As previously discussed, depletion of GSH and accumulation of endogenous oxidants and ultimate formation of potent oxidizing species (eg, toxic lipid peroxides) may be contributory factors in HD-induced cytotoxicity.<xref ref-type="bibr" rid="B1">1</xref> Topically applied HD has been shown to negatively affect antioxidant enzymes in blood cells and body tissues of rats.<xref ref-type="bibr" rid="B116">116</xref> Several antioxidants have been shown to protect liver and lung from oxidative damage following inhalation or percutaneous exposure to HD in a mouse model.<xref ref-type="bibr" rid="B117">117</xref> It has been suggested that administration of antioxidants may be protective and useful.<xref ref-type="bibr" rid="B118">118</xref> Thus, initial antioxidant treatment aimed at affecting the progression of lesions that is instituted during the erythema phase may prove to be of benefit. The effectiveness and role of the interruption of the inflammatory cascade by the inclusion of topical and systemic antioxidant agents as well as a determination of the optimal timing for such therapy are important and intriguing avenues for investigation.</p><p>Placement of an occlusive or semiocclusive dressing will likely prove to be helpful in promoting autolytic debridement and preventing desiccation. Debridement will play a central role in improving the healing of cutaneous HD lesions, and beginning the process early may be beneficial. How soon following exposure these dressings can be applied remains to be determined. While maintaining a moist environment has long been known to facilitate wound healing,<xref ref-type="bibr" rid="B119">119</xref>–<xref ref-type="bibr" rid="B122">122</xref> caution needs to be observed since very early occlusion that builds up moisture levels in the skin will exacerbate the lesion. In addition, there is a period following exposure to sulfur mustard at which off-gassing of unbound HD occurs following a vapor<xref ref-type="bibr" rid="B104">104</xref> or liquid exposure in weanling pigs (J. S. Graham et al, unpublished data, 2001). The existence of a reservoir of unbound HD in human skin was previously discussed. These studies have suggested that off-gassing can continue for 24 to 36 hours, given a large exposure. Limiting the escape of this unbound HD by using an occlusive dressing may exacerbate the lesion. Placement of any occlusive dressing should probably be postponed for at least 24 hours following exposure. Keeping clothing off the exposed area, thereby not allowing vapors to build up, may also be of benefit.</p><p>The potential need for replacement fluid therapy (RFT) and caution in avoiding overhydration was previously discussed. Requirements for RFT can be studied experimentally. A scientific study using an appropriate animal model needs to be conducted to determine whether body weight and percent TBSA affected should be taken into consideration when trying to determine the total volume of fluid required. It is also important to determine which fluids are most appropriate, what infusion rates are needed, and when to commence RFT. Using instrumented pigs, large areas of injury can be induced, followed by close monitoring over time of important physiological parameters. The effects of different RFT protocols can then be determined. To generate large TBSA injuries, exposures with partial-body vapor chambers can be used. Alternatively, liquid HD can be diluted to volumes greater than that allowed under current surety regulations and fast-wicking material used to spread the liquid out over a large area. Previous research has enumerated a number of stable vehicles that are suitable for cutaneous exposure with liquid HD.<xref ref-type="bibr" rid="B108">108</xref></p><p>Management of casualties presenting with intact blisters varies according to the situation and level of care. Avoidance of infection is paramount so that the depth of injury does not increase. Small blisters should not be disrupted until deroofing can be done under controlled conditions. Larger blisters that are already flaccid may require deroofing or collapsing under sterile conditions. Current data does not permit recommendation of a single best approach. Once the necessary conditions and skills are reached, intervention should be more aggressive, with the dual goals of avoiding wound infection and optimizing recovery. As detailed below, aggressive management of the cutaneous HD injury, as opposed to that conducted on thermal or toxic epidermal necrolysis blisters, will require removal of the wound edges into normal-appearing skin along the periphery of the lesion and debridement of the base of the blister through the damaged BMZ into healthy dermis. Accurately determining depth of injury will influence this step. These patients will also require pain management and close observation for the systemic effects of HD exposure. Absence of thorough removal of damaged tissues will greatly slow healing and will enhance scarring and contracture in all but the most minor injuries.</p><p>For patients presenting with intact frank blisters, it may be beneficial to aspirate the blister fluid with a sterile needle and syringe and allow the roof of the blister to act as a sterile dressing until a physician can remove it. Reattachment of blister roofs has been noted to occur in the treatment of vitiligo via epidermal grafting using the tops of suction blisters<xref ref-type="bibr" rid="B123">123</xref>,<xref ref-type="bibr" rid="B124">124</xref> and in experimental suction blisters in humans following aspiration of blister fluid.<xref ref-type="bibr" rid="B125">125</xref> The roofs of HD blisters, however, are not expected to reattach to the blister floor owing to HD-induced damage to basal cells and BMZ components. Sloughing is expected to eventually occur. For patients presenting with ruptured HD-induced blisters, careful removal of the blister roof with scissors, application of an antibiotic ointment, and placement of a sterile dressing is warranted. For both of these scenarios, more complete debridement will be necessary for large lesions, as described later.</p></sec><sec id="S5-3"><title>Treatment of established partial-thickness lesions: overview of the approach</title><p>The remaining sections of this review describe the recommendations of the US-UK working group for the treatment of established partial-thickness cutaneous HD lesions. The general approach that is being taken is to perform adequate debridement of partial-thickness injuries, then treat the lesions like chronic cutaneous ulcers or partial-thickness thermal burns using contemporary medical approaches.</p><p>Before treatment can begin, the extent and severity of cutaneous HD injuries must be determined. Following assessment of injury and deroofing of frank blisters, adequate wound debridement needs to be performed, followed by 1 or more treatment adjuncts. Examples of adjuncts under consideration are dressings, growth factors, skin substitutes, topical nutritional support, and Vacuum Assisted Closure™ (V.A.C.®), Kinetic Concepts, Inc (KCI), San Antonio, Tex.</p></sec><sec id="S5-4"><title>Injury assessment</title><p>Before HD injuries can be appropriately treated, assessment of the injuries must be made. Total body surface area of the injuries should be established and depth of injury determined. Total body surface area can be determined using Wallace's Rule of Nines and the Lund and Browder chart for estimating burn severity, as previously discussed.<xref ref-type="bibr" rid="B77">77</xref>,<xref ref-type="bibr" rid="B78">78</xref> Determination of injury depth is a much more challenging task. Accurate depth assessment is important because it dictates how aggressive treatment needs to be to minimize or prevent cosmetic and functional deficits.</p><p>In thermal burns, depth of injury is typically assessed by physical examination. Surface appearance, the pinprick test to assess pain, the “blanch-capillary return test” to evaluate microcirculation, and surface temperature difference between burned and unburned skin are often used in diagnosis of depth.<xref ref-type="bibr" rid="B126">126</xref> Using these methods, diagnosing very superficial and very deep burns is relatively easy for the experienced burn surgeon. Burns of intermediate depth are often problematic in determining the need for grafting. Determining depth of HD injuries is more challenging. First, the full extent of cutaneous injury can take several days to manifest. Second, superficial appearances do not accurately predict depth of injury or need for grafting. While the presence of blisters in thermal burns is generally associated with superficial dermal injuries, blistering in HD injuries can occur in deep dermal/full-thickness injuries because of the unique nature of the agent and the unique progression of the injury.</p><p>Noninvasively examining cutaneous blood flow using available bioinstrumentation can greatly assist the physician in determining depth of injury. Laser Doppler perfusion imaging and indocyanine green fluorescence imaging may prove to be very valuable tools in prognosticating optimal wound healing of both thermal burns and cutaneous HD injuries.</p><p>Laser Doppler flowmetry and laser Doppler perfusion imaging (LDPI) have been used for prolonged, noninvasive monitoring of tissue viability and wound healing and for the assessment of peripheral vascular disease, inflammation, ischemia, reperfusion, skin graft acceptance (take), and burn depth.<xref ref-type="bibr" rid="B127">127</xref>–<xref ref-type="bibr" rid="B142">142</xref> Laser Doppler perfusion imaging may prove useful in delineating the areas of HD damage that need to be debrided, thereby avoiding areas with sufficient blood flow. Brown et al<xref ref-type="bibr" rid="B101">101</xref> found that laser Doppler perfusion images of vesicant vapor burns on the backs of swine correlated well with histopathological findings (thrombosis and necrosis of subepidermal capillaries) between 1 hour and 7 days postexposure and suggested that clinical management decision making for treatment of early vesicant burns could be aided by LDPI. Chilcott et al<xref ref-type="bibr" rid="B102">102</xref> used several noninvasive bioengineering methods to monitor wound healing in a large white pig model for 7 days following exposure to HD and Lewisite vapors. They concluded that while reflectance colorimetry and TEWL measurements could provide quantitative, noninvasive methods for determining efficacy of candidate treatment regimens, neither is comparable to the prognostic capabilities of LDPI. Graham et al<xref ref-type="bibr" rid="B83">83</xref> found LDPI to be useful in examining blood flow in grafted and ungrafted sites following treatment of deep dermal/full-thickness liquid HD injuries in a weanling swine model (Fig <xref ref-type="fig" rid="F1">1</xref>). Laser Doppler perfusion imaging is currently rather time consuming if there are multiple sites to be evaluated and/or large images to be collected at high resolution. The length of scanning procedures could be decreased by increasing scanning speed (thus decreasing flux resolution), decreasing the size of the scan area, and/or decreasing the number of lines scanned within the scanning area (scan resolution). Improvements in the technology that will speed up LDPI without compromising image resolution are being developed.</p><p>Indocyanine green fluorescence imaging has also shown promise in determining burn depth based on microcutaneous blood flow. It is a minimally invasive procedure that requires the placement of an intravenous line. Indocyanine green (ICG) is US Food and Drug Administration–approved for use in humans to determine cardiac output, hepatic function and blood flow, and for ophthalmic angiography. The fluorescence of intravenous ICG has been shown to estimate burn depth in small animals.<xref ref-type="bibr" rid="B143">143</xref> In contrast to fluorescein fluorescence,<xref ref-type="bibr" rid="B144">144</xref> ICG fluorescence is capable of distinguishing superficial and deep partial-thickness burns from full-thickness burns. The fluorescence intensity of ICG decreases exponentially with burn depth for burns of similar age.<xref ref-type="bibr" rid="B145">145</xref> Indocyanine green fluorescence was used to estimate burn depth in a porcine model.<xref ref-type="bibr" rid="B146">146</xref> An imaging system with a diagnostic algorithm was developed at the Wellman Laboratories of Photomedicine, Boston, Mass, that accurately diagnosed burns that healed within 21 days with minimal scarring from those that took longer to heal by secondary means. Measurements were made on burns created 2, 24, 48, and 72 hours prior to imaging. The algorithm was shown to be dependent on the age of the burn but not on location. This technology showed promise in plastic surgical applications<xref ref-type="bibr" rid="B147">147</xref>,<xref ref-type="bibr" rid="B148">148</xref> and for accurate determination of thermal burn depth in humans.<xref ref-type="bibr" rid="B147">147</xref>,<xref ref-type="bibr" rid="B149">149</xref> Indocyanine green fluorescence imaging also shows promise in diagnosing depth of HD injury (Fig <xref ref-type="fig" rid="F2">2</xref>; J. S. Graham et al, unpublished data, 1999). The advantage that this technology has over LDPI is the speed of image capture. Multiple images over large areas can be captured in a relatively short period of time. Images are typically collected 5 to 10 minutes after ICG injection to allow uptake and distribution. The dye is then excited (eg, 780 nm) and the resultant fluorescence emission (eg, 825 nm) immediately captured and saved by a computer and analyzed for burn-to-normal skin fluorescence ratio. Indocyanine green binds strongly to plasma globulins, limiting both extravasation within burn-injured vascular epithelia and extravascular transport to areas nearby.<xref ref-type="bibr" rid="B145">145</xref> Large signals are thought to be the result of vasodilation and hyperemia, and smaller signals are thought to be attributable to vascular occlusion and edema.<xref ref-type="bibr" rid="B143">143</xref>,<xref ref-type="bibr" rid="B145">145</xref> With this technique, live streaming video can also be captured immediately after injection, allowing the physician to watch the dye flowing through viable tissue and around nonviable tissue in real time.</p><p>It is of fundamental importance that any noninvasive or minimally invasive technique used to assess lesion severity be fully validated in order that results may be correctly interpreted. For example, a recent study has demonstrated that TEWL may not necessarily correlate with skin “barrier function.”<xref ref-type="bibr" rid="B150">150</xref></p></sec><sec id="S5-5"><title>Debridement</title><p>Experimental approaches to vesicant wound debridement have included powered dermabrasion,<xref ref-type="bibr" rid="B99">99</xref>,<xref ref-type="bibr" rid="B106">106</xref>,<xref ref-type="bibr" rid="B107">107</xref> sharp surgical excision,<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B83">83</xref>,<xref ref-type="bibr" rid="B151">151</xref> laser debridement,<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B83">83</xref>,<xref ref-type="bibr" rid="B100">100</xref>,<xref ref-type="bibr" rid="B107">107</xref>,<xref ref-type="bibr" rid="B109">109</xref>,<xref ref-type="bibr" rid="B152">152</xref> and enzymatic debridement.<xref ref-type="bibr" rid="B152">152</xref></p><p>Powered dermabrasion has been shown to speed up the reepithelialization process of cutaneous HD injuries.<xref ref-type="bibr" rid="B99">99</xref>,<xref ref-type="bibr" rid="B106">106</xref>,<xref ref-type="bibr" rid="B107">107</xref> Kjellstrom et al<xref ref-type="bibr" rid="B151">151</xref> found sharp surgical excision with primary suturing of the skin defect to be effective in decreasing healing time of HD vapor lesions in guinea pigs. Powered dermabrasion, pulsed CO<sub>2</sub> laser ablation, and erbium:yttrium-aluminum-garnet (Er:YAG) laser ablation have been shown to accelerate the rate of healing of full-thickness cutaneous Lewisite vapor burns in swine without the need for split-thickness skin grafting.<xref ref-type="bibr" rid="B107">107</xref>,<xref ref-type="bibr" rid="B109">109</xref> Eldad et al<xref ref-type="bibr" rid="B152">152</xref> found that excimer laser ablation and debrase (Debridase, MediWound Ltd, Yavne, Israel) enzymatic debridement were efficacious in improving the healing of partial-thickness nitrogen mustard burns in a guinea pig model.</p><p>Because of positive results achieved by laser and enzymatic debridement of vesicant injuries, our future research efforts will concentrate on the use of usthese methods to debride partial-thickness cutaneous HD injuries prior to the application of treatment adjuncts. The treatment regimen that is found to be most efficacious in the pig will ultimately be recommended for use in treating human casualties. Because humans form frank blisters, unlike the pig, these regimens would follow deroofing of any frank blisters present.</p></sec></sec><sec id="S6"><title>LASER DEBRIDEMENT OF CUTANEOUS VESICANT WOUNDS</title><p>Laser debridement of cutaneous vesicant wounds has proven to be an effective method of improving the rate of wound healing in pig models. Graham et al<xref ref-type="bibr" rid="B100">100</xref> showed that viability, thickness, and organization of the epidermis were all significantly improved by partial-thickness pulsed CO<sub>2</sub> laser debridement of small, mild to moderately severe cutaneous HD vapor injuries. Laser debridement followed by skin grafting was as efficacious in improving the wound healing of deep HD burns as sharp surgical tangential excision followed by grafting (the gold standard in human deep dermal/full-thickness thermal burns medicine).<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B83">83</xref> Middermal debridement by sharp excision or laser ablation without grafting produced less desirable results but was better than no treatment.<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B83">83</xref> A 4-fold improvement in reepithelialization of Lewisite injuries was achieved at 1 week following laser dermabrasion, with almost 100% reepithelialization by 3 weeks.<xref ref-type="bibr" rid="B109">109</xref> It is not apparent why these full-thickness Lewisite injuries (10 cm<sup>2</sup>) did not require grafting, as did HD injuries (12.6 cm<sup>2</sup>)<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B83">83</xref> or as would a full-thickness thermal burn. There are differences in biochemical action and rates of spontaneous reepithelialization between Lewisite and HD injuries.<xref ref-type="bibr" rid="B109">109</xref> Further studies need to be conducted to fully examine the comparative healing of deep Lewisite, HD, and thermal injuries.</p><p>Laser debridement offers additional benefits including hemostatic control during surgery, minimal risk of exposure to aerosolized pathogens, and time efficiency. Another major advantage to the use of lasers is the ability to control the amount of normal perilesional skin that is removed. Eldad et al<xref ref-type="bibr" rid="B152">152</xref> noted that it is technically difficult to control the amount of tissue to be removed by surgical tangential excision and that laser ablation of nitrogen mustard burns in a guinea pig model enabled both controlling the amount of tissue to be removed and minimizing blood loss. Minimizing the amount of tissue removed will be of a cosmetic benefit to the patient.</p><sec id="S6-1"><title>Types of lasers available</title><p>Pulsed CO<sub>2</sub> lasers have been used for a variety of dermatological applications, including skin resurfacing,<xref ref-type="bibr" rid="B153">153</xref>–<xref ref-type="bibr" rid="B155">155</xref> excision of burn eschar,<xref ref-type="bibr" rid="B156">156</xref>–<xref ref-type="bibr" rid="B159">159</xref> preparation of adequate graft beds,<xref ref-type="bibr" rid="B160">160</xref> and conservative ablation of skin lesions.<xref ref-type="bibr" rid="B161">161</xref> They are designed to promote rapid healing by minimizing laser-induced residual thermal damage,<xref ref-type="bibr" rid="B153">153</xref>,<xref ref-type="bibr" rid="B154">154</xref>,<xref ref-type="bibr" rid="B160">160</xref>,<xref ref-type="bibr" rid="B162">162</xref>–<xref ref-type="bibr" rid="B166">166</xref> and offer precise, micrometer-depth removal of tissue.<xref ref-type="bibr" rid="B166">166</xref> They vaporize tissue rapidly and efficiently, with minimal blood loss.<xref ref-type="bibr" rid="B159">159</xref>,<xref ref-type="bibr" rid="B166">166</xref>,<xref ref-type="bibr" rid="B167">167</xref> However, because of the low average power available from a pulsed laser system, these lasers may be inefficient while performing full-thickness debridement of deep burns or while debriding burns covering large body surface areas. Continuous wave (cw) lasers can be quite efficient in removing tissue but tend to create significant amounts of thermal damage, sometimes creating more damage than the initial burn being treated. Domankevitz and Nishioka<xref ref-type="bibr" rid="B167">167</xref> demonstrated that under appropriate conditions, a scanned cw CO<sub>2</sub> laser could ablate tissue with a zone of residual thermal injury less than 200 μm, making it useful for cutaneous surgery and debridement of burn wounds prior to skin grafting. Residual thermal damage of cw lasers can be minimized if the laser is scanned over the surface rapidly enough that the amount of time the laser spends on any 1 point mimics a short laser pulse.<xref ref-type="bibr" rid="B167">167</xref>,<xref ref-type="bibr" rid="B168">168</xref> Use of such lasers has proven efficacious in pigs<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B169">169</xref> and humans.<xref ref-type="bibr" rid="B170">170</xref> Glatter et al<xref ref-type="bibr" rid="B169">169</xref> usused a prototype cw CO<sub>2</sub> laser in a thermal burn model in pigs and found that long-term scarring, based on Vancouver scar assessments, was equivalent at 6 months postsurgery in both laser-ablated + grafted and sharply excised + grafted burns. In addition, they noted no significant difference in engraftment rates between the 2 methods of debridement. Graham et al<xref ref-type="bibr" rid="B45">45</xref> had similar success in engraftment rates using this same laser to treat HD injuries in pigs. In an initial clinical trial, Sheridan et al<xref ref-type="bibr" rid="B170">170</xref> used a similar, commercially available system to perform full-thickness ablations of thermal burns in children. They found that no bleeding occurred in laser-ablated sites, that engraftment rates for both laser-ablated sites and sharply excised sites were equally high, and that there were no significant difference in Vancouver scar scores at an average follow-up of 32.0 ± 5.2 weeks.</p><p>There are a number of lasers manufactured in the United States, Canada, and Europe that could be considered for routine debridement of vesicant injuries. Acland and Barlow<xref ref-type="bibr" rid="B171">171</xref> have provided a review on the current uses of lasers in dermatological practice and a list of the types of lasers used for specific procedures. They list CO<sub>2</sub> and Er:YAG lasers as being the most appropriate for cutaneous resurfacing. While rapid-scanning, high-powered cw CO<sub>2</sub> lasers would provide time-efficient ablation of damaged tissue,<xref ref-type="bibr" rid="B45">45</xref> they are no longer commercially available. Pulsed CO<sub>2</sub> lasers such as the UltraPulse (Lumenis Inc, Santa Clara, Calif) are in common use in dermatology and plastic surgery and have proved effective in improving wound healing of cutaneous vesicant injuries.<xref ref-type="bibr" rid="B100">100</xref>,<xref ref-type="bibr" rid="B109">109</xref> Er:YAG lasers are also commercially available and have been used for a wide variety of procedures, ranging from facial resurfacing to burn debridement.<xref ref-type="bibr" rid="B171">171</xref>–<xref ref-type="bibr" rid="B175">175</xref> They have been shown to be particularly useful in the debridement of partial-thickness burns<xref ref-type="bibr" rid="B175">175</xref> and in the management of deep Lewisite injuries.<xref ref-type="bibr" rid="B109">109</xref> Unlike the Gaussian beam profiles created by CO<sub>2</sub> lasers, Er:YAG laser beams tend to be uniform and produce uniform depths of ablation.<xref ref-type="bibr" rid="B175">175</xref> One commercially available unit, the Sciton PROFILE (Sciton Inc, Palo Alto, Calif), can be configured as a high-powered, dual-mode long-pulse Er:YAG laser that allows independent control of both depth of coagulation and depth of ablation. This versatility would be very advantageous to any clinic or hospital that treats burns and a variety of dermatological disorders. Er:YAG lasers will play a central role in future HD wound healing studies at our facilities.</p></sec></sec><sec id="S7"><title>ALTERNATIVE METHODS OF DEBRIDEMENT UNDER CONSIDERATION</title><p>There are alternatives to using a laser to debride vesicant injuries. Sharp surgical tangential excisions and powered dermabrasion have proved effective.<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B99">99</xref>,<xref ref-type="bibr" rid="B106">106</xref>,<xref ref-type="bibr" rid="B107">107</xref>,<xref ref-type="bibr" rid="B152">152</xref> Curettage, cryotherapy, larval therapy, and enzymatic debridement may be effective, lower-cost alternatives.</p><p>Sharp surgical tangential excisions have been effective in the treatment of sulfur mustard<xref ref-type="bibr" rid="B45">45</xref> and nitrogen mustard<xref ref-type="bibr" rid="B152">152</xref> injuries. Powered dermabrasion has been shown to speed up the reepithelialization process of cutaneous sulfur mustard<xref ref-type="bibr" rid="B99">99</xref>,<xref ref-type="bibr" rid="B106">106</xref> and Lewisite<xref ref-type="bibr" rid="B107">107</xref> injuries. There are drawbacks with this method, however, including lack of uniform depth control and risk of aerosolizing pathogens.</p><p>Scraping, using dermal curettes, may be a viable option for removing desiccated material from the surface of the skin. Cryotherapy using ice, liquid nitrogen, or Peltier coolers may also be efficacious if applied early. The therapeutic effects of cooling pig skin soon after exposure to HD vapor has recently been reported.<xref ref-type="bibr" rid="B176">176</xref> It may be possible to superficially freeze HD lesions before they proceed to vesication, thereby retarding the activity of the HD and progression of the lesion. The frozen tissue could then be removed using appropriate means, and the wound bed dressed with an antibiotic and sterile dressing until healed. Caution would have to be observed in the length of time the cooling agent is in direct contact with the skin and its potential to induce hypopigmentation taken into consideration. Finally, the use of larval therapy (maggots), while unconventional, has undergone a renaissance in the past few years and has proven to be very effective in debriding and improving the healing rate of hard-to-heal wounds (eg, chronic leg and foot ulcers).<xref ref-type="bibr" rid="B177">177</xref>–<xref ref-type="bibr" rid="B184">184</xref> The success of this approach may warrant study as a possible treatment of small TBSA sulfur mustard injuries.</p><p>A final alternative under consideration for debridement of HD injuries is enzymatic debridement. These enzymes are categorized as proteolytics, fibrinolytics, and collagenases, and are designed to dissolve necrotic tissue from wounds.<xref ref-type="bibr" rid="B185">185</xref> They are often used to debride chronic wounds (eg, decubitus ulcers, venous stasis ulcers, arterial insufficiency ulcers, diabetic foot ulcers). Many have been found to be safe and effective in removing devitalized tissue and accelerating healing in burns.<xref ref-type="bibr" rid="B186">186</xref>–<xref ref-type="bibr" rid="B193">193</xref> Any burn eschar present is typically cross-hatched to allow the agent to penetrate into the wound. Other agents, such as the bacterial proteolytic enzymes streptokinase and streptodornase, have given disappointing results in deep burns because they do not break down the collagen that separates vital from nonvital tissue.<xref ref-type="bibr" rid="B194">194</xref> Use of fibrinolysins may impair wound healing of HD lesions, as fibrin is an early matrix protein that is essential for wound healing. In addition, fibrinolysins are typically combined with deoxyribonuclease (DNase) and as such will also digest DNA in the dividing fibroblasts, which play a role in healing.<xref ref-type="bibr" rid="B185">185</xref> Some effective enzymes have produced better results than others, with enzyme concentration, skin moisture level, and the presence of certain antibacterial agents affecting results. Secondary dressings are needed to keep the wound moist and to allow these agents to work.<xref ref-type="bibr" rid="B185">185</xref> Klasen<xref ref-type="bibr" rid="B194">194</xref> offers an excellent review of the use of enzymatic debridement agents in burns. The most popular and effective agents on the market today are collagenases (eg, Collagenase Santyl ointment, Ross Products Division, Abbott Laboratories Inc, Columbus, Ohio) and papain/urea combinations (eg, Accuzyme and Panafil, Healthpoint Ltd, Fort Worth, Tex; and Gladase Papain-Urea Debriding Ointment, Smith & Nephew Inc, Largo, Fla). In addition, a promising proteolytic enzyme extracted from the stem of the pineapple plant is in Phase II clinical trials in the United States and Europe for the treatment of deep partial- and full-thickness burns (Debrase Gel Dressing, MediWound Ltd, Yavne, Israel). Enzymatic debridement of HD injuries is a promising and cheaper alternative to laser debridement, albeit more time consuming. Research is planned for determining which available enzymatic debridement product is most efficacious in debriding partial-thickness HD injuries. The specific application regimen, the time required to reach adequate debridement, and potential adverse effects (eg, conversion to a deeper injury, infection) need to be determined in an appropriate animal model.</p><p>Burn wound sepsis and bacteremias have been noted in burn patients undergoing enzymatic debridement.<xref ref-type="bibr" rid="B185">185</xref>,<xref ref-type="bibr" rid="B194">194</xref> Concomitant use of a topical antibiotic that does not interfere with the action of the enzyme under study may be warranted as a preventative measure.</p></sec><sec id="S8"><title>EXTENT OF DEBRIDEMENT REQUIRED</title><p>In addition to vesication and death of epidermal keratinocytes, HD exposure results in sublethal damage to keratinocytes along the periphery of the gross lesion. Damage to the basement membrane zone and underlying collagen in the papillary dermis has also been noted. Deroofing frank blisters followed by timely removal of this adjacent and subjacent damage will likely improve the rate of reepithelialization.</p><p>Nonlethal damage is clearly noted at the periphery of cutaneous HD lesions and has been reported previously.<xref ref-type="bibr" rid="B11">11</xref>,<xref ref-type="bibr" rid="B105">105</xref>,<xref ref-type="bibr" rid="B195">195</xref> Nikolsky's sign,<xref ref-type="bibr" rid="B196">196</xref> characterized by separation and loss of the epidermis from the dermis when the skin is pressed with a sliding or twisting motion, has been demonstrated in weanling pig skin following HD vapor exposure.<xref ref-type="bibr" rid="B11">11</xref>,<xref ref-type="bibr" rid="B195">195</xref> These weakened areas of the dermal-epidermal junction occurred along the periphery of the gross lesions and are indicative of sublethally damaged basal cells and/or altered proteins of extracellular matrices of the BMZ. Sublethally injured cells at the periphery of an HD lesion and in hair follicles and other adnexal structures may be partly responsible for the slow rate of reepithelialization seen in these injuries. Rice et al<xref ref-type="bibr" rid="B106">106</xref> suggested that the level of damage to cellular DNA at the margins of HD lesions may be sufficient to delay or prevent effective replication of those keratinocytes. Removal of these sublethally damaged keratinocytes at the margins of the lesions by debridement beyond the visible borders of the lesion will likely speed up the reepithelialization process.</p><p>As previously discussed, HD induces damage to the BMZ at the level of the lamina lucida.<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B39">39</xref> The floor of the blister retains portions of the damaged BMZ and needs to be removed to provide an adequate scaffold over which keratinocytes feeding the reepithelialization process can migrate. Thus, at minimum, debridement needs to proceed down into the papillary dermis after removal of the blister roof. Beyond the BMZ, dermal collagen itself is affected by HD exposure and can itself impede the wound healing process.<xref ref-type="bibr" rid="B97">97</xref>,<xref ref-type="bibr" rid="B106">106</xref>,<xref ref-type="bibr" rid="B197">197</xref> Brown and Rice<xref ref-type="bibr" rid="B197">197</xref> reported coagulation and hypereosinophilia of the papillary dermis in Yucatan minipig skin 12 to 24 hours following saturated HD vapor exposure, with the deeper reticular dermis unaffected. Rice et al<xref ref-type="bibr" rid="B106">106</xref> and Lindsay and Rice<xref ref-type="bibr" rid="B97">97</xref> suggested that following exposure to HD, papillary dermal collagen is altered and may no longer function normally as a healthy scaffold over which epidermal cells can migrate.</p><p>The question of how deep to debride needs to be addressed. Ablative lasers that create less than 160 ± 60 μm of residual thermal damage permit optimal skin graft take and healing.<xref ref-type="bibr" rid="B160">160</xref> Domankevitz and Nisioka<xref ref-type="bibr" rid="B167">167</xref> concluded that lasers that induce residual thermal damage zones of less than 200 μm are useful for cutaneous surgery and burn wound debridement prior to skin grafting. Lam et al<xref ref-type="bibr" rid="B109">109</xref> were able to improve wound healing of full-thickness cutaneous Lewisite injuries in pigs by partial-thickness laser debridement. Graham et al<xref ref-type="bibr" rid="B45">45</xref> were also able to improve wound healing of deep cutaneous HD injuries in pigs by partial-thickness debridement without grafting, albeit not to the extent attained by full-thickness debridement followed by grafting. These studies thus indicate that retaining some amount of damaged dermal tissue does not significantly impede wound healing. Complete debridement of partial-thickness injury, therefore, will likely not be required. Debridement of partial-thickness HD injury into the papillary dermis or upper reticular dermis will likely be adequate.</p><sec id="S8-1"><title>Dressings</title><p>Following wound debridement of HD injuries, an appropriate dressing will be needed to promote moist wound healing. Beneficial effects of such dressings include prevention of tissue dehydration and cell death, accelerated angiogenesis, increased breakdown of dead tissue and fibrin (eg, pericapillary fibrin cuffs), significant reduction in pain, and potentiation of the interaction of growth factors with their target cells.<xref ref-type="bibr" rid="B122">122</xref> Helfman et al<xref ref-type="bibr" rid="B119">119</xref> and Singhal et al<xref ref-type="bibr" rid="B185">185</xref> have provided overviews of the various types of occlusive and semiocclusive dressings. Hydrocolloids, hydrogels, foam dressings, alginates, and transparent film dressings are commercially available from a large number of manufacturers. As foam dressings and alginates are designed to control moderate to heavy exudates, they will likely not be needed for covering debrided cutaneous HD injuries. Silver impregnated dressing materials may be of great potential benefit in treating these wounds owing to their antimicrobial efficacy<xref ref-type="bibr" rid="B198">198</xref>–<xref ref-type="bibr" rid="B200">200</xref> and demonstrated ability to enhance rates of reepithelialization.<xref ref-type="bibr" rid="B201">201</xref>,<xref ref-type="bibr" rid="B202">202</xref> A number of these dressing materials are currently employed in burn and chronic wound care, while other more advanced silver dressings are in various stages of development. Thus, research on finding the most appropriate dressing during wound healing of these lesions should concentrate on hydrocolloids, hydrogels, thin films, and silver-impregnated dressings. Testing these products in a pig model should be adequate, since similar responses to occlusive and semiocclusive dressings on wound healing have been noted in pigs and humans.<xref ref-type="bibr" rid="B203">203</xref></p><p>Weak attachment of the neoepidermis to the underlying dermis has been noted in human HD casualties<xref ref-type="bibr" rid="B44">44</xref> and experimentally exposed weanling pigs (J. S. Graham et al, unpublished data, 2004). Once the lesions have fully reepithelialized, protective dressings may be needed to avoid or minimize damage as a result of friction with clothing or bedding.</p></sec><sec id="S8-2"><title>Growth factors</title><p>During cutaneous wound healing, growth factors play dominant roles in regulating cell proliferation, differentiation, and synthesis of extracellular matrix.<xref ref-type="bibr" rid="B91">91</xref> Epidermal growth factor (EGF), transforming growth factor-β (TGF-β), platelet-derived growth factor (PDGF), insulin-like growth factor (IGF), keratinocyte growth factor (KGF), hepatocyte growth factor (HGF), granulocyte-macrophage colony-stimulating factor (GM-CSF), and fibroblast growth factors (FGFs) play important and critical roles in the healing of cutaneous wounds.<xref ref-type="bibr" rid="B203">203</xref>–<xref ref-type="bibr" rid="B241">241</xref> Reviews of the effects of these growth factors on wound healing have been previously published.<xref ref-type="bibr" rid="B203">203</xref>–<xref ref-type="bibr" rid="B208">208</xref> Improved wound healing has been reported for topical applications of EGF,<xref ref-type="bibr" rid="B203">203</xref>–<xref ref-type="bibr" rid="B220">220</xref>,<xref ref-type="bibr" rid="B222">222</xref>,<xref ref-type="bibr" rid="B223">223</xref> PDGF,<xref ref-type="bibr" rid="B203">203</xref>–<xref ref-type="bibr" rid="B208">208</xref>,<xref ref-type="bibr" rid="B214">214</xref>,<xref ref-type="bibr" rid="B219">219</xref>,<xref ref-type="bibr" rid="B224">224</xref>–<xref ref-type="bibr" rid="B229">229</xref> KGF,<xref ref-type="bibr" rid="B206">206</xref>,<xref ref-type="bibr" rid="B207">207</xref>,<xref ref-type="bibr" rid="B231">231</xref>–<xref ref-type="bibr" rid="B233">233</xref> and IGF-I.<xref ref-type="bibr" rid="B235">235</xref>,<xref ref-type="bibr" rid="B236">236</xref> There have been some negative reports on the effectiveness of EGF<xref ref-type="bibr" rid="B221">221</xref> and KGF<xref ref-type="bibr" rid="B230">230</xref> in improving wound healing of experimental split-thickness skin wounds in humans and esophagogastric anastomotic wounds in rats, respectively. Epidermal growth factor has been shown to improve the healing of graft donor sites,<xref ref-type="bibr" rid="B210">210</xref> corneal burns,<xref ref-type="bibr" rid="B217">217</xref>,<xref ref-type="bibr" rid="B220">220</xref> and cutaneous burns,<xref ref-type="bibr" rid="B218">218</xref>,<xref ref-type="bibr" rid="B219">219</xref> whereas PDGF and KGF have been shown to improve the healing of burns<xref ref-type="bibr" rid="B219">219</xref> and skin-grafted lesions.<xref ref-type="bibr" rid="B224">224</xref>,<xref ref-type="bibr" rid="B233">233</xref></p><p>Human leptin is a 146–amino acid residue, nonglycosylated polypeptide involved in body weight regulation. It is released from white adipose tissue and exerts its effect via receptors in the hypothalamus. While not a growth factor per se but rather characterized as a satiety-regulating cytokine, leptin has been shown to be a potent mitogenic stimulus to keratinocytes during skin repair.<xref ref-type="bibr" rid="B237">237</xref>,<xref ref-type="bibr" rid="B238">238</xref></p><p>Results from these reports suggest that topical application of healing-enhancing factors alone or in combination may be beneficial in improving the wound healing of cutaneous HD injuries following wound debridement. When to commence such treatment following agent exposure needs careful consideration and experimental testing. During the early phases of wound healing, chemokines and cytokines regulate the chemotaxis and activation of inflammatory cells, along with synthesis of proteases and protease inhibitors.<xref ref-type="bibr" rid="B91">91</xref> Early application of growth factors would be ineffective in a milieu of proteases. Application of such growth factors or combinations thereof will likely require a delay of 3 to 7 days following HD exposure until the inflammatory response has subsided. Concomitant use of protease inhibitors or a dressing that binds matrix metalloproteases and protects growth factors (eg, PROMOGRAN* Matrix Wound Dressing, Johnson & Johnson Wound Management Worldwide, Somerville, NJ) may be necessary. A number of these growth factors are commercially available for efficacy testing in animal models (EGF, PDGF, leptin). Of these 3, only PDGF has been approved for human use by the US Food and Drug Administration. Recombinant human EGF (β-urogastrone) is available from Roche Diagnostics Corp, Indianapolis, Ind. Regranex Gel, a recombinant human PDGF-BB, is available from Johnson & Johnson Wound Management Worldwide, Somerville, NJ. Recombinant human leptin is available from R&D Systems Inc, Minneapolis, Minn. Amgen Inc (Thousand Oaks, Calif) has recently completed Phase III clinical trials of a recombinant human KGF (Palifermin) that significantly reduces the duration and incidence of radiation- and chemotherapy-induced oral mucositis.</p><p>In addition to these growth factors, topical application of extracellular matrix components such as fibronectin, which supports fibroblast, keratinocyte, and endothelial cell adhesion and movement,<xref ref-type="bibr" rid="B242">242</xref>,<xref ref-type="bibr" rid="B243">243</xref> may be of benefit following wound debridement. Fibronectin can serve as a template for collagen deposition<xref ref-type="bibr" rid="B242">242</xref> and is a key component of the provisional matrix during wound repair.<xref ref-type="bibr" rid="B243">243</xref> Exogenous application of intact fibronectin has been shown to be beneficial in helping to close human skin and corneal wounds.<xref ref-type="bibr" rid="B244">244</xref> The ability of keratinocytes to spread on fibronectin may require the presence of serum or epibolin/vitronectin.<xref ref-type="bibr" rid="B244">244</xref> Addition of another adhesive glycoprotein, laminin, may actually be detrimental to the reepithelialization process. When human keratinocytes are placed in apposition with collagen, they attach and begin migrating.<xref ref-type="bibr" rid="B244">244</xref> This migration is inhibited by the addition of laminin, which acts as a major cell adhesion factor for keratinocytes. Laminin 5 has been shown to inhibit human keratinocyte migration and strongly promotes keratinocyte attachment.<xref ref-type="bibr" rid="B244">244</xref> It is believed to anchor the keratinocytes to the substratum, via α<sub>3</sub>β<sub>1</sub>-integrin receptors.<xref ref-type="bibr" rid="B244">244</xref></p></sec><sec id="S8-3"><title>Skin substitutes</title><p>Skin substitutes may provide an excellent temporary wound dressing for debrided HD injuries. Permanent wound closure can only be achieved by spontaneous reepithelialization or by the provision of autologous skin by means of skin grafting. The use of skin substitutes to temporarily restore the multiple functions of normal skin may be of substantial benefit in the management of cutaneous HD injuries.</p><p>For a skin substitute application to be successful, the same conditions required for successful autograft “take” must be created and maintained. The selection of the most suitable and effective temporary skin substitute will require a critical assessment of its comparative attributes when applied to HD wounds as well as the issues of cost, ease of use, availability, and consistency of results. Skin substitutes are widely used in human thermal burns management and can be (1) temporary or permanent; (2) epidermal, dermal, or composite; and (3) biologic or synthetic.<xref ref-type="bibr" rid="B245">245</xref>–<xref ref-type="bibr" rid="B253">253</xref> They have also been shown to be effective in speeding up time to closure of chronic leg and foot ulcers,<xref ref-type="bibr" rid="B254">254</xref>–<xref ref-type="bibr" rid="B269">269</xref> surgical excision sites,<xref ref-type="bibr" rid="B269">269</xref> and partial-thickness donor sites.<xref ref-type="bibr" rid="B269">269</xref> They may be a source of growth factors and are generally semiocclusive in nature. They can provide barrier function; add tensile strength to the wound; are generally flexible and pliable; markedly reduce pain, inflammation, and drainage; and provide a moist wound healing environment. They do not control deep bacterial infections; can seal bacteria in; and, being a biologic, they can transmit infection. Hence, the wound surface must not be infected for application of a skin substitute. A number of skin substitutes are available on the market and should be tested for their efficacy in improving wound healing of cutaneous HD injuries. Marketed products currently under consideration include (1) living bilayered skin substitutes (APLIGRAF [Organogenesis Inc, Canton, Mass], designed for the treatment of venous leg ulcers and diabetic foot ulcers; and OrCel [Ortec International Inc, New York], designed for the management of split-thickness donor site wounds and for the treatment of epidermolysis bullosa), (2) bilayered composites consisting of a synthetic epidermal analog and a biologic (collagen-based) dermal analog (TransCyte [Smith & Nephew Inc, Hull, United Kingdom] and Biobrane [Bertek Pharmaceuticals Inc, Morgantown, Wva], both designed for partial-thickness wounds), (3) complex weaves of biopolymers that produce a thin protective membrane (Silon-TSR [Bio Med Sciences Inc, Allentown, Pa], designed for use on partial-thickness burns, donor sites, and laser-resurfaced skin), and (4) acellular dermal matrices designed to aid in the natural healing of partial-thickness injuries of limited depth (SkinTemp, BioCore Medical Technologies Inc, Silver Spring, Md). Permanent skin substitutes that are designed for treating deep injuries and require application of a thin epithelial autograft will likely be inappropriate for use in treating partial-thickness HD injuries (eg, AlloDerm [LifeCell Corporation, Blanchburg, NJ] and INTEGRA dermal regeneration template [Integra LifeSciences Holdings Corporation, Plainsboro, NJ]).</p><p>Cryopreserved and glycerol-preserved cadaver skin has been used as a temporary dressing in the treatment of burns for a number of years.<xref ref-type="bibr" rid="B270">270</xref>–<xref ref-type="bibr" rid="B277">277</xref> Similarly, xenografts from a variety of animal species, especially pig, have been used as temporary cover to treat burns.<xref ref-type="bibr" rid="B278">278</xref>–<xref ref-type="bibr" rid="B280">280</xref> While pig skin is antigenic and would ultimately get sloughed, portions of the dermis may become incorporated in a healed wound and elicit an unwanted granular response.<xref ref-type="bibr" rid="B278">278</xref></p><p>Cultured epithelial allografts and autografts have been used for about 2 decades as a treatment for chronic ulcers and thermal burns.<xref ref-type="bibr" rid="B248">248</xref>,<xref ref-type="bibr" rid="B281">281</xref>–<xref ref-type="bibr" rid="B312">312</xref> Keratinocytes can be harvested from skin biopsies and grown to confluence by the method originally described by Rheinwald and Green.<xref ref-type="bibr" rid="B313">313</xref> Large amounts of stratifying epidermis can thus be grown in the laboratory in short periods of time and used to restore defects in the epidermis.<xref ref-type="bibr" rid="B314">314</xref> Such grafts can be used immediately or cryopreserved and used at a later date. In addition to their usefulness in improving the healing of deep ulcers and burns, they have shown efficacy in improving the rate of reepithelialization of partial-thickness burns<xref ref-type="bibr" rid="B286">286</xref>,<xref ref-type="bibr" rid="B290">290</xref>,<xref ref-type="bibr" rid="B300">300</xref> and split-thickness skin graft donor sites.<xref ref-type="bibr" rid="B289">289</xref>,<xref ref-type="bibr" rid="B291">291</xref>,<xref ref-type="bibr" rid="B307">307</xref> There is no evidence that cultured allografts survive permanently on the wound bed.<xref ref-type="bibr" rid="B289">289</xref> Kaawach et al<xref ref-type="bibr" rid="B286">286</xref> showed that allografted cells were not present between 8 and 100 days postgrafting and suggested that the newly formed epithelium was of host origin. Cultured keratinocyte allografts speed healing by providing cover and producing growth factors and extracellular matrix proteins.<xref ref-type="bibr" rid="B291">291</xref> Because these coverings can be produced in large quantities and would thus be more readily available than cadaver skin, their application in the treatment of debrided partial-thickness HD injuries should be considered. Cultured epidermal autografts (CEAs) would be safer to use, from the perspective of disease transmission, and would not require donor-screening procedures. They do, however, require small punch biopsies to be collected from the patient and a lag time of about 2 weeks to grow the graft material. Several laboratories in the United States perform this service for their local burn centers (eg, Living Skin Bank, University Hospital, SUNY, Stony Brook, NY). Genzyme Corporation (Cambridge, Mass) has shown that CEA (Epicel) can be commercially produced. Despite their theoretical usefulness, CEAs are rather limited in their clinical effectiveness because they are unable to withstand even very low levels of bacterial wound contamination and do not provide a durable epithelial surface. Wounds covered by this modality are unstable and are subject to frequent epithelial disruption as a result of minor mechanical trauma. Durability has been increased by placing the CEA on a scaffolding of widely meshed autograft.<xref ref-type="bibr" rid="B315">315</xref> Alternatively, CEA placed over deepithelialized allograft (ie, engrafted allodermis) has also proved successful.<xref ref-type="bibr" rid="B284">284</xref></p><p>Finally, application of keratinocytes in suspension has shown to improve epidermal wound healing in pig<xref ref-type="bibr" rid="B316">316</xref>,<xref ref-type="bibr" rid="B317">317</xref> and mouse<xref ref-type="bibr" rid="B318">318</xref>,<xref ref-type="bibr" rid="B319">319</xref> models. Reconstitution of the dermal-epidermal junction was significantly enhanced in an athymic mouse model by suspending the cells in a fibrin-glue matrix.<xref ref-type="bibr" rid="B319">319</xref> Using a pig model, Currie et al<xref ref-type="bibr" rid="B320">320</xref> recently compared the effects of keratinocyte cell sprays with and without fibrin glue. No differences in mean epithelial area or quality of epithelium were noted at 3 weeks. Keratinocyte suspension technology shows promise in that it does not require the length of time necessary to produce cultured epidermal sheets. Use of this technology has proven efficacious in the treatment of thermal burns in humans.<xref ref-type="bibr" rid="B321">321</xref> A small biopsy is collected and the cells cultured and expanded in a clinical laboratory, then placed into a syringe-like spraying mechanism and sprayed onto the wound 2 to 5 days following biopsy. This technology is currently available (CellSpray and CellSpray XP, Clinical Cell Culture, Bentley, Western Australia). These products are designed for use in partial, deep partial, and full-thickness burns, donor sites, scar treatment, chronic ulcers, pigment loss, and cosmetic skin rejuvenation following laser resurfacing, dermabrasion, or chemical peels. A similar spray-on product in development that delivers allogeneic keratinocytes, fibroblasts, and fibrin to wounds has recently shown positive results in Phase II trials in Europe and the United States (Allox, IsoTis OrthoBiologics, Irvine, Calif). An innovative medical device (ReCell, Clinical Cell Culture, Bentley, Western Australia) has been developed that will allow rapid harvesting of cells from a thin split-thickness biopsy followed by spray application onto wounds within 30 minutes of collecting the biopsy, without the need of culturing the keratinocytes in a clinical laboratory. This single-use device is designed for injuries up to 2% TBSA. It may prove beneficial in the treatment of small TBSA HD injuries and is worthy of laboratory investigation.</p></sec><sec id="S8-4"><title>Topical nutritional support</title><p>Boyce et al<xref ref-type="bibr" rid="B322">322</xref> noted that application of topical nutrients supports keratinocyte viability during graft vascularization of cultured skin substitutes and inhibits wound contraction. There are a large number of “cosmeceutical” products on the market designed to enhance the appearance, feel, flexibility, and function of skin by supplying moisturizing and nutritive substances. Amino-Plex Spray (biO<sub>2</sub> Cosmeceuticals International Inc, Beverly Hills, Calif) is such a product that is designed to increase oxygen in cells, stimulate ATP synthesis, improve glucose transportation, stimulate collagen formation, and promote angiogenesis. It is a mixture of over 100 low-molecular-weight ingredients, including amino acids, trace minerals, nucleotides, nucleosides, oligopeptides, electrolytes, glycosaminoglycans, and glycolipids. According to the company, this product has been shown to reduce irritation and improve results in laser resurfacing, chemical peels, microdermabrasion, hair transplantation, and hair removal. A new product (Oxy-Mist, biO<sub>2</sub> Cosmeceuticals International Inc, Beverly Hills, Calif) combines the ingredients of Amino-Plex Spray with micellized vitamin E and sterile mineralized water and uses medical-grade oxygen as the delivery source. The manufacturer reports that used after facial resurfacing with a pulsed CO<sub>2</sub> or erbium laser, Oxy-Mist has been clinically shown to accelerate reepithelialization, minimize pain, and decrease the period of postlaser erythema. The oxygen itself likely contributed to the improved healing noted, as both hyperbaric<xref ref-type="bibr" rid="B323">323</xref> and topical<xref ref-type="bibr" rid="B324">324</xref> oxygen therapies have been shown to facilitate wound healing. Because of its reported benefits in dermatology following laser resurfacing procedures, it would be advantageous to determine whether these topical nutritional products will improve the healing of debrided HD injuries.</p></sec><sec id="S8-5"><title>Vacuum assisted closure</title><p>Application of topical negative pressure in the management of chronic wounds and burns has gained popularity in the last 5 years.<xref ref-type="bibr" rid="B325">325</xref>–<xref ref-type="bibr" rid="B332">332</xref> Also known as Vacuum Assisted Closure (V.A.C.), the procedure involves placing an open-cell foam into the wound bed (cut to conform to the shape of the wound), sealing it with an adhesive drape and applying subatmospheric pressure (125 mm Hg below ambient) that is transmitted via an evacuation tube by a computerized vacuum pump.<xref ref-type="bibr" rid="B326">326</xref>,<xref ref-type="bibr" rid="B327">327</xref> The procedure is becoming widely used for the closure of chronic wounds such as stage III and IV pressure ulcers; venous, arterial, and neuropathic ulcers; and subacute and acute wounds such as dehisced incisions, split-thickness meshed skin grafts, and muscle flaps.<xref ref-type="bibr" rid="B326">326</xref>,<xref ref-type="bibr" rid="B327">327</xref> V.A.C. is also gaining popularity in the management of complex orthopedic wounds.<xref ref-type="bibr" rid="B329">329</xref>,<xref ref-type="bibr" rid="B330">330</xref> This methodology increases local blood perfusion and nutrient delivery to the wound, accelerates the rate of granulation tissue formation, and decreases wound tissue bacterial levels.<xref ref-type="bibr" rid="B326">326</xref>,<xref ref-type="bibr" rid="B327">327</xref> Per the manufacturer's recommendations, wounds must be debrided of all necrotic tissue prior to application of V.A.C., and it is contraindicated with the presence of nonenteric and unexplored fistulas, osteomyelitis (untreated), exposed organs or blood vessels, or malignancy in or around the wound. The dressings are typically changed every 1 to 4 days until wound closure. V.A.C. has been shown to be effective in preventing progression of partial-thickness burns to a deeper injury in a swine model,<xref ref-type="bibr" rid="B333">333</xref> likely the result of helping to deliver oxygen and nutrients to the zone of stasis. The method has also been shown to increase the rate of skin graft donor site reepithelialization in pigs and humans<xref ref-type="bibr" rid="B331">331</xref> and is a safe and effective method for securing split-thickness skin grafts, providing improved graft survival.<xref ref-type="bibr" rid="B332">332</xref> Following debridement of partial-thickness HD injuries, V.A.C. may prove efficacious in significantly speeding the reepithelialization process in these lesions. Recently, the US Food and Drug Administration approved the use of V.A.C. in treating partial-thickness burns. The expedited closure of HD wounds by means of a mechanical force is an area that merits further consideration and investigation. Several V.A.C. therapy systems are available from Kinetic Concepts Inc, San Antonio, Tex. One lightweight portable system is available for ambulatory care.</p></sec><sec id="S8-6"><title>Use of noninvasive bioengineering methods to assess treatment efficacy</title><p>During efficacy testing of candidate treatment regimens, it is important to examine a number of parameters besides reepithelialization. While coverage of the wound by a new epithelium is important, there are a number of other skin characteristics that are important from a functional and cosmetic point of view. Surface contour and general appearance, epidermal hydration, epidermal barrier function, pH, mechanical properties, cutaneous blood flow, transcutaneous oxygen tension, neural supply/sensory function, and hair growth are all important characteristics that bear examination. While routine histopathology, immunohistochemistry, and electron microscopy are all valued tools in determining morphology and understanding the pathophysiology of HD wound development and healing, they do not directly measure physiological parameters or function. For these, a variety of noninvasive bioengineering methods are available. In support of HD wound healing research, laboratories in the United States and United Kingdom have used reflectance colorimetry to evaluate erythema, skin hue, chroma, and lightness<xref ref-type="bibr" rid="B11">11</xref>,<xref ref-type="bibr" rid="B27">27</xref>,<xref ref-type="bibr" rid="B83">83</xref>,<xref ref-type="bibr" rid="B100">100</xref>,<xref ref-type="bibr" rid="B102">102</xref>; LDPI to examine cutaneous blood flow, depth of injury, neovascularization, and skin graft viability<xref ref-type="bibr" rid="B83">83</xref>,<xref ref-type="bibr" rid="B101">101</xref>; torsional ballistometry to evaluate the mechanical properties of skin firmness and elasticity<xref ref-type="bibr" rid="B83">83</xref>; evaporimetry to examine transepidermal water loss as a way to evaluate skin hydration/barrier function<xref ref-type="bibr" rid="B83">83</xref>,<xref ref-type="bibr" rid="B102">102</xref>; 2-dimensional and 3-dimensional high-frequency (20-MHz) ultrasound to examine edema formation<xref ref-type="bibr" rid="B11">11</xref> and scar tissue thickness (J. S. Graham et al, unpublished data, 2002); and image analysis to evaluate wound size, shape morphometry, and wound contraction.<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B100">100</xref> While not yet used in HD wound healing research efforts, instrumentation is also available for evaluating surface contour, pH, and sensory function.</p></sec></sec><sec id="S9"><title>SUMMARY</title><p>The toxicity of sulfur mustard has been widely described.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B3">3</xref>–<xref ref-type="bibr" rid="B16">16</xref> Cutaneous HD injuries can take several months to heal, may necessitate lengthy hospitalizations, and can result in significant cosmetic and/or functional deficits. There are currently no standardized or optimized methods of casualty management that prevent or minimize deficits and provide for speedy wound healing.</p><p>Research laboratories in the United States, United Kingdom, and Canada have developed concepts for medical countermeasures to vesicant agents. The initial step in protecting a person from the deleterious effects of HD is to eliminate contact with the agent. Protective gear and topical skin protectants have been designed for this purpose. Should the agent come in contact with the skin, it needs to be removed within 2 minutes to fully prevent damage.<xref ref-type="bibr" rid="B15">15</xref> Decontamination is generally performed by physical removal. Sulfur mustard is not painful on contact and the exposed person may not be aware of the exposure until symptoms begin to appear after a latent period. Pharmacological approaches are being studied for their efficacy in minimizing or preventing damage. At this time, it seems clear that the earliest possible application of anti-inflammatory agents, including cold packs, topical and systemic steroids, and nonsteroidal anti-inflammatory drugs, is beneficial. Should HD come in contact with the skin, were decontamination not performed timely, and should pharmacological intervention be absent or ineffective, a chemical casualty will be produced that requires medical attention. Casualty management now comes into play. There are no antidotes to HD. Therapy therefore rests on management of symptoms and consequences of exposure with the intent to reduce long-term morbidity. Historically, blister aspiration and/or deroofing (epidermal removal), physical debridement, irrigation, topical antibiotics, and sterile dressings have been the main courses of action in the medical management of cutaneous HD injuries. New strategies to relieve symptoms, prevent infections, and promote healing have been formulated. Deep cutaneous HD injuries will require aggressive surgical intervention, including skin grafting, if cosmetic and functional deficits are to be avoided. Our future research efforts will concentrate on partial-thickness injury that will not require such aggressive approaches.</p><p>Assessment of the injuries must occur early in the process. Total body surface area of the injuries should be established and depth of injury determined. Laser Doppler perfusion imaging and ICG fluorescence imaging show promise in prognosticating optimal wound healing of HD injury on the basis of examination of microcutaneous blood flow. Following assessment of HD injury, adequate wound debridement needs to be performed. At the minimum, debridement needs to proceed into normal-appearing skin along the periphery of the lesion and down through the base of the blister (eg, damaged BMZ) into the papillary dermis. Debridement is then followed by 1 or more treatment adjuncts. Such adjuncts under consideration are dressings, growth factors, skin substitutes, topical nutritional support, and Vacuum Assisted Closure.</p><p>The ultimate goal is to determine the most efficacious treatment regimen to be applied in the clinical management of HD casualties. The ideal regimen should return damaged skin to optimal appearance and normal function in the shortest time. Improved treatment will result in a better cosmetic and functional outcome for the patient, and will enable the casualty to return to normal activities sooner.</p></sec>
Collagenase Promotes the Cellular Responses to Injury and Wound Healing In Vivo	<p><bold>Objective:</bold> This study focuses on the growth-promoting and migration-enhancing role that Clostridial collagenase plays in vitro and in vivo. <bold>Methods:</bold> For in vitro studies, biosynthesized extracellular matrices were treated with purified Clostridial collagenase, nonspecific proteases, or buffer controls. Keratinocytes were subsequently plated upon these matrices in the presence or absence of Clostridial collagenase and/or heparin-binding epidermal-like growth factor, and cell proliferation and migration were quantified. To examine the effects of Clostridial collagenase in vivo, we performed a double-blind study of full-thickness wounds on the backs of Yucatan Micropigs, testing the effects of purified Clostridial collagenase, Regranex (PDGF-BB), and Solosite (carboxymethyl cellulose) on wound healing. <bold>Results:</bold> <italic>In vitro studies:</italic> Matrix pretreatment with Clostridial collagenase stimulates a 2-fold increase in proliferation and postinjury migration; when Clostridial collagenase and/or heparin-binding epidermal-like growth factor are added to the growth media, there is an additional doubling of growth and migration, yielding approximately 5-fold enhancement of keratinocyte proliferation and migration. Papain-urea treatment under similar conditions results in a 50% decrease in cell number over a 1-week time course. <italic>In vivo studies:</italic> By all parameters measured, including granulation tissue formation, inflammation, re-epithelization, and time to wound closure, purified Clostridial collagenase was superior (analysis of variance, <italic>P</italic> > .05) to other treatments tested. <bold>Conclusion:</bold> On the basis of these findings, we concluded that Clostridial collagenase stimulates keratinocyte cellular responses to injury in vitro and may represent a novel therapeutic approach for promotion of wound healing in vivo.</p>	<contrib contrib-type="author"><name><surname>Riley</surname><given-names>Kathleen N.</given-names></name><degrees>BS</degrees><xref rid="aff1" ref-type="aff"/></contrib><contrib contrib-type="author"><name><surname>Herman</surname><given-names>Ira M.</given-names></name><degrees>PhD</degrees><xref rid="aff1" ref-type="aff"/><xref ref-type="corresp" rid="cor1"/></contrib>	Journal of Burns and Wounds	<p>The keratinocyte basement membrane serves as a scaffold and a macromolecular signaling matrix responsible for regulating cell behavior during development, adult life, and wound healing. Synthesized and organized by keratinocytes, matrix components play a pivotal role in orchestrating epithelial proliferation,<xref ref-type="bibr" rid="B1">1</xref> adhesion,<xref ref-type="bibr" rid="B2">2</xref> and migration,<xref ref-type="bibr" rid="B3">3</xref> including the cellular responses to injury. Glycoproteins and proteoglycans not only provide a physical scaffold for epithelial cells but also can bind and sequester signaling molecules, including members of the epidermal growth factor (EGF),<xref ref-type="bibr" rid="B4">4</xref> fibroblast and keratinocyte growth factor,<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B6">6</xref> and transforming growth factor families.<xref ref-type="bibr" rid="B7">7</xref> Keratinocytes transduce signals from these matrix-bound or soluble growth factors through their high-affinity tyrosine and serine/threonine kinase growth factor receptors, thereby completing this biochemical signaling cascade.</p><p>Keratinocyte responsiveness to the extracellular matrix involves controlling the expression, activation, and localization of the receptors that transduce matrix-associated signals. For example, the activation and relocation of integrin receptors in response to injury is crucial to the initiation of wound healing. In the wound area, keratinocytes come in contact with the newly exposed dermal matrix, which contains Type I collagen and laminin 5 not found in the intact keratinocyte basement membrane, typically composed of collagen IV, laminin 1, fibronectin, and other molecules.<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B9">9</xref> The newly exposed dermal matrix components bind and activate the α2β1 integrin receptor, causing its relocation to the site of injury, thus allowing keratinocytes to adhere to the exposed dermal matrix and begin migration into the wound area.<xref ref-type="bibr" rid="B10">10</xref> Expression of αVβ5 integrin is also induced at the wound edge, and likewise promotes migration.<xref ref-type="bibr" rid="B11">11</xref> Interestingly, knockout mice lacking the β1 integrin subunit show abnormal keratinocyte morphology and are highly deficient in wound healing because of defects in adhesion, migration, and proliferation.<xref ref-type="bibr" rid="B12">12</xref></p><p>In addition to the modifications seen in the integrin receptor repertoire during wound healing, injured keratinocytes release growth factors and matrix metalloproteinases (MMPs) to stimulate their integrin-mediated migration.<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref> Growth factors, such as heparin-binding epidermal growth factor (HbEGF), are produced by keratinocytes in a membrane-tethered form. In response to injury, HbEGF is cleaved and released by extracellular MMP-3, or Stromelysin I.<xref ref-type="bibr" rid="B15">15</xref> Keratinocytes also produce MMP-1, MMP-2 and MMP-9, all of which help to remodel the matrix, allowing cells to migrate into and close the wound.<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B16">16</xref>–<xref ref-type="bibr" rid="B18">18</xref> MMP production is stimulated by the presence of target substrates and growth factors,.<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B16">16</xref>,<xref ref-type="bibr" rid="B19">19</xref>–<xref ref-type="bibr" rid="B21">21</xref> and MMP-1 can also be bound by integrin α2β1 for targeted collagen degradation.<xref ref-type="bibr" rid="B22">22</xref> When healing is completed and the wound is closed, the remodeled matrix signals through integrins to downregulate keratinocyte MMP and growth factor production.<xref ref-type="bibr" rid="B9">9</xref></p><p>Under normal circumstances, the dynamic exchange between keratinocytes and matrix results in complete wound closure and regeneration of the dermis and epidermis. However, some wounds fail to heal normally and become chronic. Chronic wounds exhibit a healing profile different from that for normal acute wounds,<xref ref-type="bibr" rid="B23">23</xref>,<xref ref-type="bibr" rid="B24">24</xref> remaining in an inflamed state for protracted periods of time. Chronic wounds are frequently exacerbated by intrinsic factors such as excess inflammatory cytokines and poor vascularization/ischemia.<xref ref-type="bibr" rid="B25">25</xref> The deregulation of integrins<xref ref-type="bibr" rid="B26">26</xref> and overabundance of MMPs,<xref ref-type="bibr" rid="B27">27</xref>–<xref ref-type="bibr" rid="B29">29</xref> as well as the lack of tissue-specific inhibitors of MMPs,<xref ref-type="bibr" rid="B30">30</xref> can also give rise to chronic wounds. These wounds almost always require clinical intervention in order for healing to occur.</p><p>Because of the fluctuating balance of activating and inhibiting signals that are induced during wound healing, we question whether exogenous application of factors known to regulate normal healing might accelerate the acute healing process, or activate the healing of chronic wounds. We have previously shown that collagenase from <italic>Clostridium</italic> bacteria promotes the migration and proliferation of vascular endothelial cells and keratinocytes<xref ref-type="bibr" rid="B31">31</xref> after injury, with an efficiency more than twice that of mammalian collagenase. While mammalian enzymes cleave collagen at one site to produce one-quarter and three-quarter size fragments,<xref ref-type="bibr" rid="B32">32</xref> Clostridial collagenase cleaves all 3 helical domains to produce several breakdown products.<xref ref-type="bibr" rid="B33">33</xref>,<xref ref-type="bibr" rid="B34">34</xref></p><p>There has also been a great deal of interest in the papaya-derived protease papain, initially reported as a wound treatment 4 decades ago.<xref ref-type="bibr" rid="B35">35</xref> Research is still ongoing into the precise role this enzyme may have in promoting healing,<xref ref-type="bibr" rid="B36">36</xref> though it is currently classified as an enzymatic debriding agent. We have examined its effects on keratinocytes in vitro to determine whether its clinical efficacy may be due to the promotion of keratinocyte proliferation and/or motility. We have also revisited the Clostridial collagenase experiments on keratinocytes, and expanded our investigation to include an in vivo wound healing assay on miniature swine. In our in vivo study, we have compared Clostridial collagenase with 2 commercially available preparations, Regranex, the active ingredient of which is PDGF-BB, and Solosite, a cross-linked carboxymethyl cellulose–based hydrogel that maintains a moist healing environment, as well as a sterile gauze negative control. Overall, we hope to clarify the role that Clostridial collagenase may play as a wound healing agent.</p><sec sec-type="methods" id="S1"><title>METHODS</title><sec id="S1-1"><title>Cells</title><p>Epithelial keratinocytes isolated from normal human epidermal keratinocyte (NHEK) were purchased from Clonetics (Walkersville, Md) and cultured in complete Keratinocyte Growth Medium (KGM, Clonetics) in 175 cm<sup>2</sup> tissue culture flasks (Costar) as previously described.<xref ref-type="bibr" rid="B31">31</xref> Cells were passaged with trypsin, quenched with medium containing 5% bovine calf serum, rinsed in phosphate-buffered saline, and replated in KGM at a ratio of 1:4 for cell propagation and at specified cell numbers for all experiments. Cells were fed every other day by complete media exchange with fresh KGM. In all experiments where soluble factors were added to the growth media, these factors were added at the time of cell plating, and replenished as cells were fed on alternate days. HbEGF was obtained from Oncogene, San Diego, Calif.</p></sec><sec id="S1-2"><title>Matrix Preparations</title><p>Matrix preparations were made as previously described.<xref ref-type="bibr" rid="B31">31</xref> Briefly, capillary endothelial cells at 10 days postconfluence were released from their matrix with DOC buffer (0.5% sodium deoxycholate, 0.02 M Tris-Cl [pH 8.0], 0.015 M NaCl, 0.001 M EGTA [pH 7.0], 0.001 M phenylmethyl sulfonyl fluoride), and the remaining matrix was washed with phosphate-buffered saline. Enzymes used for matrix digestions were diluted in calcium-buffered saline, for a total of 250 μL per well in a 24-well plate. Digestions were performed at 37°C for 1 hour in a cell culture incubator. Following digestion or control treatment with calcium-buffered saline alone, the matrix was first washed with phosphate-buffered saline and then allowed to equilibrate in KGM for 15 minutes at 37°C. The KGM was removed, and NHEK cells were released from flasks with trypsin, counted, and plated onto the matrix in fresh KGM. Earlier published experiments indicate that the enzyme doses used for matrix treatment do not affect the plating efficiency of the keratinocytes.<xref ref-type="bibr" rid="B31">31</xref></p></sec><sec id="S1-3"><title>Enzymes</title><p>Enzymes used for extracellular matrix digestion were purified collagenase from <italic>Clostridium</italic>, crude collagenase, and clostripain (Advanced Biofactures Corp, Lynbrook, NY), as in previous experiments.<xref ref-type="bibr" rid="B31">31</xref> A mixture of papain and 1% urea (Beckton Dickinson, Sparks, Md) was also tested,<xref ref-type="bibr" rid="B36">36</xref>–<xref ref-type="bibr" rid="B39">39</xref> with or without the addition of chlorophyllin (Beckton Dickinson).<xref ref-type="bibr" rid="B40">40</xref>–<xref ref-type="bibr" rid="B42">42</xref> All treatment doses are given in standard units of enzymatic activity.</p></sec><sec id="S1-4"><title>Cell proliferation assay</title><p>Growth assays were performed as previously described.<xref ref-type="bibr" rid="B31">31</xref> Briefly, NHEK cells were released from the matrix with trypsin, diluted in Isoton II electrolyte buffer (Beckman, Fullerton, CA) for counting in a Coulter Counter model ZF (Coulter Electronics, Miami, Fla). Experimental conditions were plated in duplicate, and each sample was counted twice. Each condition was tested in at least 3 separate experiments. Data were recorded manually and analyzed in Microsoft Excel (Microsoft, San Jose, Calif).</p></sec><sec id="S1-5"><title>Cell migration assay</title><p>For motility studies, matrix was prepared on 10 mm<sup>2</sup> glass cover slips (Corning, Big Flats, NY) as described above. NHEK cells were plated at confluence on the resultant matrix and allowed to attach overnight. Cell monolayers were injured with a fire-polished pasteur pipette to create a narrow scratch wound. The wounded populations were monitored through time-lapse imaging, as previously described.<xref ref-type="bibr" rid="B43">43</xref> Relative motility was calculated by comparing the change in area covered by cells in the same sized viewing field over the same period of time for different treatment conditions.</p></sec><sec id="S1-6"><title>Wound healing study</title><p>A full-thickness wound healing study on swine was conducted by North American Science Associates (NAMSA), Northwood, Ohio, in accordance with the National Institutes of Health guidelines for the treatment of laboratory animals. Treatments tested were (1) <italic>Clostridium</italic> bacterial collagenase in ointment, (2) Regranex gel, (3) Solosite gel, and (4) dry sterile gauze as a negative control. The study was conducted on 8 healthy female adult Micro-Yucatan Miniature swine (<italic>Sus scrofa domesticus</italic>), selected on the basis of their use in other published research and their similarity of wound healing to humans. Each animal received eight 2-cm diameter full-thickness wounds on its back while under general anesthesia. Pressure was applied to stop bleeding before any treatment or dressing was affixed on Day 0. Two wounds on each pig were dressed with each of the 4 treatment conditions, with locations (cranial to caudal) randomized. Clostridial collagenase was prepared at 1000 U/g of ointment. Standard commercial preparations of Regranex and Solosite were used. Ointments were applied at 0.25 g/d, according to the directions for use of the commercial products. Treated wounds were covered with OpSite Flexgrid, a transparent wound dressing. The dry control wounds were covered with sterile gauze. All bandages and dressings were changed daily. The gauze dressings were removed carefully by a veterinarian, who noted minimal disruption to the wounds due to moisture from wound exudation. Observations were recorded daily for each wound, such as the diameter, color, exudation, and epithelization, in a double-blinded manner. Wound area was calculated from tracings made on transparent plastic during daily examinations. Following daily observations, the wound area was gently rinsed with sterile saline and dried with sterile gauze to permit adhesion of the new dressing. Vigorous flushing of the wound itself was not performed.</p><p>At the conclusion of the study, each wound was excised, with a 1 to 2 cm border of intact skin, for histological processing. Each sample was fixed in 10% neutral buffered formalin, embedded, sectioned, and stained with hematoxylin and eosin. Two sections were taken from each wound sample: one across the central diameter of the wound and one parallel section from the side halfway to the edge of the original defect. Epithelization was recorded from the central section of each wound as the percentage of the defect covered by well-defined dermal epithelium.</p></sec></sec><sec sec-type="results" id="S2"><title>RESULTS</title><sec id="S2-1"><title>In vitro proliferation</title><p>To simulate the in vivo basement membrane of keratinocytes, we used the biomatrix generated by confluent capillary endothelial cells, which has been shown to contain fibronectin, laminin, and collagen of types I, III, and IV.<xref ref-type="bibr" rid="B44">44</xref> Treatment of the endothelial-produced biomatrix with purified bacterial collagenase prior to plating caused an approximate doubling in keratinocyte proliferation as compared to cells on untreated matrix (Table <xref ref-type="table" rid="T1">1</xref>). The addition of HbEGF<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B31">31</xref>,<xref ref-type="bibr" rid="B45">45</xref> to the culture media was also tested in proliferation assays. The addition of 0.1 ng/mL HbEGF to cells on untreated matrix caused a 1.4-fold increase in proliferation over controls after 7 days, but when added to cells growing on collagenase-digested matrix, proliferation nearly tripled. Treatment of the matrix with crude collagenase, which contains other nonspecific proteases, caused only a slight increase in cell number over controls. Clostripain, a purified nonspecific protease, produced an even smaller increase in proliferation. Pretreatment of the matrix with urea and papain,<xref ref-type="bibr" rid="B36">36</xref>,<xref ref-type="bibr" rid="B38">38</xref>,<xref ref-type="bibr" rid="B46">46</xref> with or without chlorophyllin, diminished cell proliferation to half of control numbers. We next asked whether addition of collagenase to the growth media would affect proliferation, alone or in combination with matrix pretreatment (Table <xref ref-type="table" rid="T1">1</xref>). On untreated matrix, incubating cells with 16 U/mL soluble collagenase in the growth medium resulted in a 1.6-fold increase in proliferation. When cells were grown on matrix pretreated with collagenase, and incubated with the same dose of soluble enzyme, total proliferation was 4.5 times that for control, and nearly double the result seen with treated matrix alone.</p></sec><sec id="S2-2"><title>In vitro motility</title><p>To assess the effect of bacterial collagenase on cell motility, we used digital imaging microscopy to monitor the rate at which a confluent cell monolayer is able to close a narrow wound. In all conditions, keratinocytes at the wound edge first elaborate a membrane fan, then detach the trailing edge, and finally pull the nucleus and cell body forward toward the leading edge. As this happens, cells further from the wound edge spread to fill the newly available space. On matrix pretreated with collagenase, the cells at the leading edge extend a much longer membrane fan before translocating the rest of the cell, thus covering a greater distance in a shorter period of time (Figs <xref ref-type="fig" rid="F1">1C</xref> and <xref ref-type="fig" rid="F1">1D</xref>). Overall, keratinocyte motility was increased nearly 8-fold over controls when assayed on treated matrix, in the presence of HbEGF and soluble collagenase (Fig <xref ref-type="fig" rid="F1">1E</xref>). Time-lapse digital images can be viewed as Quicktime movies at <ext-link ext-link-type="uri" xlink:href="http://www.journalofburnsandwounds.com">http://www.journalofburnsandwounds.com</ext-link>.</p></sec><sec id="S2-3"><title>In vivo</title><p>In order to determine whether the enhancement of proliferation and migration seen in vitro would translate into the promotion of healing in vivo, a double-blinded study of full-thickness wounds on Yucatan Micropigs was conducted by NAMSA. Superficial observations of wound color and exudation did not vary markedly among the treatment conditions. Granulation tissue developed more slowly in control wounds than in treated wounds, but no obvious difference was noted between treatment conditions. By analysis of the wound area as a percentage of the original wound, it was found that collagenase accelerates overall healing (Fig <xref ref-type="fig" rid="F2">2</xref>). The differences among the treatment conditions became statistically significant (analysis of variance, <italic>P</italic> < .05) beginning on Day 6 of the trial, when the wound area began to decrease. Differences between collagenase and Solosite or collagenase and Regranex became significant (Tukey test, <italic>P</italic> < .05) from Day 7 onward. The difference between the collagenase condition and the control condition was significant (Tukey test, <italic>P</italic> < .05) from Day 4. The area of the wound covered by new epithelial tissue in the terminal histology samples was also greater with collagenase treatment than with Regranex, Solosite, or the dry control (Fig <xref ref-type="fig" rid="F3">3</xref>). This improved re-epithelization was accompanied by a decrease in observed granulation tissue in the collagenase-treated wounds (not shown). The improved condition of the collagenase-treated wounds can be seen in photographs taken on Day 10 after wounding (Fig <xref ref-type="fig" rid="F4">4</xref>).</p></sec></sec><sec sec-type="discussion" id="S3"><title>DISCUSSION</title><p>Through our experiments, we have reaffirmed that pretreatment of a biosynthesized extracellular matrix by Clostridial collagenase promotes human keratinocyte responses to injury. Collagenase acts through multiple signaling pathways, increasing proliferation and migration in vitro. It is effective via pretreatment of the matrix and/or its presence in the growth media. Interestingly, the effects of collagenase are strongly potentiated by inclusion of HbEGF, which is prevalent in in vivo wounds.<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref> Through a combination of matrix pretreatment and addition of soluble collagenase and/or HbEGF in the growth media, keratinocyte proliferation can be potentiated 5-fold. Keratinocyte migration following injury in vitro is similarly enhanced. Papain-urea, alone or in combination with chlorophyllin, currently in clinical use for enzymatic debridement of wounds,<xref ref-type="bibr" rid="B36">36</xref> inhibited keratinocyte proliferation in vitro, over a wide range of tested doses. Importantly, purified Clostridial collagenase also promoted the cellular responses to injury in vivo, resulting in reduced granulation tissue, increased rates of re-epithelization, and a shorter interval to wound closure.</p><p>Pretreatment of biosynthesized extracellular matrix,<xref ref-type="bibr" rid="B47">47</xref> which is analogous to the human keratinocyte basement membrane, with Clostridial collagenase promoted cell proliferation (Table <xref ref-type="table" rid="T1">1</xref>) and migration (Fig <xref ref-type="fig" rid="F1">1</xref>). These results are supported by earlier work from this laboratory on vascular endothelial cells migrating in response to injury.<xref ref-type="bibr" rid="B44">44</xref>,<xref ref-type="bibr" rid="B48">48</xref> In addition, endogenous collagenase has been shown to promote keratinocyte migration in vitro.<xref ref-type="bibr" rid="B16">16</xref>,<xref ref-type="bibr" rid="B49">49</xref> Collectively, these results point to the important role that collagenase and other MMPs play in modulating cellular responses to injury and wound healing.</p><p>Inclusion of Clostridial collagenase in the growth media, alone or in addition to matrix pretreatment, was also able to stimulate the keratinocyte response to injury (Table <xref ref-type="table" rid="T1">1</xref>). Furthermore, HbEGF stimulated keratinocyte proliferation in vitro, as well as potentiated the migration-enhancing effects of Clostridial collagenase (Table <xref ref-type="table" rid="T1">1</xref>). It has been shown in vitro that stromelysin, or MMP-3,<xref ref-type="bibr" rid="B15">15</xref> is able to cleave HbEGF at the juxtamembrane site, releasing the mature, soluble form of the growth factor to signal in a paracrine or juxtacrine fashion. Interestingly, MMP-3 cleaves HbEGF successfully, but MMP-2 and MMP-9 do not. Other in vitro studies have shown that membrane-type MMP-1 can cleave and activate pro-αV, -3, and -5 integrins.<xref ref-type="bibr" rid="B49">49</xref>,<xref ref-type="bibr" rid="B50">50</xref> Through its processing of the αV integrin subunit, MT-MMP-1 prevents the suppression of the collagen-binding α2β1 receptor by αVβ3,<xref ref-type="bibr" rid="B51">51</xref> with the overall effect promoting migration on Type I collagen. These data indicate that membrane-associated growth factors and receptors can be released and/or activated through MMP action. Similarly, by releasing and activating endogenous promoters of growth and migration, Clostridial collagenase may be able to induce the keratinocyte responses to injury.</p><p>Our in vivo wound healing study on swine extends our in vitro findings that Clostridial collagenase promotes epithelial cell proliferation and migration by revealing that direct application of purified Clostridial collagenase to full-thickness wounds promotes healing and improves wound health as compared to other commercial preparations tested or untreated controls. There are numerous reports indicating that several types of collagenase have already been tested and are in clinical use as enzymatic debriding agents.<xref ref-type="bibr" rid="B52">52</xref>–<xref ref-type="bibr" rid="B56">56</xref> From the studies using preparations of Clostridial collagenase,<xref ref-type="bibr" rid="B57">57</xref>,<xref ref-type="bibr" rid="B58">58</xref> there is ample evidence to indicate that collagenase not only successfully debrides wounds but also hastens wound closure.<xref ref-type="bibr" rid="B53">53</xref> Altogether, these results point to a promising role for collagenase in promoting acute and chronic wound healing.</p></sec>
Management of an Acute Thermal Injury With Subatmospheric Pressure	<p><bold>Objective:</bold> This article reports the first application of subatmospheric pressure management to a deep, partial-thickness human thermal burn. <bold>Methods:</bold> After cleaning the wound, the decision was made to treat the hand and distal forearm with subatmospheric pressure (V.A.C., KCI, Inc, San Antonio, Tex). The sponge was applied directly to the burned skin without additional interface at approximately 6 hours after injury. The dressing was maintained at a continuous negative pressure of 125 mm Hg over the next 40 hours, with interruption only for routine clinical evaluation at 5, 16, and 24 hours after initiation of treatment. This was accomplished by opening the dressing without completely changing it. The treatment was tolerated well by the patient, requiring no excessive pain medication. After the subatmospheric pressure treatment was stopped, the wound appeared to be of indeterminate depth and the patient was started on twice daily applications of silver sulfadiazine. <bold>Results:</bold> The clinical impression at this time was that the hand burn had not progressed but had stabilized and had minimal edema. He was followed as an outpatient and returned to work by 8 weeks. At approximately 4 weeks postinjury, his skin not only was functional but also appeared more normal, with less hyperemia than adjacent areas treated with topical antibacterials. <bold>Conclusion:</bold> The present case does not prove that subatmospheric pressure treatment prevents burn wound progression. However, when combined with the previously reported laboratory studies it suggests the need for further research. Currently, a prospective, randomized, blinded, controlled multicenter trial is underway to evaluate the clinical importance of these observations.</p>	<contrib contrib-type="author"><name><surname>Molnar</surname><given-names>Joseph A.</given-names></name><degrees>MD, PhD, FACS</degrees><xref rid="aff1" ref-type="aff"/><xref ref-type="corresp" rid="cor1"/></contrib><contrib contrib-type="author"><name><surname>Simpson</surname><given-names>Jordan L.</given-names></name><degrees>BS</degrees><xref rid="aff1" ref-type="aff"/></contrib><contrib contrib-type="author"><name><surname>Voignier</surname><given-names>Denise M.</given-names></name><degrees>CMA</degrees><xref rid="aff1" ref-type="aff"/></contrib><contrib contrib-type="author"><name><surname>Morykwas</surname><given-names>Michael J.</given-names></name><degrees>PhD</degrees><xref rid="aff1" ref-type="aff"/></contrib><contrib contrib-type="author"><name><surname>Argenta</surname><given-names>Louis C.</given-names></name><degrees>MD</degrees><xref rid="aff1" ref-type="aff"/></contrib>	Journal of Burns and Wounds	<p>Management of acute thermal injury is often frustrated by the phenomenon of burn wound progression. In this circumstance, heat-damaged tissue that is alive at presentation becomes progressively nonviable until the skin is found to be nonsalvageable and requires excision and grafting. The etiology of this process is unclear. It has been best described by Jackson as a zone of stasis where with increased vascular permeability, progressive edema, increased blood viscosity, and vascular thrombosis the tissue dies.<xref ref-type="bibr" rid="B1">1</xref> While limiting burn wound progression would be of clear benefit to the burn patient, no clinical studies have shown a way to prevent it.<xref ref-type="bibr" rid="B2">2</xref></p><p>The V.A.C. (K.C.I., Inc, San Antonio, Tex) consists of an open cell polyurethane ether foam with an embedded evacuation tube. The foam is sealed to the wound with an adherent drape, and subatmospheric pressure is applied to the evacuation tube. Previous studies have demonstrated the effectiveness of this device in helping to control edema and speed up the vascularization of wounds.<xref ref-type="bibr" rid="B3">3</xref> Morykwas et al have also demonstrated in a swine model of thermal injury that the maximum depth of cell death could be decreased with application of subatmospheric pressure.<xref ref-type="bibr" rid="B4">4</xref></p><p>We report the first application of subatmospheric pressure management to a deep, partial-thickness human thermal burn.</p><sec sec-type="cases" id="S1"><title>CASE REPORT</title><p>On August 2, 1995, a 26-year-old male electrician received a flash burn to his right upper extremity and face when exposed to the heat from a high-voltage electrical arc (Fig <xref ref-type="fig" rid="F1">1</xref>).</p><p>The hand and digits were pale in color and dry on the dorsum, suggesting deep, partial-thickness burns (Fig <xref ref-type="fig" rid="F2">2</xref>). The injury was progressively more superficial proximally on the extremity. The clinical impression of 3 surgeons with experience in burn care was that the distal portion of the burn, ie, hand and forearm, would require excision and grafting.</p><p>After cleaning the wound, the decision was made to treat the hand and distal forearm with subatmospheric pressure (V.A.C., KCI, Inc, San Antonio, Tex) and to apply silver sulfadiazine more proximally. The sponge was applied directly to the burned skin without additional interface at approximately 6 hours after injury. The dressing was maintained at a continuous negative pressure of 125 mm Hg over the next 40 hours, with interruption only for routine clinical evaluation at 5, 16, and 24 hours after initiation of treatment. This was accomplished by opening the dressing without completely changing it. The treatment was tolerated well by the patient, requiring no excessive pain medication. After the subatmospheric pressure treatment was stopped, the patient was started on twice daily applications of silver sulfadiazine.</p><p>The clinical impression at this time was that the hand burn had not progressed but had stabilized and had minimal edema (Fig <xref ref-type="fig" rid="F3">3</xref>). However, it was now of indeterminate depth.</p><p>The patient was started on hand therapy, and the hand was kept elevated. The wound continued to epithelialize until it was clinically healed by Day 10, but the patient had received significant fingernail injury that persisted until the nail was completely replaced.</p><p>The patient was discharged home on the 12th postinjury day. He was followed as an outpatient and returned to work by October. At approximately 4 weeks, his skin not only was functional but also had an excellent cosmetic result (Fig <xref ref-type="fig" rid="F4">4</xref>). In addition, the skin on the hand appeared more normal with less hyperemia than the skin of the shoulder, despite the fact that the hand had received the deepest burn consistent with the mechanism of surgery (Fig <xref ref-type="fig" rid="F5">5</xref>).</p></sec><sec sec-type="discussion" id="S2"><title>DISCUSSION</title><p>Fifty years ago, Jackson put forth a paradigm for an understanding of the pathogenesis of burn wound progression.<xref ref-type="bibr" rid="B1">1</xref> He described the wound as consisting of 3 concentric zones of injury. The most severe of these is the zone of coagulation. It is irreversibly damaged and represents nonsalvageable dead tissue. To the other extreme is the zone of hyperemia. This tissue is minimally injured, resulting in an inflammatory response, and will usually heal spontaneously. In between is the zone of stasis. This is characterized by increased vascular permeability, edema, and progressive blood viscosity, leading to thrombosis and additional tissue death. It is this zone of stasis that represents the deep second-degree burn that is clearly viable tissue when the patient arrives but subsequently goes on to die and requires excision and grafting much in the manner of a third-degree or full-thickness burn.</p><p>While initially Jackson thought that such capillary stasis and burn wound progression was an inevitable consequence of the original injury, Order et al<xref ref-type="bibr" rid="B5">5</xref> demonstrated reopening of the circulation in second-degree burns in a rat model more than a decade later. This led Jackson and others to consider the possibility of prohibiting the progression so as to minimize the potential need for surgery and perhaps to save lives.<xref ref-type="bibr" rid="B6">6</xref> However, in order to control this process, it would be necessary to understand the mechanism.</p><p>Evaluation of the microcirculatory changes due to thermal injury has demonstrated the complex nature of the response. Early after the injury, endothelial cells swell, resulting in capillary narrowing and decreased flow.<xref ref-type="bibr" rid="B7">7</xref> The swelling of the endothelial cells contributes to capillary leak but may also be the result of free-radical mechanisms.<xref ref-type="bibr" rid="B2">2</xref>,<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B9">9</xref> The capillary leak allows margination of cellular elements of the blood, platelet aggregation, and stimulation of inflammatory mediator response. This process begins in the first 3 to 24 hours depending on the severity of injury and continues for up to 48 hours after burn.<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B9">9</xref> While much of the emphasis of microcirculatory clotting has been on the arteriole, it appears that venous occlusion may occur first, resulting in secondary arteriolar clotting.<xref ref-type="bibr" rid="B10">10</xref>–<xref ref-type="bibr" rid="B13">13</xref></p><p>In addition to inflammation and progressive thrombosis, more direct mechanisms may cause progressive tissue damage. Zawacki and others have shown that dehydration due to the loss of the outer protective layers may contribute to burn wound progression.<xref ref-type="bibr" rid="B14">14</xref>–<xref ref-type="bibr" rid="B16">16</xref> Systemic hypoperfusion, infection, malnutrition, and inadequate immune response are all important causes of worsening of the burn wound, and proper resuscitation and metabolic support limit this process.<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref></p><p>Efforts to limit burn wound progression have primarily concentrated on pharmacologic interventions in the thrombosis or inflammatory response. Robson et al showed that application of 1% methylprednisolone acetate to a guinea pig model of burn wounds decreased loss of dermal appendages and increased dermal perfusion, presumably by interfering with white blood cell adherence.<xref ref-type="bibr" rid="B17">17</xref> However, this was not confirmed by subsequent investigators using clobestasol propionate.<xref ref-type="bibr" rid="B18">18</xref> Use of monoclonal antibodies to prevent leukocyte adherence in a burn model did decrease burn size, speed up reepithelialization, produce thinner eschar, spare more hair follicles, and have greater patency of vessels than controls.<xref ref-type="bibr" rid="B19">19</xref>,<xref ref-type="bibr" rid="B20">20</xref> While Ehrlich found that a lazaroid could prevent burn wound progression, Melikian et al could not find an effect of the free-radical mechanism on burn wound progression using dimethyl sulfoxide, allopurinol, or polyethylene glycol-superoxide dismutase (PEG-SOD; a superoxide scavenger).<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B21">21</xref></p><p>Ehrlich has also demonstrated the importance of the clotting mechanism in this process by the use of ancrod, a protease derived from pit vipers that converts fibrinogen to a nonclotting molecule.<xref ref-type="bibr" rid="B22">22</xref> By giving this to rats 3 days before creating experimental burns, he was able to limit the size of the burn. Heparin has also been used anecdotally to treat clinical burns by preventing worsening by thrombosis.<xref ref-type="bibr" rid="B23">23</xref> However, the most consistent effect has been seen with the use of the nonsteroidal anti-inflammatory drug ibuprofen. While early reports suggested a thromboxane mechanism to prevent burn wound progression, more recent studies have suggested that it works by blocking a plasmin inhibitor that would normally block fibrinolysis in the burn wound.<xref ref-type="bibr" rid="B2">2</xref>,<xref ref-type="bibr" rid="B24">24</xref>–<xref ref-type="bibr" rid="B26">26</xref> Finally, a topical form of ibuprofen (flurbiprofen) when applied to an acute burn model within 4 hours of injury has also been suggested to have a positive effect on vascularity.</p><p>Despite these efforts, no definitive technique has been shown to be clinically effective in minimizing burn wound progression. This is because the pharmacologic methods described are either toxic or contraindicated or must be administered before or too early after injury to be effective. Since the average burn patient arrives at the hospital 3 hours after injury,<xref ref-type="bibr" rid="B27">27</xref> therapeutic interventions must take this into consideration. The ideal technique to stop burn wound progression would allow for a 3 or more hour delay before hospital treatment, would have no systemic effects, and would not interfere with other treatment methods.</p><p>Morykwas et al evaluated the effect of subatmospheric pressure on acute burn wounds in a swine model.<xref ref-type="bibr" rid="B4">4</xref> He applied a relative negative pressure of 125 mm Hg in an artificially closed space to experimental wounds with control wounds on the same animal. When applied within 12 hours after injury, a significant improvement was found as measured by the maximum depth of cell death. Based on this evaluation, treatment periods as short as 6 hours were efficacious. In fact, application periods as long as 5 days were not significantly different from application periods as short as 6 or 12 hours. In addition, histologic evaluation demonstrated decreased inflammatory response in wounds treated with subatmospheric pressure as compared to controls.</p><p>In the present study, subatmospheric pressure treatment was applied to an upper extremity in a patient with a flash burn that extended from his fingertips to over his shoulder. This is the first clinical application of subatmospheric pressure to an acute human burn injury. The treatment period was approximately 2 days, and it was applied approximately 6 hours after injury. Despite the clinical impression of 3 surgeons experienced in burn care that this would ultimately require excision and grafting, this was avoided. The wound healed without complication, and applying subatmospheric pressure to the acutely burned tissue did no harm. In addition, the skin that initially appeared the deepest burned on the hand and the forearm healed in a manner that was less hyperemic with superior skin quality to the skin of the shoulder that received less injury.</p><p>Unfortunately, there is no absolute method of evaluation that the burn surgeon may use to ascertain the depth of the burn at this early time point. Recent reports using scanning laser Doppler have had some interest but are not without error and not widely used to make such decisions.<xref ref-type="bibr" rid="B28">28</xref> The evaluation of burn depth remains primarily a clinical decision. It is impossible to know with certainty if this patient would have healed as well with alternative treatments such as silver sulfadiazine. Nonetheless, the severe nail damage, as seen in Figure <xref ref-type="fig" rid="F4">4</xref>, suggests that excision and grafting would have been necessary.</p><p>A more intriguing issue is how subatmospheric pressure treatment improves the healing of the burn wound as seen in the laboratory studies of Morykwas et al<xref ref-type="bibr" rid="B4">4</xref> and as suggested by this clinical case. One possible way is that the device removes acute inflammatory mediators, such as free radicals and cytokines, that are involved with burn wound progression.<xref ref-type="bibr" rid="B2">2</xref>,<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B20">20</xref>,<xref ref-type="bibr" rid="B21">21</xref> While this has not been proven in burns, it is clear from the studies in crush injury that certain toxins may be removed from acute wounds.<xref ref-type="bibr" rid="B29">29</xref> It is also possible that decreasing edema is an important mechanism to speed up the healing of the acute burn. With edema there is a decrease in vascular density, increased diffusion distance, possible vasospasm, thrombosis, and stasis in the microcirculation. Present observations suggest that subatmospheric pressure treatment does decrease wound edema by yet uncertain mechanisms. Finally, subatmospheric pressure treatment may provide an ideal environment for the healing wound by providing the damaged skin with the ideal water vapor pressure to avoid desiccation.<xref ref-type="bibr" rid="B14">14</xref>–<xref ref-type="bibr" rid="B16">16</xref></p><p>The present case does not prove that subatmospheric pressure treatment prevents burn wound progression. However, when combined with the previously reported laboratory studies it suggests the need for further research. Currently, a prospective, randomized, blinded, controlled multicenter trial is underway to evaluate the clinical importance of these observations.</p></sec>
The Role of Anabolic Hormones for Wound Healing in Catabolic States	<p><bold>Objective:</bold> The purpose of this paper is to present an overview of the interrelationship between hormones, nutrition, and wound healing. <bold>Methods:</bold> The data on various hormones and their effects on specific elements of nutrition and wound healing are reviewed. <bold>Results:</bold> The key anabolic hormones are human growth hormone, insulin-like growth factor-1, insulin, and testosterone and its analogs. Although each has specific metabolic actions, there is also a very important hormone-hormone interaction. A deficiency of these hormones occurs in acute and chronic catabolic states, resulting in lean mass loss and impairing the healing process. <bold>Conclusion:</bold> There is a well-recognized interrelationship between hormones, nutrition, and wound healing. The anabolic process of protein synthesis, with new tissue formation, requires the action of anabolic hormones. Exogenous administration of these agents has been shown to maintain or increase lean body mass as well as directly stimulate the healing process through their anabolic and anticatabolic actions.</p>	<contrib contrib-type="author"><name><surname>Demling</surname><given-names>Robert H.</given-names></name><degrees>MD</degrees><xref rid="aff1" ref-type="aff"/><xref ref-type="corresp" rid="cor1"/></contrib>	Journal of Burns and Wounds	<p>There are a number of key hormones involved with energy production, anabolism or protein synthesis, and catabolism or protein breakdown. The balance of anabolic and catabolic hormones affects wound healing both indirectly by the status of overall net protein synthesis and directly by improving the wound healing process.<xref ref-type="bibr" rid="B1">1</xref>–<xref ref-type="bibr" rid="B4">4</xref> A decrease in normal anabolic hormone activity and an increase in catabolic hormone activity occurs with the “stress response” to injury and also with aging and chronic illness.</p><p>The altered hormonal environment can lead to both a significant increase in catabolism, with net tissue breakdown, and a decrease in the overall anabolic activity required to preserve lean mass and maintain the healing process. The stress response to injury also produces an alteration in the normally protective protein sparing as seen in the normal and starved states aimed at preserving lean body mass.</p><p>The metabolic pathways, which generate energy to meet daily demands and for new protein synthesis, are very tightly regulated in normal or starved humans.<xref ref-type="bibr" rid="B5">5</xref>–<xref ref-type="bibr" rid="B8">8</xref> Macronutrients in the form of fat and carbohydrates are channeled into production of energy,<xref ref-type="bibr" rid="B5">5</xref>–<xref ref-type="bibr" rid="B11">11</xref> while the majority of protein consumed is used for protein synthesis, restoring and maintaining lean body mass. Lean mass, the metabolically active body compartment containing all the protein plus water, in the body includes muscle, skin, and the immune system, all of which are composed of protein. Normally, only 5% of consumed protein is used for energy. However, if anabolic activity decreases, as with stress or with aging or chronic illness, there is an escape of protein from the protein synthesis compartment to the energy compartment. Up to 25% of available protein substrate is burned for energy.<xref ref-type="bibr" rid="B3">3</xref>–<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B12">12</xref> A protein deficiency can occur and with the increased energy demands of a wound, a protein energy malnutrition can quickly evolve, especially in high-risk groups like the elderly with preexisting lean mass loss.<xref ref-type="bibr" rid="B10">10</xref>,<xref ref-type="bibr" rid="B11">11</xref> Morbidity, especially impaired immunity and impaired healing, is directly proportional to the degree of lean mass loss.<xref ref-type="bibr" rid="B13">13</xref>–<xref ref-type="bibr" rid="B15">15</xref> Impaired healing therefore is the result of both an inadequate intake of protein substrate and actual shunting of protein substrate away from the wound to be used instead for the restoration of lost lean mass.</p><sec id="S1"><title>THE RATIONALE FOR HORMONAL MODIFICATION</title><p>It is now well recognized that the hormonal environment, so critical to wound healing, can be beneficially modified.<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B16">16</xref>,<xref ref-type="bibr" rid="B17">17</xref> In general, restoration or improvement in net protein synthesis, required in wound healing, is the result of 2 processes. The first is an attenuation of the catabolic hormonal response to injury. Any hormonal manipulation that decreases the rate of catabolism would appear to be beneficial for wound healing. All the anabolic hormones have been shown to have anticortisol activity.<xref ref-type="bibr" rid="B1">1</xref>–<xref ref-type="bibr" rid="B4">4</xref> This effect decreases the catabolic response of cortisol but does not alter its protective anti-inflammatory response. A blockade of cortisol's anti-inflammatory properties could lead to an excessive “autodestructive” inflammatory process.</p><p>The second process is the accentuation of anabolic activity both generally and specific to the wound. A number of clinical and basic science studies have demonstrated the ability of exogenous delivery of anabolic hormones to increase net nitrogen retention and overall protein synthesis. Wound healing has also been reported to be improved. However, it remains difficult to sort out how much of the response is a result of an overall systemic anabolic effect and how much is due to direct effect on wound healing.<xref ref-type="bibr" rid="B18">18</xref>,<xref ref-type="bibr" rid="B19">19</xref></p><p>There are 4 major anabolic hormones that indirectly or directly affect wound healing. They are human growth hormone (HGH), insulin-like growth factor-1 (IGF-1), insulin, and testosterone (and its analogs) (Table <xref ref-type="table" rid="T2">2</xref>). As will be described later, each hormone has a specific mode of action but there are considerable interrelationships among these 4 hormones. In subsequent sections, the individual anabolic hormones will be discussed. It is important to emphasize that for any anabolic hormone to stimulate protein synthesis adequate calories for energy and protein for substrate must be provided. In the case of the hypermetabolic state seen in the stress response, a high caloric (30 cal/kg daily) and high protein (1.5 g/kg daily) intake is necessary.<xref ref-type="bibr" rid="B20">20</xref>,<xref ref-type="bibr" rid="B21">21</xref></p></sec><sec id="S2"><title>HUMAN GROWTH HORMONE</title><sec id="S2-1"><title>Actions</title><p>HGH is a potent endogenous anabolic hormone produced by the pituitary gland in daily doses of 0.5 to 0.8 mg in children and young adults. Growth hormone is a large polypeptide that contains 2 receptor-binding sites. There are a number of growth hormone–binding proteins, and growth factor–binding sites are found on a large variety of tissues, especially liver. The production of HGH decreases rapidly with increasing age. The levels are at their peak during the growth spurt. Starvation and intense exercise are 2 other potent stimuli while acute or chronic injury or illness suppress HGH release, especially in the elderly.<xref ref-type="bibr" rid="B22">22</xref>–<xref ref-type="bibr" rid="B27">27</xref> The amino acids glutamine and arginine, when given in large doses, have also been shown to increase HGH release.</p><p>HGH has a number of metabolic effects. The most prominent is its anabolic effect. HGH increases the influx of amino acids into the cell and decreases the efflux. Cell proliferation is accentuated as is overall protein synthesis and new tissue growth. HGH also stimulates IGF-1 production by the liver, and some of the anabolism seen with HGH is that produced by IGF-1, another anabolic agent.<xref ref-type="bibr" rid="B26">26</xref>–<xref ref-type="bibr" rid="B30">30</xref> Other effects, listed in Tables <xref ref-type="table" rid="T3">3</xref> and <xref ref-type="table" rid="T4">4</xref>, include its effects on glucose and fat metabolism.</p><p>The effect on increasing fat metabolism is beneficial in that fat is preferentially used for energy production and amino acids are preserved for use in protein synthesis.</p><p>Recent data indicate that insulin provides some of the anabolic effects of HGH therapy. At present, the issue as to the specific anabolic effects attributed to HGH versus IGF-1 and insulin remains unresolved.</p></sec><sec id="S2-2"><title>Clinical results of HGH use</title><p>Clinical studies have in large part focused on the systemic anabolic and anticatabolic actions of HGH.<xref ref-type="bibr" rid="B31">31</xref>–<xref ref-type="bibr" rid="B33">33</xref> Populations where HGH has been shown to be beneficial include those with severe burn and trauma, those with HIV infection with wasting, and the frail elderly (Table <xref ref-type="table" rid="T5">5</xref>). In addition, HGH is being used to slow down the aging process. Increases in lean mass, muscle strength, and immune function have been documented in clinical use. HGH is approved for use only in short statured children and is an orphan drug when used for improving protein synthesis. Increased anabolic activity from HGH requires intake of a high-protein, high-energy diet.</p></sec><sec id="S2-3"><title>Wound healing effect</title><p>As to its direct wound healing effects, skin is a target tissue for HGH, both directly through HGH receptors on the surface of epidermal cells and indirectly through the action of IGF-1.<xref ref-type="bibr" rid="B30">30</xref>,<xref ref-type="bibr" rid="B34">34</xref> Exogenously administered HGH has been shown to increase skin thickness in normal humans.<xref ref-type="bibr" rid="B35">35</xref></p><p>Other effects on the wound include increased rate of re-epithelialization of skin graft donor sites in adults and children with severe burns or trauma (Table <xref ref-type="table" rid="T6">6</xref>).<xref ref-type="bibr" rid="B36">36</xref>–<xref ref-type="bibr" rid="B38">38</xref></p><p>In addition, HGH has been shown to increase wound collagen content, granulation tissue and wound tensile strength, and the local production of IGF-1 by fibroblasts. These data are mainly derived from animal studies.<xref ref-type="bibr" rid="B39">39</xref>–<xref ref-type="bibr" rid="B41">41</xref></p></sec><sec id="S2-4"><title>Complications</title><p>Significant complications can occur with the use of HGH. The anti-insulin effects are problematic in that glucose is less efficiently used for fuel and increased plasma glucose levels are known to be deleterious.</p><p>Increased insulin requirements occur. Complications are listed in Table <xref ref-type="table" rid="T7">7</xref>.<xref ref-type="bibr" rid="B42">42</xref>,<xref ref-type="bibr" rid="B43">43</xref> It is important to also point out the findings of a multicenter European study of critically ill patients receiving HGH. In this study of mainly critically ill postoperative cardiac patients, mortality was 2-fold greater in those treated with HGH compared to those treated with placebo.<xref ref-type="bibr" rid="B43">43</xref></p></sec><sec id="S2-5"><title>Summary</title><p>In summary, HGH used in conjunction with adequate nutrition and protein intake clearly results in increased anabolic activity and will positively impact wound healing by increasing net protein synthesis in catabolic states. There is some data that HGH can directly improve wound healing. However, the impact of IGF-1 and insulin on the effects of HGH remains undefined.</p></sec></sec><sec id="S3"><title>INSULIN-LIKE GROWTH FACTOR-1</title><sec id="S3-1"><title>Actions</title><p>IGF-1 is a large polypeptide that has hormone like properties.<xref ref-type="bibr" rid="B44">44</xref>–<xref ref-type="bibr" rid="B46">46</xref> IGF-1, also known as somatomedin-C, has metabolic and anabolic properties very similar to those of insulin.</p><p>Although produced by a variety of wound cells, such as fibroblasts and platelets, the main source of production is liver where its synthesis is initiated by HGH. The IGF receptor is expressed in many different tissues and the active peptide is bound in plasma by IGF-binding proteins. IGF increases systemic nitrogen retention and protein synthesis.<xref ref-type="bibr" rid="B47">47</xref>–<xref ref-type="bibr" rid="B49">49</xref> However, its anabolic activity is difficult to distinguish from that of HGH as HGH needs to be present in order for IGF-1 to be produced and to have anabolic actions. The combination of HGH and IGF-1 delivery results in a synergistic anabolic effect.<xref ref-type="bibr" rid="B49">49</xref></p><p>The effects of HGH on wound healing are also considered to be due in part to IGF-1.<xref ref-type="bibr" rid="B49">49</xref> IGF-1 production is also dependent on normal levels of circulating androgens.<xref ref-type="bibr" rid="B45">45</xref> There is therefore a close interrelationship between all of the anabolic hormones. IGF-1 levels decrease with aging and also with a major insult such as trauma or sepsis. The decrease in IGF-1 levels increases the net nitrogen losses caused by wounds.</p></sec><sec id="S3-2"><title>Clinical results of IGF-1 use</title><p>Properties of IGF-1 are summarized in Table <xref ref-type="table" rid="T8">8</xref>. Its metabolic properties include increased protein synthesis, decrease in blood glucose, and attenuation of stress-induced hypermetabolism, the latter 2 properties being quite different from those of HGH.<xref ref-type="bibr" rid="B50">50</xref>–<xref ref-type="bibr" rid="B52">52</xref> The attenuation of stress-induced hypermetabolism is a very favorable property of IGF-1. Clinical trials, using an IGF-1 infusion, have focused on demonstrating increased anabolic activity.<xref ref-type="bibr" rid="B53">53</xref>,<xref ref-type="bibr" rid="B54">54</xref> Increased protein synthesis and nitrogen retention has been reported in burns, head injury, and HIV-induced catabolic states.</p></sec><sec id="S3-3"><title>Wound healing effect</title><p>The wound healing effects of IGF-1 are described in Table <xref ref-type="table" rid="T9">9</xref>.<xref ref-type="bibr" rid="B31">31</xref>,<xref ref-type="bibr" rid="B32">32</xref>,<xref ref-type="bibr" rid="B55">55</xref>,<xref ref-type="bibr" rid="B56">56</xref> IGF-1 is considered to be a wound healing stimulant, increasing cell proliferation and collagen synthesis. In addition, IGF-1 infusion has been shown to reverse diabetes and corticosteroid-induced impairment in wound healing. It is important to point out that these properties have largely been reported in animal studies. However, the increase in overall anabolism should benefit a wound.</p></sec><sec id="S3-4"><title>Complications</title><p>Hypoglycemia is the main complication. IGF-1 in general appears to have fewer side effects than do HGH. However, it is usually given as a continuous infusion because of its very short half-life. This factor limits its clinical usefulness.</p></sec><sec id="S3-5"><title>Summary</title><p>IGF-1 used in conjunction with adequate nutrition significantly increases net anabolic activity. Its direct effect on wound healing is less clear. The attenuation of stress-induced hypermetabolism is a significant advantage. Clinical use is hampered by the need for a continuous infusion.</p></sec></sec><sec id="S4"><title>INSULIN</title><sec id="S4-1"><title>Actions</title><p>The hormone insulin is known to have anabolic activities in addition to its effect on glucose and fat metabolism.<xref ref-type="bibr" rid="B57">57</xref> In a catabolic state, exogenous insulin administration has been shown to decrease proteolysis in addition to increasing protein synthesis.<xref ref-type="bibr" rid="B33">33</xref>,<xref ref-type="bibr" rid="B39">39</xref>–<xref ref-type="bibr" rid="B41">41</xref></p><p>The anabolic activity appears to mainly affect the muscle and skin protein in the lean body mass compartment (Table <xref ref-type="table" rid="T10">10</xref>). An increase in circulating amino acids produced by wound amino acid intake increases the anabolic and anticatabolic effect in both normal humans and populations in a catabolic state.<xref ref-type="bibr" rid="B39">39</xref>–<xref ref-type="bibr" rid="B41">41</xref>,<xref ref-type="bibr" rid="B58">58</xref></p></sec><sec id="S4-2"><title>Clinical results of insulin use</title><p>A number of clinical trials, mainly in burn patients (Table <xref ref-type="table" rid="T11">11</xref>), have demonstrated the stimulation of protein synthesis and decreased protein degradation and net nitrogen uptake, especially in skeletal muscle.<xref ref-type="bibr" rid="B58">58</xref>–<xref ref-type="bibr" rid="B63">63</xref> An increase in anabolic activity is also evident in diabetic patients who are provided more insulin. The positive effect of insulin on protein synthesis decreases with aging.<xref ref-type="bibr" rid="B64">64</xref> This response is different from that of HGH and anabolic steroids where age does not appear to blunt the anabolic response.</p></sec><sec id="S4-3"><title>Wound healing effect</title><p>There is less data on the actions of insulin on wound healing over and above its systemic anabolic effect.<xref ref-type="bibr" rid="B61">61</xref>,<xref ref-type="bibr" rid="B65">65</xref>–<xref ref-type="bibr" rid="B67">67</xref> Increases in skin protein content have been demonstrated with a chronic insulin infusion. Increased re-epithelialization of skin graft donor sites was reported in one clinical trial in burn patients. Several animal studies have demonstrated increased collagen production with insulin, and increasing insulin administration to diabetic mice improved all phases of healing. However, the effects of insulin on wound healing have not been well studied in humans.</p></sec><sec id="S4-4"><title>Complications</title><p>The main complication is hypoglycemia. There does not appear to be any fluid retention or hypermetabolism with its use.</p></sec><sec id="S4-5"><title>Summary</title><p>In summary, hyperinsulinemia in catabolic patients and in normal humans increases net protein synthesis and decreases protein breakdown. An infusion of glucose is required to avoid hypoglycemia. In turn, inadequate insulin intake in diabetic patients leads to progressive lean mass loss and hyperglycemia appears to accentuate the nitrogen loss. Although some positive data are present on insulin improving healing, the data are limited.</p></sec></sec><sec id="S5"><title>TESTOSTERONE</title><sec id="S5-1"><title>Actions</title><p>Testosterone, whose basic structure is a steroid ring, is a natural endogenous androgen.<xref ref-type="bibr" rid="B68">68</xref>–<xref ref-type="bibr" rid="B70">70</xref> It is synthesized primarily in the Leydig cells of the testes in men and by the ovaries and adrenal glands in women. Healthy adult men produce 3 to 10 mg of testosterone a day, yielding plasma concentrations ranging from 300 to 1000 μg/dL.<xref ref-type="bibr" rid="B71">71</xref>–<xref ref-type="bibr" rid="B73">73</xref> It acts on the cells' androgenic receptors found mainly in skin, muscle, and male sex glands. It has both androgenic or masculinizing properties and anabolic properties. Androgenic effects are present to some degree in all anabolic steroids. Androgenic effects include development of male sex glands, determination of male hair growth pattern, increased libido, and assertiveness (Table <xref ref-type="table" rid="T12">12</xref>). Most testosterone analogs or anabolic steroids have androgenic properties much lower than those of testosterone itself.<xref ref-type="bibr" rid="B57">57</xref>,<xref ref-type="bibr" rid="B74">74</xref>,<xref ref-type="bibr" rid="B75">75</xref> The anabolic properties were defined in the 1930s. These include an increase in muscle size, synthesis, and strength. Increased skin thickness has also been noted with administration of testosterone to hypogonadal men. The importance of testosterone is evidenced by the complications seen with low testosterone levels, which include sarcopenia or lost lean mass, increased rate of development of osteoporosis, anemia, thinning of skin and weakness, and impaired wound healing (Tables <xref ref-type="table" rid="T13">13</xref> and <xref ref-type="table" rid="T14">14</xref>).<xref ref-type="bibr" rid="B71">71</xref>,<xref ref-type="bibr" rid="B76">76</xref>,<xref ref-type="bibr" rid="B77">77</xref></p><p>The native molecule was first used in the 1950s to correct a debilitated state, correct anemia, and increase calcium deposition in bones as well as to treat hypogonadal states.<xref ref-type="bibr" rid="B68">68</xref>–<xref ref-type="bibr" rid="B70">70</xref>,<xref ref-type="bibr" rid="B76">76</xref>,<xref ref-type="bibr" rid="B77">77</xref> The testosterone molecule is rapidly metabolized by the liver such that the half-life is only about 20 minutes. Adjustments were made to the molecule to increase its time of action, the most popular being testosterone enthanate.</p><p>Decreased production, leading to a hypogonadal state, occurs with increasing age as well as with injury or infection, especially severe trauma and chronic illness such as HIV infection and chronic wounds.</p><p>A hypogonadal state is seen in many patient populations, including those in an acute severe injury state, those with infection or more chronic states such as aging, and those with chronic obstructive pulmonary disease and other chronic illnesses.<xref ref-type="bibr" rid="B71">71</xref>,<xref ref-type="bibr" rid="B72">72</xref>,<xref ref-type="bibr" rid="B76">76</xref></p></sec><sec id="S5-2"><title>Clinical results of testosterone use</title><p>Testosterone administration is used mainly to correct a hypogonadal state, while testosterone analogs, which have a much greater anabolic activity, are used to increase anabolism (Table <xref ref-type="table" rid="T14">14</xref>).<xref ref-type="bibr" rid="B70">70</xref>–<xref ref-type="bibr" rid="B74">74</xref>,<xref ref-type="bibr" rid="B76">76</xref>–<xref ref-type="bibr" rid="B78">78</xref> Clinical studies have demonstrated a significant increase in net protein synthesis, especially in muscle and skin, with high doses of testosterone delivered parenterally.<xref ref-type="bibr" rid="B74">74</xref>,<xref ref-type="bibr" rid="B75">75</xref></p></sec><sec id="S5-3"><title>Wound healing effect</title><p>It is clear that testosterone is needed for the wound healing process since decreased levels impede healing.<xref ref-type="bibr" rid="B57">57</xref>,<xref ref-type="bibr" rid="B79">79</xref>,<xref ref-type="bibr" rid="B80">80</xref> Adequate testosterone levels are required for IGF-1 production, IGF-1 being a wound healing agent. However, there is no good data that an increase in testosterone levels above normal improves wound healing. This is not the case with a number of anabolic steroids that have been shown to increase the rate of wound healing even in the absence of hypogonadism. These agents will be discussed next.</p></sec><sec id="S5-4"><title>Complications</title><p>The major complications are the androgenic side effects. Some fluid retention has been reported with high doses. A decrease in high-density lipoproteins has also been reported with the use of large doses.</p></sec><sec id="S5-5"><title>Summary</title><p>Testosterone is a necessary androgen for maintaining lean mass and wound healing. A deficiency leads to catabolism and impaired healing. The use of large doses exogenously increases net protein synthesis, but a direct effect on wound healing has not yet been demonstrated.</p></sec></sec><sec id="S6"><title>ANABOLIC STEROIDS</title><sec id="S6-1"><title>Actions</title><p><italic>Anabolic steroids</italic> refer to the class of drugs produced by modification of testosterone.<xref ref-type="bibr" rid="B68">68</xref>–<xref ref-type="bibr" rid="B70">70</xref>,<xref ref-type="bibr" rid="B76">76</xref>,<xref ref-type="bibr" rid="B77">77</xref> These drugs were developed in order to take clinical advantage of the anabolic effects of testosterone while decreasing androgenic side effects of the naturally occurring molecule. Modifications in the steroid ring were made because of the short half-life of testosterone and its masculinizing properties. Modifications included a 17α methyl derivative for oral use and a 17β ester configuration for parenteral use. These changes markedly increased its half-life and decreased its androgenic properties (Table <xref ref-type="table" rid="T15">15</xref>).</p><p>The mechanisms of action of testosterone analogs are also through activation of the androgenic receptors, which are found in highest concentration in myocytes and skin fibroblasts. Some populations of epithelial cells also contain these receptors. Androgenic receptors were first isolated in the 1960s.<xref ref-type="bibr" rid="B68">68</xref>–<xref ref-type="bibr" rid="B70">70</xref></p><p>Stimulation leads to a decrease in efflux of amino acids and an increase in influx into the cell. Activation of intracellular DNA and DNA polymerase also occurs with androgenic receptor stimulation. A decrease in fat mass is also seen because of the preferential use of fat for fuel. There are no metabolic effects on glucose production.</p><p>All anabolic steroids increase overall protein synthesis and new tissue formation, evidenced by an increase in skin thickness and muscle formation<xref ref-type="bibr" rid="B71">71</xref>,<xref ref-type="bibr" rid="B72">72</xref> All these agents also have anticatabolic activity, decreasing the protein degradation caused by cortisol and other catabolic stimuli.<xref ref-type="bibr" rid="B73">73</xref></p><p>In addition, all anabolic steroids also have androgenic or masculinizing effects.</p><p>The quality of a testosterone analog is determined by the ratio of androgenic to anabolic activity, the lower the better. A low value indicates very little masculinizing effects compared to a very potent anabolic effect (Table <xref ref-type="table" rid="T16">16</xref>).</p><p>The anabolic steroid oxandrolone also happens to have the greatest anabolic and least androgenic side effects in the class of anabolic steroids.<xref ref-type="bibr" rid="B78">78</xref> It is the only steroid in which a carbon atom within the phenanthrene nucleus has been replaced by another element, namely oxygen. In addition, oxandrolone is cleared by the kidney and not the liver, so hepatotoxicity is rare.</p><p>Oxandrolone is the only approved anabolic steroid for restoration of lost body weight and lost lean mass.</p></sec><sec id="S6-2"><title>Clinical result of anabolic steroids use (effect on lean body mass)</title><p>Most of the recent studies on anabolic steroids and lean body mass have used the anabolic steroid oxandrolone (Table <xref ref-type="table" rid="T17">17</xref>).<xref ref-type="bibr" rid="B57">57</xref>,<xref ref-type="bibr" rid="B74">74</xref>,<xref ref-type="bibr" rid="B75">75</xref>,<xref ref-type="bibr" rid="B79">79</xref> Oxandrolone is a 17β-hydroxy-17α-methyl ester of testosterone and is cleared primarily by the kidney. Hepatotoxicity is minimal, even at doses higher than the 20 mg/d recommended by the Food and Drug Administration. Oxandrolone has potent anabolic activity, being up to 13 times that of methyltestosterone. In addition, its androgenic effect is considerably less than that of testosterone, minimizing this complication common to other testosterone derivatives. The increased anabolic activity and decreased androgenic (masculinizing) activity markedly increases its clinical value. Oxandrolone is given orally, with 99% bioavailability. It is protein-bound in plasma with a biologic life of 9 hours.</p><p>The anabolic steroids, especially oxandrolone, have been successfully used in the trauma and burn patient population to decrease lean mass loss in the acute phase of injury as well as more rapidly restore the lost lean mass in the recovery phase.<xref ref-type="bibr" rid="B43">43</xref>–<xref ref-type="bibr" rid="B46">46</xref>,<xref ref-type="bibr" rid="B57">57</xref>,<xref ref-type="bibr" rid="B74">74</xref>,<xref ref-type="bibr" rid="B75">75</xref>,<xref ref-type="bibr" rid="B79">79</xref>,<xref ref-type="bibr" rid="B81">81</xref> A significant attenuation of catabolism and increase in lean mass have also been reported in those with HIV infection, in the chronic obstructive pulomonary disease population, and in the spinal cord injury population.<xref ref-type="bibr" rid="B79">79</xref>,<xref ref-type="bibr" rid="B80">80</xref>,<xref ref-type="bibr" rid="B82">82</xref> There are several studies demonstrating an increase in the healing of chronic wounds. However, significant lean mass gains were also present.</p><p>It is important to point out that in all of the clinical trials where lean mass gains were reported, a high-protein diet was used. In most studies, a daily protein intake of 1.2 to 1.5 g/kg was used, 0.8 g/kg being the recommended daily intake in a healthy adult.</p></sec><sec id="S6-3"><title>Wound healing properties</title><p>The effects of anabolic steroids on wound healing appear to be, in large part, due to a general stimulation of overall anabolic activity. However, there is also increasing evidence of a direct stimulation of all phases of wound healing by these agents (Tables <xref ref-type="table" rid="T18">18</xref> and <xref ref-type="table" rid="T19">19</xref>).<xref ref-type="bibr" rid="B83">83</xref>–<xref ref-type="bibr" rid="B87">87</xref></p><p>Falanga et al<xref ref-type="bibr" rid="B87">87</xref> reported a stimulation of collagen synthesis with the anabolic steroid stanazol. Erlich et al<xref ref-type="bibr" rid="B88">88</xref> reported a 10-fold increase in the messenger RNA for collagen synthesis in a human fibroblast culture with oxandrolone. Tenenbaum et al<xref ref-type="bibr" rid="B64">64</xref> reported increased synthesis of bone, collagen, matrix, and epidermis in a wound of the oral cavity stimulated with oxandrolone. Demling<xref ref-type="bibr" rid="B79">79</xref> reported a marked increase in healing of a cutaneous wound in rats treated with oxandrolone compared to controls. A 50% increase in wound collagen as well as a doubling of tensile strength was noted at 3 weeks with oxandrolone. Histology also revealed more densely packed collagen with more fibroblasts and mononuclear cells. Anabolic steroids have also been shown to release the transforming growth factor beta by human fibroblasts. The mechanism of improved wound healing with the use of anabolic steroids is not yet defined. Stimulation of androgenic receptors on wound fibroblasts may well lead to a local release of growth factors.</p></sec><sec id="S6-4"><title>Complications</title><p>In addition to androgenic activity, a number of potential side effects exist for this class of drugs. Some fluid retention will occur initially but is usually transient.<xref ref-type="bibr" rid="B57">57</xref>,<xref ref-type="bibr" rid="B71">71</xref>–<xref ref-type="bibr" rid="B73">73</xref>,<xref ref-type="bibr" rid="B77">77</xref>,<xref ref-type="bibr" rid="B78">78</xref> Liver toxicity has been reported, ranging from a transient increase in aminotransferases to jaundice, liver failure, and rarely a liver tumor.<xref ref-type="bibr" rid="B89">89</xref> The potential for liver change varies among anabolic steroids.<xref ref-type="bibr" rid="B89">89</xref> Oxandrolone appears to be the safest. A recent 1-year study in elderly men given oxandrolone demonstrated only transient increases in aminotransferases.</p><p>A change in the lipid profile has been reported.<xref ref-type="bibr" rid="B87">87</xref> Several studies have demonstrated a decrease in high-density lipoproteins, potentially increasing the risk of atherosclerosis. The lipid response differs among the drugs in this class.<xref ref-type="bibr" rid="B90">90</xref></p><p>Anabolic steroids have been reported to increase the potency of coumadin, and coumadin dose often has to be decreased. Finally, this class of drugs is contraindicated in patients with prostate cancer as this tumor is stimulated by androgenic receptors.<xref ref-type="bibr" rid="B77">77</xref>,<xref ref-type="bibr" rid="B78">78</xref></p></sec><sec id="S6-5"><title>Summary</title><p>Anabolic steroids are analogs of testosterone modified to increase anabolic and decrease androgenic side effects. All these agents have been shown to increase lean body mass. In addition, there appears to be a direct wound healing effect. Side effects include liver dysfunction. Oxandrolone appears to be the most anabolic and the safest anabolic steroid.</p></sec></sec><sec id="S7"><title>SUMMARY AND FUTURE DIRECTION</title><p>Anabolic hormones are necessary to maintain the necessary protein synthesis required for maintaining lean body mass including wound healing, assuming the presence of adequate protein intake. However, endogenous levels of these hormones are decreased in acute and chronic illness and with increasing age, especially in the presence of a large wound.</p><p>The corroborating data for the use of anabolic hormones are excellent for more rapidly restoring protein synthesis and lean mass with lean mass loss. A high-calorie, high-protein diet is required.</p><p>Since lost lean mass, caused by the stress response, aging, and malnutrition, retards wound healing, the ideal use of these agents is to more effectively restore anabolic activity. All these agents can cause complications, specific to the hormone used, which needs to be appreciated.</p><p>There is also data that indicate a direct wound healing stimulating effect for some of these hormones. However, more clinical data needs to be obtained before a recommendation can be made to use anabolic hormones to increase the rate of wound healing in the absence of a catabolic state or in the absence of an existing lean mass loss. Oxandrolone is currently the agent of choice, unless contraindicated with the presence of prostate cancer, as this agent is safe, easy to administer, and does not have the metabolic side effects of HGH, IGF-1, and insulin.</p><p>Three areas of research and development are indicated at this point. The first area is to better define the effect of all of these anabolic hormones on the various stages of wound healing. This information is needed in order to determine the indications for the use of the available anabolic hormones. It is quite possible that combination therapy would be more beneficial if it is determined that these agents have different modes of action. The second area is the development of analogs of anabolic hormones that appear to have the most beneficial wound healing effects. The analogs would be developed to maximize wound-healing activity and minimize complications. The third area would be the development of a topical form of the anabolic hormones that demonstrate the most beneficial wound healing effects. The topical form would provide a direct wound healing benefit without the potential complications of systemic use.</p></sec>
Case Studies	<p><bold>Objective:</bold> The purpose of this report is to present the results of a preliminary treatment regimen for hypertrophic scars combining topical 2% salicylic acid cream (Avosil) with an overlay of hydrogel dressing (Avogel). <bold>Methods:</bold> The study group consisted of 3 patients with symptomatic hypertrophic scars: 2 presternal and 1 on the inner thigh. Scars were divided into 3 equal-size areas: (1) untreated control, (2) hydrogel alone, and (3) 2% salicylic acid with hydrogel cover. Treatments were applied every 8 to 12 hours and a Velcro appliance was employed to cover the area during treatment. The total length of treatment was 60 days. <bold>Results:</bold> At the end of the 60-day treatment protocol, the area treated with 2% salicylic acid and hydrogel was asymptomatic. In contrast, the hydrogel-treated and untreated control areas remained erythematous and symptomatic for burning pain and pruritis. <bold>Conclusion:</bold> This small study suggests the efficacy of combined salicylic acid and hydrogel therapy in the treatment of hypertrophic scars. More extensive studies of scar treatment with salicylic acid and hydrogel are needed. These studies must be larger in scope to carefully document the spectrum of patient responses and should include methods for evaluating alterations in the levels of different inflammatory mediators.</p>	<contrib contrib-type="author"><name><surname>Danielson</surname><given-names>John R.</given-names></name><degrees>MD, PhD</degrees><xref rid="aff1" ref-type="aff">a</xref><xref ref-type="corresp" rid="cor1"/></contrib><contrib contrib-type="author"><name><surname>Walter</surname><given-names>Robert J.</given-names></name><degrees>PhD</degrees><xref rid="aff2" ref-type="aff">b</xref></contrib>	Journal of Burns and Wounds	<p>Thirty years ago, Larson et al observed that elastic pressure wrap dressings were effective in reducing scar hypertrophy in healing burn scars.<xref ref-type="bibr" rid="B1">1</xref> Since that report, 20 to 24 mm Hg pressure garments have become the mainstay of scar prevention. The mechanism of action of pressure dressings is uncertain, particularly in view of the fact that they remain effective following several weeks of daily use despite a loss of elasticity and a diminution of pressure exertion. Biochemical assays of tissue healing beneath pressure garments show moderate decreases in wound metabolism and increased collagenase activity.<xref ref-type="bibr" rid="B2">2</xref> However, pressure garments are poorly tolerated because of the general discomfort and restriction of movement that accompanies their use.</p><p>In the past decade, another topical device, silicone gel sheeting, has been found to be useful clinically in controlling scar formation.<xref ref-type="bibr" rid="B3">3</xref>–<xref ref-type="bibr" rid="B5">5</xref> As with elastic garments, the mechanism of action has not been convincingly elucidated but postulated mechanisms include induction of scar hypoxia, increased hydration of the epidermis covering the scar,<xref ref-type="bibr" rid="B6">6</xref> and increased scar temperature.<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B8">8</xref> The effect does not depend on the composition of the gel sheeting used, and it is clear that release of silicone into the scar is not the mode of action. Recently, several reports have shown that hydrogel sheeting is just as effective as silicone but has fewer adverse side effects.<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B10">10</xref> Hydrogel sheeting has been approved by the Food and Drug Administration and is considered to be substantially equivalent to silicone for treatment of hypertrophic scars. Thus, the effect of the topical gel sheeting seems independent of its composition. Hydrogels have the added advantages of being useful as drug-delivery vehicles and of having higher heat capacities for buffering scar temperature.<xref ref-type="bibr" rid="B8">8</xref></p><p>Inflammation is critically important for healing and scarring. Dang et al<xref ref-type="bibr" rid="B11">11</xref> and O'Kane<xref ref-type="bibr" rid="B12">12</xref> recently reviewed the biology of scar formation. Both reviews underscore the importance of inflammation as the essential epigenetic stimulator of scarring. Therefore, we performed a pilot investigation of topical salicylic acid (2%) treatment in conjunction with hydrogel sheeting to determine whether there is an added beneficial effect from using a topical nonsteroidal anti-inflammatory drug in scar management. Here, we report our experience with the first patients.</p><sec sec-type="methods" id="S1"><title>METHODS</title><p>The first 2 cases were part of an ongoing pilot study at St. Mary's Hospital designed to document the effect of topically applied salicylic acid on symptomatic hypertrophic scars. Both patients had tender and pruritic hypertrophic scars following median sternotomy. For study inclusion, the length of the scar needed to be sufficient to accommodate a control area and 2 experimental areas (each of equal size). The control area was left untreated while one experimental area was covered with Avogel hydrogel (Avocet Polymer Technologies, Inc, Chicago, Ill) and a second area of equal length was treated with 2% salicylic acid in Eucerin (Beiersdor, Wilton, Conn) and covered with Avogel hydrogel. The study regimen called for treatments to be performed for 8 to 12 hours per day employing a Velcro appliance specially constructed to cover the scar during treatment and to prevent salicylic acid from spreading into the control areas. Exclusion criteria included the use of anti-inflammatory agents during the previous 2 weeks or known allergies to anti-inflammatory agents. The scars were assessed at 2-week intervals for 60 days. Direct measurements, histories, and digital photographs were obtained. An end stage was achieved when the experimental region of the scar was flat and asymptomatic and when erythema had receded.</p><p>The third patient was treated in a similar fashion by a Chicago area plastic surgeon for an 11-month-old hypertrophic scar from a burn injury on the inner thigh. This case is included to demonstrate that similar effects have been observed independently. The protocol used to treat this patient is described below.</p><sec id="S1-1"><title>Ethics</title><p>The Institutional Review Board of St. Mary's Hospital approved this study, and all participants gave informed consent. The procedures followed were in accordance with the ethical standards as set forth by the institution and with the Helsinki Declaration of 1975, as revised in 1983.</p></sec></sec><sec sec-type="results" id="S2"><title>RESULTS</title><p>All patients experienced significant symptomatic relief from tenderness and pain within days in the area of the scar treated with 2% salicylic acid and hydrogel. Some relief in symptoms was recorded in the area treated with hydrogel alone, but no symptomatic relief was experienced in the untreated area. Noticeable decreases in scar width and height were noted in all patients over the 60-day study period in the area treated with salicylic acid and hydrogel. Comparatively small changes in scar dimensions were noted with hydrogel alone, and no change was noted in the control region. Toward the end of the study period, scar erythema was greatly decreased in the area treated with hydrogel and salicylic acid in all patients.</p></sec><sec sec-type="cases" id="S3"><title>CASE REPORTS</title><sec id="S3-1"><title>Case #1</title><p>This patient is a 49-year-old man (Fig <xref ref-type="fig" rid="F1">1</xref>) who underwent open-heart surgery 6 months prior to presentation at our clinic. The patient's scar was 22 cm long, varied from 5 to 8 mm in width, and was elevated by 2 to 3 mm throughout. He complained of tenderness to touch and pain. After starting on the treatment protocol, all symptoms abated in the area treated with 2% salicylic acid and hydrogel within 2 weeks. During subsequent weeks, the scar height of the experimentally treated area decreased dramatically such that the scar became completely flattened to the level of the adjacent tissue. The control area remained unchanged, and the experimental area treated with hydrogel alone did not resolve as markedly as the area treated with 2% salicylic acid and hydrogel together. Furthermore, the degree of erythema decreased in the area treated with 2% salicylic acid and hydrogel, whereas the scar color remained unchanged in the area treated with hydrogel alone. At the end of the 60-day treatment protocol, the area treated with 2% salicylic acid and hydrogel was asymptomatic. In contrast, the hydrogel-treated and untreated control areas remained erythematous and symptomatic for burning pain and pruritis.</p></sec><sec id="S3-2"><title>Case #2</title><p>This patient is a 65-year-old woman (Fig <xref ref-type="fig" rid="F2">2</xref>) who underwent open-heart surgery 2½ years prior to presentation at our clinic. Before the study, she had applied vitamin E topically to the scar for a short period of time without effect. She complained of pain and tenderness, and the scar was very elevated and erythematous. The patient's scar was 14 cm long, varied from 4 to 6 mm in width, and was elevated by 3 to 4 mm. Symptomatic relief from pruritis and tenderness occurred within 48 hours of applying salicylic acid and hydrogel. One week after beginning the study regimen, she noted a marked decrease in symptoms and decreased erythema in this area. Regression in the size of the scar was less impressive, but symptomatic relief was significant. After 60 days of treatment, her symptoms in the area treated with 2% salicylic acid and hydrogel had completely abated. The hydrogel-treated and untreated control areas remained symptomatic.</p></sec><sec id="S3-3"><title>Case #3</title><p>This patient is a 42-year-old firefighter who suffered a flame burn injury to the skin on the left inner thigh 11 months before presentation. He was treated as an outpatient at one of the burn centers in Chicago where he received topical anti-microbials and dressing changes. The wound healed by epithelialization and hypertrophic scarring. The patient was referred to the University of Chicago Scar Clinic because the burn scar had become quite hypertrophic and was unresponsive to compression stocking therapy. A photograph of the scar taken at that time is shown in Figure <xref ref-type="fig" rid="F3">3</xref> (left). Using a visual analog scale of scar pruritus, the patient scored itch intensity at 6/10 (where 10 = unbearable itching) and itch frequency at 6/10 (where 10 = itching at all times). With time, the scar continued to progressively increase in thickness.</p><p>Therapy was initiated with Avogel hydrogel with 3% salicylic acid in a cream base applied to the scar each night and left for 8 to 10 hours. The use of compression garment therapy was discontinued. One month later, treatment regimen was adjusted such that 2% acetylsalicylic acid in Eucerin ointment (Beiersdor, Wilton, Conn) with hydrogel was used to achieve better symptomatic relief. On 7-week follow-up (Fig <xref ref-type="fig" rid="F3">3</xref>, right), the patient showed a decrease in scar redness and thickness and a 50% reduction in the intensity of pruritus. Scar symptoms were exacerbated by work activity, in which heavy clothing abraded the scar, but on 16-week follow-up, the scar manifested considerable involution and the intensity of itching was scored as 3/10. The patient continues to be treated with salicylic acid (left overnight for 8 hours).</p></sec></sec><sec sec-type="discussion" id="S4"><title>DISCUSSION</title><p>Limiting inflammation is paramount in the control of scar growth and scar-associated symptoms. The inflammatory response can be regulated at several different physiological levels. Three common ways that inflammation can be controlled are by inhibiting cytokine production via cyclooxygenase (COX) regulation, inhibiting histamine binding to its receptors, and inhibiting the NF-κB signal that upregulates inflammation. The most widely used anti-inflammatory agents belong to the broad category of nonsteroidal anti-inflammatory drugs. They inhibit either prostaglandin production or NF-κB generation or both. It is noteworthy that steroids also exhibit anti-inflammatory effects by inhibiting inflammatory response gene promoters NF-κB, AP-1 (activator protein-1), and NF-AT (nuclear factor of activated T lymphocytes).<xref ref-type="bibr" rid="B13">13</xref></p><p>The most common mode of prostaglandin synthesis blockade is inhibition of cell membrane–bound COX. There are 2 isoforms of COX expressed in human cells, COX-1 and COX-2.<xref ref-type="bibr" rid="B14">14</xref> COX-1 is expressed constitutively throughout the body, especially in stomach and kidneys. On the other hand, COX-2 is expressed constitutively in the brain and kidneys.<xref ref-type="bibr" rid="B15">15</xref> COX-2 is highly inducible at sites of inflammation, playing an important role in fibrosis. Many adverse side effects of COX inhibition are minimized by use of specific COX-2 inhibitors, making them agents of choice for prolonged usage. The therapeutic value of these agents in many rheumatologic diseases is well established. However, their potential value in management of hypertrophic scarring has been a more recent consideration.<xref ref-type="bibr" rid="B16">16</xref> NF-κB is a rapid response transcription factor that is involved in stress responses and is central to the inflammatory reaction.<xref ref-type="bibr" rid="B17">17</xref> NF-κB is involved in the upregulation of both cell membrane receptors to inflammatory peptides and the production of cytokines, chemokines, and growth factors. NF-κB activation can be inhibited by several different agents, including cyclosporine, tacrolimus, antioxidants, and salicylates (including aspirin). Salicylic acid inhibits NF-κB expression by blocking the dissociation of IκB (the inactivator of NF-κB) from NF-κB in the cytoplasm and thus decreases the amount of inflammation that occurs.<xref ref-type="bibr" rid="B17">17</xref> At concentrations of 2% to 5%, salicylates are commonly used to control skin inflammation and are routinely used in over-the-counter acne remedies.</p><p>Antihistamines are commonly used only to control the symptoms of scar pruritus. However, they have other important effects that may function to reduce scarring. Antihistamines, particularly H<sub>1</sub> blockers, inhibit the inflammatory response, resulting in reduced scar formation and reduced discomfort. Patients scratch the inflamed scar less frequently, which probably reduces scar growth rate. Finally, antihistamines are well known to inhibit collagen synthesis.<xref ref-type="bibr" rid="B18">18</xref> Benadryl and Atarax are the most commonly used antihistamines for scar management. In the past few years, we have preferred the use of long-acting, nondrowsy formulations such as loratadine (Claritin; Schering, Kenilworth, NJ) or fexofenadine (Allegra; Aventis, Kansas City, Mo), which have the advantages of sustained action and fewer central nervous system side effects. In recent studies, topically applied aspirin has been found to decrease histamine-induced wheal and flare reactions.<xref ref-type="bibr" rid="B19">19</xref> However, topically applied salicylic compounds did not diminish serotonin-induced scratching behavior in rats.<xref ref-type="bibr" rid="B20">20</xref></p><p>From the above, it is clear that more extensive studies of scar treatment with salicylic acid and hydrogel are needed. These studies must be larger in scope to carefully document the spectrum of patient responses and should include methods for evaluating alterations in the levels of different inflammatory mediators.</p></sec>
International Journal of Integrated Care in PubMed Central	Could not extract abstract	<contrib contrib-type="author"><name><surname>Schrijvers</surname><given-names>Guus</given-names></name><role>Editor-in-Chief of IJIC</role><aff>Head of the Julius Centre of the University of Utrecht Medical Centre</aff></contrib><contrib contrib-type="author"><name><surname>Manten</surname><given-names>Erika</given-names></name><role>Managing editor IJIC</role></contrib>	International Journal of Integrated Care	<p>We are very proud to announce that starting June 2006, articles, editorials and book reviews published in this journal can also be found in the database of PubMed Central.</p><p>In addition, abstracts of the articles, with direct links to full text, can be found in the PubMed database.</p><p>PubMed Central (PMC) is an electronic archive of full-text journal articles, offering free access to its contents. PMC contains over half a million articles, most of which (like IJIC) have a corresponding entry in PubMed.</p><p>One of the aims of PMC is to improve access to biomedical information for health professionals, researchers and the public. Participation in PMC will increase the visibility and impact of our journal.</p><p>By depositing our content in PubMed Central we also get the benefit of a permanent and freely accessible archive, managed by the United States National Library of Medicine (NLM).</p><p>Furthermore, the integration of our full text with PubMed and the numerous other databases in NCBI's<xref ref-type="fn" rid="fn1">1</xref> Entrez system will increase our exposure.</p><p>We feel that this is an important moment in the development of our journal. Integrated care is gaining importance in the field of healthcare. We have worked hard over the past five years to create a quality, open access journal. Being accepted by PMC is a wonderful acknowledgement and proof that IJIC has achieved a prominent place among health services research journals.</p><p>For further information regarding our journal please visit <ext-link ext-link-type="uri" xlink:href="http://www.ijic.org">http://www.ijic.org</ext-link> or contact the managing editor at <email>[email protected]</email>.</p><p content-type="signature">Guus Schrijvers</p><p content-type="signature-affiliation">Editor-in-Chief of IJIC Head of the Julius Centre of the University of Utrecht Medical Centre</p><p content-type="signature">Erika Manten</p><p content-type="signature-affiliation">Managing editor IJIC</p>
Brugada-Like Electrocardiographic Pattern	Could not extract abstract	<contrib contrib-type="author"><name><surname>Francis</surname><given-names>Johnson</given-names></name><degrees>MD, DM</degrees><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Antzelevitch</surname><given-names>Charles</given-names></name><degrees>PhD, FACC</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<p>The Brugada syndrome is characterized by a ST-segment elevation in the right precordial leads associated with right bundle branch block (RBBB) pattern and a propensity for life-threatening ventricular arrhythmias in the absence of structural heart disease [<xref ref-type="bibr" rid="R1">1</xref>]. Mutations in a cardiac sodium channel gene have been linked to this syndrome [<xref ref-type="bibr" rid="R2">2</xref>].</p><p> The mechanism underlying the RBBB and ST-segment elevation in right precordial leads in patients with the Brugada syndrome is thought to be an outward shift of the ionic currents during early repolarization causing a marked accentuation of the action potential notch in right ventricular epicardial but not endocardial cells. The outward shift of current ultimately leads to loss of the action potential dome causing marked abbreviation of the action potential in the right ventricular epicardial cells [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R5">5</xref>].</p><p> Local pressure applied to the right ventricular wall has also been reported to induce an ECG pattern similar to the Brugada syndrome. Tarin et al reported a patient with a mediastinal tumor and electrocardiographic findings similar to those described in the Brugada syndrome. This ECG pattern disappeared after tumor removal, thus suggesting that it was probably caused by compression of the right ventricular outflow tract by the mass [<xref ref-type="bibr" rid="R6">6</xref>]. Another case of pericardial fluid and "tumour" compressing the right ventricle with Brugada-like ECG pattern in a patient with rheumatoid arthritis has also been reported [<xref ref-type="bibr" rid="R7">7</xref>] During surgery the "tumour" was found to be organised haemopericardium. After the surgery the patient was well and had a normal ECG.</p><p> Nakazato et al describes a similar situation in this issue of the journal [<xref ref-type="bibr" rid="R8">8</xref>]. Compression of the right ventricular outflow tract by an abnormal infective mass, with/without focal pericardial inflammation was thought to be the mechanism of Brugada-like ST elevation in their patient.</p><p> The ability of local pressure to give rise to an ST segment elevation has been demonstrated experimentally in the arterially perfused right ventricular wedge preparation (Antzelevitch & Dumaine, 2002) [<xref ref-type="bibr" rid="R9">9</xref>]. Focal pressure was shown to cause loss of the action potential dome at some right epicardial sites but not others. The potential for this mechanism to give rise to closely coupled phase 2 reentrant extrasystoles and VT was also demonstrated in this experimental study.</p><p>Several other instances of Brugada-like patterns have been reported. Ortega-Carnicer et al [<xref ref-type="bibr" rid="R10">10</xref>] noted transient Brugada-type electrocardiographic abnormalities in renal failure which was reversed by dialysis. They reported a patient with a previous history of epilepsy treated with psychotropic drugs (with a sodium channel blocking effect) and chronic renal failure on haemodialysis who developed hyperkalaemia and ECG findings resembling Brugada syndrome. These ECG changes disappeared after haemodialysis when the potassium became normal. They concluded that hyperkalaemia along with cardiac membrane active drugs may cause ECG changes mimicking the Brugada syndrome.</p><p> Transient Brugada pattern has been observed repeatedly after recreational use of cocaine [<xref ref-type="bibr" rid="R11">11</xref>]. Intravenous administration of procainamide and subsequent intravenous propranolol followed by noradrenaline failed to reproduce the Brugada in this case. Electrophysiologic study performed in the presence of the Brugada ECG pattern showed no inducible arrhythmias. Yet another report describes a patient in whom a typical Brugada ECG pattern developed in relation to fever but could not be reproduced at normal temperature on administration of flecainide [<xref ref-type="bibr" rid="R12">12</xref>]. This case suggests that in some patients a Brugada-like ECG may only manifest during a febrile state.</p><p>All these reports of Brugada-like ECG pattern give us a better insight into the genesis of this pattern and possibly localize the abnormality to the right ventricular outflow tract. Heterogenous response of repolarization across the ventricular wall in the right ventricular outflow tract is thought to be responsible for accentuation of ST segment elevation in the right precordial leads [<xref ref-type="bibr" rid="R13">13</xref>].</p>
Surgical treatment of permanent atrial fibrillation during cardiac surgery using monopolar and bipolar radiofrequency ablation	Could not extract abstract	<contrib contrib-type="author"><name><surname>Geidel</surname><given-names>Stephan</given-names></name><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Ostermeyer</surname><given-names>Jorg</given-names></name><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Lass</surname><given-names>Michael</given-names></name><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Boczor</surname><given-names>Sigrid</given-names></name><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Kuck</surname><given-names>Karl-Heinz</given-names></name><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>In cardiac surgery strategies to treat permanent atrial fibrillation (pAF) effectively are of greatest interest because pAF is a frequent problem that substantially deteriorates the prognosis [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R3">3</xref>]. The data concerning the incidence of pAF following the classification of the ACC/AHA/ESC Practice Guidelines [<xref ref-type="bibr" rid="R4">4</xref>] that had been documented to persist for a period of at least 6 months among all patients scheduled for open heart surgery in our institution between February 2001 and April 2003 are outlined in <xref ref-type="table" rid="T1">Table 1</xref>.</p><p>Since Cox has demonstrated that AF can be definitely eradicated [<xref ref-type="bibr" rid="R5">5</xref>], efforts were made to achieve alternative and particularly less complex methods using surgical catheter ablation techniques during cardiac surgery. A broad spectrum of these methods has been presented recently [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R14">14</xref>]. Our experience with radiofrequency (RF) ablation procedures to treat pAF during open heart surgery is reported.</p></sec><sec sec-type="materials\|methods" id="s2"><title>Material and methods</title><p>Etiology of heart valve disease was assessed by clinical history, intraoperative valve examination and histological analysis. Patients hospitalized for heart valve surgery with pAF (≥6months) underwent combined intraoperative AF treatment with monopolar RF ablation. Since March 2003 a bipolar RF ablation procedure was performed in CABG cases with pAF. Patients with any other form of AF (intermittent or pAF <6months), emergency operation, severe reduced left ventricular function (EF≤25%), acute bacterial endocarditis, cachexia [Body Mass Index (BMI) ≤24], severe intracavitary thrombosis or extreme left atrial size were excluded (LA diameters of ≥72mm were assessed to be our limit for the procedure).</p><sec id="s2a"><title>Surgical procedure</title><sec id="s2a1"><title>Heart valve cases</title><p>The surgical procedure has been described detailed elsewhere [<xref ref-type="bibr" rid="R15">15</xref>]. To create endocardial RF ablation lesions two almost identical systems were used: either the Thermaline® device or (since 01/02) the Cobra® device (both Boston Scientific Corporation, San Jose, USA). Monopolar RF ablation was performed using 100W RF power for 120 sec, the local temperature was set at 70°C. The first lesion line completed the isolation of the right pulmonary veins (RPVs) from the inferior to the superior RPV using the left atriotomy. Isolation of the left pulmonary veins (LPVs) was performed with a semicircular ablation line close to the inferior, and another one around the superior LPV. These were connected by a transverse lesion across the posterior wall of the LA (<xref ref-type="fig" rid="F1">Figure 1A</xref>). Arrangements to avoid thermic esophageal injury were: (1) cachectic patients were excluded, (2) a dry compress was passed behind the LA before delivery of RF energy, (3) the transesophageal echocardiogram (TEE) probe was removed during the ablation procedure, (4) a flexible ablation probe was used and adapted to the tissue without pressure, (5) local temperature was set at 70°C and (6) monopolar RF ablation was performed precisely under direct view during conventional open heart valve surgery only.</p></sec><sec id="s2a2"><title>CABG cases</title><p>The bipolar AtriCure device (AtriCure Inc., Cincinnati, USA) was used for an almost identical lesion pattern (<xref ref-type="fig" rid="F1">Figure 1B</xref>). The device consists of a hand piece, a footswitch, connecting cables and an ablation and sensing unit (ASU) that (1) delivers RF energy while simultaneously measuring the tissue conductance and (2) uses a temperature sensing mechanism (range of 45 - 55°C). During ablation the tissue is impacted between two jaws of the hand piece and energy is delivered by footswitch. The ablation was finished when the ASU monitor indicated that the tissue conductance was at least 3 sec below 2.5 Millisiemens. After start of cardiopulmonary bypass isolation of the RPVs and the LPVs was performed by grasping the adjacent atrial tissue. Then a purse-string suture with a tourniquet was set at the posterior wall of the LA. The distal jaw was inserted through a small incision in direction of the LPVs and RF ablation was performed after clamp-closure, then the distal jaw was inserted in direction of the RPVs and the connection line was completed (<xref ref-type="fig" rid="F1">Figure 1B</xref>).</p></sec></sec><sec id="s2b"><title>Perioperative management, follow-up and statistical analysis</title><p>The perioperative management has been described detailed elsewhere [<xref ref-type="bibr" rid="R15">15</xref>]. Standard 12-lead electrocardiogram (ECG) and transthoracic echocardiogram (TTE) were routinely performed on admission (evaluation of LA diameter) and before discharge. Administration of amiodarone was started before end of cardiopulmonary bypass and given for 3 months after surgery. Early recurrence of AF was DC cardioverted after saturation with amiodarone. Patients with CABG, mitral valve (MV) repair or bioprosthesis got cumarine for 3 months, patients with mechanical valves lifelong anticoagulation. All patients were restudied 3, 6, 9, 12, 18 and 24 months after surgery by standard 12-lead ECG and clinical examination. Quantitative preoperative and operative data were normally distributed and described by arithmetic mean ± standard deviation; qualitative distributed data were presented as absolute frequencies. For pAF and sinus rhythm (SR) the relative frequency among all patients and some subgroups were calculated. Qualitative characteristics were compared using the exact Fisher Chi-Square-Test. All p-values were two-tailed and interpreted nominal that is not adjusted for multiple comparisons. P-values <0.05 were considered to be statistically significant. Analysis was performed with SPSS for Windows 11.5.1.</p></sec></sec><sec id="s3"><title>Results</title><sec id="s3a"><title>Heart valve cases</title><p>Fifty-nine patients underwent surgical pAF ablation procedures associated with primary valve operations (<xref ref-type="table" rid="T2">Table 2</xref>). From these patients 45 had MV, 13 AV and 1 AV+MV surgery; 13 were excluded. The main group suffered from rheumatic heart valve disease (39/59). Mean LA diameter was 55.6±7.1mm, 32 patients had a small (<56mm) and 27 patients a large LA (≥56mm).There were two cases of hospital mortality (3.4%; 1 cardiac, 1 non-cardiac). Mean follow-up time up to April 2003 was 14.1 ± 9.9 months (<xref ref-type="table" rid="T3">Table 3</xref>). One-year-survival was 91.7%; late mortality was related to cancer disease (n=2) and sudden death (n=1). At late follow-up approximately 75% of all patients and almost 90% of those with a preoperative LA-diameter of <56mm were in stable SR. Early postoperative recurrence of AF during the first 3 months (32/59) was DC cardioverted in 13 of 25 cases, 6 turned to SR spontaneously later. Preoperative duration of AF was not predictive for long term results.</p></sec><sec id="s3b"><title>CABG cases</title><p>In 3 bipolar RF ablation cases (<xref ref-type="table" rid="T2">Table 2</xref>) no severe complication occurred. At end of surgery 2 patients were externally paced in DDD-mode, one was in SR. Early postoperative return of AF was DC cardioverted in 1 of 2 cases; 2 of 3 patients left the hospital in SR.</p></sec></sec><sec id="s4"><title>Discussion</title><sec id="s4a"><title>Rationale of surgical practices to treat pAF</title><p>Reliable and effective treatment strategies for pAF are of greatest interest in cardiac surgery because pAF causes two times higher rates of death, five times higher risk of stroke, reduced cardiac output and the need of systemic anticoagulation with the danger of bleeding [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R3">3</xref>]. The incidence among patients scheduled for open heart surgery is particularly high among heart valve cases and of special importance in MV disease [<xref ref-type="bibr" rid="R15">15</xref>]. That is why most of the recently reported surgical AF ablation data were observed following heart valve (particularly MV) surgery [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R14">14</xref>]. The permanent form of AF is one of two AF-types (permanent and intermittent) and normally the final mode in which non-permanent AF cumulates [<xref ref-type="bibr" rid="R4">4</xref>]: patients with pAF have atrial fibrillation all of the time without any episode of SR. The variant clinical presentations of permanent and intermittent AF have unfortunately caused heterogeneous nomenclatures in different countries with currently used misleading terms (chronic, paroxysmal or persistent AF) and the resulting difficulty to compare the success-rates of different groups.</p><p>During the past years mainly three different energy sources have been favoured for surgical practice to create atrial lesions to cure these AF-types; RF energy has been established for variant treatment strategies and became the most widely used energy source for AF surgery [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R12">12</xref>]. The use of microwave [<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R14">14</xref>] and cryoablation [<xref ref-type="bibr" rid="R6">6</xref>] has been described alternatively. In these investigations [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R12">12</xref>] RF energy was used without exception in monopolar fashion for endocardial or/and epicardial ablation techniques with more or less comparable and successful results. However, two theoretical deficiencies remain with the application of monopolar RF energy: (1) transmurality of the created lesions is not definitely guaranteed and (2) rare but fatal complications because of too deep lesions have occurred [<xref ref-type="bibr" rid="R16">16</xref>,<xref ref-type="bibr" rid="R17">17</xref>]. Experiences with possibly reliable bipolar RF ablation techniques that guarantee lesion transmurality and continuity and definitely bar extra-cardiac tissue injury are therefore of great interest.</p><p>Many surgeons who used ablation techniques in the past followed more or less closely the principles of the Maze procedure which are PV isolation, reduction of atrial size and block of re-entrant circuits by complex incisions [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R12">12</xref>]. An ideal lesion pattern should combine (1) slight invasiveness, (2) simplicity, (3) high reproducibility and (4) saving of time with excellent success rates in almost all cases. The question if such a lesion pattern is actually applied in practice must be answered negative. The mechanisms of AF initiation and maintenance obviously vary and are connected with individual electrophysiological and pathological atrial tissue changes in cases with AF and particularly pAF [<xref ref-type="bibr" rid="R18">18</xref>-<xref ref-type="bibr" rid="R20">20</xref>]. It is possible that individual patients with pAF probably need individual surgical ablation procedures. The relevant questions are: (1) does a lesion pattern exist which can be recommended as a basis and can be completed easily according to the individual pathological or electrophysiological substrate and (2) can patients be defined who need specific additional ablation procedures?</p><p>Referring to this Haissaguerre has described an important pathophysiologic finding: he demonstrated that the initiation of AF originates from rapidly firing foci predominantly located inside the PVs [<xref ref-type="bibr" rid="R21">21</xref>]. According to that the concept was developed that isolation of the PVs creating transmural encircling RF ablation lesions around the LPVs and the RPVs should be a sufficient basis for surgical ablation procedures. To save LA function and to bar potential generation of foci the maze pattern of multiple incisions was reduced to a short connection line between both. During valve surgery RF ablation was performed from the endocardial side with direct view on the atrial tissue to guarantee continuity of the lesions. A new bipolar RF approach was recently used in primary CABG cases with pAF as an endocardial proceeding would have been of obvious disadvantage because of greater invasiveness and prolonged aortic cross clamping time in these cases. Amiodarone was normally given to reduce postoperative recurrence of AF [<xref ref-type="bibr" rid="R22">22</xref>].</p></sec><sec id="s4b"><title>Evaluation of the results</title><p>Our results conform to basic experiences of other research [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R23">23</xref>,<xref ref-type="bibr" rid="R24">24</xref>]: AF wavelets sustained by foci located inside the PVs were blocked by the created lesions, the described antiarrhythmic protection supported SR [<xref ref-type="bibr" rid="R22">22</xref>] during the unstable initial stage, which was approximately 3 months. Early recurrence of AF after ablation surgery obviously occurred due to the fact that the refractory period of the atrium was still shortened. In the case of AF recurrence DC cardioversion was still recommended, so the influence on the long-term results must be further clarified. Our data indicate that the preoperative LA size is of significant concern for the success of the described method. However, it remains uncertain whether the LA size itself is the critical issue. It can be expected that rather cellular, structural morphologic and in parallel electrophysiological changes of the atrial tissues are more marked in cases with progressive enlargement and hypertrophy of the atria [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R20">20</xref>]; this electrical and anatomic atrial remodeling is supposed to be the reason for what has been decscribed as AF begats AF [<xref ref-type="bibr" rid="R25">25</xref>]. It must be conceded that an additional effect may be that in cases with small LA the encircling lesions possibly encompassed larger portions of the LA in comparison to its overall size than in patients with large LA; in that cases re-entrant circuits in the LA tissue could be interrupted more frequently. Besides LA reduction has been described to be of advantage for restoring SR in patients with chronic AF and large LA following MV surgery [<xref ref-type="bibr" rid="R26">26</xref>].</p><p>Even if our patients were approximately a decade older compared to patients of other studies [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R10">10</xref>] surgery was tolerated well, particularly no case of esophageal injury [<xref ref-type="bibr" rid="R16">16</xref>,<xref ref-type="bibr" rid="R17">17</xref>] was observed. Also confirmed was the increase of SR during the following months after surgery [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R11">11</xref>]. It can be suggested that a 3 months administration of amiodarone is appropriate but should be handled flexible. The described procedures are of slight invasiveness and easy to perform. The lesion pattern can be created with monopolar RF energy and with a bipolar approach as well. However, it will have to be shown, if the results with bipolar RF energy are as encouraging as the described data in valve patients. As for MV surgery the LA has to be opened anyhow we continue to recommend for the present an endocardial ablation technique for these patients.</p></sec><sec id="s4c"><title>Limitations</title><p>For rhythm evaluation only 12-lead ECG was used. We consider to complete the follow-up data by performing a 24-hour-ECG registry to assess the possibility of nonpermanent AF. The data were not evaluated under randomized conditions.</p></sec></sec><sec id="s5"><title>Conclusion</title><p>We suppose that the described concept provides a successful treatment of pAF in patients undergoing open heart surgery; it fulfils all demands of an effective and easy to handle method. It simplifies the treatment of pAF and can be recommended in patients undergoing cardiac surgery. The advantages compared to other techniques are: (1) the procedure is easy to practice and (2) atrial tissue trauma is extremely slight. Our data indicate that particularly in patients with small LA restoration of stable SR in 90% of the cases can be achieved.</p></sec>
The Lead Extractor's Toolbox: A Review Of Current Endovascular Pacemaker And ICD Lead Extraction Techniques	Could not extract abstract	<contrib contrib-type="author"><name><surname>Bracke</surname><given-names>FA</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>Recently introduced pacemaker leads float freely within the veins and myocardium. Later on, fibrous encapsulation of the lead develops [<xref ref-type="bibr" rid="R1">1</xref>]. These adhesions not only occur at the lead tip but are commonly found anywhere along the whole length of the lead at sites where the lead is in contact with the vein or the myocardium [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R4">4</xref>].</p><p>These adhesions hamper lead removal as tight scar tissue can withhold the leads during traction. This not only occurs at the level of the flings and tines of passive fixation leads but at any level of the lead body, especially at sites of unequal diameter for example electrodes and defibrillator coils. Further the lead tip is often larger than the lead body due to the fixation mechanism and adhering scar tissue and can become impacted on withdrawal in the narrow canal provided by the fibrous envelope.</p><p>Force applied to leads is limited by the tensile strength of the insulation and conductor coils of the leads. They may severe with forceful traction, and denuded indwelling lead fragments have a higher incidence of thrombo-fibrotic complications and may maintain infection [<xref ref-type="bibr" rid="R5">5</xref>-<xref ref-type="bibr" rid="R8">8</xref>]. Force is also limited by the impact of traction on the veins and myocardium. Unopposed traction can lead to invagination of the myocardium, myocardial rupture, arrhythmia, hypotension or avulsion of a tricuspid valve leaflet [<xref ref-type="bibr" rid="R9">9</xref>-<xref ref-type="bibr" rid="R13">13</xref>].</p><p>Therefore, additional tools have been developed to assist in freeing the lead body from the adhesions as well as the lead tip from the myocardium, to prevent laceration of the myocardium and to provide enough room for the lead to be withdrawn whilst preventing disintegration of the lead. We describe the technical aspects of current endovascular techniques, the results, the complications and shortly discuss the indications.</p></sec><sec id="s2"><title>Technique of extraction</title><sec id="s2a"><title>Traction</title><p>Direct traction without additional tools is the most basic technique for lead extraction. Rosenheck et al. added 5- 10 times rotation of the lead with simultaneous gentle traction [<xref ref-type="bibr" rid="R14">14</xref>]. Another modification consists of prolonged graded traction [<xref ref-type="bibr" rid="R15">15</xref>-<xref ref-type="bibr" rid="R17">17</xref>]. For this purpose, increasing weights are connected to the proximal end of the lead. They are guided over a pulley mounted on the bed of the patient or, to keep the patient ambulatory, fixed under tension to the skin using rubber bands and adhesive tape.</p></sec><sec id="s2b"><title>Locking stylets</title><p>To avoid disintegration of the lead during traction, a locking stylet is introduced into the central lumen of the lead. It consists of a straight non-expandable wire that can be locked into the coil close to the tip of the lead. The force exerted via a locking stylet is almost directly applied at the tip, bypassing most of the conductor and the insulation. Locking mechanisms differ between manufacturers and recently introduced devices can be unlocked and repositioned if necessary.</p><p>There are limitations to the use of a locking stylet. If the conductor is broken or distorted it is not possible to introduce the stylet. Excessive force can dislocate the stylet or the distal conductor coil can still unwind or even disconnect from the electrode. Similar as with direct traction it does not provide a solution for either the risk of invagination of the myocardium or for removing the bulbous tip through the fibrous sheaths along the body of the lead.</p></sec><sec id="s2c"><title>Conventional intravascular counter traction</title><p>To overcome the limitations of a locking stylet, two telescoping synthetic sheaths can be alternately advanced over the lead. The fibrous bindings within the veins or myocardium can be mechanically disrupted, and at the same time enough room is created to remove the lead [<xref ref-type="bibr" rid="R18">18</xref>-<xref ref-type="bibr" rid="R19">19</xref>]. It is necessary to use a locking stylet: the leads are often to fragile to withstand the traction necessary to align the sheaths with the lead and to counter the forces applied to advance the sheath.</p><p> Once the distal electrode is reached counter traction is applied: the larger bore outer sheath is positioned and held against the myocardium to prevent its inversion during traction on the locking stylet. The force is thus concentrated at a small area of the scar tissue without gross displacement of the myocardium.</p><p> Although counter traction prevents invagination of the myocardium, perforation of the myocardium is still possible. The lead tip may have been incorporated into the myocardium, inevitably leading to a perforation after dislocation of the tip. Further, the possibility of increasing the force using counter-traction can lacerate the myocardium especially in the thin-walled atrium [<xref ref-type="bibr" rid="R20">20</xref>-<xref ref-type="bibr" rid="R21">21</xref>].</p></sec><sec id="s2d"><title>Laser assisted extraction</title><p>A laser sheath replaces the inner sheath with laser assisted lead extraction [<xref ref-type="bibr" rid="R22">22</xref>-<xref ref-type="bibr" rid="R25">25</xref>]. The laser sheath consists of optic fibers spirally warped between the inner and outer tubing of the sheath (<xref ref-type="fig" rid="F1">Figure 1</xref>). At the tip of the device the fibers are arranged in a ring from which the pulsed laser light is emitted to ablate the tissue (<xref ref-type="fig" rid="F2">Figure 2</xref>). Lasing energy is delivered with a 308 nm XeCl excimer laser (Spectranetics CVX-300) which emits pulsed light at a maximum fluence of 60 mJ/mm2 and a 40 - 80 Hz repetition rate. The ablation mechanism combines photochemical destruction of cellular structures with explosive photo-thermal vaporization of cellular water, which creates transient micro-bubbles to mechanically disrupt the tissue. As the penetration depth of 308 nm light in vascular tissue is approximately 100 micron, it is completely absorbed in the tissue immediately in front of the tip. This results in an ablation depth, depending on the applied force, between 2 and 15 microns per pulse in the experimental setting [<xref ref-type="bibr" rid="R26">26</xref>]. Force exerted on the sheath increases the mechanical effect of the micro-bubbles entrapped beneath the tip of the device in creating microscopic tears. The ablation results in a shearing of the fibrous scar, often leaving a rim around the lead. It has to be understood that the blunt tip of the laser sheath is not suited for direct mechanical disruption of the fibrous scar and that applying more force than necessary to assure good contact with the tissue does not improve efficacy but increases the risk of complications.</p><p>As with conventional sheaths the outer sheath helps to align the laser sheath with the lead and facilitates its handling by reducing friction with the surrounding tissue. Lasing can only be applied until the distal electrode to prevent myocardial perforation. Therefore, counter-traction is still necessary to dislocate the tip. The synthetic outer sheath is used for counter-traction as the inner diameter of the laser sheath is often too small to accommodate the lead tip and its attachments. It has to be noted that the laser has no effect on the insulation of the leads. To accommodate for different sizes of leads 12, 14 and 16 F sheaths are available.</p></sec><sec id="s2e"><title>Electrosurgical sheaths</title><p>An emerging new technique is the bipolar electrosurgical dissection sheath. Two electrodes are mounted on the beveled tip of a synthetic sheath (similar to a conventional sheath) and radiofrequency energy is applied between them providing bipolar point dissection of scar tissue. Interestingly, directional dissection is theoretically possible by rotating the sheath to orientate the bipole. Similar as with a laser sheath, counter traction is used to dislocate the tip.</p></sec><sec id="s2f"><title>Transfemoral approach</title><p>All previous techniques used a superior approach from the lead insertion site, but with the transfemoral approach a long 16 F sheath is introduced via the femoral vein towards the right atrium. Then, a retriever is inserted through the sheath to grab and secure the lead as close to the tip as possible (<xref ref-type="fig" rid="F3">Figure 3</xref>). The lead (with the connector cut off) is pulled inferiorly by the retriever whilst the outer sheath is advanced over the doubled up lead (<xref ref-type="fig" rid="F4">Figure 4</xref>). The proximal part of the lead is pulled down through the fibrous envelope; the distal binding sites are disrupted by the advancing outer sheath. When the sheath reaches the distal electrode, counter traction is again applied. In comparison to the superior approach, the isodiametric lead body is often readily pulled down from the binding sites not hampered by a bulbous tip. Although no locking stylet is used to reinforce the lead, the shorter distance from the retriever to the tip decreases the change of elongating the lead. The technique is especially useful to remove severed and indwelling leads or as back-up after failed superior attempts [<xref ref-type="bibr" rid="R22">22</xref>]. Bongiorni et al. first pulled the proximal end of the lead down from the femoral vein, then retrieved the lead via the internal jugular vein.</p></sec></sec><sec sec-type="results" id="s3"><title>Results</title><p>The results and complications of different extraction techniques are summarized in <xref ref-type="table" rid="T1">Table 1</xref>. Although traction and graded traction for lead extraction have been reported on for more than 30 years only Rosenheck et al. recently reported results with the use of rotational forces [<xref ref-type="bibr" rid="R14">14</xref>]. Applied on 113 leads in 81 patients this method resulted in complete removal in 86 % of the leads. In 6 % leads dislocated from the myocardium but proximal adhesions impeded removal of the lead. In 8 %, traction was unsuccessful.</p><p>Locking stylets have been reported as a stand-alone extraction tool. Alt et al. reported their experience with the VascoExtor stylet (VascoMed GmbH, Weil am Rhein, Germany) in 150 leads in 105 patients, (110 ventricular leads, 40 atrial; passive fixation with tines in 109 leads). Complete removal was possible in 81 % of cases, partial removal in 12 %. Manolis et al. reported success with the same device in 24 out of 25 leads: in 81 % with the sole use of a stylet, in 19 % with additional tools [<xref ref-type="bibr" rid="R27">27</xref>]. Results of counter traction technique using conventional sheaths includes also the use of a femoral approach and are reported in the U.S. Lead Extraction Database [<xref ref-type="bibr" rid="R20">20</xref>]. Complete removal was achieved in 86.8 % and partial removal in 7.5 % of 2,195 leads. In a paper describing their experience with lead extraction between 1994 and 1996, Byrd et al. reported complete removal in 93 % and partial removal in 5 % of 3540 leads in 2338 patients [<xref ref-type="bibr" rid="R21">21</xref>].</p><p>Klug et al. reported results of a femoral approach alone with a Needle's Eye snare in 70 leads out of 82 leads extracted in 39 patients [<xref ref-type="bibr" rid="R28">28</xref>]. Eighty-seven percent of leads were successfully extracted, 4 % incomplete and 9 % failures.</p><p>In a European multi-center study laser sheath extraction was attempted of 179 leads in 149 patients (104 atrial, 57 ventricular, one superior vena cava ICD and 17 ventricular ICD leads) [<xref ref-type="bibr" rid="R29">29</xref>]. Complete extraction was achieved in 89.5% of the leads, 6% were partially extracted and 4.5% of the extractions failed. Three out of the 8 failures were completely removed by a femoral non-laser approach, 1 with an alternative superior approach and 1 with thoracotomy.</p><p>In a United States registry of 1684 patients undergoing laser sheath extraction of 2561 leads, complete success was achieved in 90 % and partial success in 3 % of leads [<xref ref-type="bibr" rid="R30">30</xref>]. In the Excl trial of lead extraction of 287 leads in 166 patients with bipolar electrosurgical sheaths 96% of leads was completely removed, 4 % partially removed and only one lead not removed (laser sheath as an adjunct was used in 2 % of cases) [<xref ref-type="bibr" rid="R31">31</xref>].</p><p>There is only one trial comparing extraction techniques in a randomized fashion. In the Plexus trial 465 leads were randomized between the laser sheath and conventional sheaths (all investigators had experience with the conventional technique) [<xref ref-type="bibr" rid="R23">23</xref>]. Crossover from non-laser to laser was allowed in case of failure of the non-laser approach. Initial attempts at extraction were successful in 94% of leads with the laser sheath against 64 % of leads with the conventional sheaths. However, there was a 33.5 % crossover from conventional techniques to laser. This reflected the greater predictability of laser assisted lead extraction as perceived by the investigators, hence the low threshold for abandoning conventional sheaths if success was not swiftly obtained. If no crossover was allowed one should have expected a success rate in the non-laser group comparable to the published data on conventional extraction [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R21">21</xref>].</p><p>Successful lead extraction is especially dependent on implant duration: Byrd et al. reported that the risk of failed or partial doubled for every 3 year of implant [<xref ref-type="bibr" rid="R21">21</xref>]. Further, extraction is more likely to be successful with increasing physician's experience, atrial leads and infected leads [<xref ref-type="bibr" rid="R32">32</xref>].</p></sec><sec id="s4"><title>Complications</title><p>Using only traction, Rosenheck et al. reported only 1 out of 89 patients with a small pericardial effusion [<xref ref-type="bibr" rid="R14">14</xref>]. With conventional sheaths or a femoral workstation, major complications occurred in the U.S. Lead Extraction Database in 2.5 % of the 1,299 patients (hemopericardium or tamponade 1.2 %, hemothorax 0.5 % and pulmonary embolism 0.2 %, death 0.6 %) and in the experience of Byrd et al. in 1.4 % of 2338 patients (thoracotomy in 14 patients, pericardial drainage in 11 patients, transfusions in 4 pts. and 1 death) [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R21">21</xref>].</p><p>The use of femoral workstation was accompanied by 2 deaths and transient limb ischaemia in 1 patient [<xref ref-type="bibr" rid="R28">28</xref>]. In the 149 patients of the European laser registry, complications included ventricular perforation in one patient; two other perforations were related to the reimplantation of leads and required surgery [<xref ref-type="bibr" rid="R29">29</xref>]. There were no fatal complications. In the United States laser registry major complications occurred in 1.9 % of patients, mainly tamponade and haemothorax resulting in the demise of 12 patients (0.8%) [<xref ref-type="bibr" rid="R30">30</xref>]. With electrosurgical sheaths, the first results of the Excl trial in 166 patients reported tamponade in 3 patients, haemothorax and AV fistula both in 1 patient.</p><p>A direct comparison of complications between techniques is only available from the Plexus trial [<xref ref-type="bibr" rid="R23">23</xref>]. Tamponade or hemothorax occurred only 1 in 3 patients randomized to the laser group, one of whom died. Two patients in the laser group and 1 patient in the non-laser had venous thrombotic complications.</p><p>It is remarkable that the complication rates between endovascular extraction techniques are comparable. As all different extraction techniques except direct traction rely on countertraction it is not surprising that tamponade resulting from perforation after dislocation of the lead tip is common to all these techniques. A second common predilection site for perforation is the lower superior vena cava. As the vessel wall is sometimes part of the fibrous envelope surrounding the leads, it can be exposed to the mechanical, laser or radiofrequency energy of the sheaths. Electrosurgical sheaths have the possibility of directional application of energy which should avoid the risk of damage to the vessel wall but this has yet to be proven. A femoral approach has an advantage in this respect as the proximal isodiametric part of the lead is most often simply pulled down through the binding sites avoiding applying sheaths in the vena cava superior.</p><p>The risk of complications has been associated with physician's experience, number of leads extracted and female sex [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R30">30</xref>]. Although implant duration was not linked with complications in any of the cited papers, major complications reported for extraction of Accufix leads (Telectronics Pacing Systems Inc. Englewood, Co, USA) increased from 2 % at one-year implant duration to 8.3 % with an implant duration of more than 5 years as reported by the Accufix Resarch Institute (data available on <ext-link ext-link-type="uri" xlink:href="http://www.accufix.com"> www.accufix.com</ext-link>).</p></sec><sec sec-type="conclusions" id="s5"><title>Conclusion</title><p>In practice, probably no technique is sufficient by itself to address all lead extractions. The necessity to use tools to extract leads largely depends on the time from implant. Within six months from implant it is rare that traction alone will not suffice. However, from that time on extraction tools are necessary in an increasing number of procedures [<xref ref-type="bibr" rid="R32">32</xref>].</p><p>Powered sheaths, laser or electrosurgical, provide the best chance of extracting the entire lead, but have a risk of laceration of the veins along the proximal part of the lead. The most versatile technique is probably a femoral workstation and retriever system. It is the only technique suited to remove severed and indwelling leads and is very useful as back-up after a failed or stalled superior venous approach [<xref ref-type="bibr" rid="R22">22</xref>].</p><p>Although the complication rate of these extraction techniques seems low, their nature is life-threatening and hence extraction is only acceptable for indications that are vital or have a serious morbidity. Ergo, these procedures should be limited to the operation room with adequate cardio-surgical stand-by. The risk of emergency surgery itself is important to consider in advance. In some cases elective surgery is a preferable when risks of endocardial extraction are too high, even in the course of a procedure. There will always be a trade-off between perseverance to obtain successful extraction and complication rate.</p><p>All considerations mentioned result that in our opinion only infected pacing systems qualify for primary lead extraction [<xref ref-type="bibr" rid="R33">33</xref>]. In high-risk patients or in the absence of systemic infection alternative, albeit less efficient, treatments may be tried before lead extraction is performed. In case of non-functional or superfluous leads there is no proof that abandoning leads has a risk that justifies lead extraction [<xref ref-type="bibr" rid="R10">10</xref>] [<xref ref-type="bibr" rid="R34">34</xref>]. It is only defendable to extract non-functional leads that have a high chance of success without complications: an implant time less than one year, younger patients and no serious comorbidity.</p></sec>
Sinoatrial Reentry Tachycardia: A Review	Could not extract abstract	<contrib contrib-type="author"><name><surname>Simmers</surname><given-names>TA</given-names></name><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Sreeram</surname><given-names>N</given-names></name><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>History and Electrophysiology</title><p>The concept of reentry within the sinus node is by no means new. In their 1943 report, Barker and co-workers postulated that “…a circus rhythm could be accommodated in auricular muscle and in one of the specialized nodes at known rates of conduction and with cycle lengths such as occur in paroxysmal tachycardia” [<xref ref-type="bibr" rid="R1">1</xref>]. Lack of invasive electrophysiology at that time and subsequent failure to appreciate the heterogeneity of supraventricular arrhythmias, left their astute observation in the realm of conjecture. It was not until 1968 that Han, Malozzi and Moe finally demonstrated the existence of sinoatrial echoes; [<xref ref-type="bibr" rid="R2">2</xref>]. In a superfused isolated rabbit right atrial preparation, they examined the response to premature extrastimuli with an 18 electrode grid at the sinus node and surrounding atrial tissue. They found that critically timed extrastimuli led to early re-excitation of the atrium, supposedly due to sinoatrial reentry. Sequential microelectrode measurements of sinus node and atrial transmembrane potentials were performed in an attempt to provide a temporal and anatomical map of the observed phenomenon. Based on findings, the authors concluded that sinoatrial reciprocation was caused by entrance block at one site, slow conduction within the node with disparate refractoriness, and re-excitation of the atrium at the original area of entrance block. They went on to suggest that repetition of this phenomenon may conceivably form the basis for clinically relevant tachycardia. In vivo confirmation of sinoatrial echoes was inferred from work by Childers; [<xref ref-type="bibr" rid="R3">3</xref>] and Paulay; [<xref ref-type="bibr" rid="R4">4</xref>]. In the former study, programmed electrical stimulation was performed in dogs. Three responses to premature atrial stimuli were noted: complete interpolation (i.e. the subsequent sinus complex was on time), incomplete interpolation (i.e. the subsequent beat was delayed), and sinus echo. In the latter, the first post-extrastimulus complex was electrocardiographically the same as sinus, but earlier than expected. The authors attributed this to sinoatrial reentry, and went on to describe each of the three responses to premature atrial beats in a 70-year-old man in an elegant electrocardiographic deduction of underlying mechanisms. An in vivo canine model was also employed in the report by Paulay et al. While confirming Childers’ findings, Paulay offered two additional pieces of evidence implicating the sinus node/perinodal area: firstly, timing of the echo was independent of stimulation site (they performed pacing at the sinus node, Bachmann’s bundle, right and left atrial appendages and low atrial septum), and secondly echoes were eliminated by crushing the sinus node. As has long been the case for AV nodal reentry tachycardia, it is apparent that considerable discussion also revolves around the issue of whether sinoatrial echoes do or do not involve perinodal atrial tissue as well as the sinus node itself. In 1979, Allessie and Bonke published work suggesting confinement of the re-entrant circuit to the node per se [<xref ref-type="bibr" rid="R5">5</xref>]. The experimental set-up was similar to earlier work by Han [<xref ref-type="bibr" rid="R2">2</xref>], with the significant difference that resolution of the measuring electrodes was superior: 32 unipolar surface atrial electrograms were recorded during initiation of sinoatrial echoes as opposed to the 18 used by Han. In addition, up to 130 transmembrane sinus node potentials were recorded sequentially in cases of sustained reciprocation. These detailed observations led to the conclusion that sinus node reentry was confined to a circuitous pathway in an extremely small area (1-2mm) with low conduction velocities (2.5cm/s) within the node itself. It is of note that the same authors pointed out the inability to induce sustained arrhythmia in this model, and later concluded that sinoatrial reentry tachycardia (SART) may not be feasible without border zone atrial tissue. Given the limitations inherent to in vivo study in man, it may be fair to continue to use “sinus node reentry” and “sinoatrial reentry” synonymously for the time being. In the wake of in vivo studies in dogs, and reports of sinus node reentry in man by Paulay [<xref ref-type="bibr" rid="R6">6</xref>] and Childers [<xref ref-type="bibr" rid="R3">3</xref>], Narula first described sustained SART in two patients in 1974 [<xref ref-type="bibr" rid="R7">7</xref>]. Out of 300 patients undergoing electrophysiologic study, he observed sinus node reentry beats in 20, and sustained tachycardia in two. The criteria for the diagnosis of SART proposed by Narula are still valid: 1. atrial activation and P-wave morphology are the same or highly similar to sinus rhythm, with activation from high to low right atrium, 2. the arrhythmia is inducible with atrial extrastimuli at specific coupling intervals, independent of AV nodal conduction intervals and site of stimulation, and 3. the arrhythmia can be terminated by atrial stimuli. SART is thus, by definition, a paroxysmal arrhythmia.</p></sec><sec id="s2"><title>Clinical Presentation</title><p>Symptoms of SART vary widely. Patients may present with paroxysmal palpitations, dyspnoea, dizziness, (near-) syncope, chest discomfort and other symptoms [<xref ref-type="bibr" rid="R8">8</xref>]. Given the electrocardiographic similarity to sinus tachycardia and the on average lower heart rates than, for example, circus movement using a concealed bypass [<xref ref-type="bibr" rid="R9">9</xref>], a degree of under-diagnosis is likely. Patient’s symptoms can in fact be so reminiscent of anxiety disorders that a psychiatric diagnosis is initially entertained and referral to a cardiologist unnecessarily delayed. Again, electrocardiography is usually typical, with sinus rhythm morphology P waves and RP:PR ratio of greater than one. In one report, however, prolonged AV nodal conduction leading to superposition of the P wave on the preceding T wave or ST segment was shown to cause confusion with a diagnosis of AVNRT in 27% of a cohort of 65 patients undergoing EP study for symptomatic paroxysmal SVT [<xref ref-type="bibr" rid="R10">10</xref>]. In general, there are only two serious electrocardiographic differential diagnoses: inappropriate sinus tachycardia, and atrial reentry tachycardia arising near the sinus node [<xref ref-type="bibr" rid="R11">11</xref>]. The former differs primarily in its non-paroxysmal character, with an often elevated resting rate and gradual but excessive acceleration as a reaction to even mild exercise; the latter, also being a paroxysmal arrhythmia, can be extremely difficult to differentiate from sinoatrial reentry. Reaction to vagal manoeuvres and adenosine may provide clues to the correct diagnosis, as both tend to terminate sinoatrial but not intraatrial reentry tachycardia. Differentiation between SART, inappropriate sinus tachycardia and intraatrial reentry tachycardia is detailed in <xref ref-type="table" rid="T1">Table 1</xref>. Reported incidence of SART varies widely, although it is by any standards a relatively uncommon arrhythmia. A study by Wellens and co-workers demonstrated SART in only 1.8% of 379 patients undergoing diagnostic EP study [<xref ref-type="bibr" rid="R12">12</xref>]; however, the large proportion of patients included with VT and atrial flutter may cause underestimation of overall prevalence in the general population. At the other end of the spectrum, Gomes et al reported an incidence of 16.9% in a study mentioned earlier [<xref ref-type="bibr" rid="R10">10</xref>]; mean age of SART patients was 60 years and all but one had concomitant organic heart disease, which in this case may lead to overestimation of overall prevalence. The finding that SART in adults is associated with a relatively high prevalence of heart disease, be it hypertensive, ischemic or valvular, has been reported by several other authors [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R13">13</xref>-<xref ref-type="bibr" rid="R16">16</xref>], as much as 100% in the 7 patients reported by Wu et al [<xref ref-type="bibr" rid="R9">9</xref>]. Interestingly, although numbers are small this does not seem to hold true in the paediatric population. Probably the first case of SART reported in a child, an otherwise healthy 10-year-old, was published by Pahlajani, Miller and Serrato in 1975 [<xref ref-type="bibr" rid="R17">17</xref>]. The following year, Gillette reported SART in 5 of 35 children aged below 18 years undergoing diagnostic EP study for symptomatic SVT [<xref ref-type="bibr" rid="R18">18</xref>]; three had no organic heart disease, and two an ASD (of which one previously operated). Work by both Garson [<xref ref-type="bibr" rid="R19">19</xref>] and Blaufox [<xref ref-type="bibr" rid="R20">20</xref>] further underlines the relationship between recent cardiac surgery and SART in children, with 70% in the latter study having recently undergone corrective or palliative surgery mainly for hypoplastic left heart syndrome.</p></sec><sec id="s3"><title>Treatment</title><p>In contrast to inappropriate sinus tachycardia and atrial tachycardias, SART rarely responds well to β-blockers. Digoxin, calcium channel blockers such as verapamil and amiodarone are the drugs of choice [<xref ref-type="bibr" rid="R10">10</xref>]. In drug refractory cases, aversion to long-term pharmacotherapy or in the event of severe symptoms, a more curative invasive approach may be warranted. Surgical modalities are no longer preferred since the advent of endocardial catheter ablation, but are nonetheless interesting from an historical viewpoint. In 1984, Yee and co-workers described management of what was probably inappropriate sinus tachycardia by subtotal right atrial exclusion in a 27-year-old woman [<xref ref-type="bibr" rid="R21">21</xref>]. This method entailed the amputation of a large part of the right atrium from all surrounding structures, severing the sinus node artery, and subsequently suturing the excluded portion back. This resulted in a stable junctional rhythm and palliation of symptoms in the patient described. More recently, catheter ablation has been demonstrated to be a feasible method with which to isolate the offending sinus node [<xref ref-type="bibr" rid="R22">22</xref>,<xref ref-type="bibr" rid="R23">23</xref>]. Hendry et al attempted surgical treatment, again of inappropriate sinus tachycardia rather than SART, by simply excising the sinus node area of the right atrium [<xref ref-type="bibr" rid="R24">24</xref>]. Of the three patients described, only one was cured; the other two subsequently developed new arrhythmias. Conceptually much closer to present techniques of catheter ablation, Kerr and co-workers described epicardial cryoablation at the site of earliest atrial activation at the superior vena cava - right atrial junction in a 42-year-old woman with SART [<xref ref-type="bibr" rid="R25">25</xref>]. This terminated the tachycardia, followed by which a stable atrial escape was seen to originate 3cm inferior and posterolaterally to the ablation site. Additional cryoablation was required at a second surgery 4 days later due to early recurrence, after which the patient remained free of SART during 14 months follow-up. The literature on endocardial radiofrequency catheter ablation for SART is obviously more recent, and also limited to small numbers of cases. Kay [<xref ref-type="bibr" rid="R24">24</xref>], Sperry [<xref ref-type="bibr" rid="R15">15</xref>], Lesh [<xref ref-type="bibr" rid="R16">16</xref>] and Poty [<xref ref-type="bibr" rid="R26">26</xref>] all reported 100% acute success of ablation for SART at the site of earliest atrial activation, each in no more than 4 patients. Poty et al describe characteristics of successful ablation sites in slightly more detail, referring to fragmentation of local atrial electrograms as a marker of success. They also used unipolar electrograms for mapping purposes, in contrast to the bipolar mode described by other authors. The advantages of the unipolar recording mode for endocardial mapping have been described elsewhere in detail for a wide variety of arrhythmias [<xref ref-type="bibr" rid="R27">27</xref>-<xref ref-type="bibr" rid="R33">33</xref>]. In the event of fragmented, multiphasic bipolar electrograms as in the case of SART as reported by Poty and others [<xref ref-type="bibr" rid="R26">26</xref>,<xref ref-type="bibr" rid="R34">34</xref>-<xref ref-type="bibr" rid="R36">36</xref>], one of the principle advantages of the unipolar mode is reliable determination of local activation time. Sanders [<xref ref-type="bibr" rid="R34">34</xref>] and Ivanov [<xref ref-type="bibr" rid="R35">35</xref>] reported the two largest series of radiofrequency catheter ablation for SART, 10 patients each with a 100% success rate. Sanders et al reiterate the importance of ablation at the site of earliest atrial activation during tachycardia, but go further in identifying specific characteristics of successful versus unsuccessful sites. They found that atrial activation at 35ms or more before onset of the P-wave in the surface ECG and more than 20ms before high right atrium, and fragmented electrograms (they described an average of 87ms) were associated with successful ablation. A wide range of timing has been reported to represent "early" atrial activation, from -20ms [<xref ref-type="bibr" rid="R35">35</xref>] to -100ms [<xref ref-type="bibr" rid="R15">15</xref>]; however 25-35ms pre-P is usually cited, as in the Sanders report. The youngest patient to undergo ablation for SART to date was a 2-month-old infant reported by Simmers, Sreeram and Wittkampf. While the patient demonstrated all typical electrophysiologic findings of SART, there was markedly less prematurity of local activation relative to the P-wave during tachycarida than usually reported in adults [<xref ref-type="bibr" rid="R36">36</xref>]. It is conceivable that this difference was due to the fact that the patient weighed only 3000g. While complications of ablation for SART are rare, two issues deserve attention: firstly, right hemidiaphragmatic paralysis. Due to the close relation between the phrenic nerve and ablation site in some individuals, it is of the utmost importance to perform high output pacing (10mA or more) at any putative ablation site, to identify phrenic stimulation and thus risk of thermal injury on applying radiofrequency current. Secondly, ablation near the junction between right atrium and superior vena cava (SVC) may entail a risk of SVC stenosis. In a recent report on ablation for inappropriate sinus tachycardia by Leonelli and co-workers [<xref ref-type="bibr" rid="R37">37</xref>], 3 of the 35 patients developed SVC stricture. Whether this may also apply to SART is unclear; ablation for IST usually requires more ablation lesions than for SART, and most importantly all 3 cases reported by Leonelli had undergone pacemaker implantation or upgrade during the same procedure. <xref ref-type="fig" rid="F1">Figures 1</xref> and <xref ref-type="fig" rid="F2">2</xref> demonstrate typical findings during mapping and ablation for SART.</p></sec><sec sec-type="conclusions" id="s4"><title>Conclusion</title><p>Sinoatrial reentry tachycardia is a relatively uncommon arrhythmia in the literature, but may be prone to underdiagnosis due to electrocardiographic similarity to sinus tachycardia and misinterpretation of symptoms as psychosomatic. Awareness of the diagnosis, and hence differentiation of SART from inappropriate sinus tachycardia, atrial tachycardia or non-cardiac diagnoses paves the way for adequate therapy. While medication (digoxin, verapamil and amiodarone) may be successful or even desirable under certain conditions, radiofrequency catheter ablation offers curative therapy in most cases with few adverse effects.</p></sec>
Factors Influencing Lesion Formation During Radiofrequency Catheter Ablation	<p> In radiofrequency (RF) ablation, the heating of cardiac tissue is mainly resistive. RF current heats cardiac tissue and in turn the catheter electrode is being heated. Consequently, the catheter tip temperature is always lower - or ideally equal - than the superficial tissue temperature. The lesion size is influenced by many parameters such as delivered RF power, electrode length, electrode orientation, blood flow and tissue contact. This review describes the influence of these different parameters on lesion formation and provides recommendations for different catheter types on selectable parameters such as target temperatures, power limits and RF durations.</p>	<contrib contrib-type="author"><name><surname>Eick</surname><given-names>Olaf J</given-names></name><degrees>PhD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><sec id="s1a"><title>Lesion Generation due to Radiofrequency Energy Application - theoretical considerations</title><p>In RF ablation, the heating of cardiac tissue is mainly resistive. RF current is applied to the tissue via a metal electrode at the tip of the catheter, with a large skin electrode serving as indifferent electrode. The current density patterns in the tissue are determined by electrode size and geometry, electrode contact and local tissue properties. Also, of course, the current density will be proportional to the current (I) delivered by the RF generator, which, for constant resistance (R) of the electrode-tissue volume conductor is proportional to the square root of the RF power (P = I<sup>2</sup> R).</p><p>Current flow through a resistive medium causes heating, which is proportional to the square of the current or, locally, to the square of the local current density. The temperature increase will locally be proportional to the energy applied per second (local power), and increases proportional to the heat capacity of the local medium. In addition, when temperature differences between adjacent areas develop because of differences in local current density or local heat capacity, heat will conduct from "hotter" to "colder" areas, causing the temperature of the former to decrease and that of the latter to increase. Additionally, heat loss to the blood pool at the surface and to intramyocardial vessels determine the temperature profile within the tissue.</p><p>The heating occurs especially in the proximity of the active electrode due to its relatively small surface area causing locally high current density as compared to the site of the indifferent electrode. Typically, living tissue will be permanently destroyed at temperatures of approximately 45° to 50° C sustained for several seconds [<xref ref-type="bibr" rid="R1">1</xref>].</p><p>The tissue surface is cooled by the blood flow and thus the highest temperature during radiofrequency delivery occurs slightly below the surface.</p></sec></sec><sec id="s2"><title>Parameters Determining Lesion Formation</title><sec id="s2a"><title>Impact of target temperature - In Vitro Study</title><p>The impact of the target temperature on lesion size was evaluated by the author in an in vitro study. A 4-mm tip catheter (RF Marinr, Medtronic EPSystems, Minneapolis, MN) was positioned in a parallel orientation to porcine epicardium either with 0.5 N or 1.0 N contact force. A total of 48 lesions was produced with different target temperatures of 50, 60, 70 and 80° C (Atakr II, Medtronic, 100 W max. output, 60 s duration). Each setting was repeated 6 times and the average values were used for evaluation. The results are given in <xref ref-type="table" rid="T1">Table 1</xref>.</p><p>With increasing target temperature the delivered power, tip temperature and lesion dimension increased. Lesions created with a target temperature of 50° C were very small and nearly unrecognizable.</p><p>These results indicate that the lesion size could be well predicted by measuring tip temperature. However, the in vitro experiments were performed under stable flow conditions. In the in vivo situation, the flow is dependent on the ablation site and also varying over the heart cycle. These cooling effects have a strong impact on the catheter tip temperature and thus on the delivered power and also on lesion size.</p></sec><sec id="s2b"><title>Impact of Flow - In Vitro Study</title><p>Petersen et al. evaluated the impact of convective cooling on lesion dimension [<xref ref-type="bibr" rid="R2">2</xref>]. In vitro strips of porcine left ventricular myocardium during different levels of convective cooling were ablated. All applications were performed with a 4-mm tip catheter in parallel orientation to the tissue and automatic power adjustment (Atakr, 75W) to reach and maintain a target temperature of 80° C for 60 seconds. The results are shown in <xref ref-type="table" rid="T2">Table 2</xref>.</p><p>With increasing flow, i.e. increasing convective cooling, lesion depth, width and volume increased due to increasing power consumption to reach and maintain target temperature. Note, that the tip temperature was not different and is thus a poor indicator for lesion size if the flow condition is not stable.</p></sec><sec id="s2c"><title>Impact of Flow - In Vivo Study</title><p>Petersen et al. induced a total of 13 lesions in 6 pigs either at the left ventricular apex or at the mid-septum of the left ventricle [<xref ref-type="bibr" rid="R3">3</xref>]. A 4-mm tip catheter was used for RF delivery with a target temperature of 80° for 60 s (Atakr, 50 W maximum output). The results are shown in <xref ref-type="table" rid="T3">Table 3</xref>.</p><p>Lesions induced at the apical site were significantly smaller than those at the septal site. The blood flow is considered to be higher at the septum as compared to the apex, thus more power is consumed to maintain target temperature at the septum resulting in larger lesions. Again the tip temperature is a poor (or even contradicting) parameter for lesion size. Comparing the lesion volumes of the in vivo study with those of the in vitro study, the flow at the apex might be between 0.0 m/s and 0.1 m/s whereas the flow at the septum might me be more in the order of 0.2 m/s or higher.</p></sec><sec id="s2d"><title>Catheter Orientation - In Vitro Study</title><p>Using the same experimental setup Petersen et al. investigated the impact of electrode orientation on lesion size [<xref ref-type="bibr" rid="R4">4</xref>]. A 4-mm tip catheter was used for ablation with a target temperature of 80° C for 60 s either in a parallel or perpendicular orientation on porcine tissue stripes. The contact pressure between electrode tip and tissue was 10g and the flow 0.1m/s. The results are shown in <xref ref-type="table" rid="T4">Table 4</xref>.</p><p>Perpendicular electrode orientation yielded larger lesion volume using less power than parallel electrode orientation. The target temperature was approached in all applications with a mean electrode tip temperature of 75±2° C.</p><p>Changing the 4-mm tip electrode orientation from parallel to perpendicular decreases the proportion of the electrode tip area that is in contact with the tissue and increases the proportion of the tip area that is exposed to the convective cooling of the surrounding fluid. The larger lesion volume in the perpendicular electrode orientation suggests that cooling by flow around the electrode has greater impact than contact area. However, one would expect a significantly higher power delivery in the group with larger lesion volume. This can be explained by the fact, that only a minor fraction of the energy delivered by the generator is used for the lesion production itself. [<italic>In a first approximation the power required to produce a lesion with a certain volume can be calculated as follows: Power = 1/t * lesion volume * density * specific heat * ∆T = 1/60 * 0.25 * 1.05 * 3.72 * (80-37) = 0.7 W. That means that only 0.7 W are required to produce a lesion of 250 mm3 with a target temperature of 80°C for 60 s ! Assumption: Tissue Density = 1.05g/cm3, specific heat = 3.72 J/(g*°C), t = 60 seconds, ∆T = target temperature - body temperature = 43°C.</italic>] The major part is dissipated as electrical heating of the intracavitary blood, convective heat losses from electrode to blood, electrical heating of tissue outside the lesion volume and in electrical resistance of catheter and skin and fat layers at the indifferent electrode.</p></sec><sec id="s2e"><title>Catheter Orientation - In Vivo Study</title><p>Chan et al. published results based on in vivo data that are in conflict with those described by Petersen [<xref ref-type="bibr" rid="R5">5</xref>] and a likely explanation will be offered at the end of this paragraph.</p><p>In 26 dogs 144 lesions were created either in a parallel or perpendicular orientation in the right atrium with a target temperature of 75° C for 60 s using different tip lengths. The orientation was confirmed by fluoroscopy and intravascular ultrasound. For reasons of comparison only the results for the 4-mm tip catheter are given in <xref ref-type="table" rid="T5">Table 5</xref>.</p><p>The lesion volume was larger for the parallel orientation as compared to the perpendicular orientation which is in contradiction with the results from the in vitro study performed by Petersen et al.. However, all lesions were markedly smaller than in the previously summarized studies mainly due to the small lesion depth. The lesions were produced in the right atrium and the atrial wall is rather thin. The lesions would have been deeper in thicker tissue-note that 10 out of 14 for the perpendicular orientation and 9 out of 14 for the parallel orientation were transmural.</p><p>Chugh et al. who performed similar experiments in the left ventricle confirmed this [<xref ref-type="bibr" rid="R6">6</xref>]. They analysed 103 lesions in 20 dogs produced in the left ventricle also with a target temperature of 70° C for 60 s using different tip lengths. <xref ref-type="table" rid="T6">Table 6</xref> summarizes the results for the 4-mm long catheter tip.</p><p>The lesion depth was markedly higher than that of the atrial applications and similar to the values that Petersen et al. reported. Although there was a trend that the lesions produced in the parallel orientation were larger the difference did not reach statistical significance.</p><p>Based on these studies one may conclude that lesion depth is only little affected by catheter tip orientation using 4-mm long tip catheters but that lesions are slightly longer in the parallel orientation as compared to the perpendicular orientation.</p></sec><sec id="s2f"><title>Impact of Electrode Tip length</title><p>Petersen et al. evaluated the impact of electrode tip length on lesion size in 34 pigs [<xref ref-type="bibr" rid="R2">2</xref>]. Lesions were produced in the left ventricle in the posterior mid-septum, the left anterior free wall and in the apex using 2, 4, 6, 8, 10 and 12-mm long catheter tips. A target temperature of 80° C for 60 s was chosen with a maximum available power output of 75 W (Atakr II). The results are given in <xref ref-type="table" rid="T7">Table 7</xref>.</p><p>The lesion volume increased with increasing tip length for tip lengths between 2 and 10-mm. The lesion volume produced with an 8-mm long tip was about twice as big than that with a 10-mm tip catheter and even 3 times as big than that produced with a 4-mm tip catheter. Further increase in tip length did not result in further increase in lesion volume. With a very long catheter tip a large part is exposed to the blood flow and more energy is dissipated into the blood stream. Note that the average temperature decreased with increasing tip length and was thus a poor (or no) indicator for lesion volume. It is the amount of power that is effectively delivered to the tissue that determines lesion size. In addition, the depth did not differ between 4 and 8-mm tip catheters, the produced lesions are only wider but not necessarily deeper. Also, the applied average power was "only" 49 W for lesions created with an 8-mm tip catheter. Limiting the wattage to 50 W may reduce the likelihood of coagulum formation without compromising lesion size.</p><p>Langberg et al., who also produced lesions in the left ventricle using either 4-mm, 8-mm or 12-mm tip catheters, confirmed these results in part [<xref ref-type="bibr" rid="R7">7</xref>]. The target temperature was also 80° C for 60 s with a maximum available power of 100 W. The results are given in <xref ref-type="table" rid="T8">Table 8</xref>.</p><p>The power required to achieve a steady state temperature of 80° C was directly proportional to electrode size. The lesions produced by the 8-mm tip electrode were nearly twice as deep and four times as large as those made with a conventional 4-mm tip electrode. Lesions produced by the 12-mm tip electrode were intermediate in size and sometimes associated with charring and crater formation. Langberg et al. stated furthermore that ablations with larger tip electrodes caused a drop in arterial pressure and more ventricular ectopy than those with a 4-mm tip electrode.</p><p>The main difference to the results published by Petersen et al. is that lesions produced with the 8-mm tip catheter were also much deeper than those produced with the 4-mm tip catheter in the Langberg study. In concordance to the Petersen study, the use of a very large electrode did not further increase lesion size, and the tip temperature was even negatively correlated with lesion size.</p><p> The clinical relevance of the fact that 8-mm tip catheters produce larger lesions was demonstrated by Tsai et al. [<xref ref-type="bibr" rid="R8">8</xref>]. In a prospective, randomized study they compared 4-mm with 8-mm tip electrodes for linear ablation of typical atrial flutter in 104 patients. They reported that the 8-mm electrode catheter achieved higher complete isthmus block rate (92% vs. 67%, p<0.05) with fewer pulses (2±1 vs. 3±1, p<0.05), shorter procedure time (24±15 vs. 31±12 minutes, p<0.05), and shorter fluoroscopy time (14±10 vs. 23±15 minutes, p<0.05).</p></sec><sec id="s2g"><title>Impact of RF Duration</title><p>Simmers et al. evaluated the relation between RF duration and lesion size and published their results in 1994 [<xref ref-type="bibr" rid="R9">9</xref>]. In 11 dogs a total of 46 lesions was produced, 31 at left ventricular sites and 15 at right ventricular sites. They applied a constant power of 25 W for either 5, 10, 20, 30 and 60 seconds using a 7F, 4-mm tip electrode catheter (Mansfield-Webster, Watertown, MA, USA). The results are given in <xref ref-type="table" rid="T9">Table 9</xref>.</p><p>These results indicate that the lesion is predominantly generated within the first 10 seconds of energy delivery and reaches a maximum after 30 s. Further extension of RF delivery during power controlled RF delivery does not seem to further increase lesion size.</p></sec><sec id="s2h"><title>Impact of Electrode-Tissue Contact - In Vitro Study</title><p>It is difficult to evaluate the impact of electrode-tissue contact in the in vivo situation since the contact pressure between electrode and tissue cannot be assessed by fluoroscopy and a direct indicator for tissue contact is lacking at present. Some studies have been published where the influence of the electrode-tissue contact has been investigated in a well controllable in vitro environment. However, the conclusions out of these studies need to be drawn carefully to avoid misleading interpretations.</p><p>The author produced in vitro ablations on porcine myocardium with a 7F, 4-mm tip electrode with different electrode-tissue contact forces and a target temperature of 70° C for 30 s (50 max. output) [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R11">11</xref>]. A thermostat maintained a moderate flow. <xref ref-type="fig" rid="F1">Figures 1</xref> and <xref ref-type="fig" rid="F2">2</xref> illustrate the results. With increasing contact force between tissue and electrode the electrode tip temperature increased and the average power decreased (<xref ref-type="fig" rid="F1">Figure 1</xref>). <xref ref-type="fig" rid="F2">Figure 2</xref> illustrates the behavior of applied power and lesion depth. With increasing contact force the lesion depth increased until a plateau was reached. Further increase in contact force decreased lesion depth (although tip temperature was further increasing).</p><p>It is the amount of RF delivered effectively into the tissue that determines tissue heating and thus lesion generation. With increasing electrode-tissue contact a higher amount of RF power can be effectively brought into the tissue resulting in increasing lesion depth. At a certain <italic>moderate</italic> contact force further increase in contact force results in progressively smaller lesions because less amount of RF power is required to reach target temperature. Between 0.2 and 0.9 N (which is likely to be the bandwith in the clinical environment) the lesion depth is not much affected by the contact force and increasing contact force is balanced by decreasing applied power. These experiments were performed under stable flow conditions and in conclusion it seems likely that the flow around the electrode is of greater impact on lesion size than that of the electrode-tissue contact.</p></sec><sec id="s2i"><title>Impact of Irrigation</title><p>Irrigation has been introduced to avoid overheating at the tissue-electrode interface, thus allowing the delivery of higher amounts of RF power for a longer duration to create relatively large lesions.</p></sec><sec id="s2j"><title>Irrigated Power controlled RF delivery</title><p>Skrumeda et al. compared lesions created with a standard 4-mm tip catheter (RF Marinr) with those created with an irrigated tip catheter (RF Sprinklr) in animal experiments [<xref ref-type="bibr" rid="R12">12</xref>]. Three ablation protocols were conducted in canine left ventricles. Protocol I: standard ablation was performed in temperature controlled mode at 70° C and 90° C (120 sec). Protocol II: irrigated ablation was conducted with 30 and 50 W (30 and 120 sec). Protocol III: irrigated ablation was performed at 20 W with very long RF durations (5 and 10 min). The maximal available RF power output was 50 W. The results are given in <xref ref-type="table" rid="T10">Table 10</xref>.</p><p>With a standard electrode, lesions were larger with a target temperature of 90° C as compared to those created with a target temperature of 70° C, however, a coagulum was observed in 95% of applications with a target temperature of 90° C. The largest irrigated lesions were formed using 50 W (986±357 mm3) but were associated with craters in 54% and coagulum in 27% of the applications, respectively. Large lesions without craters and coagulum were created with irrigation using 20 W for 10 minutes (602±175 mm3). Skrumeda et al. concluded that irrigated ablation created larger lesions than standard ablation and that large lesions may be created without craters using moderate power and long duration !</p></sec><sec id="s2k"><title>Irrigated Temperature Controlled RF delivery</title><p>Petersen et al. compared lesions produced by standard temperature controlled RF delivery (TC) with those produced by either power controlled RF delivery (PC) with a high irrigation flow rate (20 ml/min) or temperature controlled RF delivery with a low irrigation flow rate (1 ml/min) [<xref ref-type="bibr" rid="R16">16</xref>]. The results are given in <xref ref-type="table" rid="T11">Table 11</xref>.</p><p>Petersen et al. demonstrated that lesion size and tissue temperatures were significantly higher during irrigated tip ablation compared to standard temperature controlled RF delivery (p<0.05). Lesion volume correlated positively with tissue temperature (r=0.87). The maximum recorded tissue temperature was always 1 mm from the ablation electrode. Crater formation only occurred at tissue temperatures of greater than 100° C.</p><p>Based on the results of this in vitro study it may be concluded that irrigated temperature controlled RF delivery yields relatively large lesions without crater formation if a moderate target temperature between 60 and 70° C and a low irrigation flow rate of 1ml/min are chosen. A target temperature of greater than 70° C may result in tissue overheating and crater formation.</p></sec><sec id="s2l"><title>Impact of irrigation flow rate</title><p>Weiss et al. investigated the influence of different flow rates on lesions produced on the thigh muscle in six sheep [<xref ref-type="bibr" rid="R17">17</xref>]. A total of 43 lesions was created with an irrigated tip catheter (RF Sprinklr) in a power controlled mode with 30 W target power for 30 s. A constant contact pressure of 0.1 N was maintained with a perpendicular orientation to the tissue. The results are given in <xref ref-type="table" rid="T12">Table 12</xref>.</p><p>The tissue temperatures at 7-mm depth, the lesion depth and width were not significantly different between the 3 different flow rates. The diameter measured at the surface was significantly smaller following RF applications with an irrigation flow rate of 20 ml/min due to increased cooling at the surface, which resulted also in lower tissue temperatures at a depth of 3.5 mm. Neither audible pops nor thrombus formation was observed in all applications. Based on these results a flow rate of 10 ml/min may be recommended when operating an irrigated catheter in the power controlled mode with a target power of about 30 W. The application of more than 30 W may require a higher flow rate to avoid excessive heat development at the superficial tissue layers.</p></sec><sec id="s2m"><title>Irrigation during Atrial Flutter Ablation</title><p>Jais et al. published the results of a prospective randomized comparison of irrigated tip versus conventional tip catheters for ablation of atrial flutter [<xref ref-type="bibr" rid="R13">13</xref>].</p><p>Cavotricuspid ablation was performed with a conventional (n=26) or an irrigated tip catheter (n=24). RF was applied for 60 seconds with a temperature-controlled mode: 65°C to 70°C up to 70 W with a conventional catheter or 50°C up to 50 W (with a 17 ml/min saline flow rate) with the irrigated tip catheter. Complete bidirectional isthmus block was achieved for all patients. Four patients crossed over from conventional to irrigated tip catheters. The number of applications, procedure duration, and x-ray exposure were significantly higher with the conventional than with the irrigated tip catheter: 13±10 versus 5±3 pulses, 53±41 versus 27±16 minutes, and 18±14 versus 9±6 minutes, respectively. No significant side effects occurred, and the coronary angiograms of the first 30 patients after ablation was unchanged.</p><p>Jais et al. concluded that irrigated tip catheters were found to be more effective than and as safe as conventional catheters for flutter ablation, facilitating the rapid achievement of bidirectional isthmus block.</p></sec><sec id="s2n"><title>Irrigation for difficult Accessory Pathways</title><p>Yamane et al. used an irrigated tip catheter for the ablation of accessory pathways resistant to conventional catheter ablation [<xref ref-type="bibr" rid="R14">14</xref>].</p><p>Among 314 accessory pathways in 301 consecutive patients, conventional ablation failed to eliminate accessory pathway conduction in 18 accessory pathways in 18 patients (5.7%), 6 of which were located in the left free wall, 5 in the middle/posterior-septal space, and 7 inside the coronary sinus (CS) or its tributaries. Irrigated tip catheter ablation was subsequently performed with temperature control mode (target temperature, 50°C), a moderate saline flow rate (17 ml/min), and a power limit of 50 W (outside CS) or 20 to 30 W (inside CS) at previously resistant sites. Seventeen of the 18 resistant accessory pathways (94%) were successfully ablated with a median of 3 applications using irrigated tip catheters. A significant increase in power delivery was achieved (20.3±11.5 versus 36.5±8.2 W; P<0.01) with irrigated tip catheters, irrespective of the accessory pathway location, particularly inside the CS or its tributaries. No serious complications occurred.</p><p>Yamane et al. concluded that irrigated tip catheter ablation is safe and effective in eliminating accessory pathway conduction resistant to conventional catheters, irrespective of the location.</p></sec><sec id="s2o"><title>Irrigation for Ventricular Tachycardia (VT)</title><p>Nabar et al. used irrigated tip catheters for ablation of ventricular tachycardias that were resistant to conventional catheter ablation [<xref ref-type="bibr" rid="R15">15</xref>].</p><p>Eight patients (6 men, age 59±12 years) in whom the clinical target VT (cycle length 430±97 msec) could not be ablated using a conventional 4-mm tip RF ablation catheter underwent additional attempts to ablate this VT using an irrigated tip catheter. Ablation of the clinical target VT using an irrigated tip catheter was attempted from the left ventricle in 6 (septal, posterobasal, and inferior: 2 each) and from the right ventricle in 2 patients (both septal), by entrainment, activation, or pace mapping. A mean of 6±5 (range 2 to 15) pulses was delivered. Target VT ablation was successful in 5 patients (63%). After successful ablation, at a mean follow-up of 6.5±4 months and while taking antiarrhythmic drugs, all 5 patients were free of VT recurrences. Nabar et al. concluded that the clinical target VT could be ablated using an irrigated tip catheter in 5 (63%) of the 8 patients in whom ablation using a conventional RF catheter was unsuccessful.</p></sec></sec><sec id="s3"><title>Recommendations</title><sec id="s3a"><title>4-mm Tip Catheters</title><p>The target temperature for 4-mm tip catheters should be less than 80° C. Since tissue temperature can be markedly higher than tip temperature a higher target temperature may increase the incidence of tissue overheating associated with crater formation and coagulum formation. The lesion size is poorly correlated to tip temperature in the in vivo situation. In high flow areas the tip is cooled and more RF power is delivered to the tissue to reach target temperature resulting in relatively large lesions and vice versa. Consequently, in high flow areas in the heart the difference between tip temperature and tissue temperature is large and a lower target temperature should be considered (e.g., 60° C) whereas in low flow areas the tissue temperature is much better reflected by the tip temperature and a higher target temperature could be considered (e.g., 80° C).</p><p>The duration could be limited to 30 seconds for non-irrigated 4-mm tip electrodes. The lesion is formed within the first 30 seconds predominantly. A longer duration does not create larger lesions.</p></sec><sec id="s3b"><title>8-mm Tip Catheters</title><p>A larger portion of 8-mm tip catheters is exposed to the blood and thus cooled by the blood flow and a relatively large difference between tip temperature and tissue temperature can be expected. Consequently, a moderate target temperature (e.g., 60° C) should be chosen and the RF power may be limited to 50-60W to avoid tissue overheating and coagulum formation.</p></sec><sec id="s3c"><title>Irrigated Tip Catheters: Irrigation Flow rate</title><p>An irrigation flow rate of 10ml/min may be selected in a power controlled mode with a delivered power of up to 30 W. The irrigation flow rate should be increased to 15-20 ml/min when more than 30 W are delivered to avoid excessive heat development at the superficial tissue layers.</p></sec><sec id="s3d"><title>RF Power and RF duration</title><p>The RF duration in power controlled mode with irrigated tip catheters should be considered to be longer than 30 s. Instead of increasing the power to achieve the desired effect (which increases the likelihood of crater formation) the duration could be increased. Skrumeda demonstrated lesions of similar size with 20 W for 300 s as with 50 W for 30 s. Consequently, a moderate power of 20-35W with relatively long RF duration of 60-300 seconds should be considered to achieve relatively large lesions with a limited risk of crater formation.</p></sec></sec>
Cardiac Resynchroniztion Therapy In Heart Failure: Recent Advances And New Insights	<p>Newer non-pharmacological therapies for heart failure are being evaluated for patients of congestive heart failure (CHF). Mechanical support with left ventricular assist devices and heart transplantation are reserved for the minority of patients who have severely decompensated heart failure. Despite these therapeutic advances, it is generally accepted that current therapies do not adequately address the clinical need of patients with heart failure, and additional strategies are being developed. Cardiac resynchronization therapy (CRT) is a new modality that involves synchronization of ventricular contraction and has shown a lot of promise in managing symptomatic patients of CHF who are on optimal medical therapy and have interventricular conduction delay (IVCD). It has improved exercise tolerance and NYHA functional class in such patients in sinus rhythm and a recent meta-analysis has also shown mortality benefits in CHF. Recently benefits of CRT have also been observed in CHF patients who do not have wide QRS complexes on electrocardiogram (EKG). It has also been shown to benefit drug refractory angina in CHF. Recent studies have also focused on the combined use of CRT and implantable cardioverter defibrillator (ICD) and it has shown encouraging results. Our aim in this descriptive review is to define practice guidelines and to improve clinicians' knowledge of the available published clinical evidence, concentrating on few randomized controlled trials.</p>	<contrib contrib-type="author"><name><surname>Bhatia</surname><given-names>V</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Bhatia</surname><given-names>R</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Dhindsa</surname><given-names>S</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Virk</surname><given-names>A</given-names></name><degrees>MD, MPH</degrees><xref ref-type="aff" rid="aff3">‡</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>Approximately 30 percent of patients with cardiomyopathy have IVCD such as left or right bundle-branch block, leading to loss of coordination of ventricular contraction [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>]. This dyssynchronous pattern of ventricular contraction is believed to contribute to the pathophysiology of heart failure, reducing the already diminished contractile reserve of the heart [<xref ref-type="bibr" rid="R3">3</xref>]. Specifically, dyssynchronous contraction exacerbates inefficient use of energy by the heart (a process termed mechanoenergetic-uncoupling [<xref ref-type="bibr" rid="R4">4</xref>]). The finding of IVCD has been associated with clinical instability and an increased risk of death in patients with heart failure [<xref ref-type="bibr" rid="R5">5</xref>-<xref ref-type="bibr" rid="R8">8</xref>].</p><p>Accordingly, the idea that cardiac-pacing technology might be used to restore the synchrony of ventricular contraction has been of theoretical interest for over a decade. Pacing modalities that utilize biventricular (BiV) or left ventricular (LV) stimulation to optimize cardiac pump function through synchronization of ventricular contraction are referred to as resynchronization or ventricular resynchronization therapies [<xref ref-type="bibr" rid="R2">2</xref>]. Resynchronization therapies can be present in a single device, in a device equipped with bradycardia pacing support, or incorporated into an ICD [<xref ref-type="bibr" rid="R9">9</xref>].</p></sec><sec id="s2"><title>CHF and IVCD</title><p>The most common causes for an IVCD in patients with heart failure are delayed left ventricular activation and left bundle branch block (LBBB). Impaired left ventricular function is also seen in otherwise normal subjects with isolated LBBB [<xref ref-type="bibr" rid="R10">10</xref>]. Approximately 20 to 30 percent of patients with symptomatic heart failure have an IVCD [<xref ref-type="bibr" rid="R2">2</xref>]. In a study done by Farewell et al [<xref ref-type="bibr" rid="R11">11</xref>] patients with a hospital diagnosis of "heart failure" were investigated. These patients did not undergo cardiac catherterization. The criteria for inclusion were severe heart failure (NYHA Class III or IV), heart failure due to dilated cardiomyopathy, QRS duration > 120 ms, or the presence of LBBB or RBBB. Using these criteria, approximately 10 percent of an unselected group of heart failure patients who are admitted to a typical district general hospital in United Kingdom during a calendar year would be candidates for biventricular pacing. A recent study done by Erdogan et al [<xref ref-type="bibr" rid="R12">12</xref>] estimated that biventricular pacing might be considered as an adjunct to standard heart failure therapy in 5-10 patients per year per 100,000 residents in industrial countries.</p><p> In Europe Resynchronization therapy is approved for symptomatic heart failure that occurs in the setting of IVCD or BBB. This approval was granted on the basis of several studies of acute resynchronization therapy and data compiled in approximately 150 patients receiving BiV or LV stimulation for three months as part of two controlled studies (InSync [<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R14">14</xref>] and PATH-CHF [<xref ref-type="bibr" rid="R15">15</xref>]). In the United States, resynchronization therapy with or without an ICD is approved for patients with NYHA class III-IV heart failure on the basis of the chronic studies described below, which were all performed with a control group randomly assigned to no resynchronization therapy [<xref ref-type="bibr" rid="R16">16</xref>-<xref ref-type="bibr" rid="R22">22</xref>].</p><p>There is another setting in which resynchronization might be important. It is estimated that approximately 8 to 15 percent of patients with advanced heart failure have pacemakers implanted for symptomatic bradycardia. Such patients have an increased risk of mortality or urgent transplantation due to progressive pump dysfunction; in one series, the risk at one year was 49 versus 15 percent in patients without a pacemaker) [<xref ref-type="bibr" rid="R23">23</xref>]. This difference may be due in part to the dyssynchronous contraction caused by right ventricular (RV) based pacing.</p><p>Whether such patients would derive long-term benefit from "upgrading" these devices to resynchronization therapies by the addition of a LV lead is currently under investigation. Initial data in patients with severe heart failure, prior atrio-ventricular (AV) junction ablation for rate control of AF, and chronic RV pacing has shown that there are significant benefits by upgrading from RV to BiV pacing [<xref ref-type="bibr" rid="R24">24</xref>].</p></sec><sec id="s3"><title>Effect on Contractile Function</title><p>Hemodynamic data acquired in patients with heart failure and bundle branch block (BBB) during acute or chronic BiV or LV stimulation have consistently shown improvements in measures of contractile response, such as force of contraction, cardiac output, left ventricular ejection fraction (LVEF), and pulmonary artery pressure, when compared to normal sinus rhythm or RV pacing [<xref ref-type="bibr" rid="R25">25</xref>,<xref ref-type="bibr" rid="R32">32</xref>]. CRT has been shown to decrease the functional mitral regurgitation in advanced systolic heart failure [<xref ref-type="bibr" rid="R33">33</xref>,<xref ref-type="bibr" rid="R34">34</xref>]. In contrast to other therapies that increase myocardial contractility, BiV and LV stimulation appear to modestly reduce myocardial energy demands and myocardial oxygen consumption [<xref ref-type="bibr" rid="R3">3</xref>]. The magnitude of acute systolic improvement by CRT is mainly due to resynchronization rather than due to change in myocyte function. An increased mechanical efficiency without increase in oxygen demand can be effective in drug refractory angina in CHF. A study done by Gasparini et al [<xref ref-type="bibr" rid="R35">35</xref>] showed the beneficial effects of CRT, during a mean follow-up of 9 months, in increasing the angina threshold in severely symptomatic patients with CHF and coronary artery disease (CAD) not amenable to cardiac revascularization. This study suggested that CRT increases the ischemic threshold in CHF patients on the long term, by markedly reducing the incidence of drug refractory anginal episodes, and by increasing a previously profoundly reduced exercise capacity. In another series of 18 patients with dilated cardiomyopathy (DCM) and an IVCD, aortic and LV pressures, dp/dt, and pressure-volume measurements were obtained during stimulation at single RV endocardial sites, at single LV epicardial sites, or during BiV pacing [<xref ref-type="bibr" rid="R36">36</xref>]. There was an improvement in systolic pressures with LV free wall or BiV stimulation, primarily due to an improvement in systolic function; there was no benefit on diastolic filling pressure or relaxation and RV apical or septal stimulation did not produce any hemodynamic changes. The markers of sympathetic activation, such as serum norepinephrine and heart rate variability, often vary directly with the severity of heart failure, these markers have not predictably changed in patients in whom resynchronization therapy appears to improve contractile function [<xref ref-type="bibr" rid="R37">37</xref>-<xref ref-type="bibr" rid="R40">40</xref>]. The improvement in mechanical synchrony appears to be the mechanism for reverse remodeling [<xref ref-type="bibr" rid="R41">41</xref>].</p></sec><sec id="s4"><title>Reverse Remodeling</title><p>Based upon echocardiography, preliminary data from the MIRACLE trial [<xref ref-type="bibr" rid="R16">16</xref>] suggested that BiV pacing is associated with reverse remodeling in patients with heart failure. BiV pacing produced an improvement in cardiac structure and function with a significant reduction in mitral regurgitation jet area and left ventricular mass, both signs of reverse remodeling [<xref ref-type="bibr" rid="R42">42</xref>]. Reverse remodeling was also observed in the CONTAK CD, PATH-CHF, and VIGOR CHF trials, in which BiV produced a significant reduction in left ventricular end-systolic and end-diastolic dimensions on echocardiography [<xref ref-type="bibr" rid="R38">38</xref>,<xref ref-type="bibr" rid="R43">43</xref>]. In the PATH-CHF trial, baseline left ventricular end-diastolic volumes were significantly smaller in those who exhibited reverse remodeling with BiV pacing compared to those who did not have a reduction in left ventricular volume [<xref ref-type="bibr" rid="R43">43</xref>].</p></sec><sec id="s5"><title>Clinical Trials</title><p>There are a number of trials evaluating the role of resynchronization therapy in patients with heart failure due to systolic dysfunction. The usual inclusion criteria include symptomatic heart failure that is stable on medical therapy, New York Heart Association (NYHA) class II to IV, left ventricular ejection fraction (LVEF) <35 percent, QRS duration >120 to 140 ms, and, in some trials, an indication for an ICD.</p></sec><sec id="s6"><title>MIRACLE Trial</title><p>In this trial16 453 patients with moderate-to-severe symptoms of heart failure associated with an LVEF of 35 percent or less and a QRS interval of 130 ms or more were studied. They were randomly assigned to a cardiac-resynchronization group (228 patients) or to a control group (225 patients) for six months, while conventional therapy for heart failure was maintained. The primary end points were the NYHA functional class, quality of life, and the distance walked in six-minutes. As compared with the control group, patients assigned to cardiac resynchronization experienced an improvement in the distance walked in six-minutes (+39 vs. +10 m, P=0.005), functional class (P<0.001), quality of life (-18.0 vs. -9.0 points, P= 0.001), time on the treadmill during exercise testing (+81 vs. +19 sec, P=0.001), and ejection fraction (+4.6 percent vs. -0.2 percent, P<0.001). In addition, fewer patients in the group assigned to cardiac resynchronization than control patients required hospitalization (8 percent vs. 15 percent) or intravenous medications (7 percent vs. 15 percent) for the treatment of heart failure (P<0.05 for both comparisons). Implantation of the device was unsuccessful in 8 percent of patients and was complicated by refractory hypotension, bradycardia, or asystole in four patients (two of whom died) and by perforation of the coronary sinus requiring pericardiocentesis in two others.</p></sec><sec id="s7"><title>MUSTIC Trial</title><p>The MUSTIC (Multisite Stimulation in Cardiomyopathies) trial is a single-blind randomized, controlled crossover study involving 131 patients who were divided into two groups based upon their underlying rhythm [<xref ref-type="bibr" rid="R17">17</xref>-<xref ref-type="bibr" rid="R19">19</xref>]. Group one included 67 patients with NYHA class III heart failure, QRS duration >150 ms with stable sinus rhythm and no conventional indications for pacemaker therapy [<xref ref-type="bibr" rid="R17">17</xref>]. The patients were randomly assigned to BiV pacing or no BiV pacing for three months, after which the pacing modes were switched; a total of 48 patients completed both phases of the study (MUSTIC SR). Exercise tolerance, as measured by the six-minute walk distance, increased by 23 percent after BiV pacing (399 versus 326 m, p<0.001). Other significant improvements included a 32 percent increase in quality of life, an 8 percent increase in peak oxygen consumption, and a two-thirds reduction in hospitalizations. Furthermore, BiV pacing was preferred by 85 percent of patients. At the end of the six-month crossover phase, the patients were programmed to the phase they preferred or, if there was no preference, according to the physician's judgment; almost all patients ended up with BiV pacing. The benefits with BiV pacing compared to baseline were maintained at 12 months19. Group two included 59 patients with heart failure and chronic atrial fibrillation (AF) with a wide QRS complex that required a permanent pacemaker because of a slow ventricular rate (MUSTIC AF). These patients were randomly assigned to either single site RV pacing or BiV pacing in the same fashion as in group one [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>]. Only 37 patients completed the six-month crossover trial, which limits any conclusions that can be drawn [<xref ref-type="bibr" rid="R17">17</xref>]. Using an intention-to-treat analysis, there were no significant differences in exercise tolerance or peak oxygen consumption. In contrast, when only the 37 patients who completed the study were evaluated, biventricular pacing was associated with a significant increase in six-minute walking distance (9.3 percent, 32 meters) and peak oxygen consumption (13 percent, 1.7 mL/kg per min). At the end of the six-month crossover phase, 33 of 37 patients (89 percent) preferred BiV pacing. Among 33 patients followed at one year, significant improvements persisted in both six-minute walking distance and peak oxygen consumption.</p></sec><sec id="s8"><title>CRT and Implantable Cardioverter Defibrillator (ICD) Therapy</title><p>Recently there has been some trials evaluating the combined use of CRT and ICD in patients of heart failure.</p><sec id="s8a"><title>I</title><p>InSync trial [<xref ref-type="bibr" rid="R44">44</xref>] demonstrated the efficacy and safety of implanting a combined device in 362 patients with class III and IV heart failure who also required an ICD. Patients were randomly assigned to have BiV pacing turned on or off; the ICD was active in all patients. InSync ICD Italian Registry [<xref ref-type="bibr" rid="R45">45</xref>] studied InSync ICD model 7272, a dual chamber ICD combined with CRT. In this registry, CRT combined with ICD implantation has been feasible with few device or left pacing lead related complications and this was found to be concordant with previous reports [<xref ref-type="bibr" rid="R9">9</xref>]. The clinical benefits match those obtained in recipients of biventricular pacemakers, both in LVEF and NYHA functional class [<xref ref-type="bibr" rid="R16">16</xref>].</p></sec><sec id="s8b"><title>II</title><p>VENTAK CHF/CONTAK CD [<xref ref-type="bibr" rid="R22">22</xref>] enrolled 581 patients with heart failure, most of whom had an ischemic cardiomyopathy, who also had an indication for an ICD. The majority of patients were male, had NYHA class II to IV heart failure, and a QRS duration >120 ms. The study utilized an ICD system designed to provide BiV pacing. All patients had an ICD and either BiV pacing or no pacing, each for six months.</p></sec><sec id="s8c"><title>III</title><p>The COMPANION trial [<xref ref-type="bibr" rid="R46">46</xref>] is a study of resynchronization therapy with and without an ICD in patients with NYHA class III-IV heart failure who had a hospitalization for heart failure within the year prior to enrollment. Nearly half of all patients enrolled had a non-ischemic etiology of heart failure. Patients were randomly assigned to optimal medical therapy, resynchronization alone, or resynchronization with an ICD. The trial was discontinued in November 2002 due to a significant benefit in the combined end point of total hospitalizations and mortality among the device treated patients.</p></sec></sec><sec id="s9"><title>Effects Of CRT and ICD Therapy on Arrhythmic Burden</title><p>Preliminary reports suggest that BiV pacing has an anti-arrhythmic effect [<xref ref-type="bibr" rid="R43">43</xref>,<xref ref-type="bibr" rid="R47">47</xref>]. The anti-arrhythmic effect has been attributed to improved hemodynamics. A low mean number of ventricular arrhythmic episodes were observed in the whole population and in the patients without Class I indications in InSync Italian registry [<xref ref-type="bibr" rid="R45">45</xref>]. It was found that the patients without standard ICD indications sustained serious arrhythmic events, confirming their high risk of death.</p></sec><sec id="s10"><title>Pacing Sites</title><p>Short-term studies have suggested that the lateral wall is a preferred site of LV stimulation to achieve effective CRT [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R25">25</xref>]. However, this choice may be limited by technical difficulties like high capture threshold from the presence of scar or fibrosis, particularly in patients with CAD, determining high LV pacing threshold, unfavorable coronary venous anatomy with narrow and tortuous coronary sinus tributary (CST), phrenic nerve stimulation or pacing lead instability. Gasparini et al [<xref ref-type="bibr" rid="R48">48</xref>] conducted a study to evaluate the effects of different pacing sites in patients treated with CRT. The data from this study revealed that, during long term follow-up, the most important clinical and echocardiographic parameters improved significantly in the patients, independently of the stimulation site. This was the case when considering each CST separately, or when dividing patients between "lateral" and "septal" sites, in the entire population and in the subgroups of patients without CAD. Hence in the presence of major technical difficulties preventing stimulation of the lateral LV, alternative-pacing sites, particularly the basal anterior LV wall, may be suitable to offer effective CRT to these patients. Tissue Doppler echocardiography has also been used to determine the optimal pacing site for BiV pacing [<xref ref-type="bibr" rid="R47">47</xref>] and to document an improvement in LV function, manifest by an increase in LV and interventricular synchrony, a shortened isovolumic contraction time, and an increased diastolic filling time [<xref ref-type="bibr" rid="R41">41</xref>,<xref ref-type="bibr" rid="R50">50</xref>].</p></sec><sec id="s11"><title>QRS Duration And Dyssynchrony</title><p>Wide QRS duration, possibly with LBBB, has been proposed as an independent predictor of total mortality in CHF patients [<xref ref-type="bibr" rid="R51">51</xref>-<xref ref-type="bibr" rid="R53">53</xref>]. In addition it has been considered a key criteria for selecting CHF patients for CRT, as wide QRS has been suggested to be associated with marked RV to LV and intra-LV dyssynchrony [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R25">25</xref>,<xref ref-type="bibr" rid="R54">54</xref>]. However, the correlation between the QRS width and regional electromechanical LV dyssynchrony has not been completely clarified [<xref ref-type="bibr" rid="R47">47</xref>,<xref ref-type="bibr" rid="R55">55</xref>,<xref ref-type="bibr" rid="R56">56</xref>] and a high prevalence of left ventricular systolic and diastolic asynchrony has been found in patients with congestive heart failure and normal QRS duration [<xref ref-type="bibr" rid="R57">57</xref>]. Hence for a given QRS width there is a considerable scatter in response to CRT responsive patients with narrow complexes and less responsive ones with wide complexes exist [<xref ref-type="bibr" rid="R32">32</xref>]. A study was done by Gasparini et al [<xref ref-type="bibr" rid="R58">58</xref>] to assess in a large cohort of patients the role of baseline QRS width (<150 / ≥150 ms) on clinical and echocardiographic parameters, hospitalization rates, and survival after CRT. In this study 158 CHF patients (121 men, mean age 65 years, mean LVEF 0.29, mean QRS width 174 ms) underwent successful BiV implantation and were then followed for a mean time of 11.2 months. According to the basal QRS duration, patients were divided in two groups with wide QRS group (≥150 ms, 128 patients, 81 percent) and the narrow QRS (<10 ms, 30 patients, 19 percent).</p><p> In the wide QRS group, following results were noted: <list list-type="order"><list-item><p>LVEF improved from 20 percent to 39 percent (P < 0.0001)</p></list-item><list-item><p>Six-minute walk test from 311 to 463 m (P<0.0001)</p></list-item><list-item><p>NYHA Class III-IV patients decreased from 86 percent to 8 percent (P<0.0001).</p></list-item></list></p><p> In the narrow QRS group, following were the results: <list list-type="order"><list-item><p>LVEF improved from 30 percent to 38 percent (P<0.0001).</p></list-item><list-item><p>Six-minute walk test 370 to 506 m (P<0.0001).</p></list-item><list-item><p>NYHA Class III-IV patients decreased from 60 percent to zero percent (P<0.0001).</p></list-item></list></p><p>The data showed that in wide and narrow QRS patients, BiV pacing significantly improved clinical parameters (NYHA lass, six-minute walk test, quality of life, and hospitalization rate) and main echocardiographic indicators. Furthermore, narrow QRS patients had a better survival rate, rapidly regained left ventricular function, and only a few patients remained in a higher NYHA class during follow-up. These patients should not be excluded "a priori" from CRT.</p><p>This study highlights the important point at what level a QRS has to be considered "wide enough" to be proposed to benefit from CRT. For example this was >150 ms for the MUSTIC study17, >130 ms in the MIRACLE study16, and >120 ms in Comparison of medical therapy, pacing, and defibrillation in chronic heart failure (COMPANION) trial [<xref ref-type="bibr" rid="R46">46</xref>]. It is evident from these differences in opinion that there is still no consensus on just how "wide" QRS should be for an efficacious CRT. Moreover the duration of QRS alone no longer seems to be defining parameter for patients with either inter-or-intraventricular dyssynchrony, given that recent studies have shown that even patients with a QRS<150 ms or without LBBB can suffer from significant dyssynchrony [<xref ref-type="bibr" rid="R53">53</xref>,<xref ref-type="bibr" rid="R54">54</xref>].</p></sec><sec id="s12"><title>Are There Any New Markers of Asynchrony?</title><p>The results of the recent investigations have prompted a reappraisal of the apparent correlation between conduction disorders and cardiac dyssynchronization. In a tissue Doppler study of 104 patients with BBB, Garrigue et al [<xref ref-type="bibr" rid="R56">56</xref>] observed that 35 percent of patients with LBBB had no interventricular dyssynchronization, and 20 percent had no left ventricular dyssynchronization. Despite fulfilling the "classic" criteria of wide QRS and LBBB, these patients are hardly candidates for CRT. Conversely, a sizable number of patients with right bundle branch block (RBBB) may present with mechanical anomalies, which may be corrected by CRT. Therefore, new markers of asynchrony are desirable, more directly related to cardiac mechanical function than the EKG. Among several methods available, angioscintigraphy with phase analysis of the contraction isochrones was the first, though its cumbersome implementation, high cost, and limited availability in routine clinical practice have prevented its widespread application [<xref ref-type="bibr" rid="R59">59</xref>,<xref ref-type="bibr" rid="R60">60</xref>].</p><p>A recent study done by Cazeau S et al [<xref ref-type="bibr" rid="R61">61</xref>] explored the value of an echocardiographic model to identify cardiac electromechanical dyssynchrony parameters (EDP) in candidates for CRT and their potential correction after implantation. The study included 66 CRT recipients of CRT NYHA functional class III or IV who had one or more AV, interventricular or intra-ventricular dyssynchrony criteria. An immediate improvement was observed in 85 percent of the population with partial or total correction of their EDP. However the modification in EDP differed considerably between recipients of de novo CRT systems and patients with previously implanted standard pacing systems upgraded with the implantation of a left ventricular lead. EDP measurements appear to identify candidates for CRT and to confirm the success of system implantation. This is the first report of a selection of candidates for CRT based on mechanical instead of electrical criteria. An overlap certainly exists between patients presenting with a wide QRS and patients with disorders of cardiac synchronization. However the echocardiographic method, which distinguishes three different types of synchronization offers a finer analysis of the anomalies amenable to resynchronization.</p></sec><sec id="s13"><title>Etiology of CHF and CRT</title><p>The mechanisms of CHF in patients with DCM are complex and multiple. A study done by Gasparini et al et al [<xref ref-type="bibr" rid="R61">61</xref>] examined the importance of underlying cardiac pathology on the outcome of CRT, hypothesizing that myocardial infarction scar and the non-contractile segment represent limitations to the ability to resynchronize cardiac contraction in patients with CHF associated with DCM. The results of this study showed that the functional capacity improved significantly during CRT in CAD and non-CAD patients. LVEF and NYHA class in non-CAD patience showed a significantly greater improvement. However, changes in quality-of-life were similar in both groups. The mechanisms of slow myocardial conduction associated with asynergic contraction in patients with DCM vary with the underlying pathology et al [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R62">62</xref>,<xref ref-type="bibr" rid="R63">63</xref>]. In patients with DCM not due to CAD, ventricular asynergy may be associated with interventricular or intra-ventricular conduction delays. Interventricular asynergy is most often associated with LBBB. A progressive remodeling of myocardial collagen matrix well documented in familial cardiomyopathies may impair intraventricular conduction. Disruption of collagen network, by altering the cellular architecture, impairs intraventricular conduction and the coordinated mechanical response of the ventricles. The consequences are QRS prolongation and waste of mechanical work. In patients with CAD, beside ventricular remodeling, ventricular asynergy may be associated with segmental wall- motion abnormalities as result of myocardial infarction scars, or of ischemic non-contractile segments. Segmental wall motion abnormalities affect intraventricular conduction and the coordinated mechanical response of the ventricles. CRT may correct conduction delay in remodeled dilated myocardial segments, but has no effect on extensive myocardial scars or ischemic segments. CRT can only recruit and coordinate a fraction of the myocardial mass to increase ventricular mechanical work in patients with CAD. Although significant benefits were observed in both groups after CRT, myocardial infarction scars limit the mechanical benefits of QRS narrowing and resynchronization. Hence the benefits of CRT should not be denied to patients with severe CHF on the basis of underlying cardiac pathology including patients with severe LV dysfunction associated with CAD and wide QRS et al [<xref ref-type="bibr" rid="R64">64</xref>].</p></sec><sec id="s14"><title>Pacing in AF</title><p>Paroxysmal or persistent AF occurs in up to 30 percent of patients with HF [<xref ref-type="bibr" rid="R65">65</xref>]. Rate control can be achieved with pharmacological therapy. In patients refractory to such therapies these objectives can be achieved with radiofrequency ablation of the AV node and pacemaker therapy with traditional RV-based pacemakers. Initial data regarding "upgrading" from RV to BiV pacing using an LV lead to achieve cardiac resynchronization in heart failure patients with chronic AF who have undergone radiofrequency AV nodal ablation followed by standard RV pacing is promising [<xref ref-type="bibr" rid="R24">24</xref>].</p></sec><sec id="s15"><title>Recommendations</title><p>BiV pacing is an effective approach to the therapy of patients with heart failure and IVCD and studies suggest that BiV pacing can improve exercise tolerance and NYHA functional class in such patients in sinus rhythm. A meta-analysis was done by Bradley et al [<xref ref-type="bibr" rid="R66">66</xref>] of the available studies to determine the effect of CRT on mortality in CHF. 11 reports of 4 randomized trials with 1634 total patients were included in this meta-analysis. It was found that cardiac resynchronization reduces mortality from progressive heart failure in patients with symptomatic left ventricular dysfunction. This finding suggests that cardiac resynchronization may have a substantial impact on the most common mechanism of death among patients with advanced heart failure. Cardiac resynchronization also reduces heart failure hospitalization and shows a trend toward reducing all-cause mortality.</p><p>More data on effect of CRT on mortality in CHF is still awaited. Although suggestive, the data are insufficient to prove efficacy in patients in AF [<xref ref-type="bibr" rid="R18">18</xref>]. As a result of the MIRACLE trial, the FDA has approved BiV pacing as a treatment for moderate to severe heart failure. Potential concerns include the small risk of serious complications during implantation as noted in MIRACLE [<xref ref-type="bibr" rid="R16">16</xref>] and lack of data concerning the long-term effects of cardiac resynchronization as noted by the 2001 Task Force of the ACC/AHA [<xref ref-type="bibr" rid="R67">67</xref>].</p><p>At present, it seems reasonable to consider BiV pacing in patients with a low LVEF and prolonged QRS duration who remain symptomatic (NYHA class III or IV HF) despite optimal medical therapy [<xref ref-type="bibr" rid="R68">68</xref>]. The 2002 task force of the ACC/AHA/NASPE gave a class IIa recommendation (weight of evidence in favor of efficacy) to BiV pacing in medically refractory, symptomatic NYHA class III or IV patients with idiopathic dilated or ischemic cardiomyopathy, prolonged QRS interval (130 ms), LV end-diastolic diameter greater than or equal to 55 mm and ejection fraction less than or equal to 30 percent.</p><p>It is not known if these devices should routinely incorporate a defibrillator. The MADIT II trial showed a significant survival benefit from ICD placement in patients who have had a previous myocardial infarction and have an LVEF 30 percent [<xref ref-type="bibr" rid="R69">69</xref>]. Combination therapy with an ICD and BiV might therefore be beneficial in such patients who have QRS prolongation.</p></sec>
Congenital Junctional Ectopic Tachycardia: Presentation And Outcome	Could not extract abstract	<contrib contrib-type="author"><name><surname>Sarubbi</surname><given-names>Berardo</given-names></name><degrees>MD PhD</degrees></contrib><contrib contrib-type="author"><name><surname>Vergara</surname><given-names>Pasquale</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>D'Alto</surname><given-names>Michele</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Calabro</surname><given-names>Raffaele</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>Junctional ectopic tachycardia (JET) is a rare type of supraventricular arrhythmia. Even if its management has improved in recent years, it remains a great challenge for the cardiologist. Two are the possible clinical presentations of this arrhythmia: as a primary idiopathic disorder during infancy, configuring the so called "congenital" JET, or more often as a transient phenomenon immediately after surgery for congenital heart disease, giving rise to the "post-operative" variety.</p><p>The congenital form, firstly described as a distinct entity by Coumel et al. in 1976 [<xref ref-type="bibr" rid="R1">1</xref>], usually occurs in the first six months of life presenting as a persistent sustained form, lasting up to 90% of the time. Its clinical presentation may be dramatic, being associated in up to 60% of cases with cardiomegaly and/or heart failure.</p><p>Congenital JET is hampered by high mortality. Secondary dilated cardiomyopathy, ventricular fibrillation and sudden cardiac death have also been reported [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R3">3</xref>].</p></sec><sec id="s2"><title>Clinical Presentation</title><p>Clinical presentation of congenital JET usually ranges from birth to 4 weeks of age. Sporadic cases of history of intrauterine tachycardia, with sometimes hydrops fetalis, have been reported in patients who at birth showed overt JET [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R5">5</xref>].</p><p>Usually, high percentage of patients present, at the time of referring, echocardiographyc evidences of impaired left ventricle function and clinical signs of congestive heart failure. However, it has been noted that age of tachycardia presentation is not related to the occurrence of congestive heart failure or impaired left ventricular function. On the contrary, clinical status seems related to the ventricular rate at presentation [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>The incidence of congenital heart disease in presence of congenital junctional ectopic tachycardia has been reported sporadic [<xref ref-type="bibr" rid="R6">6</xref>].</p></sec><sec id="s3"><title>Diagnostic Assessment</title><p>The diagnosis of junctional ectopic tachycardia usually is based on electrocardiographic evidence of a narrow complex tachycardia (heart rate ranging from 160 bpm to 300 bpm) and atrio-ventricular dissociation. (<xref ref-type="fig" rid="F1">Figure 1</xref>) Sometimes, it is possible to observe slight variation of heart rate during the 24 hours and irregularity in the RR due to sinus capture beats. Distinct P waves can be seen only in the tracing of the patients with slower heart rate following drug treatment.</p><p>Prenatal diagnosis of JET has been also reported, with the superior vena cava/ascending aorta Doppler echocardiography approach [<xref ref-type="bibr" rid="R4">4</xref>]. Doppler wave showed AV dissociation with aortic ejection occurring at smaller interval than venous retrograde flow due to atrial contractions and "a" waves occurring at a very short interval after aortic ejection, during periods of 1:1 AV relationship tachycardia.</p></sec><sec id="s4"><title>Electrophysiological Mechanism</title><p>Electrophysiological mechanism of JET is thought to be abnormal automaticity within the His bundle [<xref ref-type="bibr" rid="R7">7</xref>]. JET typically presents a "warming-up" phase at its onset and a "cooling-down" phase at its termination. It is usually unresponsive to overdrive pacing or DC shock. When there is retrograde conduction, adenosine infusion can rule out the diagnosis of AV reciprocating tachycardia, because of tachycardia cycle lengthening without termination of the arrhythmia [<xref ref-type="bibr" rid="R8">8</xref>].</p><p>Tachycardia is not inducible by programmed stimulation in baseline status, but can be elicited after sympathetic stimulation with isoproterenol. Programmed atrial and ventricular stimulation usually are not able to terminate tachycardia, that can be only transiently suppressed by atrial and/or ventricular pacing, exhibiting a resetting curve-response to extrastimuli during tachycardia.</p><p>JET can coexist with the occurrence of other types of arrhythmias. Scheinman et al. [<xref ref-type="bibr" rid="R9">9</xref>] reported the development of JET after catheter ablation of persistent junctional reciprocating tachycardia in one patient, but a differentiation between a masking effect of the latter tachycardia on JET and a procedure related genesis of JET itself could not be argued.</p></sec><sec id="s5"><title>Familiar occurrence</title><p>Even more interest is attributed to this arrhythmia by the fact that it is a familiar condition. In a multicenter study, Villain reports a familiar history of JET positive in 50% of the children. [<xref ref-type="bibr" rid="R2">2</xref>]. In our experience up to three members of the same family (2 sisters, 1 cousin) were affected by JET with a family history of JET recognised in up to 55.6% of the patients [<xref ref-type="bibr" rid="R6">6</xref>]. These characteristics suggest a possible Mendelian inheritance but, at the moment, a specific gene has not been identified.</p></sec><sec id="s6"><title>Outcome</title><p>Congenital JET is hampered by high mortality, with up to 34% mortality rate. The exact mechanism related to death in congenital JET is still unclear. Most are sudden death, that has been found also in children with "well compensated" tachycardia [<xref ref-type="bibr" rid="R2">2</xref>].</p><p>Sporadic case of sudden death has been attributed to a dramatic evolution to paroxysmal complete AV block (2; 10). Probably, the dramatic clinical course of these patients is the extreme expression of the pathological process occurring in the His bundle region, causing JET in a first time and finally the complete loss of AV conduction.</p><p>Necropsy reports underlined the importance of His-bundle fibbers severe disruption [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R12">12</xref>] or focal degeneration and split2. In other cases the arrhythmia has been attributed to extensive abnormalities involving the whole atrio-ventricular junction [<xref ref-type="bibr" rid="R13">13</xref>] or to an intra-His bundle tumour originating from Purkinje cells [<xref ref-type="bibr" rid="R14">14</xref>].</p><p>Henneveld reported a case of an 8 months old girl affected by a rapid ventricular rate JET, who developed two years later a complete heart block with rapidly increasing left ventricular end diastolic diameter [<xref ref-type="bibr" rid="R10">10</xref>]. The patient was not assuming therapy at the time of the onset of complete heart block.</p><p>In other cases it can not be excluded that a pharmacological treatment can be responsible of the depression of ventricular excitability, leading to a slow unsatisfactory ventricular escape rhythm in case of AV block and sudden cardiac death. However, in some cases it can not be excluded the potential proarrhythmic effect of the drugs used to control the arrhythmia [<xref ref-type="bibr" rid="R15">15</xref>]. In our series, two patients developed symptomatic non-sustained life-threatening ventricular tachycardia as a spontaneous desincronization of junctional ectopic tachycardia, due to a proarrhythmic effect of the combination of used drugs (Amiodarone + Propafenone and Amiodarone + Flecainide) [<xref ref-type="bibr" rid="R6">6</xref>].</p></sec><sec id="s7"><title>Pharmacological Treatment</title><p>Treatment is indicated in infants with symptoms, reduced ventricular function or rapid rates [<xref ref-type="bibr" rid="R16">16</xref>'<xref ref-type="bibr" rid="R17">17</xref>]. The management of infants with slow JET (less than 150 beats/min) without symptoms appear to be debated. However, necessity to monitor accurately these asymptomatic patients is undoubted. It has been reported a case in which an asymptomatic patient at birth with junctional ectopic tachycardia with a ventricular rate of 140 beats/min returned at 6 weeks of age in cardiovascular collapse due to a ventricular rate of 300 beats/min [<xref ref-type="bibr" rid="R2">2</xref>].</p><p>The pharmacological treatment of JET is hampered by a high rate of failures. Several drugs have been tested, but a comprehensive analysis of efficacy has not been performed yet.</p><p>The largest study available is a multicenter study performed in nineties by European and American teams who reviewed dates of 26 pts treated with digoxin, propranolol and/or amiodarone [<xref ref-type="bibr" rid="R2">2</xref>]. The largest single centre experience of pharmacological treatment has been recently reported, with the use of Amiodarone, Flecainide or Propafenone alone or in combination [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>Digoxin did not affect ventricular rate in any patients and has been demonstrated to be not completely safe in patients with congenital JET. Sporadic cases have been reported in which patients affected by congenital JET and severe cardiac failure developed ventricular fibrillation or faster tachycardia (up to 400 beats/min) during progressive digoxin loading [<xref ref-type="bibr" rid="R2">2</xref>].</p><p>Amiodarone seems to have the highest response rates: it has been shown to be effective alone in decreasing the ventricular rate to less than 150 beats/min in a percentage between 50-70% of the cases [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R6">6</xref>].</p><p>Sporadic reports of medical treatment efficacy exist also about Propafenone [<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R18">18</xref>], Flecainide or Encainide [<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R19">19</xref>] and Sotalol [<xref ref-type="bibr" rid="R20">20</xref>]. However, Propafenone has been shown to result particularly effective in preventing or controlling the tachycardia only in patients with lower heart rate [<xref ref-type="bibr" rid="R6">6</xref>]</p><p>Other studies have also tested phenytoin, which was able to control ventricular rate but caused ataxia, ajmaline, which intravenous infusion was followed by ventricular tachycardia and verapamil which caused cardiovascular collapse [<xref ref-type="bibr" rid="R2">2</xref>].</p><p>Personal experience has shown that in those cases not responding to a single drug regimen, as association of antiarrhythmic agents with different electrophysiological effects (Amiodarone + a Ic antiarrhythmic drug) may control otherwise untreatable congenital JET. Otherwise the association Propafenone or Flecainide plus Amiodarone could reduce the Amiodarone dose keeping an high efficacy of the pharmacological therapy [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>Recently Dorman et al. have described a protective effect of magnesium supplementation on JET in paediatric patients undergoing surgery for congenital heart defects [<xref ref-type="bibr" rid="R21">21</xref>]. This could be due to the stabilization of the membrane potential and the reduction of the automaticity, resulting in a low rate of development of the arrhythmia. Further studies are needed to evaluate the potential role of magnesium in the treatment of the congenital form of JET.</p><p>All pharmacological therapies are burdened by a high risk of toxicity as they have to be maintained at high dosage for long time in young patients.</p><p>Congestive heart failure is a frequent complication of JET with high ventricular rate and its management is even more difficult. In these cases it has been underlined the negative inotropic effects of antiarrhythmic drugs and the limitations in the use of sympathomimetic agents. Isoproterenol, dopamine, dobutamine and amrinone all increase the JET rate. Sodium Nitroprusside has to be avoided too, since low blood pressure will reflexly increase adrenergic tone.</p></sec><sec id="s8"><title>Non pharmacological treatment</title><p>A definitive treatment could be the removal of the arrhythmogenic area. Surgical His ablation has been attempted in some critical patients, with contrasting results [<xref ref-type="bibr" rid="R2">2</xref>]. In one case cautery surgical ablation led to the development of ventricular tachycardia. Radiofrequency catheter ablation of JET was reported for the first time in the eighties [<xref ref-type="bibr" rid="R22">22</xref>,<xref ref-type="bibr" rid="R23">23</xref>]. It required, in the first procedures, the elimination of AV conduction tissue with the following pacemaker implantation. In nineties radiofrequency catheter ablation technique was able to control the tachycardia with the preservation of the normal AV conduction [<xref ref-type="bibr" rid="R9">9</xref>]. Lesions were carried out in the region between the coronary sinus and the anterior septum, moving progressively anteriorly to a site where a small His deflection was recorded. In some cases His-bundle localization can be uncommon, in view of the fact that a His-bundle has been found on the left sided of the interventricular septum [<xref ref-type="bibr" rid="R2">2</xref>]. Recent experiences have proven the feasibility of the selective modification of the AV junction without the complete destruction of the conduction system even in small children, less than one year old [<xref ref-type="bibr" rid="R24">24</xref>].</p><p>Very controversial is still the matter of prophylactic pacemaker implant to prevent sudden death. In the multicenter study [<xref ref-type="bibr" rid="R2">2</xref>] most part of patients did not undergo to pacemaker implantation, but four patients with well compensated JET died suddenly and in one of those a slow ventricular rate was recorded during the terminal event.</p><p>Walsh suggested to evaluate AV conduction with transesophageal electrophysiological test in all patients with congenital JET and consider pacemaker insertion only if impaired conduction can be demonstrated by atrial stimulation, or by the observation of spontaneous AV block on ECG and Holter monitoring [<xref ref-type="bibr" rid="R25">25</xref>].</p></sec><sec sec-type="conclusions" id="s9"><title>Conclusions</title><p>Since JET therapy relay on different medical options, a staged treatment protocol is advisable: a pharmacological approach should be tested and carried on, also with a multi drug treatment, while catheter ablation can be a definitive option in critically ill patients, when drugs fail or when tachycardia becomes chronic and the patient is dependent on a potentially toxic therapy.</p></sec>
Arrhythmogenic Right Ventricular Cardiomyopathy/Dysplasia<xref ref-type="fn" rid="fn1">1</xref>	<p>Arrhythmogenic right ventricular cardiomyopathy/dysplasia (ARVC/D) is characterized by the patchy replacement of myocardium by fatty or fibrofatty tissue. These changes lead to structural abnormalities including right ventricular enlargement and wall motion abnormalities that can be detected by echocardiography, angiography, and cine MRI. ARVC/D is a genetically heterogeneous disorder, since it has been linked to several chromosomal loci. Myocarditis may also be a contributing etiological factor. Patients are typically diagnosed during adolescence or young adulthood. Presenting symptoms are generally related to ventricular arrhythmias. Concern for the risk of sudden cardiac death may lead to the implantation of an intracardiac defibrillator. An ongoing multicenter international registry should further our understanding of this disease.</p>	<contrib contrib-type="author"><name><surname>Indik</surname><given-names>Julia H</given-names></name><degrees>MD PhD</degrees></contrib><contrib contrib-type="author"><name><surname>Marcus</surname><given-names>Frank I</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>Arhythmogenic right ventricular cardiomypathy/dysplasia (ARVC/D) is a disorder in which the right ventricular myocardium is interspersed with fatty or fibrofatty tissue. This pathological process principally affects the right ventricle, but in advanced cases can lead to biventricular involvement. Linkage analysis has identified several loci on different chromosomes, indicating this disease is genetically heterogeneous. Specific mutations have been identified in the genes for desmoplakin [<xref ref-type="bibr" rid="R1">1</xref>], plakoglobin [<xref ref-type="bibr" rid="R2">2</xref>] and the ryanodine receptor gene [<xref ref-type="bibr" rid="R3">3</xref>]. The predominant presenting symptoms are due to ventricular arrhythmias, including palpitations, sustained ventricular tachycardia, or uncommonly, sudden cardiac death. Although ARVC/D is an unusual disorder, it has been described as the most common cause of sudden death in young athletes under the age of 35 in the Veneto area in Italy [<xref ref-type="bibr" rid="R4">4</xref>] . Heart failure is infrequent but may occur due to severe right or biventricular enlargement.</p><sec id="s2a"><title>a) Structural Abnormalities and Pathophysiology</title><p>Anatomically, ARVC/D is characterized by replacement of myocardium with fatty and fibrous tissue that primarily affects the epicardium and the mid-myocardium, with relative sparing of the endocardium (<xref ref-type="fig" rid="F1">Figure 1</xref> and <xref ref-type="fig" rid="F2">2</xref>). This process most commonly affects the posterior and inferior areas of the right ventricular inflow tract adjacent to the tricuspid valve [<xref ref-type="bibr" rid="R5">5</xref>], but it also affects the anterior infundibulum and the apex, thus forming what is known as the "triangle of dysplasia" [<xref ref-type="bibr" rid="R6">6</xref>]. Loss of myocardium may result in aneurysm formation, commonly in the basal inferior wall, underneath the tricuspid valve [<xref ref-type="bibr" rid="R7">7</xref>]. Fibro-fatty replacement occurs in a segmental, patchy fashion and adjacent areas, particularly, the septum, are generally spared. Therefore, endomyocardial biopsies, which usually are performed from septal tissue, may be non-diagnostic. ARVC/D is distinguished from Uhls anomaly, a congenital disorder in which right ventricular myocardium is absent, resulting in a paper-thin right ventricular wall [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R9">9</xref>]. Recently, apoptotic cells have been reported in endomyocardial biopsy specimens [<xref ref-type="bibr" rid="R10">10</xref>-<xref ref-type="bibr" rid="R13">13</xref>]. It is unknown how the apoptotic pathway may be initiated in ARVC/D, but may relate to altered intracellular calcium concentration [<xref ref-type="bibr" rid="R1">1</xref>]. Inflammatory infiltrates consisting of lymphocytes may also be found on biopsy specimens, suggestive of a focal myocarditis, and may explain sporadic cases of ARVD/C.</p><p>Diffuse hypokinesis or regional wall motion abnormalities of the right ventricle can be seen with contrast ventriculography, considered to be the most reliable test to assess structural abnormalities. RV angiography should be performed in four views: RAO 30, LAO 60, AP and lateral. Affected areas will show hypokinesis, akinesis, or dyskinesis. Wall motion abnormalities can also be observed by echocardiography but particular attention must be focussed on imaging the basal posterior and inferior walls near the tricuspid valve [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R14">14</xref>]. MRI is emerging as a useful non-invasive tool to identify morphologic abnormalities such as myocardial fat, right ventricular chamber size and shape, and wall motion abnormalities with newer cine techniques. However, with conventional MR technology, the identification of myocardial fat was not found to be a reliable predictor of the presence of ARVC/D [<xref ref-type="bibr" rid="R15">15</xref>].</p></sec><sec id="s2b"><title>b) Electrocardiographic Abnormalities</title><p>Electrocardiographic changes include inverted T waves in the right precordial leads beyond V1 (<xref ref-type="fig" rid="F3">Figure 3</xref>) in the absence of right bundle branch block [<xref ref-type="bibr" rid="R16">16</xref>,<xref ref-type="bibr" rid="R17">17</xref>]. Right ventricular late potentials in the form of epsilon waves may be found on the routine 12 lead ECG. If not present overtly, they may be identified by recording the ECG at double standard speed (50mm/sec) and amplitude (20mm/mV) and utilizing a 40 Hz filter [<xref ref-type="bibr" rid="R5">5</xref>] (<xref ref-type="fig" rid="F4">Figure 4</xref>) . Evidence of right ventricular parietal block is a QRS duration that is longer in the right (leads V1, V2 and V3) than in the left (V4, V5, V6) precordial leads. Suggested parameters to verify this include a localized QRS duration greater than 110ms in the right precordial leads, or a maximum QRS duration in leads V1, V2 or V3 or more than 25ms above the QRS duration in lead V6 in the presence of incomplete or complete right bundle branch block [<xref ref-type="bibr" rid="R18">18</xref>]. The signal averaged electrocardiogram is often abnormal and supports the diagnosis of ARVC/D, but if normal, does not exclude the diagnosis. The sensitivity of the signal averaged electrocardiogram can be improved by using a 25 Hz high pass filter and paying particular attention to the Z lead [<xref ref-type="bibr" rid="R19">19</xref>].</p></sec><sec id="s2c"><title>c) Arrhythmias</title><p>The ventricular arrhythmias generally arise from the right ventricle, and therefore have a left branch block morphology. Right ventricular outflow tract tachycardia (RVOT), must be excluded as this diagnosis carries a benign prognosis and is not hereditary5 . RVOT tachycardia has a left bundle branch block morphology. The QRS points to the right and inferior (positive in leads II, III, AVF and negative in AVL); this morphology can also be seen in ventricular tachycardia due to ARVC/D. Patients with ARVC/D may also experience supraventricular arrhythmias [<xref ref-type="bibr" rid="R20">20</xref>] including atrial flutter [<xref ref-type="bibr" rid="R21">21</xref>-<xref ref-type="bibr" rid="R23">23</xref>].</p></sec><sec id="s2d"><title>d) Genetics</title><p>Thirty to fifty percent of patients with ARVC/D will have evidence of familial disease [<xref ref-type="bibr" rid="R25">25</xref>,<xref ref-type="bibr" rid="R26">26</xref>]. The variable clinical expression and course is believed to be at least in part due to genetic heterogeneity. Inheritance is autosomal dominant, with the exception of Naxos disease, which is autosomal recessive. Linkage analysis in families with ARVD has revealed several loci for this disorder on chromosomes 1, 2, 3, 10, 14, and 17 [<xref ref-type="bibr" rid="R26">26</xref>-<xref ref-type="bibr" rid="R31">31</xref>] (see <xref ref-type="table" rid="T1">Table 1</xref>). The mutation for ARVD8 has been identified in a gene for desmoplakin on chromosome 6, a component of desmosomes that are cellular adhesion proteins responsible for cellular binding [<xref ref-type="bibr" rid="R1">1</xref>]. This is the first gene identified to cause a major form of the disease, with autosomal dominant inheritance. The locus on chromosome 1 leads to a form of ARVC/D that is notable for exertional polymorphic, instead of monomorphic, ventricular tachycardia [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R32">32</xref>] and has been further identified to encode a cardiac ryanodine receptor gene that is responsible for calcium release from the sarcoplasmic reticulum. This genetic mutation has also been implicated in familial catecholaminergic polymorphic ventricular tachycardia [<xref ref-type="bibr" rid="R33">33</xref>,<xref ref-type="bibr" rid="R34">34</xref>]. Naxos disease discovered on the Greek island, Naxos, is characterized by palmoplantar keratosis and woolly hair in addition to RV dysplasia. Naxos disease has been mapped to chromosome 17 involving a mutation in the gene that encodes plakoglobin, a component of desmosomes and adherens junctions responsible for maintaining tight adhesions between cells [<xref ref-type="bibr" rid="R2">2</xref>]. Another mutation in the gene for desmoplakin that also causes an autosomal recessive syndrome with woolly hair and a dilated left ventricle has been reported in patients from Ecuador [<xref ref-type="bibr" rid="R35">35</xref>]. Therefore, ARVD can result from a variety of genetic mechanisms, and in particular, appears to be a disorder of proteins involved in cellular adhesion and calcium release.</p><p>Not all cases of ARVD are inherited since mutations may occur sporadically. Nonetheless, because of the genetic basis for this disorder, it is recommended that first degree relatives of ARVD patients be screened with an ECG and echocardiogram [<xref ref-type="bibr" rid="R5">5</xref>]. Other non-invasive tests to identify affected family members include signal-averaged ECG, Holter monitor and exercise stress testing.</p></sec></sec><sec id="s3"><title>Other Contributory Factors</title><p>A large number of ARVC/D patients are noted to be athletes, and sudden cardiac death is more likely during exertion [<xref ref-type="bibr" rid="R4">4</xref>]. Athletic activity, possibly through stretch of myocardial fibers or elevated catecholamines may accelerate the degeneration of the ventricular myocardium and leads to ventricular arrhythmias. Also in support of the importance of other etiologic factors are the reports of Coxsackie virus B3 biopsy specimens [<xref ref-type="bibr" rid="R36">36</xref>], and recently, enterovirus and adenovirus by polymerase chain reaction in 7 of 12 specimens from sporadic cases [<xref ref-type="bibr" rid="R37">37</xref>]. This raises the possibility that the myocardium in ARVD is particularly vulnerable to inflammatory processes. A superimposed myocarditis could result in lymphocytic infiltrates and fibrofatty replacement of the myocardium [<xref ref-type="bibr" rid="R38">38</xref>,<xref ref-type="bibr" rid="R39">39</xref>]. It is not known whether the detection of viral DNA or RNA indicates that right ventricular cardiomyopathy is a result of viral myocarditis or this cardiomyopathy predisposes the heart to viral infection. This process may only affect the RV, but can affect both ventricles as well. Patients with left ventricular involvement tend to be older and more symptomatic from both arrhythmias and heart failure, consistent with the view that the disease is progressive [<xref ref-type="bibr" rid="R40">40</xref>].</p><p>Since there is no single "gold standard" diagnostic test to reliably verify this condition, a combination of diagnostic criteria have been established by the Task Force of the Working Group on Myocardial and Pericardial Disease [<xref ref-type="bibr" rid="R41">41</xref>]. Major criteria include demonstration of severe wall motion abnormalities, fibrofatty replacement by biopsy, epsilon waves and family history of histologically confirmed disease. Minor criteria include milder alterations of ventricular function, inverted T waves in the right precordial leads, late potentials, ventricular tachycardia with left bundle branch morphology or frequent premature ventricular complexes (>1000 PVCs in a 24 hour period), or family history of suspected or clinically diagnosed ARVC/D. A diagnosis can then be made if at least two major, 1 major and 2 minor, or 4 minor criteria from different categories are fulfilled Disease [<xref ref-type="bibr" rid="R41">41</xref>].</p><p>A prospective study of the course of disease in 37 affected families with 365 relatives in Italy showed that 41% could be diagnosed with certainty to have ARVC/D while in another 11% the diagnosis was uncertain Disease [<xref ref-type="bibr" rid="R42">42</xref>]. Patients were diagnosed mostly during adolescence and early adulthood. After a mean follow-up of 8.5 years, an additional 15 subjects developed structural abnormalities or ventricular arrhythmias. In an evaluation of relatives of 67 ARVC/D patients in the United Kingdom, 28% of patients had relatives that satisfied the Task Force criteria, but another 20% had relatives with only minor ECG, echo or Holter abnormalities suggestive of early or mild disease Disease [<xref ref-type="bibr" rid="R43">43</xref>]. Thus familial disease may have been present in up to 48% of index cases, which is plausible for a disease with autosomal dominant inheritance. It has been suggested by these authors that the standard diagnostic criteria should be modified to identify ARVC/D in relatives of patients with confirmed disease if they satisfy any one of the following abnormalities: i) T-wave inversion in the precordial leads ii) abnormal SAECG iii) ventricular tachycardia with left bundle branch morphology during ECG, Holter monitoring or exercise testing iv) greater than 200 PVCs in a 24 hour period or v) any mild echocardiographic abnormalities such as RV dilatation with preserved LV function Disease [<xref ref-type="bibr" rid="R42">42</xref>] .</p></sec><sec id="s4"><title>Treatment/Management</title><p>Therapy with beta blockers Disease [<xref ref-type="bibr" rid="R44">44</xref>], sotalol [<xref ref-type="bibr" rid="R44">44</xref>] or amiodarone [<xref ref-type="bibr" rid="R45">45</xref>] may be effective in suppressing ventricular arrhythmias and possibly in preventing sudden cardiac death. Implantation of an ICD may also be indicated to prevent sudden death. Catheter ablation of ventricular tachycardia may be useful in patients with refractory symptoms despite antiarrhythmic therapy. However, ventricular arrhythmias may recur from other areas. Patients should also be advised not to perform vigorous exercise or engage in competitive sports. Surgical disarticulation of the right ventricular free wall from its attachments to the left ventricle and septum can prevent the electrical propagation of ventricular arrythmias from the right to the left ventricle [<xref ref-type="bibr" rid="R46">46</xref>,<xref ref-type="bibr" rid="R47">47</xref>]. This was an effective means to prevent sudden death prior to the availability of the ICD, but resulted in severe right ventricular failure. Management of heart failure includes standard medical therapy with consideration of heart transplantation if severe ventricular, especially biventricular, dysfunction is present.</p><p>It is important to emphasize that risk factors for sudden death are not well characterized. A history of syncope, right or left ventricular abnormalities seen on radionuclide angiography and QRS dispersion greater than 40ms are independent predictors of sudden death [<xref ref-type="bibr" rid="R48">48</xref>]. However, the absence of these factors cannot provide absolute assurance that a patient or relative is free of risk. The diverse phenotypic expression of ARVC/D is still being appreciated as new genetic loci are discovered. Furthermore, the risk of sudden death is probably variable among different types of ARVC/D. In Newfoundland, Canada, ARVD5 was reported to cause sudden death in 44% of affected males, while females had a more benign course with no sudden deaths. Thus, determining prognosis is complicated by the genetic heterogeneity of this disorder.</p><p>It is unclear how to advise family members of individuals that have died suddenly with ARVD. The eight-year actuarial probability of developing ventricular arrhythmias in asymptomatic family members with normal echocardiograms was found to be 3%, with no sudden deaths [<xref ref-type="bibr" rid="R50">50</xref>]. The potential risks associated with the implantation and long-term management of an intracardiac defibrillator in family members with suspected disease but minimal right ventricular dysfunction may far outweigh its potential benefits. Counseling of family members needs to be individualized.</p><p>The Multidisciplinary Study of Right Ventricular Dysplasia is a multicenter study funded by the National Institutes of Health and National Heart, Lung, and Blood Institutes that is enrolling patients with a recent diagnosis of ARVC/D into a North American registry. A large registry of patients is needed to answer important questions related to diagnosis of the disorder, in particular in family members, as well as clinical management of affected patients and risk stratification for sudden death. This study will also facilitate genetic research to identify other specific mutations and their role in the pathogenesis of this disease [<xref ref-type="bibr" rid="R51">51</xref>,<xref ref-type="bibr" rid="R52">52</xref>].</p></sec>
Do All Children with Congenital Complete Atrioventricular Block Require Permanent Pacing ?	Could not extract abstract	<contrib contrib-type="author"><name><surname>Balmer</surname><given-names>Christian</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Bauersfeld</surname><given-names>Urs</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<p>With an incidence of 1 in 20,000 live born infants [<xref ref-type="bibr" rid="R1">1</xref>], congenital complete atrioventricular block (CCAVB) is a rare disease. The aetiology is not completely understood. However, CCAVB may be isolated or combined with congenital heart diseases in up to 53% of affected individuals [<xref ref-type="bibr" rid="R2">2</xref>]. Isolated CCAVB is in up to 98% of the children associated with positive autoimmune antibodies in the maternal serum (anti-Ro/SS-A and anti-LA/SS-B) [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R4">4</xref>]. Interestingly, these antibodies are not specifically directed against the conduction system but also against normal myocardial cells and may cause myocarditis [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R6">6</xref>]. Affection of the conduction system can occur at different levels [<xref ref-type="bibr" rid="R7">7</xref>]. Histologically, the atrioventricular node tissue may be replaced by fibrous fatty tissue with variable involvement of the distal conduction system [<xref ref-type="bibr" rid="R8">8</xref>].</p><p>The onset of clinical symptoms in patients with CCAVB is already antenatally in up to 28% [<xref ref-type="bibr" rid="R9">9</xref>], but can also occur only later in life. This is due to a variable degree of heart block and heart rate. Most of the symptoms are related to the slow heart rate: hydrops foetalis, heart failure of the neonate, exercise intolerance of the child. Longer pauses may cause praesyncope, syncope (classical Adams Stokes attacks) or even sudden cardiac death. Whether or not Cardiomegaly is mainly the result of a chronic compensatory increased stroke volume secondary to the slow heart rate is somewhat controversial [<xref ref-type="bibr" rid="R10">10</xref>]. Cardiomegaly may also be a distinct disease in a subgroup of patients because it does not necessarily resolve with pacemaker (PM) therapy [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R12">12</xref>]. Morbidity and mortality of CCAVB do not seem to correlate with antibody status or associated cardiac lesions [<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R14">14</xref>]. There are case reports, that fetal CCAVB can be improved with steroids [<xref ref-type="bibr" rid="R15">15</xref>]. Rarely, CCAVB resolves spontaneously [<xref ref-type="bibr" rid="R16">16</xref>-<xref ref-type="bibr" rid="R19">19</xref>]. In most patients, the degree of conduction abnormality will either persist or worsen over time.</p><p>There is no causal therapy for CCAVB. Permanent pacing is the only possible therapy. The aim of PM therapy is to prevent sudden cardiac death and to restore an adequate heart rate to avoid any secondary effects of longstanding bradycardia.</p><p>In our clinical practice, the ACC/AHA/NASPE guidelines for PM implantation proved to be useful [<xref ref-type="bibr" rid="R20">20</xref>]. Although one has to keep in mind that because of the rareness of the disease and the lack of large randomized trials, the level of evidence is usually low and most of these recommendations are based on experts opinions only.</p><p>The following patients with CCAVB should undergo permanent pacing (<xref ref-type="fig" rid="F1">1</xref>): First, all patients with advanced second- or third- degree AV block with symptomatic bradycardia, exercise intolerance or low cardiac output. In children, symptoms can be subtle. In young children there may be poor growth and development, sleep disturbances including night - terror and bed wetting, frequent naps compared to peers and preference for a more sedentary behavior. School children may show poor school performance secondary to tiredness, the need to nap or to go to bed very early after school, irritability, and inability to keep up with peers during exercise [<xref ref-type="bibr" rid="R13">13</xref>]. Patients and parents tend to accept and adapt to minor symptoms and recognize them only in retrospect. Often, restored normal heart rates and the chronotropic competence after PM implantation lead to an exciting and motivating gain in energy and perseverance, which can change the personality of the child from lazy and lethargic to energetic and quick.</p><p>Second, patients with ECG changes believed to be associated with an increased risk for life threatening arrhythmias such as an unstable escape rhythm, sinus tachycardia, wide QRS escape rhythm, prolonged QT interval or ventricular arrhythmia [<xref ref-type="bibr" rid="R20">20</xref>]. The level of evidence is best for the latter two [<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R21">21</xref>]. Long RR pauses are acceptable up to 3 seconds while awake or 5 seconds while sleeping [<xref ref-type="bibr" rid="R13">13</xref>]. Ambulatory ECG monitoring with frequent episodes of junctional exit block, flat junctional response or associated tachyarrhythmias have been noted in patients with adverse outcomes [<xref ref-type="bibr" rid="R22">22</xref>] but are not generally accepted as risk factors for sudden death.</p><p>Third, patients with cardiomegaly [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R23">23</xref>] or ventricular dysfunction [<xref ref-type="bibr" rid="R20">20</xref>]. In some patients, cardiomegaly is part of a dilated cardiomyopathy which can still progress in spite of PM therapy [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R12">12</xref>]. In most patients a decrease in left ventricular size and an improved myocardial function has been demonstrated after PM implantation [<xref ref-type="bibr" rid="R23">23</xref>].</p><p>The lower limits of an acceptable escape rhythm might be around 55 bpm in neonates and 50 bpm in patients older than one year with the patient being awake and at rest [<xref ref-type="bibr" rid="R20">20</xref>]. In a prospective study of 27 patients with CCAVB and a mean age of 6.8 years 8 of 13 patients with mean daytime rates below 50 bpm in ambulatory ECG monitoring had cardiac complications such as sudden cardiac death (n=3), syncope, praesyncope or excessive fatigue [<xref ref-type="bibr" rid="R22">22</xref>].</p><p>The indication for a PM implantation is also given in patients with CCAVB and significant congenital heart diseases (i.e. those with cyanotic heart disease, single ventricle physiology, complete AVSD or significant AV valve disease) [<xref ref-type="bibr" rid="R13">13</xref>]. Again, this recommendation is not strictly evidence based but in these children, the haemodynamic status is often jeopardized from many reasons and restoring a normal heart rate and AV synchrony is often followed by a general improvement and well being.</p><p>Unfortunately, a large number of patients with CCAVB do not fall in the categories mentioned above and need an individual assessment. In most of these asymptomatic patients without specific changes in ECG or echocardiography we would still follow a permanent pacing strategy. We have to balance carefully the risks and disadvantages of a PM implantation against the risk of sudden cardiac death. The acceptance of a permanent PM by patients as well as parents is usually extremely good. There are however some specific age and size related particularities.</p><p>We try to achieve a haemodynamic situation as close as possible to the normal heart rhythm. Therefore we tend to implant dual chamber PM in most of the children. The miniaturizing of pulse generators nowadays allows for implanting dual chamber pacing systems in babies down to 1600g [<xref ref-type="bibr" rid="R24">24</xref>]. In infants and small children we do have very good experiences with bipolar steroid-eluting epicardial pacing leads, attached to the right atrium and right ventricle by a median sternotomy (<xref ref-type="fig" rid="F2">2</xref>) or fixed to the left atrium and left ventricle after a lateral thoracotomy. Since all of the children will need their PM for a lifetime, we push the age limit for transvenous pacing system towards preschool age in order to avoid intimal trauma and obstruction of large veins in order to preserve them for later PM - systems. It is necessary to leave enough length to the leads to allow for growth of the patient. We prefer devices with automatic threshold measurement and output adjustment (AutoCapture) to increase patient safety and to extend device service life since most of the patients will be paced continuously at relatively high rates [<xref ref-type="bibr" rid="R25">25</xref>]. Although some children are still asymptomatic during childhood most of the patients with CCAVB will have PM therapy when reaching adulthood [<xref ref-type="bibr" rid="R19">19</xref>].</p><p>Patients with CCAVB without PM do need a close, at least yearly, follow-up with 12 lead ECG, 24 hour Holter monitoring and echocardiography. Despite the absence of symptoms in childhood, 50% of the patients will develop symptoms in adulthood and 10% will die prematurely [<xref ref-type="bibr" rid="R17">17</xref>-<xref ref-type="bibr" rid="R19">19</xref>]. With regards to the ECG it is important to know that not only the AV conduction can worsen over time but also prolonged QTc interval can appear the first time in adulthood in patients who previously had normal QTc [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R19">19</xref>].</p><p>Patients with CCAVB may present with similar symptoms but in fact it is a mixed group of diseases with a variety of yet not clearly defined aetiologies. Therefore it is very difficult to predict the natural history. There is a lack of evidence with regards to risk factors predicting sudden cardiac death or syncope in patients with CCAVB. With newer PM technologies the indications for PM therapy in children have been extended over the last years so that patients without symptoms and without specific signs in ECG or echocardiography may also benefit from the advantages of a permanent PM.</p>
Brugada-like Precordial ST Elevation on ECG by Anterior Mediastinal Infective Mass Lesion	Could not extract abstract	<contrib contrib-type="author"><name><surname>Nakazato</surname><given-names>Yuji</given-names></name></contrib><contrib contrib-type="author"><name><surname>Ohmura</surname><given-names>Takayasu</given-names></name></contrib><contrib contrib-type="author"><name><surname>Shimada</surname><given-names>Issei</given-names></name></contrib><contrib contrib-type="author"><name><surname>Daida</surname><given-names>Hiroyuki</given-names></name></contrib>	Indian Pacing and Electrophysiology Journal	<p>Several causes are known to induce the right precordial ST elevation mimicking Brugada syndrome. Right ventricular outflow area is assumed to be responsible for such ECG changes. We experienced a case of anterior mediastinal infective mass lesion with a Brugada-like ECG.</p><p>A 52-year-old female, who has pulmonary stenosis and recurrent episodes of right ventricular heart failure, complained of high fever, abdominal discomfort, and edema. On physical examination, jugular vein dilation, hepatomegaly, and facial and leg edema were noted. Leucocytosis was also noted on blood examination. An ECG showed right ventricular hypertrophy, incomplete right bundle branch block pattern and marked ST elevation on precordial leads mimicking Brugada syndrome. Magnetic resonance imaging revealed an abnormal mass shadow located on the anterior mediastinum and compressing the right ventricle (<xref ref-type="fig" rid="F1">Figure 1A</xref>). Trans-thoracic echocardiography also showed the high echogenic mass lesion at the anterior side of right ventricle and the vicinity of pulmonary valve. After treatment with antibiotics, the mass lesion gradually shrunk. Concomitantly, the ST elevation disappeared with improvement of inflammatory markers (<xref ref-type="fig" rid="F1">Figure 1B</xref>). The symptoms suggesting right ventricular failure were also ameliorated. The mechanism of Brugada-like ST elevation in this patient was considered to be compression, by the abnormal infective mass, of the right ventricular outflow tract with/without focal pericardial inflammation.</p>
The fallacies of QT correction	Could not extract abstract	<contrib contrib-type="author"><name><surname>Lokhandwala</surname><given-names>Yash</given-names></name><degrees>MD, DM</degrees><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Toal</surname><given-names>SC</given-names></name><degrees>MD, DM</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<p>“Not to correct QT, but how to, that is the question”. The QT interval is a reflection of the action potential in the cardiac cells. Homogenous or heterogenous changes in the action potential duration lead to alteration of QT interval (in addition to morphological changes of T & U waves) [<xref ref-type="bibr" rid="R1">1</xref>]. Such changes can be due to change in heart rate & autonomic tone. They can also be markers of abnormal repolarization, depolarization or both as a result of electrolyte disturbances, cardiac diseases, drugs and congenital long QT syndromes [<xref ref-type="bibr" rid="R2">2</xref>].</p><p>Repolarization disorders are responsible for life threatening arrhythmias like torsades de pointes [<xref ref-type="bibr" rid="R2">2</xref>]. The purpose of heart rate correction is to obtain a standardized value that would have been measured in the same subject if the heart rate was 60 beats per minute (QTc). Thus this QTc value will now become independent of the heart rate and measure replarization changes. It will thus be a surrogate marker of the risk of torsade de pointes.</p><p>The concept of QTc appeared in 1920, when Bazett introduced his square root formula [<xref ref-type="bibr" rid="R3">3</xref>]. This formula obtained from data on 39 young men has been questioned because it overcorrects QT at fast heart rate and undercorrects at low heart rate [<xref ref-type="bibr" rid="R4">4</xref>]. Thus at slow heart rate, which is one of the predisposing factors of torsade initiation, Bazett correction can easily mask substantial QT prolongation by under correcting. This can hide the proarrhythmic toxicity of drugs slowing heart rate. An alternative, cube root correction of Fridericia, corrects better than Bazett but again is not reliable at fast heart rates. Compared to these non linear correction formulae, linear regression correction obtained from large population data, like the Framingham heart study linear correction are still better [<xref ref-type="bibr" rid="R4">4</xref>].</p><p>However the QT does not adapt to changes in heart rate immediately. It takes more than 2 minutes for the QT to adapt (QT/RR hysteresis). Hence correction needs to be done at steady heart rates [<xref ref-type="bibr" rid="R1">1</xref>]. The concept of heart rate correction ignores the dynamicity of QT/RR relationship [<xref ref-type="bibr" rid="R6">6</xref>]. The QT interval is also under autonomic control. Therefore different modes of heart rate changes, e.g. fast heart rate due to parasympathetic withdrawal versus sympathetic overactivity, lead to different direct and reflex effects on QT prolongation. Hence the standard QTc correction formula will not be representative of the actual repolarization milleu. Ideally each individual should have his own correction derived from multiple ECGs at different heart rates and conditions to get the ideal correction constant. Since this is not feasible a compromise can be made by using formula which have been developed and validated in a large population based cohort like the Framingham study [<xref ref-type="bibr" rid="R1">1</xref>].</p><p>Alternatively, a table of lower and upper limits of QT interval for different RR cycle lengths can be used by clinicians for references purpose [<xref ref-type="bibr" rid="R4">4</xref>]. This method obviates the need of using any QT correction formula. The reliability of this model results from its derivation in a large population based sample. Inspite of all the fallacies of Bazett’s correction, it is still being used clinically. The reason probably lies in its simplicity and the fact that all clinical data signaling risk of torsade are derived from this formula only. But in the near future we feel that the Bazett’s correction will be replaced by a better formula like Fridericia. We also need to be aware that in addition to prolonged QT, an abnormally short QT can also carry dangerous implications of arrhythmogenecity [<xref ref-type="bibr" rid="R5">5</xref>]. The QT also represents depolarization events and hence in the presence of depolarization abnormalities like Left Bundle Branch Blocks (LBBB) and Preexcitation syndromes, QTc will not be representative of repolarization abnormalities and should not be commented upon. An alternative interval, the “JT” has been proposed in such cases [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>In light of all the above fallacies a well designed study looking at different correction formulae and defining one which clinically can risk stratify patients at risk of torsade best, is the need of the hour. There are limitations not only in correcting QT, but also in how it is measured and the reproducibility of these measurements. Therefore, it is possible that in the future the measurement of QT interval may just become an adjunctive, being replaced by more objective and reproducible signs of repolarization abnormalities. Till than ask not why QT, but how to correct QT.</p>
Optical Mapping Of Cardiac Arrhythmias	Could not extract abstract	<contrib contrib-type="author"><name><surname>Arora</surname><given-names>Rishi</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Das</surname><given-names>Mithilesh K</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Zipes</surname><given-names>Douglas P</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Wu</surname><given-names>Jianyi</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Recent Advances In Mapping Of Cardiac Arrhythmias</title><p>The concept of mapping rhythmic activation of the heart dates back to the beginning of last century, with initial descriptions of reentry in turtle hearts [<xref ref-type="bibr" rid="R1">1</xref>], to the first systematic mapping of sinus rhythm and then atrial flutter by Lewis et al [<xref ref-type="bibr" rid="R2">2</xref>]. Barker et al [<xref ref-type="bibr" rid="R3">3</xref>] were the first to map the human heart. Initial mapping was primarily performed using single probes to record activation in different regions of the heart. The 1960’s and 70’s saw the development of computerized mapping of the human heart, e.g. in the cure of Wolf-Parkinson-White syndrome as well as in the study of Langendorff preparations [<xref ref-type="bibr" rid="R4">4</xref>]. In fact, most of the recent advances in cardiac mapping have focused on improvements in multisite recordings within the heart, with the ability to simultaneous record electrical activation from several hundreds of sites having contributed significantly to our understanding of atrial and ventricular arrhythmias.</p><p>Despite these recent advances, multisite contact mapping suffers from several limitations, including the technical problems associated with amplification, gains, sampling rates, signal-to-noise ratio, and the inability to see signals during high-voltage shocks. In addition, an intrinsic limitation of current mapping techniques is their inability to provide information about repolarization characteristics of electrically active cells, thereby limiting our ability to study entire action potentials. In fact, intracellular microelectrode recordings are still considered the gold standard for the study of action potential characteristics in whole tissue. Microelectrode techniques are limited however, by an inability to record action potentials from several sites simultaneously, thereby precluding their use in high-density activation mapping.</p><p>In part due to the above-mentioned limitations, the last few years have seen the development and use of voltage-sensitive dyes as a means to map not only activation, but repolarization as well. Voltage-sensitive dyes, when excited, provide an optical signal that mimics an action potential and thus allows the visualization of both activation and recovery processes in any region under view. This allows one to precisely evaluate the propagation of a wave of excitation and to measure its wavelength visually.</p><p>Optical mapping techniques use imaging devices such as a photodiode array or a charge-coupled device video camera with the heart being illuminated and either continuously or spatially scanned. The basis for these techniques is the use of voltage-sensitive dyes that bind to or interact with cell membranes.</p></sec><sec id="s2"><title>Voltage-Sensitive Dyes</title><p>Voltage-sensitive dyes are molecules that bind to the cell membrane with high affinity. While bound to the cardiac cell membrane, the dye molecules fluoresce light in direct proportion to transmembrane voltage. Therefore, voltage-sensitive dyes function as highly localized transducers of membrane potential, transforming a change of membrane potential into a change in fluorescent intensity [<xref ref-type="bibr" rid="R5">5</xref>]. For any given constant excitation light intensity and wavelength, light is emitted by voltage-sensitive dyes over a range of wavelengths that can be represented by emission spectra; the emission spectra for a voltage-sensitive dye molecule changes with membrane potential. A suitable filter then passes light only above a certain wavelength, with the amount of light changing at different membrane potentials, thereby allowing the generation of an optical action potential. Importantly, the precise shape of the emission spectra (and therefore the optical action potential) does not correspond to any absolute voltage, and only relative potential change is detected. Also, an optical action potential represents a multicellular spatial average of transmembrane potential from cells within a volume of tissue [<xref ref-type="bibr" rid="R6">6</xref>]. Thus, as magnification decreases (1 x), a decrease in the rate of rise of the optical action potential upstroke may occur due to spatial averaging. However, at higher magnifications, the upstroke of the optical action potential approaches that measured from a single cell [<xref ref-type="bibr" rid="R7">7</xref>].</p><p>The styryl dye, Di-4-ANNEPS, is now the probe of choice for a number of laboratories that are studying complex patterns of activity in the heart. This dye works by the principle of electrochromism, whereby when provoked by excitation light, the molecule undergoes a charge shift from the ground state to the excited state; fluorescence occurs due the emission of a photon during transition from the excited to the ground state. The wavelength of the emitted photon is determined by the change in molecular energy when passing from the excited state to the ground state, less the energy lost in the process. Di-4-ANNEPS possesses several advantages, compared to most other optically sensitive dyes; it exhibits large fractional fluorescence changes during an action potential (8-15%) with low toxicity and photobleaching, and its signal amplitude and kinetics are stable for 2 to 4 hours without restaining the preparation [<xref ref-type="bibr" rid="R8">8</xref>].</p></sec><sec id="s3"><title>Optics And Photodectors</title><p>As mentioned earlier, voltage-sensitive dye must be excited by light to induce fluorescence. The most common excitation light sources are tungsten-halogen lamps, mercury arc lamps and argon ion lasers. The source of each optical action potential is derived from as few as one to as many as hundreds of cardiac cells, depending on the extent of optical magnification used. Therefore, a single optical action potential represents the average potential from a small aggregate of neighboring cells.</p><p>High-quality images can be acquired with a microscope or with photographic lenses. Fluorescence microscopes are commercially available with objective lens magnifications ranging from 4x to 100x, thereby allowing mapping of very small preparations (1 to 5 mm). Photographic lenses are better suited for magnifications under 10x, permitting one to map action potentials from larger preparations (5 to 50 mm). A simple photographic lens system is shown schematically in <xref ref-type="fig" rid="F1">Figure 1</xref>. For very small preparations, laser-scanning systems can be used with a single photodetector, without the need for a collector lens.</p><p>The most widely used photodetectors in optical mapping are photodiode arrays and charged coupled device (CCD) video cameras. Both types of photodetectors are similar in that they transduce light energy into electricity. When photons of sufficient energy strike a detector material, electron-hole pairs are created (i.e. the photoelectric effect). Photodiode arrays typically consist of several hundred individual photodiodes, and are configured to instantaneously convert photoexcited charge carriers to current flow (i.e. photocurrent). The magnitude of photocurrent is directly proportional to the light intensity falling on a single photodiode element, and is converted to a voltage signal using a current-to-voltage amplifier; this voltage signal has a amplitude that is proportionate to membrane potential. A major advantage of a photodiode array is that it generates photocurrent in response to membrane potential changes continuously, allowing one to digitally sample the action potential at very rapid rates (i.e. high sampling rates), without compromising the fidelity of the recorded action potential. As a result, these detectors are very useful for measuring details in the time course and morphology of the action potential. They are limited by the number of photodetectors on the array (typically 256 or less).</p><p>Unlike a photodiode array, a CCD camera can contain hundreds of thousands of pixels, permitting greater spatial resolution between recording sites. CCD detectors differ from photodiodes in that photoexcited charge carriers are collected within a single pixel over a finite period of time (integration time), and are read off at regular time intervals (i.e. the frame rate). Given that a CCD array contains several hundred thousand pixels and that each pixel is read sequentially, readout time is a major factor that limits the sampling rate of CCD systems. However, because of a larger number of recording sites (pixels), CCD detectors are better suited for mapping larger preparations at low magnification so that spatial resolution of 1 mm or less can be attained.</p></sec><sec id="s4"><title>Role Of Optical Mapping In The Understanding The Substrate For Cardiac Arrhythmias</title><p>In optical mapping, the time course of membrane potential is registered at every recording site. Therefore, it is possible to relate complex propagation patterns to voltage changes occurring at the cellular and subcellular level. Studies performed at the level of the single cell have shed important insights into the nature of intracellular propagation within single cardiomyocytes [<xref ref-type="bibr" rid="R9">9</xref>], as well as impulse propagation between cardiomyocytes [<xref ref-type="bibr" rid="R10">10</xref>], and have contributed greatly to our understanding of the role of cell-to-cell coupling, structural discontinuities, and tissue anisotropy in propagation of the electrical impulse [<xref ref-type="bibr" rid="R11">11</xref>]. More recently, optical mapping has also created new frontiers for investigating the physiology of very small but important structures such as the atrioventricular node [<xref ref-type="bibr" rid="R12">12</xref>].</p><p>Since reentrant cardiac arrhythmias may occur over various spatial scales, ranging from microreentry to rotors giving rise to spiral waves that encompass a whole atrium or ventricle, an effective mapping system should cover a relatively large area with closely spaced recording sites. Recent application of optical mapping techniques to the study of cardiac fibrillation has given us experimental proof of several theories of wave propagation in excitable media [<xref ref-type="bibr" rid="R13">13</xref>-<xref ref-type="bibr" rid="R17">17</xref>], such as the high-frequency reentrant sources that underlie fibrillation and generate spiraling waves that propagate throughout the ventricles in complex patterns [<xref ref-type="bibr" rid="R18">18</xref>]. Optical recording of the cardiac wavefront during reentry has demonstrated that conduction depends upon the curvature of the spiral wave, i.e., a greater degree of conduction slowing is noted in the presence of a more pronounced curvature. Cardiac excitation is enhanced when the curvature of the wavefront is less as compared to a highly convex wavefront [<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R20">20</xref>] Using optical mapping, Laurita et al [<xref ref-type="bibr" rid="R21">21</xref>] have recently demonstrated how the spatial and temporal heterogeneity in action potential duration and repolarization lead to concordant and discordant alternans and ultimately reentrant arrhythmias.</p></sec><sec id="s5"><title>Recent Insights Into The Mechanisms Of Specific Arrhythmias</title><sec id="s5a"><title>AVNRT</title><p>From our institution, Wu et al have recently defined the mechanisms underlying atrioventricular nodal conduction and the reentrant circuit of atrioventricular nodal reentrant tachycardia using optical mapping [<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R22">22</xref>]. In contrast to long-standing concepts derived from traditional multisite mapping techniques-that propose the presence of discrete dual pathways within the atrioventricular node-optical mapping data suggest that atrial inputs into the AV node are from nondiscrete pathways, which can be arbitrarily divided into the fast (FP), intermediate (IP) and slow pathways (SP) (<xref ref-type="fig" rid="F2">Figure 2C</xref>). In fact, the reentrant circuits of different types of AVNRT were observed directly. The reentrant circuit of the slow/fast type started counterclockwise with block in the FP, conduction delay in the SP connection to the compact AV node, then exit from the AV node to the FP, and rapid return to the SP through the atrial tissue located at the base of Koch’s triangle (<xref ref-type="fig" rid="F2">Figure 2D</xref>). The reentrant circuit of the fast/slow type was clockwise. In the slow/slow type, anterograde conduction was over the IP and retrograde conduction was over the SP.</p><p>The new findings derived from these mapping data include the following: (1) the reentrant circuit for the slow/slow echo beat was different from that previously proposed. The term slow/slow type of AVNRT should be corrected to intermediate/slow type because multiple SPs were not observed to be involved. Thus, it is unlikely that the same SP could be used for anterograde and retrograde conduction. (2) Atrial tissue surrounding Koch’s triangle was clearly involved in all three types of AVNRT and suggested no upper common pathway. (3) Despite different reentrant circuits, the SP was always involved, explaining why ablation of the SP is effective in all types of AVNRT.</p></sec><sec id="s5b"><title>Pulmonary Vein Electrophysiology</title><p>The pulmonary veins have recently been shown to harbor electrically active foci that can initiate and maintain atrial fibrillation [<xref ref-type="bibr" rid="R23">23</xref>] [<xref ref-type="bibr" rid="R24">24</xref>]. The electrophysiology of these focal “drivers” and “triggers” is not well understood however, in large part due to the inability to study this small region of the heart with traditional recording techniques. Very recently, we have utilized high-resolution optical mapping to study this hitherto obscure region, and have demonstrated regions of slow conduction as well as heterogeneity of repolarization within canine pulmonary veins [<xref ref-type="bibr" rid="R25">25</xref>]. In fact, the interplay of slow conduction and heterogeneous repolarization-best studied by a technique such as optical mapping, that allows simultaneous imaging of tissue activation of and repolarization-creates substrate for leading circle reentry within the pulmonary veins. <xref ref-type="fig" rid="F3">Figure 3</xref> shows an example of a reentrant beat, in response to an extrastimulus, with the entire cycle length shown as a propagation movie.</p><p>In this study, we also demonstrated focal activity within the pulmonary veins, the underlying mechanism of which appeared to be triggered activity. Both the epicardium as well as the endocardium of the pulmonary vein were studied; while reentry was seen only on the epicardial aspect of the vein, focal activity was seen only on the endocardium, close to the ostium of the vein.</p></sec></sec><sec id="s6"><title>Direct Visualization Of VT Reentrant Circuit Using Optical Mapping</title><p>Reentry is the underlying mechanism for post-infarction ventricular tachycardia (VT) and the critical element of the reentrant circuit is the protected isthmus. Both anatomic and functional reentry have been proposed as potential models for such reentrant tachyarrhythmias. In anatomic reentry, the isthmus is fixed and located between two infarction scars or between an infarction scar and an anatomic obstacle such as the mitral valve annulus. In contrast, the isthmus in functional reentry is not fixed and is bounded by two functional lines of block located in the infarction border zone.</p><p>Using optical mapping in isolated canine left ventricular preparation, the reentrant circuit of monomorphic VT after acute myocardial infarction can be directly visualized (<xref ref-type="fig" rid="F4">Figure 4</xref>). The reentrant circuit consisted of four components: 1) a isthmus with slow conduction, located within the border zone between the infarction area and the functional line of block; 2) an entrance site located at the beginning of the isthmus; 3) an exit site located at the other end of the isthmus; and 4) an outer loop consisting of a wide region of non-ischemic normal tissue, connecting the exit and entrant sites outside the border zone. These data suggest that the reentrant circuit is the combination of functional and anatomic reentry. This unique model might contribute to some of the VT seen in patients with myocardial infarction.</p></sec><sec id="s7"><title>Future Directions And Clinical Implications</title><p>As mentioned above, optical mapping was developed as a part of a quest to simultaneously map the activation as well as the recovery of electrically active tissues. The use of voltage-sensitive dyes, photodiodes, laser scanning and CCD cameras have made it possible to record electrical activity in tissue preparations as well as in a beating heart [<xref ref-type="bibr" rid="R26">26</xref>-<xref ref-type="bibr" rid="R28">28</xref>]. Several of the basic principles of clinical cardiac electrophysiology have been confirmed by optical mapping techniques, such as the demonstration that pacing at a faster cycle length shortens tissue repolarization, whereas delivery of a premature stimulus results in greater conduction slowing, resulting in a propensity towards genesis of cardiac re-entrant arrhythmias [<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R29">29</xref>,<xref ref-type="bibr" rid="R30">30</xref>].</p><p>In addition to the insights that have been gained into the understanding of specific arrhythmias, the role of optical mapping has been recently extended by Garrigue and colleagues to study the effects of pacing in intact hearts. They studied the role of voltage output, interventricular delay, and pacing sites in ventricular arrhythmia occurrence during biventricular pacing (biventricular pacing is known to alleviate heart failure by restoring electrical, and therefore mechanical, synchrony between the ventricles); they utilized optical mapping in Langendorff-perfused guinea pig hearts to measure ventricular activation time and to examine conduction patterns during multisite pacing from 4 left ventricular and 3 right ventricular sites [<xref ref-type="bibr" rid="R27">27</xref>]. Myocardial ischemia was produced by gradually halving the perfusion output over 5 minutes. The optimal biventricular pacing sites were determined to be the RV apex and the base of the LV anterior wall; these sites were associated with the most homogeneous and rapid activation pattern (28 +/- 9 vs 41 +/- 12 ms with the other configurations, P < 0.01). There was no inducible arrhythmia. The role of high pacing output in the genesis of ventricular tachycardia was demonstrated in six hearts; ventricular tachycardia could be induced with pacing from the right and left free walls with 20 ms of interventricular delay, at six times the pacing threshold output. In four hearts, biventricular pacing (simultaneous right and left ventricular pacing) at high voltage output induced ventricular fibrillation with complex three-dimensional propagation patterns, independently of the pacing sites. The authors concluded that during biventricular pacing with ischemia, pacing at high voltage output with a long interventricular delay is likely to induce ventricular arrhythmias, particularly when left and right ventricular pacing results in a conduction pattern orthogonal to the ventricular myocardial fibers orientation. This study has indicated that it may be possible to record epicardial activation with optical mapping in determining the optimum site during left ventricular epicardial lead placement via thoracotomy. These experiments have also shown the feasibility of epicardial optical mapping in an intact heart and opened the way for further clinically relevant studies in. Although the mapping of endocardial activation in intact hearts is technically difficult at present, it is possible that with the advancement of echocardiographic techniques (such as intracardiac ultrasound) as well as further developments in optical mapping techniques, that intracardiac endocardial activation mapping may indeed become a reality in the not-too-distant future. Optical mapping is also expected to further our understanding the mechanisms of triggered and automatic activity, as well as the role of heterogeneous conduction and discordance of repolarization alternans in the genesis of reentry. A better understanding of the underlying mechanism of an arrhythmia would significantly help cardiologists tailor therapeutic strategies (whether ablative or pharmaceutical) to the individual patient.</p></sec>
Genetic basis and pathogenesis of Familial WPW Syndrome	Could not extract abstract	<contrib contrib-type="author"><name><surname>Sidhu</surname><given-names>Jasvinder</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Roberts</surname><given-names>Robert</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>The Wolff-Parkinson-White (WPW) syndrome has been a known clinical entity for over fifty years. In 1967 Durrer et al postulated WPW syndrome was due to an accessory pathway, bypassing the AV node, from the atria to the ventricles [<xref ref-type="bibr" rid="R1">1</xref>]. This was later confirmed by epicardial mapping. WPW is the second most common cause of paroxysmal supraventricular tachycardias in the western world and the most common cause in China [<xref ref-type="bibr" rid="R2">2</xref>]. WPW syndrome has a prevalence of 1.5 to 3.1 per 1000 persons in western countries [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R5">5</xref>] . Patients with the WPW syndrome may present with palpitations, presyncope, syncope, or sudden cardiac death (SCD). In some patients the first and only manifestation of the disease is SCD. This is more likely to occur in the setting of atrial fibrillation with a rapid ventricular response.</p><p>Electrocardiographic findings of WPW syndrome consist of preexcitation manifested by a shortened PR interval (<120 msec), a widened QRS (>100 msec) a delta wave (abnormal initial QRS vector) and supraventricular tachycardia. The EKG finding of preexcitation is a result of early ventricular depolarization through the accessory pathway. SVT can occur with retrograde or antigrade conduction through the accessory pathway forming the basis of the so-called “macro-reentrant” arrhythmia model [<xref ref-type="bibr" rid="R6">6</xref>]. Rapid conduction through the accessory pathway in the setting of atrial fibrillation markedly increases the risk of SCD [<xref ref-type="bibr" rid="R7">7</xref>].</p><p>We recently identified the gene responsible for familial Wolff-Parkinson-White [<xref ref-type="bibr" rid="R8">8</xref>]. The gene (PRKAG2) which encodes for a protein AMPK (AMP- activated protein kinase) was identified as the causal gene. Missense (single nucleotide change) mutations in this gene were identified in families with WPW. Six such mutations have been identified. Along with preexcitation the affected families had conduction abnormalities (AV block) and cardiac hypertrophy.</p></sec><sec id="s2"><title>Genetics of WPW</title><p>Familial WPW syndrome has an autosomal dominant mode of inheritance. In autosomal dominant inheritance 50% of the offspring inherit the mutated gene and are at risk of developing the disease and affects males and females equally.</p><p>In autosomal dominant mode of inheritance only one copy of the gene has to be mutated to cause the disease as opposed to autosomal recessive where both copies must have the mutations. We first performed genetic linkage analysis to locate the chromosomal location (locus) followed by the candidate gene approach to identify the gene in a family with an inherited form of WPW. The locus was on chromosome 7 (7q3) and the gene identified to be PKRAG2 which encodes for the gamma-2 subunit (non-catalytic subunit) of AMPK. Since this initial discovery five more mutations in the same gene have been identified. All of the mutations have been missense mutations in the PKRAG2 gene. Missense mutations are mutations in which a single nucleotide is substituted by another nucleotide inducing a change of a single amino acid in the final protein product <xref ref-type="table" rid="T1">Table 1</xref>).</p><p>AMPK is a protein that is made up of three subunits an alpha subunit, beta subunit (38 kDa) and a gamma subunit (36 kDa). PKRAG2 is the gene that encodes for the gamma-2 subunit of AMPK and it contains 16 exons (the region of the gene that is translated into amino acids) and is made up of 569 amino acids.</p></sec><sec id="s3"><title>Functions of AMPK</title><p>AMPK is a heterotrimeric protein which consists of alpha, beta, and gamma subunits. The gamma subunit has been characterized as the non-catalytic subunit. There are at least three isoforms of the gamma subunit (PKRAG1, PKRAG2, and PKRAG3). PKRAG1 and PKRAG2 are highly expressed in skeletal and cardiac muscle, whereas PKRAG3 expression is restricted to skeletal muscle [<xref ref-type="bibr" rid="R9">9</xref>]. Familial WPW syndrome is caused by mutations in the gamma-2 isoform (PRKAG2) of AMPK.</p><p>AMP-activated protein kinase (AMPK) serves as an energy sensor for the cell [<xref ref-type="bibr" rid="R9">9</xref>]. In times of energy depletion and/or stress it is activated. Stressors that have been known to activate AMPK include hypoxia, ischemia, exercise, and starvation [<xref ref-type="bibr" rid="R10">10</xref>]. This enzyme is activated by high AMP/ATP ratios and deactivated by low AMP/ATP ratios [<xref ref-type="bibr" rid="R10">10</xref>]. In energy depleted states (low ATP, high AMP) it is phosphorylated by AMPKK and is transformed to its active form. This process is reversed when ATP levels are restored. ATP interferes with the AMP/AMPK interaction and diminishes the ability of AMPKK to phosphorylate AMPK. Therefore it is the ratio of AMP/ATP that determines the level of AMPK activation.</p><p>AMPK has a multitude of physiological roles. Primarily it increases fatty acid oxidation and glucose uptake during times of stress [<xref ref-type="bibr" rid="R10">10</xref>] and it also decreases glycogen synthesis. AMPK inhibits acetyl-coA carboxylase it leads to a decrease in the synthesis of malonyl CoA. Since malonyl CoA is an inhibitor of fatty acid oxidation this leads to increased fatty acid oxidation. AMPK increases GLUT-4 a protein responsible for transport of glucose from the extracellular space to the intracellular space, thereby making glucose available for ATP production. In vitro studies have demonstrated that activation of AMPK inhibits glycogen synthase thus decreasing the amount of intercellular glycogen. These actions of AMPK classify it as a catabolic enzyme that responds to stress to maintain energy homeostasis.</p></sec><sec id="s4"><title>Clinical Characteristics of WPW syndrome</title><p>WPW can present clinically with palpitations, pre-syncope, syncope or SCD (sudden cardiac death). Familial WPW syndrome was described in a large French-Canadian family in 1986 [<xref ref-type="bibr" rid="R11">11</xref>]. In this family the members that were affected showed clinical findings that consisted of preexcitation, conduction abnormalities, and cardiac hypertrophy. We were fortunate to obtain access to the family and identify the responsible gene.</p><p>Preexcitation is defined as early depolarization of the ventricles. This early depolarization is manifested on a 12-lead EKG, as a short PR interval (<120 msec), and prolonged QRS (>100 msec), with the appearance of a delta wave (slurring of the QRS). Adult cardiac conduction begins at the SA-node and proceeds through the atria, AV-node, HIS-bundle, left/right bundles, throughout the ventricles. Patients with familial WPW syndrome have an accessory pathway that bypasses the normal cardiac conduction. The accessory pathway is usually an accessory conduction bundle connecting the atria and the ventricles. Hence the normal delay that occurs at the AV-node can be avoided by conduction through the accessory pathway. The presence of an accessory pathway also provides a reentrant circuit and the risk of developing SVT (supraventricular tachycardias). The development of SVT in the presence of preexcitation constitutes the WPW syndrome [<xref ref-type="bibr" rid="R12">12</xref>].</p><p>Familial WPW syndrome differs from other forms of WPW syndrome in that affected individuals also develop cardiac conduction abnormalities usually around the fourth decade of life [<xref ref-type="bibr" rid="R8">8</xref>]. Conduction abnormalities involve AV-nodal, ventricular as well as the accessory pathway. Treatment of patients often requires placement of a permanent pacemaker. It is thought this slowing is secondary to diffuse progressive myopathic process [<xref ref-type="bibr" rid="R13">13</xref>].</p><p>Family members afflicted with familial WPW syndrome also develop thickening of the myocardium. Cardiac hypertrophy has been demonstrated in these patients by ultrasound. The development of cardiac hypertrophy secondary to pressure overload (chronic hypertension, aortic stenosis) and sarcomeric mutations (HCM) are due to increased cell growth [<xref ref-type="bibr" rid="R14">14</xref>]. The hypertrophy found in families with WPW is also attributed to increased cell growth. It has been suggested that deposition of a glycogen-like substance may contribute to cardiac hypertrophy.</p></sec><sec id="s5"><title>Pathogenesis of Familial WPW syndrome</title><p>Families with WPW syndrome exhibit a variable phenotype consisting of cardiac hypertrophy, preexcitation, and conduction abnormalities. The syndrome is caused by missense mutations (single nucleotide change) in the gene PKRAG2 which encodes for the gamma-2 subunit of AMPK. There has been an explosion of knowledge on AMPK within the last decade. Initially it was a surprise that a gene that encodes a metabolic protein was involved in the generation of cardiac abnormalities. However, given AMPK’s vast array of functions it is perhaps not so surprising. The cardiac syndrome could be the result of a derangement of one or more of AMPK’s functions.</p><p>Pathophysiologically the syndrome due to PKRAG2 resembles other glycogen storage diseases such as Pompe disease [<xref ref-type="bibr" rid="R15">15</xref>]. The triad of cardiac hypertrophy, preexcitation, and conduction abnormalities has been documented in these diseases[<xref ref-type="bibr" rid="R16">16</xref>,<xref ref-type="bibr" rid="R17">17</xref>]. Therefore, the initial focus on the PRKAG2 induced cardiac syndrome has been on glycogen storage.</p><p>Cardiac hypertrophy has been demonstrated by ultrasound in patients with familial WPW syndrome[<xref ref-type="bibr" rid="R17">17</xref>]. The accumulation of glycogen-like substance has also been seen in the myocardium of these patients[<xref ref-type="bibr" rid="R18">18</xref>]. The increased thickness of the myocardium (cardiac hypertrophy) has been attributed to the excessive deposition of glycogen-like substance as well as increased myocyte growth, which maybe somewhat different from the hypertrophy seen in HCM (hypertrophic cardiomyopathy) which is secondary only to increased myocyte growth and fibrosis.</p><p>Conduction abnormalities (heart block or slowing of the conduction system) can be explained by two well established facts. First glycogen is known to be more abundant in the conduction system compared to the myocardial muscle tissue [<xref ref-type="bibr" rid="R13">13</xref>]. Therefore excess accumulation of glycogen-like substance could lead to loss of conduction tissue and slowing or block. Secondly excessive glycogen has been known to be toxic to the conduction system which could lead to loss of function [<xref ref-type="bibr" rid="R13">13</xref>].</p><p>WPW syndrome in humans is caused by the existence of an accessory pathway. Initially there is no separation between the atria and the ventricles. During development either through apoptosis or remodeling the atria separates from the ventricle. It is possible that the normal developmental process of apoptosis and remodeling doesn’t occur in patients with familial WPW syndrome. Therefore some myocardial tissue that remains as a connection between the atrium and the ventricle [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>].</p><p>AMPK’s role in the development of the pathophysiology in familial WPW could be a result of gain in function or a loss in function. The increased amount of glycogen found in the myocardium of these patients suggests a loss of function if we believe that AMPK activity decreases glycogen stores. Hypertrophy would be explainable since a loss of function would lead to an ATP deficient state and increased cell growth as compensation for increased work load. Conduction abnormalities could arise from the deposition of glycogen-like substance in the conduction system. Preexcitation maybe a result of a developmental abnormality as explained above.</p></sec><sec id="s6"><title>Animal model of Familial WPW syndrome</title><p>Transgenic technology has been used in biomedical research over the last 20 years. The basic idea is to introduce a gene with a known phenotype into an animal in the hope the offspring will manifest the phenotype. Since the mutation for WPW syndrome was identified interest in generating a transgenic model for the disease has intensified. Recently a mouse model of this human disease has been developed (Seidman et al) [<xref ref-type="bibr" rid="R20">20</xref>]. The generation of a transgenic model usually is done by cloning the mutated gene of interest along side a powerful promoter in the expectation that the gene will be overexpressed. In Seidman’s model the gene of interest was the mutated PRKAG2 gene which encodes for the gamma2 subunit of AMPK and the promoter used was the alpha-myosin heavy chain promoter.</p><p>The mice generated recapitulated the human phenotype with preexcitation and cardiac hypertrophy. Histopathology demonstrated vacuoles that appear to be like those seen in the human phenotype. Interestingly Seidman et al demonstrated increased AMPK activity in the mutant transgenic mice versus the wild type transgenic mice. This would suggest that the missense mutation leads to a gain in function of AMPK. Electrophysiological studies done on these mice demonstrated the existence of an accessory pathway consistent with the human form of WPW. Seidman et al proposes that the glycogen deposition disrupts the annulus fibrosa which is responsible for preexcitation.</p><p>We recently developed a transgenic animal (mice) model for human familial WPW syndrome. Our model has a phenotype of preexcitation and evidence for an accessory pathway. In our model we were able to induce SVT.</p></sec><sec id="s7"><title>Future Considerations</title><p>Although WPW in its classical form is usually treated with catheter ablation and conduction abnormalities (AV block or slowing of conduction) can be handled clinically with the placement of a permanent pacemaker. There is definitely a role for genetic screening in the setting of familial WPW syndrome. Screening will allow the physician to identify individuals and families at risk for developing AV block and cardiac hypertrophy. This may become even more important in the future if novel treatments can be found for this glycogen storage disease.</p><p>The development of animal models will also give insight into the molecular mechanisms involved in the pathogenesis of this disease. This will also help understand normal development of the cardiac conduction system.</p></sec>
Transesophageal pacing : a versatile diagnostic and therapeutic tool	Could not extract abstract	<contrib contrib-type="author"><name><surname>Verbeet</surname><given-names>Thierry</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Castro</surname><given-names>José</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Decoodt</surname><given-names>Pierre</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>Transesophageal atrial pacing is feasible because of the proximity between the oesophagus and the posterior aspect of the atria [<xref ref-type="bibr" rid="R1">1</xref>]. Atrial pacing is possible through the oesophagus in almost all patients and can yield important information in many arrhythmias where single site atrial pacing is of value. It is a cheap procedure. There is no need for X Rays or cathlab.</p></sec><sec id="s2"><title>Technical considerations</title><p>Transesophageal pacing and recording is done using specialized or non specialized catheters. There are two different lead types: 1) the pill electrode, connected to a flexible wire, that the patient swallows with water. This pill electrode necessitates patient collaboration. 2) a flexible catheter that can even be used in comatose or intubated patients.</p><p>The pacing lead, in our cases a bipolar flexible catheter, is introduced into the oesophagus via the nares after local nose anesthesia with xylocaine spray. Sometimes throat anesthesia with the same spray is also performed. Xylocaine gel is used to lubricate the lead and the lead is introduced with guidewires in it in order to increase its stiffness. It is positioned into the oesophagus in order to record the posterior paraseptal atrial electrogram. There is a relationship between the site of maximal atrial amplitude and the lowest atrial pacing threshold. The optimal atrial pacing site is usually found around 40 cm from the nares. Bipolar or unipolar recordings can be made.</p><p>The strengh-duration curve of esophageal pacing is different than that of endocardial pacing. The lowest thresholds can be reached at pulse widths between 10 and 20 msec. Thresholds at these pulse widths are usually situated between 5 and 15 mA [<xref ref-type="bibr" rid="R2">2</xref>]. So transesophageal pacing necessitates a specific pacing device and external pacemakers cannot be used for that purpose.</p><p>Commercially available transesophageal pacing systems are usually restricted to burst pacing only. They frequently have an input where one can connect a programmable stimulator that will trigger esophageal pacing enabling the delivery of extrastimuli.</p><p>The tracings shown were all obtained in our center using the following equipment: a Medtronic 6992A bipolar pacing lead (<xref ref-type="fig" rid="F1">Figure 1</xref>) initially designed to pace the coronary sinus on a permanent basis, a dedicated A.P.I. Electronique Oesostim 2 transesophageal stimulator able to provide a maximal pulse width of 22 ms and a maximal current intensity of 44 mA, an implantable programmable pacemaker Medtronic Kappa KDR 700 that is connected to the ‘synchronisation input’ of the esophageal stimulator in order to deliver a maximum of 4 programmable extrabeats using the non invasive programmable stimulation features of this pacemaker model. The pacemaker is used as a triggering system because the dedicated esophagal stimulators can usually deliver bursts only. In fact any pacemaker with non invasive programmed stimulation possibilities will do. Alternatively a classic programmable stimulator can also be used instead. Intraesophageal electrograms are usually simply recorded in unipolar mode via a classic precordial EKG lead but can also be recorded in bipolar mode using limb leads (<xref ref-type="fig" rid="F2">Figure 2</xref>) [<xref ref-type="bibr" rid="R3">3</xref>].</p><p>Due to the bipolar nature of our lead, pacing and recording cannot be obtained simultaneously. If this is mandatory quadripolar catheters have to be used. The patient is asked to fasten for 4 hours before the procedure and for as long as the throat anesthesia remains.</p></sec><sec id="s3"><title>Applications</title><p>Transesophageal pacing can yield important information in many situation where invasive atrial stimulation is frequently done [<xref ref-type="bibr" rid="R4">4</xref>] [<xref ref-type="bibr" rid="R11">11</xref>].</p><p>- Sinus node evaluation</p><p>- Atrioventricular conduction evaluation: permeability, short PR, effects of drugs (<xref ref-type="fig" rid="F3">Figure 3</xref>-<xref ref-type="fig" rid="F6">6</xref>)</p><p>- Assessment of Woff-Parkinson-White syndrome: reciprocating tachycardia inducibility, anterograde refractory period, effect of drugs, ventricular rate during atrial fibrillation (<xref ref-type="fig" rid="F7">Figure 7</xref>-<xref ref-type="fig" rid="F10">10</xref>)</p><p>- Assessment of paroxysmal supraventricular tachycardia of unknown origin: mechanism analysis, counselling for radiofrequency ablation, drug evaluation.</p><p>- Assessment of palpitations of unknown origin (<xref ref-type="fig" rid="F11">Figure 11</xref>-<xref ref-type="fig" rid="F12">12</xref>)</p><p>- Assessment of relationship between atrium and ventricle: differential diagnosis ventricular tachycardia versus supraventricular tachycardia.</p><p>- Interruption of supraventricular tachycardia and atrial flutter (<xref ref-type="fig" rid="F13">Figure 13</xref>)</p><p>- Miscellaneous: P wave synchronous pacing, evaluation of myocardial ischemia</p><p>There are a few limitations to the technique: a) there is only one site of atrial pacing and recording, b) there is no ventricular pacing, c) sometimes atrial capture can be difficult to assess on the surface ekg during premature extrastimuli delivery.</p><p>Transesophageal atrial pacing can be obtained in more than 95 % of patients. Esophageal pacing usually produces a burning chest sensation that most patients tolerate. Bust pacing is always started at a low rate to educate the patient to this burning sensation and also to avoid rapid ventricular pacing which may very rarely occur (<xref ref-type="fig" rid="F14">Figure 14</xref>). Brachial plexus stimulation and phrenic nerve pacing although reported in the literature have never been seen in our center.</p><p>Between 1991 and 1997, 51 patients were treated with esophageal pacing for atrial flutter mostly of the common form. In 37 % of patients sinus rhythm was immediately established. In 16 % patients sinus rhythm was established after conversion of atrial flutter to atrial fibrillation. The technique failed in 47 % patients either because the patients remained in iatrogenic atrial fibrillation or because they went back from atrial fibrillation to flutter. This has led us to propose external defibrillation as a first line therapy for acute conversion in these patients except in selected cases [<xref ref-type="bibr" rid="R12">12</xref>].</p></sec><sec sec-type="conclusions" id="s4"><title>Conclusions</title><p>Transesophageal pacing is a versatile tool and may replace invasive atrial pacing in many circumstances. In the vast majority of cases where a preablation measurement of electrophysiological properties is needed it could avoid performing separate diagnostic and curative procedures.</p></sec>
Catheter Ablation for Atrial Fibrillation	Could not extract abstract	<contrib contrib-type="author"><name><surname>Gauri</surname><given-names>Andre J</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Knight</surname><given-names>Bradley P</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<p>Atrial fibrillation (AF), the most common arrhythmia in adults, affects 1 in 25 people over the age of 60 years and 1 in 10 over the age of 80 years [<xref ref-type="bibr" rid="R1">1</xref>]. There is considerable morbidity, mortality and economic burden associated with AF, all of which will increase with the expanding elderly population. Until recently, pharmacologic therapy with AV nodal blocking agents, antiarrhythmics and anticoagulation were the mainstay of therapy. Although electrical cardioversion is associated with a high immediate success rate, most patients have recurrences of AF with only 23% remaining in sinus rhythm one year after cardioversion [<xref ref-type="bibr" rid="R2">2</xref>]. Antiarrhythmic agents have been shown to improve sinus maintenance, but these medications have variable success and are associated with many potentially serious side effects. In addition, the recently published AFFIRM trial suggests that a pharmacological rhythm control strategy has no benefit in terms of mortality or morbidity over a rate control and anticoagulation strategy [<xref ref-type="bibr" rid="R3">3</xref>]. Over the last few years, there has been a great deal of enthusiasm regarding catheter based ablation strategies aimed at curing AF.</p><sec id="s1"><title>Percutaneous Maze Procedure</title><p>In 1959 Moe theorized that AF resulted from multiple wavelets of reentry. With this in mind, Cox performed the surgical Maze procedure in 1991 [<xref ref-type="bibr" rid="R4">4</xref>]. The Maze procedure alters the arrhythmogenic substrate by interrupting the macroreenterant circuits and reducing the critical mass of atria needed to sustain AF. This surgical approach currently is preformed in association with coronary bypass surgery and/or mitral valve repair and is successful in curing AF in 75-90% of cases [<xref ref-type="bibr" rid="R5">5</xref>]. Although the surgical procedure has been modified over time, the approach has consistently included isolation of the pulmonary veins (PV). Currently, malleable hand-held catheters are being used to create linear lesions with conventional radiofrequency energy, rather than surgical incisions.</p><p>The surgical experience over the past decade has provided evidence that the left atrium plays a significant role in the maintenance of AF and that a reduction in the left atrial mass prevents maintenance of AF. Because an open chest procedure is associated with significant morbidity, attempts have been made to replicate the Maze procedure using a percutaneous, catheter-based approach. The MECA (Multiple Electrode Catheter Ablation) trial, sponsored by Boston Scientific/EP technologies, was designed to determine the safety and feasibility of specially designed catheters with multiple large electrodes used to create circular, biatrial linear endocardial lesions to treat atrial fibrillation (<xref ref-type="fig" rid="F1">Figure 1</xref>-<xref ref-type="fig" rid="F3">3</xref>). The concept and catheter design were based on encouraging animal data [<xref ref-type="bibr" rid="R6">6</xref>]. However, the MECA study was terminated prematurely due to a relatively high complication rate and low efficacy rate [<xref ref-type="bibr" rid="R7">7</xref>]. Additional limitations of a percutaneous Maze procedure include a long, technically difficult procedure associated with long fluoroscopy times and the risk of proarrhythmia in the form of atrial tachycardias, which likely occur due to conduction gaps in the ablation lines. The role of linear atrial lesions in the treatment of atrial fibrillation remains unresolved.</p></sec><sec id="s2"><title>Importance of the Pulmonary Veins</title><p>Embryologically, the PVs form as a bud that grows from the heart towards the lungs. As a result, the PVs have a sleeve of muscle fibers that surround them. In the late 1990’s, Haissaguerre made a critical observation that the muscle fibers associated with the PVs are an important source of ectopic beats capable of triggering AF [<xref ref-type="bibr" rid="R8">8</xref>]. This discovery led to a revolution in interventional electrophysiology.</p><p>Pulmonary vein muscle tissue has unique electrophysiologic properties and appears to be able to maintain reentry within a relatively small amount of atrial muscle mass. This is likely due to the spatial complexity and a short refractory period. This area is also a common source of rapid focal discharges. <xref ref-type="fig" rid="F4">Figure 4</xref> shows an example of a rapid, irregular atrial tachycardia arising from a right upper PV that is associated with conduction block to the left atrium. The case highlights the unusual electrophysiologic observations that are made in the PVs. Although the arrhythmia is an atrial tachycardia in this case rather than AF, it is easy to see how a rapidly discharging focus in a PV could lead to AF by causing fibrillatory conduction or by initiating reentry in the atrium.</p></sec><sec id="s3"><title>Pulmonary Vein Isolation</title><p>Initially, attempts were made to ablate the ectopic foci in patients who had paroxysmal AF that appeared to arise in the PVs. Despite high initial success rates, this approach soon proved to be inadequate with recurrence rates over 60% [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R10">10</xref>]. The limitations of a focal approach include the presence of multiple PV foci, a paucity of spontaneous or inducible AF during the procedure, the subsequent development of new foci, and the difficulty in mapping triggers when AF is persistent. In addition, delivery of radiofrequency energy deep within the PV can lead to PV stenosis ( <xref ref-type="fig" rid="F5">Figure 5</xref>).</p><p>Electrical isolation of the pulmonary veins avoids many of the limitations of focal PV ablation. By isolating a PV from the atrium without ablating the entire circumference of the PV ostium, Haissaguerre and his group developed a technique referred to as ostial segmental PV isolation <xref ref-type="fig" rid="F11">Figure 11</xref>. The myocardial sleeves that connect the PV to the left atrium are a complex network of fibers. These fibers do not completely encircle the venous os, but are typically located in segments or quadrants. Identification of these muscular sleeves allows selective radiofrequency (RF) energy delivery, thereby minimizing the risk of PV stenosis.</p><p>The procedure requires transeptal catheterization. Intracardiac echocardiography can be useful to guide transeptal puncture ( <xref ref-type="fig" rid="F6">Figure 6</xref>). Identification of the PV ostia using venography or intracardiac echo is important to identify anomalies, such as a common PV ostium ( <xref ref-type="fig" rid="F7">Figure 7</xref>), and to avoid delivery of RF current in the PVs. Many centers perform a three-dimensional cardiac CT scan or MRI prior to ablation to screen for PV variants and to have a baseline imaging study in case a patient develops symptoms suggestive of PV stenosis after ablation.</p><p>In order to identify PV potentials, a multiple electrode catheter is positioned near the ostium of the PV. This can be accomplished using a basket-type catheter or a circular multipolar mapping catheter. An example of a commonly used circular decapolar catheter is shown in <xref ref-type="fig" rid="F8">figure 8</xref>. The decapolar catheter is positioned within 5 mm of the PV ostium, and RF energy is delivered at the ostial sites that display the earliest bipolar PV potentials during sinus rhythm or during coronary sinus pacing, or at sites that have the largest intrinsic deflections in unipolar recordings. Utilization of both unipolar and bipolar recordings during segmental ostial ablation is associated with increased efficacy and less RF energy delivery [<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R13">13</xref>]. The endpoint of this procedure is elimination or dissociation of pulmonary venous potentials (<xref ref-type="fig" rid="F9">Figure 9</xref> -<xref ref-type="fig" rid="F10">10</xref>). PV isolation is usually performed during sinus rhythm but can be accomplished during AF when sinus rhythm cannot be maintained initially during the procedure (<xref ref-type="fig" rid="F11">Figure 11</xref> -<xref ref-type="fig" rid="F12">12</xref>). Initial strategies consisted of ablating only the arrhythmogenic PV’s, but empiric isolation of all four PV’s has been shown to be associated with a higher success rate.</p><p>Segmental ostial ablation has been associated with high success rates in patients with paroxysmal lone atrial fibrillation. Among the first 58 patients treated at the University of Michigan, 70% were free from AF recurrences and 83% were free of symptomatic AF or had significant improvement at 5 months follow-up [<xref ref-type="bibr" rid="R14">14</xref>]. Segmental ostial ablation appears to be less successful in patients with persistent or permanent AF. Only 22% of patients with persistent AF were free from recurrent AF at 5 months [<xref ref-type="bibr" rid="R14">14</xref>]. This high failure rate suggests that the left atrium plays a more important role in patients with persistent AF compared to patients with paroxysmal AF.</p><p>Recurrent AF after segmental ablation is likely due to recovery of conduction of the PV fibers. Other causes include ectopy arising from a small ostial cuff of muscle proximal to the ablation site, or non-PV ectopic foci arising from the superior vena cava, coronary sinus, left or right atrium, ligament of Marshall, or rarely the inferior vena cava [<xref ref-type="bibr" rid="R15">15</xref>]. As a result, 10-40% of patients undergoing segmental ostial ablation require a repeat ablation procedure [<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R14">14</xref>].</p><p>In experienced centers, the risk of a major complication including PV stenosis, cardiac tamponade, stroke and death is approximately 2%. Measures that are used to minimize the risk of PV stenosis include delivery of energy only at the ostium, limitation of the power to 30-35 watts and temperature to 52° C, discontinuation of RF if no effect is seen within 20 seconds, periodic pulmonary venography, and intracardiac echo visualization and Doppler measurement. Although some degree of PV narrowing may occur following the ablation procedure, clinically relevant PV stenosis is rare. Three-dimensional computed tomography has been shown to be useful in the identification of pulmonary vein stenosis. A series from the University of Michigan found that among 58 consecutive patients who had undergone a segmental ostial isolation procedure, there were no patients with symptomatic PV stenosis [<xref ref-type="bibr" rid="R16">16</xref>]. Routine computed tomography scans performed four months following the ablation identified an average narrowing of 1.5 mm of the ostial diameter. A 28% to 61% stenosis was present at an average distance of 7.6 ± 2.2 mm from the ostium in only 3% of the 128 PVs treated. The Cleveland Clinic group recently reported the results of using routine spiral computed tomography to detect PV stenosis following PV isolation ablation procedures [<xref ref-type="bibr" rid="R17">17</xref>]. In their series of 335 patients, 18 (5%) had severe PV stenosis defined as a luminal narrowing of > 70%. However, only 10 of the patients (3%) were symptomatic with the most prevalent symptom being shortness of breath (8 patients), followed by cough (7 patients) and hemoptysis (5 patients). An awareness of this potential complication is important since therapeutic options are available. Balloon angioplasty and venous stenting have been performed in symptomatic patients.</p><p>Pulmonary vein isolation was reported initially to be associated with long procedure and fluoroscopy times. However, it has been shown that centers that have performed over 75 cases can typically complete the procedure in less than 3 hours and with less than 60 minutes of fluoroscopy [<xref ref-type="bibr" rid="R18">18</xref>].</p></sec><sec id="s4"><title>Electroanatomic Left Atrial Ablation</title><p>An alternative endocardial ablation technique to treat AF has been described by Pappone. This technique uses a circumferential electroanatomic approach [<xref ref-type="bibr" rid="R19">19</xref>]. The procedure described by Pappone involves a 3-D electroanatomic mapping system (CARTO, Biosense Webster Inc.) to map the atria and PVs. Circumferential RF lesions are then created at 5 mm from the PV ostia. This anatomic approach eliminates the need for mapping spontaneous or induced arrhythmias. The end point for ablation is a bipolar amplitude less than 0.1 mV inside the lesion and a delay of greater than 30ms across the ablation line (<xref ref-type="fig" rid="F13">Figure 13</xref>). The one-year success rates, defined as freedom from AF, obtained by the Pappone group in 251 patients (paroxysmal AF=179, permanent AF=72) treated was 80% overall, with 86% for paroxysmal AF and 68% for permanent AF [<xref ref-type="bibr" rid="R20">20</xref>]. Only 75% of the circumferential lesions surrounding the individual PVs met criteria for complete, defined as a bipolar amplitude < 0.1 mV. Interestingly, they found no relation between lesion completeness and clinical outcome. This finding led the Pappone and his group to use the term “electroanatomic remodeling” to describe the alteration in the atria substrate that occurs during this ablation technique which prevents atrial fibrillation. This approach of PV isolation plus substrate modification may explain the higher success rates obtained compared to exclusively isolating the PV and eliminating the “trigger” of AF in the segmental ostial ablation approach. Which of these two approaches is superior is a source of debate. A randomized control trial comparing segmental ostial ablation to circumferential ablation in patients with paroxysmal AF is currently underway.</p></sec><sec sec-type="conclusions" id="s5"><title>Conclusions</title><p>The last few years have marked the beginning of an exciting new era in the treatment for AF. For patients with paroxysmal AF, both segmental and circumferential ablations appear to have comparable long term success rates and low rates of complications. However, for patients with persistent or permanent AF, the circumferential ablation approach using 3D-elcctroanatomic mapping appears to be more successful. Patients with AF who are suitable candidates for catheter ablation are those with symptomatic AF despite reasonable pharmacologic efforts and minimal structural heart disease. The future of ablation therapy for AF will likely be an approach which both eliminates the trigger of AF and alters the substrate which permits maintenance of the arrhythmia. New catheter designs and alternative energy sources are currently under investigation to improve the safety, efficiency, and success rate of catheter ablation for AF.</p></sec>
Role of Atrial Fibrillation Threshold Evaluation on Guiding Treatment	<p>Atrial fibrillation could be induced reproducibly by 50Hz rapid stimulation which was given through systolic and early diastolic phase of atrial excitation. Duration of atrial fibrillation induced in this way was roughly dependent on the current amplitude of the stimulation. The minimum current that could induce long-lasting atrial fibrillation (30sec in the clinical setting, 2sec in the rabbit or rat model) was defined as atrial fibrillation threshold (AFT). AFT was larger in patients who had history of atrial fibrillation than those who did not. Anti-arrhythmic drugs raised AFT by various degrees both in experimental and clinical cases. Long-term efficacy of a drug could be predicted in a patient, measuring how much the drug increased AFT (cut-off point = 5mA increase). AFT is a useful marker to evaluate atrial vulnerability and to guide pharmacological treatment of atrial fibrillation.</p>	<contrib contrib-type="author"><name><surname>Shirayama</surname><given-names>Takeshi</given-names></name><degrees>MD, PhD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>Atrial fibrillation is a target of intensive research because fibrillation is the final frontier in the arrhythmology [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R3">3</xref>]. Now we have several therapeutic modalities for this common arrhythmia, including pharmacological and non-pharmacological approach. However, long-term success rate of these treatments would be 50% to 60 % at the best [<xref ref-type="bibr" rid="R4">4</xref>] . There are several reasons for its difficulty: 1)this arrhythmia is essentially due to aging in clinical cases, 2)the mechanism of this arrhythmia is not fully understood [<xref ref-type="bibr" rid="R5">5</xref>] , 3)there is no appropriate animal model [<xref ref-type="bibr" rid="R6">6</xref>] . We have been working on these problems for several years, and become to use atrial fibrillation threshold as a tool to evaluate the easiness to induce atrial fibrillation (atrial vulnerability) both in the experimental [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R10">10</xref>] and the clinical [<xref ref-type="bibr" rid="R11">11</xref>-<xref ref-type="bibr" rid="R13">13</xref>] settings. This is the unique parameter to evaluate atrial vulnerability quantitatively [<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R13">13</xref>] . Because very few researches have been done in this field, I will summarize the data from our laboratory, comparing with the ventricular counterpart in this review article.</p></sec><sec id="s2"><title>What is fibrillation threshold ?</title><p>Fibrillation threshold has been studied mostly in the ventricle [<xref ref-type="bibr" rid="R14">14</xref>-<xref ref-type="bibr" rid="R16">16</xref>] since Wiggers et al [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R18">18</xref>]. A single stimulus was given to induce ventricular fibrillation in the diseased and the normal hearts. For this purpose, the stimulus should be given at a critical time point with sufficient current amplitude. The time point is usually located at around the peak of T wave of ECG. The minimum current amplitude to induce fibrillation is defined as fibrillation threshold. This concept is based on a fact that fibrillation can be induced by larger current than a critical (threshold) amplitude. Recently, it is recognized that the second critical amplitude exists when current amplitude is increased incrementally. Larger current than the second critical point can not induce fibrillation, but rather terminate fibrillation [<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R20">20</xref>]. This point is called defibrillation threshold, which is now widely studied because of implantable cardioverter defibrillators in clinical use. In the combination with vulnerable period (the zone of the critical time points to induce fibrillation), there is an area in the strength-interval curve where fibrillation can be induced by a single stimulus [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R18">18</xref>]. This area is shifted rightwards along the axis of coupling interval, for example, when myocardial ischemia or anti-arrhythmic drugs are introduced [<xref ref-type="bibr" rid="R21">21</xref>] (<xref ref-type="fig" rid="F1">Figure 1</xref>). The same argument could be applied to the atrium. However, atrial vulnerable period was less defined than that of the ventricle [<xref ref-type="bibr" rid="R22">22</xref>,<xref ref-type="bibr" rid="R23">23</xref>]. Atrial T wave is barely discernable in a standard ECG.</p><p>There have been several methods to induce ventricular fibrillation [<xref ref-type="bibr" rid="R14">14</xref>,<xref ref-type="bibr" rid="R17">17</xref>]. Among them, a continuous high frequency pacing at 50-100Hz which covers a whole “vulnerable period” can induce fibrillation consistently [<xref ref-type="bibr" rid="R14">14</xref>]. Thus we applied this type of stimulation to the atrium and the minimum current to induce long-lasting atrial fibrillation (>30sec in clinical settings, >2sec in experiments) was defined as fibrillation threshold (<xref ref-type="fig" rid="F2">Figure 2</xref>). Duration of “long-lasting” atrial fibrillation was selected rather arbitrarily as the definition but it was based on the relationship between the current intensity and the duration (see Matsuo et al [<xref ref-type="bibr" rid="R9">9</xref>]). The duration was positively correlated with the current intensity until the current reached the threshold. This definition is different from conventional usage in terms of the stimulating pattern.</p></sec><sec id="s3"><title>Mechanism of induction of fibrillation</title><p>The precise mechanism of induction of fibrillation by a single pulse or train pulses is not clear. However, there are some observational studies in the ventricle [<xref ref-type="bibr" rid="R14">14</xref>,<xref ref-type="bibr" rid="R15">15</xref>,<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R19">19</xref>]: 1)localized repetitive activities were induced at the stimulating site, 2)excitations were propagated, 3)repetitive excitations were observed at many sites of the myocardium 4)excitations resulted in disorganized activity (fibrillation) because of inhomogeneity of conduction and repolarization. When high frequency stimulation is given to the ventricle at an intensity of diastolic threshold, ventricular response mimicking ventricular tachycardia could be observed. As the intensity is raised incrementally, the intervals of ventricular excitation become shorter. At and beyond a critical intensity of the stimulating current, ventricular excitations become fibrillation. Although these results are based on the ventricular study, it is probable that the same phenomenon could be observed in the atrium.</p></sec><sec id="s4"><title>Experimental methodology of induction of atrial fibrillation</title><p>Most popular method to induce atrial fibrillation would be an extra-stimulus method, that is, an extra-stimulus is given to the atrium at a critical time point after basic conditioning stimulations at a fixed rate [<xref ref-type="bibr" rid="R23">23</xref>]. Because reproducibility of this method is relatively low, statistical evaluation is often necessary after many trials [<xref ref-type="bibr" rid="R24">24</xref>] to evaluate drug efficacy or atrial vulnerability. On the other hand, continuous rapid pacing could easily induce fibrillation. Thus this method was introduced for the quantitative evaluation of atrial vulnerability (easiness to induce atrial fibrillation). We chose 50Hz for 1sec stimulation for this purpose. The increment of the current intensity is 1mA for the clinical setting, and 0.1 to 0.5mA for the experimental use.</p></sec><sec id="s5"><title>Experimental results</title><p>Atrial fibrillation threshold has been measured in the isolated hearts of guinea pig or rabbit. Sodium channel blockers [<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R11">11</xref>] (disopyramide, pilsicainide, flecainide, aprindine, lidocaine), potassium channel blockers [<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R10">10</xref>] (E-4031), amiodarone, and SD-32129 raised atrial fibrillation threshold by various degrees. Although the increase of atrial fibrillation threshold induced by anti-arrhythmic drugs was positively related to the increase of effective refractory period, E-4031 had the least potency [<xref ref-type="bibr" rid="R7">7</xref>]. This results could be explained by the fact that sodium channel blockers made refractory period longer than E-4031 did because of longer post-excitation refractoriness after a high frequency stimulation [<xref ref-type="bibr" rid="R25">25</xref>]. Note that the blocking effect of potassium channels by Ikr blockers (such as E-4031) is less prominent when the excitation frequency becomes higher (reverse use-dependent block [<xref ref-type="bibr" rid="R26">26</xref>], but see Ohler et al [<xref ref-type="bibr" rid="R27">27</xref>]). In the presence of anti-arrhythmic drugs in the perfusing solution, it was more difficult to induce repetitive atrial firing or atrial fibrillation by electrical stimulation. At the same time, atrial fibrillation threshold was increased.</p><p>Effective refractory period becomes shorter when a rapid pacing is continuously applied to the atrium. This phenomenon is called “atrial electrical remodeling” [<xref ref-type="bibr" rid="R28">28</xref>], which is accompanied by the reduction of L-type calcium current and transient outward current [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R29">29</xref>]. Expression of novel channels was also reported [<xref ref-type="bibr" rid="R30">30</xref>]. Large numbers of experimental studies have been performed regarding atrial remodeling. In these studies, single extra-stimulus or burst pacing similar to ours has been used to induce atrial fibrillation. However, atrial vulnerability was compared in terms of the duration of atrial fibrillation that was induced, but not the intensity of the stimulation. These studies did not evaluate how easy to induce atrial fibrillation, but how difficult to recover from the arrhythmia. The evaluation of the duration of the arrhythmia disclose its stability after the physiological and the biochemical changes of the atrium were implemented, but it could not be suitable to evaluate the propensity of the patients to paroxysmal attack of atrial fibrillation. In our experience, a stronger current to the atrium could induce longer atrial fibrillation until the current reached the fibrillation threshold [<xref ref-type="bibr" rid="R12">12</xref>].</p></sec><sec id="s6"><title>Clinical feasibility</title><p>Experimental data suggested that the measurement of the threshold could predict the easiness to induce atrial fibrillation. Indeed atrial fibrillation threshold was lower in the patients who had a history of atrial fibrillation (median 11mA) than normal control (median 5mA) [<xref ref-type="bibr" rid="R12">12</xref>] . When cut-off point was set at 10mA, the sensitivity and the specificity were 94%, and 60%, respectively. Effective refractory period, conduction time, or other indicators of atrial vulnerability were not different between two groups. The secondary indicators to distinguish the patients with atrial fibrillation from normal subjects were %maximum atrial fragmentation (%MAF) and fragmented activity zone (FAZ) [<xref ref-type="bibr" rid="R31">31</xref>]. Their sensitivity and specificity were 78%, 52% (%MAF), and 47%, 84% (FAZ), respectively. %MAF is the relative increase of the width of local atrial electrogram by the extra-stimulation. FAZ is the zone of coupling intervals that made local atrial electrogram longer by 50% or more.</p><p>Atrial fibrillation threshold (AFT) could be affected by the autonomic tone because isoproterenol infusion decreased AFT (unpublished data). The data obtained on the day when the catheters were inserted were lower than those obtained on the following days. Thus the results of the measurement should be interpreted cautiously. However, reproducibility of AFT was excellent (±1mA, r=0.95 on the regression line) during a short term examination (<3 hours), or if it was measured on the next day or later after the catheters were inserted.</p></sec><sec id="s7"><title>Evaluation of drug efficacy in the clinical setting</title><p>The benefit of the AFT measurement would be quantitative evaluation of the drug efficacy. Anti-arrhythmic drugs raised AFT in general, but the increase of AFT by a drug was different in the individual patient. For example, procainamide raised AFT by 5mA in a patient, but propafenone did not change AFT at all in the same patient. In this way, we could categorize anti-arrhythmic drugs into 2 groups in the individual patient, i.e. drugs that raise AFT significantly, and drugs that does not increase AFT in the patient. The effective drugs in a patient were not necessarily the same in another patient. When the cut-off point of effectiveness was set to 5mA increase, “effective” drug could prevent 88% of the patients from the recurrence of atrial fibrillation at least for 1 year whereas only 13% were free from atrial fibrillation with “ineffective” drugs [<xref ref-type="bibr" rid="R13">13</xref>]. Thus this method could be used to select effective drugs in an individual patient.</p><p>However, the prediction of the effectiveness did not mean the high efficacy of the drug in a large population. As previously reported[<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R4">4</xref>], efficacy of each drug against atrial fibrillation was around 40 to 50% in our study. AFT is a method to select beneficial drugs to the specific patient.</p><p> Serial electrophysiological test has been less recommended in case of ventricular tachycardia, because “effective drug” determined in terms of inducibility did not improve the prognosis of the patients [<xref ref-type="bibr" rid="R32">32</xref>]. On the other hand, AFT does not depend on all-or-none type determination of inducibility, but it determines how strong is the drug effect against the arrhythmia. As a matter of fact, atrial fibrillation was always induced in this method because of the definition of AFT.</p></sec><sec id="s8"><title>Rate control and rhythm control</title><p>Clearly, AFT measurement determines a drug suitable to maintain sinus rhythm. Recently, the results of the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) study were published [<xref ref-type="bibr" rid="R33">33</xref>]. This study did not show the advantage of maintaining sinus rhythm over heart rate control in terms of prognosis. This result may argue against the usefulness of AFT measurement. However, mortality of the patients who suffered from paroxysmal atrial fibrillation is relatively low. The biggest concern of the patients is often its symptom. Thus, when mortality is not different, it would be desirable to select less symptomatic life. In this sense, pharmacological treatment with class I anti-arrhythmic drugs could still be a first line therapy for “lone” atrial fibrillation.</p></sec><sec id="s9"><title>AFT in the future</title><p>AFT is a quantitative marker to select effective drugs against atrial fibrillation. The concept of this method is “custom-made” medicine. Large-scale population study would reveal overall efficacy of the drug, but a negative side is negligence of individuality or specificity of each patient. Because the efficacy of pharmacological and non-pharmacological treatment is still not high, it is important to know which method should be applied first. It is possible to measure AFT first, and if “effective” drug could be found, the patient should be treated with the regimen. If “effective” drug is not found, catheter ablation or some other non-pharmacological approach should be considered.</p></sec>
Brugada Disease: Chronology Of Discovery And Paternity. Preliminary Observations And Historical Aspects	<p>The Brugada disease, the last clinico-cardiologic entity described in the 20th century, initially called right bundle branch block syndrome with ST segment elevation from V1 to V2 or V3 and sudden cardiac death, is genetically determined in a dominant autosomal mode, and it affects the alpha subunit of the Na<sup>+</sup> channel by alteration of chromosome 3 and mutation in the SCN5A gene.</p><p>In clinical diagnosis the mentioned electrocardiographic pattern in a patient without structural heart disease and positivity in pharmacological tests are considered major criteria. As minor criteria, the following are considered: positive family history, presence of syncope with unknown origin, documented episode of VT/VF, inducibility in electrophysiologic study and positivity of genetic study.</p><p>The long-standing technology of ECG, with more than a century of existence, remains as the supplementary method with highest value in diagnosis, and currently new electrocardiographic criteria are suggested, which indicate high risk of VF.</p><p>Natural history indicates a somber diagnosis in symptomatic patients with a high index of arrhythmic SCD secondary to very fast polymorphic ventricular tachycardia bursts, which degenerate into VF. Asymptomatic individuals with only a Brugada-type electrocardiographic pattern have a low risk. The prognosis seems to depend more on clinical facts, since a positive electrophysiologic study has an accuracy of just around 50%.</p><p>We propose that this entity should be promoted to the category of disease, since it has a characteristic set of signs and symptoms, and an identified genetic defect.</p>	<contrib contrib-type="author"><name><surname>Riera</surname><given-names>Andres Ricardo Perez</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Schapachnik</surname><given-names>Edgardo</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Ferreira</surname><given-names>Celso</given-names></name><degrees>MD, PhD</degrees><xref ref-type="aff" rid="aff3">¶</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Preliminary Observations: "Prehistory"</title><p>Halfway through the 20th century (1953), Osher and Woff noticed the right bundle branch block (RBBB) electrocardiographic pattern, associated to ST segment elevation in the right precordial leads. These were considered at the time as normal variants, not having been related to sudden cardiac death (SCD) [<xref ref-type="bibr" rid="R1">1</xref>]. In 1975, Calo [<xref ref-type="bibr" rid="R2">2</xref>] reported an electrocardiographic triad that consisted of: R' wave, ST segment elevation and negative T wave in right precordial leads, which coincide with the characteristics of the electrocardiographic pattern currently known as Brugada-type, being considered a normal variant at the time.</p><p>During the 80s, the Center for Disease Control in Atlanta observed an abnormally high incidence of SCD in Asian refugees who immigrated to USA from the northeast of Thailand. This form of SCD is known in this country as Lai Tai (death during sleep). Approximately two decades later, the conclusion was reached that the entity known as "Sudden Unexplained Nocturnal Death Syndrome" (SUNDS) originates in an allele belonging to the same gene (SCN5A) as Brugada Disease [<xref ref-type="bibr" rid="R3">3</xref>].</p><p>In 1986, Prof. Pedro Brugada received his first patient with typical ECG, a Polish Caucasian child, who suffered several episodes of syncope. The boy presented as family background his sister's SCD, even though she had been treated with association of pacemaker implantation and amiodarone. In 1989, a patient with characteristic ECG was described as being a carrier of early repolarization syndrome [<xref ref-type="bibr" rid="R4">4</xref>]. In 1991, Pedro and Josep Brugada, adding 2 more cases, presented as an abstract in the NASPE meeting, a new clinical-cardiologic syndrome, typified by the association of RBBB, persistent ST segment elevation, normal QT interval and SCD [<xref ref-type="bibr" rid="R5">5</xref>].</p></sec><sec id="s2"><title>Initial Observations And Chronological Evolution: "History"</title><p>In 1992, the brothers from Catalonia (Spain), Pedro and Josep Brugada, presented the first description of the entity, adding four more patients to the initial description, making a total of eight [<xref ref-type="bibr" rid="R6">6</xref>]. This would be the last clinical-cardiologic entity to be identified in the 20th century [<xref ref-type="bibr" rid="R7">7</xref>]. One year later, Sumiyoshi et al [<xref ref-type="bibr" rid="R8">8</xref>] identified the syndrome described by the Brugada brothers as being a variant of idiopathic ventricular fibrillation (IVF) and in the same year, Proclamer et al [<xref ref-type="bibr" rid="R9">9</xref>] described a case, posing the possibility of a new arrhythmic syndrome with the characteristics described by the Brugada brothers. In the same year, Italian authors from different centers said that the alleged new entity is just a minor initial form or concealed form of right ventricle arrhythmogenic dysplasia/cardiomyopathy [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R11">11</xref>]. That same year, the Brugada brothers, in an Italian journal (G Ital Cardiol) replied to their colleagues from Padua and Nacarella, that they should not confuse issues, since in the new entity there is no underlying structural heart disease [<xref ref-type="bibr" rid="R12">12</xref>].</p><p>In 1994, Ferraccini et al [<xref ref-type="bibr" rid="R13">13</xref>] reported a case of IVF associated to RBBB and ST segment elevation, and in the same year, Bjerregaard et al [<xref ref-type="bibr" rid="R14">14</xref>] reported recurrent syncopes in a patient carrier of prominent J wave. In 1995, Tada et al [<xref ref-type="bibr" rid="R15">15</xref>] pointed out the significance of ST segment elevation in right precordial leads in patients with IVF, and Tohyou et al [<xref ref-type="bibr" rid="R16">16</xref>] analyzed the incidence of RBBB with ST segment elevation in normal population. In this same year, D'Onofrio et al [<xref ref-type="bibr" rid="R17">17</xref>] published an article wondering if the electrocardiographic pattern of RBBB associated to ST segment elevation from V1 to V3 in all the cases would correspond to a different syndrome.</p><p>In 1996, Gan-Xin Yan and Charles Antzelevitch [<xref ref-type="bibr" rid="R18">18</xref>], in an article where they approached the cellular basis of J wave in ECG, used the eponym Brugada for the first time to describe the syndrome discovered four years earlier. That same year, Kobayashi et al [<xref ref-type="bibr" rid="R19">19</xref>] called the typical electrocardiographic signs of the new entity as "Brugada type," and Miyazaki et al [<xref ref-type="bibr" rid="R20">20</xref>], pointing out to the autonomous modulation of ST segment elevation in patients with the syndrome described by the Brugada brothers, used the eponym just as Yan and Antzelevitch had done before. In the same year, Shimada also mentioned the eponym while reporting a case of the entity that presented monomorphic ventricular tachycardia (MVT) [<xref ref-type="bibr" rid="R21">21</xref>]. This is the first reference to a case of Brugada disease with MVT. In a short time, Dr. Bartolo Martini from the group of researchers from Padua claimed the paternity of the discovery, remarking that they had described the entity three years earlier, in 1989 [<xref ref-type="bibr" rid="R22">22</xref>]. In this work, the authors informed about the electrocardiographic manifestations as having a possible relationship with SCD; however, they concluded that in these cases there is a structural heart disease: right ventricle arrhythmogenic dysplasia/cardiomyopathy (RVAD), and did not acknowledge being in the presence of a new entity without an underlying organic substrate. Additionally, from the 6 patients mentioned in this article, only patient 3 presented the typical electrocardiographic pattern reported by the Brugada brothers [<xref ref-type="bibr" rid="R23">23</xref>,<xref ref-type="bibr" rid="R24">24</xref>].</p><p>In 1996, Corrado et al [<xref ref-type="bibr" rid="R25">25</xref>] presented the hypothesis of the existence of a subpopulation of RVAD that they called "concealed forms," which would present itself with the typical electrocardiographic features of this new entity proposed, including the presence of a polymorphic form of ventricular tachycardia. In the same year, such authors as Ohe and Fontaine followed the same line, which later was proved to be wrong [<xref ref-type="bibr" rid="R26">26</xref>,<xref ref-type="bibr" rid="R27">27</xref>]. Nevertheless, the alterations of the vago-sympathetic autonomous tone only constituted a triggering factor, but would not be the cause of the entity.</p><p>In 1997, Kobayashi et al [<xref ref-type="bibr" rid="R28">28</xref>] made the first description of the existence of an autonomous imbalance in Brugada disease, by means of the 123I-metaiodobenzylguanidine (MIBG) imaging techniques, which indicated the existence of a presynaptic autonomic dysfunction in the heart [<xref ref-type="bibr" rid="R29">29</xref>-<xref ref-type="bibr" rid="R31">31</xref>]. In 1997, Chinusi et al [<xref ref-type="bibr" rid="R32">32</xref>] made the first narration about the variable effect of disopiramide in ventricular arrhythmia induction in patients who are Brugada disease carriers, by acting on the Ito channel, sometimes increasing ST segment elevation, and possibly normalizing it.</p><p>In 1998, the acknowledgment of Brugada disease as an entity became unquestionable when Chen et al [<xref ref-type="bibr" rid="R33">33</xref>] identified three mutations in the SNC5A gene of chromosome 3p21-p24, which are responsible for Brugada disease and affect the alpha subunit of the Na+ channel. The mutations found were: <list list-type="order"><list-item><p>Missense or "wrong-information" mutation: this mutation affects exon 28, and in it, the amino acid glutamine is exchanged by leucine in codon 567 (L567Q) between domains I and II of the Na+ channel. This mutation determines a temporary increase in cation entrance during phase zero, with acceleration of recovery from the inactivation state.</p></list-item><list-item><p>Structural of frameshift mutation: which consists in the subtraction of a nucleotide in the SCN5A gene.</p></list-item><list-item><p>"Splice-donor" mutation: or mutation that accompanies the donor that affects intron 7 and introduces two bases of AA.</p></list-item></list></p><p>The last two cause a failure in the channel operation. In the same year, and for the first time, Nakamura et al [<xref ref-type="bibr" rid="R34">34</xref>] showed that the IC class antiarrhythmic agents, flecainide and pilsicainide, may cause ST segment elevation in the inferior wall. Later, it would be verified that ST segment elevation could manifest spontaneously in this wall in Brugada disease.</p><p>Makita et al [<xref ref-type="bibr" rid="R35">35</xref>] demonstrated the importance of the "overlooked" accessory beta 1 subunit of the Na+ channel. Thus, the authors explained that an alteration related to the alpha/beta subunits influences on the functional state of the Na+ channel, by causing a higher overlapping of activation and inactivation states, giving rise to a window current in T1620M and consequently, in VF triggering. Dr. Tagaki et al [<xref ref-type="bibr" rid="R36">36</xref>] described for the first time, employing an ultrafast computerized tomography, in 81% of a series of 26 patients, abnormalities in right ventricle wall motion, mostly located in the outflow tract (17 patients) or in inferior wall (4 patients), wondering about the functional nature of this entity.</p></sec><sec id="s3"><title>Year 1999</title><p>For the first time, Blazer et al [<xref ref-type="bibr" rid="R37">37</xref>] included Brugada disease within the chapter of ion channel diseases or channelopathies. The authors showed that the affected channels in Brugada disease are primarily, the fast Na+ channel, and secondarily, the initial K+ outflow channel or Ito channel or transient outward current in phase 1 or 4-aminipyridine-sensitive channel, and slow Ca2+ inflow channel in phase 2 or L-type ("L-type slow or long-lasting" calcium channel ICa-L type ICa2+- L). In this year, Bezzina et al [<xref ref-type="bibr" rid="R38">38</xref>] identified a single mutation in the Na+ channel, responsible for both Brugada disease and congenital long QT syndrome. In this way, the concept that both entities are allelic because they share the same locus was consolidated.</p></sec><sec id="s4"><title>Year 2000</title><p>In this year, Dr. Takanori Ikeda et al [<xref ref-type="bibr" rid="R39">39</xref>] identified the non-invasive markers of value in risk stratification in Brugada disease. The authors concluded that only high resolution ECG, and not QT interval dispersion or microvolt T wave alternans have value to identify the patients in high risk. They determined that high resolution ECG has a sensitivity of 89%, specificity of 50%, positive predictive value of 70%, and negative predictive value of 77% for the presence of late potentials (LP). The authors did not find a correlation between the degree of ST segment elevation and the HV interval. One year later, the same group of researchers confirmed in JACC the previous results [<xref ref-type="bibr" rid="R40">40</xref>]. Dr. Silvia Priori et al [<xref ref-type="bibr" rid="R41">41</xref>], in a prospective study of a numerous universe conducted with 52 families, arrived to the following conclusions: <list list-type="order"><list-item><p>asymptomatic individuals with Brugada-type electrocardiographic pattern, present a very low risk of SCD;</p></list-item><list-item><p>symptomatic with aborted SCD present a 23% of mortality rate in a mean 33-month follow-up;</p></list-item><list-item><p>the genetic mutation can be identified in a 15% of the cases;</p></list-item><list-item><p>the positive electrophysiologic study has a 50% of accuracy;</p></list-item><list-item><p>pharmacological tests have only a 35% of accuracy in asymptomatic carriers.</p></list-item></list></p><p>Nishizaki et al [<xref ref-type="bibr" rid="R42">42</xref>] showed the effect of insulin on ST segment elevation in this entity. </p></sec><sec id="s5"><title>Year 2001</title><p>In May, Dr. Ihor Gussak and Dr. Hammill developed the major and minor criteria for diagnosing this entity [<xref ref-type="bibr" rid="R43">43</xref>]. The authors proposed that the presence of a major criterion and a minor one, constitute a diagnosis. They consider as major ones the electrocardiographic pattern in a patient without structural heart disease and positivity in pharmacological test; and as minor criteria the presence of positive family history, syncope of unknown origin, documented episode of VT/VF, inducibility in the electrophysiologic study, and positivity in genetic study (yet to be defined).</p></sec><sec id="s6"><title>Year 2002</title><p>In this year, genetic studies carried out by Vatta et al [<xref ref-type="bibr" rid="R44">44</xref>] showed that unexplained nocturnal SCD syndrome, known as SUNDS (Sudden Unexplained Nocturnal Death Syndrome) and Brugada disease are phenotypically, genetically and functionally identical and allelic since both affect the same gene: SCN5A. In November, the first virtual symposium about Brugada Syndrome was held (<bold><italic>VIRTUAL SYMPOSIUM ABOUT THE BRUGADA SYNDROME: TEN YEARS OF HISTORY: 1992/2002</italic></bold>)><ext-link ext-link-type="uri" xlink:href="http://www.brugada-symposium.org"> http://www.brugada-symposium.org</ext-link> [<xref ref-type="bibr" rid="R45">45</xref>], which achieved an extraordinary worldwide impact, where relevant issues related to this entity were updated by specialists of international level, having as Honorary Presidents Pedro, Josep and Ramon Brugada; as President of the Scientific Committee Dr. Andres Ricardo Perez Riera; and as President of the Steering Committee Dr. Edgardo Schapachnik. The event approached basic research, genetics, supplementary methods for diagnosis, electrophysiology; it showed for the first time a vectorcardiogram of this entity; guidelines were discussed regarding sports practice and treatment.</p><p>The material available was organized in Lectures written by the Members of the Honorary Committee, and articles written by the most distinguished specialists in the area, reports of cases, and a round table broadcast through radio via Internet. It is currently possible to access the symposium on the Internet: <ext-link ext-link-type="uri" xlink:href="http://www.brugada-symposium.org"> http://www.brugada-symposium.org</ext-link></p><p> During the event, a change of category was proposed for this entity: from syndrome to disease, founded in that it meets the criteria that define a disease, i.e.: all conditions that affect animals or plants, causing an interruption or modification of their performance as a response to environmental factors, specific agents, genetic defects, or the combination of these, and when at least two of these criteria are present: <list list-type="order"><list-item><p>known etiologic agent;</p></list-item><list-item><p>identified genetic defect;</p></list-item><list-item><p>set of characteristic signs and symptoms or compatible anatomic alterations, or</p></list-item><list-item><p>a combination of the previous items.</p></list-item></list></p></sec><sec id="s7"><title>Year 2003</title><p>Atarashi et al [<xref ref-type="bibr" rid="R46">46</xref>] suggested new electrocardiographic criteria that indicate high risk of VF in Brugada disease in a universe of 60 patients extracted from the Japanese register of Brugada syndrome. Thus, the authors established that: <list list-type="order"><list-item><p>S wave of duration = or > than 80msec in V1 has a predictive value of 40.5% and negative predictive value of 100% for VF with a 100% sensitivity;</p></list-item><list-item><p>ST segment elevation in V2 of 80msec or > measured from the J point has a positive predictive value of 37.8% and negative predictive value of 100% for VF with a 100% sensitivity.</p></list-item></list></p><p>Both criteria are highly specific indicators for VF in this entity.</p></sec><sec sec-type="conclusions" id="s8"><title>Conclusion</title><p>We go through "pre-history" and history since 1953, when Osher and Woff noticed an electrocardiographic pattern that resembled an acute myocardial lesion, which they attributed to a normal variant associated to RBBB and ST segment elevation in the right precordial leads. The historical steps followed to relate this pattern with SCD are described in people frequently in a productive age in life, predominantly in males, and without verifiable structural heart disease.</p><p>The original observation by the Spanish/Catalan Professor, Dr. Pedro Brugada in 1986, in a Pole caucasian and white child, with positive family background of SCD in a first-degree relative, ended in the initial description of the entity in 1992, called "A distinct clinical and electrocardiographic syndrome: right bundle branch block, persistent ST segment elevation with normal QT interval and sudden cardiac death." The impact of the discovery was so important, that in a few years the eponym: Brugada Syndrome, was employed nearly unanimously, by most investigators.</p></sec>
Selecting Dual Chamber or Single Chamber Implantable Defibrillators: What is the Golden Rule?	Could not extract abstract	<contrib contrib-type="author"><name><surname>Santini</surname><given-names>Massimo</given-names></name><degrees>MD, FESC, FACC</degrees></contrib><contrib contrib-type="author"><name><surname>Ricci</surname><given-names>Renato</given-names></name><degrees>MD, FESC</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<sec id="s1"><title>Introduction</title><p>During the last years, the implantation rate of dual chamber defibrillators (ICD) significantly increased worldwide. In 1999, the proportion of dual chamber ICD implants reached 30% in Europe and 50% in U.S.A. [<xref ref-type="bibr" rid="R1">1</xref>]. According to manufacturer data, in Italy, the ratio between implanted single chamber and dual chamber units decreased from 1.97 in 1999 to 1.86 in 2000 and 1.50 in 2001. Technological progress, demonstration of reliability and clinical efficacy of the new devices, combined with their smaller size, contributed to their wide acceptance. Nowadays, the matter to be debated is if all the patients in whom the atrium can be sensed and paced should receive a dual chamber ICD or if device selection should be individually evaluated according to different clinical profiles. As a matter of fact, criteria to identify the patients who may benefit more from dual chamber ICD have not been already defined. The theoretical advantages of dual chamber ICD include: improved discrimination between supraventricular and ventricular tachycardias, optimal treatment of symptomatic bradycardias (pre-existing, drug-induced or late developing), hemodynamic and antiarrhythmic benefits..</p></sec><sec id="s2"><title>Discrimination between supra-ventricular and ventricular arrhythmias</title><p>Superiority of dual chamber detection algorithms versus single chamber ICD in discriminating supraventricular from ventricular tachycardia has been a matter of debate since their introduction for clinical implantation. This issue is particularly challenging if we take into account that the addition of enhanced criteria in the third generation single chamber ICDs, such as tachycardia sudden onset and stability and ventricular electrogram width and morphology, significantly increased single chamber ICD specificity in tachycardia discrimination [<xref ref-type="bibr" rid="R2">2</xref>-<xref ref-type="bibr" rid="R5">5</xref>]. On the other hand, the weak point of such enhanced criteria is represented by decreased sensitivity in ventricular tachycardia detection [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R6">6</xref>]. Dual chamber ICDs have the capability of detecting atrial activity and matching atrial and ventricular patterns. Clinical studies using dual chamber ICDs showed specificity values as high as 80-90% combined with 100% sensitivity [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R12">12</xref>]. Nevertheless, inappropriate detection and therapy may still happen, mainly for "difficult arrhythmias". Hintringer et al [<xref ref-type="bibr" rid="R13">13</xref>] performed a comparison of the detection algorithms of four dual chamber ICD when dealing with a wide spectrum of tachyarrhythmias. In spite of some differences due the algorithms themselves, typical and atypical junctional tachycardia, orthodromic atrio-ventricular tachycardia and 1:1 atrial flutter represented the most challenging patterns. As a consequence, new more sophisticated algorithms are being developed to deal with this task. Appropriate sensing of atrial activity, either during sinus rhythm or during atrial tachyarrhythmias, and appropriate rejection of ventricular activity by the atrial lead is a critical issue to be dealt with. Deisenhofer et al [<xref ref-type="bibr" rid="R14">14</xref>] undertook a prospective, randomized study to compare the incidence of inappropriates therapies in patients treated with VVI-ICDs and DDD-ICDs. They enrolled 92 patients and concluded that DDD-ICD and VVI-ICD were equally safe and effective to treat life-threatening ventricular arrhythmias. Although DDD-ICDs theoretically allow better rhythm classification, the applied detection algorithms did not offer benefits in avoiding inappropriate therapies during supraventricular tachyarrhythmias. As a matter of fact, in the Deisenhofer series 75% of inappropriate therapies in the DDD-ICD group were due to atrial sensing problems, either oversensing or undersensing. This finding stresses the need of careful positioning of the atrial lead during implantation in order to combine optimal atrial electrogram amplitude with far field rejection. It has been suggested that positioning the atrial lead in the lateral atrial wall and selecting bipolar leads with short tip-to-ring distance may reduce far field incidence [<xref ref-type="bibr" rid="R15">15</xref>] [<xref ref-type="bibr" rid="R16">16</xref>]. In spite of that, atrial sensing problems may intermittently appear during the follow-up, also when they were not present at implant or during post-implant testing. A possible explanation for frequent intermittent atrial sensing problems may be the special filter settings in the atrial sensing channels of DDD-ICD, which differs substantially from those of DDD pacemakers. In fact, most detection algorithms in DDD-ICDs need correct and continuous atrial sensing with only short or even no blanking times. This may be difficult when taking into account low voltage atrial electrogram during atrial fibrillation and large ventricular far fields during paced ventricular beats.</p></sec><sec id="s3"><title>Hemodynamic issues</title><p>A emerging key point in debating optimal device selection is represented by the impact of single chamber and dual chamber ICD implantation on hemodynamics.</p><p>It has been demonstrated that, in patients with sinus bradycardia and/or atrio-ventricular conduction disturbances, physiologic pacing by sequential dual chamber stimulation and optimized, individually programmed, atrio-ventricular delay may offer major improvement in hemodynamics and clinical outcome, mainly when heart failure coexists [<xref ref-type="bibr" rid="R17">17</xref>-<xref ref-type="bibr" rid="R19">19</xref>]. On the other hand, during the last years it has been demonstrated that asynchronous ventricular activation induced by apical right ventricular pacing may induce major interventricular and/or intraventricular dysynchrony, which may deteriorate hemodynamics and impair myocardial metabolism [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R21">21</xref>]. Isovolumic contraction time and isovolumic relaxation time lenghtening may critically shorten diastolic filling, impairing cardiac output. Furthermore, delayed activation of the left ventricular lateral wall may lead to late contraction which happens after aortic valve closure, so that not only it does not contribute to stroke volume, but also impair diastolic filling [<xref ref-type="bibr" rid="R22">22</xref>].</p><p>Concern about the potential deleterious effect of unnecessary right ventricular pacing in ICD population is even greater than in pacemaker patients, when considering the higher prevalence of heart failure and left ventricular dysfunction in patients who need ICD implantation [<xref ref-type="bibr" rid="R23">23</xref>,<xref ref-type="bibr" rid="R24">24</xref>]. The recently published DAVID [<xref ref-type="bibr" rid="R25">25</xref>] (Dual Chamber and VVI Implantable Defibrillator) Trial dealt with this issue. Objective of the study was to determine the efficacy of dual chamber pacing compared with backup ventricular pacing in patients with standard indications for ICD implantation but without indications for antibradycardia pacing. The design of the study was a single-blind, parallel group, randomized, multicenter clinical trial. Five hundred and six patients candidates for ICD with left ventricular ejection fraction of 40% or less, no indications for antibradycardia pacing and no persistent atrial tachyarrhythmias, received a dual chamber ICD and were randomly assigned to have the ICD programmed to ventricular back-up pacing at 40/min or to dual chamber rate responsive pacing at 70/min. Main outcome measurement was the combined end point of time to death or first hospitalization for congestive heart failure. The study was early stopped by the Data and Safety Monitoring Board because the conditional power for the original alternative (DDDR-70 being better than VVI-40) was less than 10%. The VVI-40 group had fewer occurrences of the composite end point than DDD-70 group: one-year survival free was 83.9% versus 73.3% (relative hazard 1.61; 95% confidence interval 1.06-2.44, p< 0.03). Although the VVI-40 patients had fewer events, the component end points, either death or heart failure hospitalizations, did not individually reach statistical significance.</p><p>Some criticisms have been pointed out about the DAVID Trial. The study definitely demonstrated that right ventricular pacing is deleterious in patients with left ventricular dysfunction, but it cannot be concluded that single chamber ICD with back-up ventricular pacing is the most useful device for patients with heart failure. In spite of a large enrolled population, the follow-up was very short because of early stopping of the study. Only a minority of patients completed one year follow-up. Programming dual chamber rate responsive pacing at 70/min without long AV delay does not seem the best choice for patients without any indication for antibradycardia pacing. Unnecessary apical right ventricular pacing is probably the key to explain the higher event rates in DDD-70 arm. The Authors stressed that dual chamber pacing could be beneficial in heart failure patients since it may allow a wider use of drugs such as beta blockers which depress sinus and AV node function. As a matter of fact, there were no differences in drug regimen between the VVI and DDD arms after randomization as well as after 6-month follow-up. Finally, patients with atrial tachyarrhythmias were excluded from the study, so introducing a limitation in the clinical value of the study.</p></sec><sec id="s4"><title>Atrial fibrillation in ICD patients</title><p>Atrial fibrillation prevention and early treatment by dual chamber devices, mainly if equipped with atrial antitachycardia functions, may represent a major benefit in patients with heart failure. ICD patients actually show a high incidence of atrial tachyarrhythmias. It has been reported that 20% of them had atrial fibrillation before implantation and that during the life-span of the defibrillator more than 50% may develop atrial fibrillation [<xref ref-type="bibr" rid="R26">26</xref>]. Atrial fibrillation may lead to inappropriate ventricular shocks [<xref ref-type="bibr" rid="R27">27</xref>], ventricular arrhythmia induction [<xref ref-type="bibr" rid="R28">28</xref>], may impair hemodynamics and induce thromboembolic events or acute myocardial infarction, and has been identified as an individual predictor of poor prognosis [<xref ref-type="bibr" rid="R29">29</xref>,<xref ref-type="bibr" rid="R30">30</xref>]. Atrial antitachycardia functions available in some last generation dual chamber ICDs (pacing prevention algorithms and antitachycardia pacing) have been demonstrated to be effective in preventing and early treating atrial tachyarrhythmias. In our own experience [<xref ref-type="bibr" rid="R31">31</xref>], related to 112 patients receiving an ICD because of life threatening ventricular arrhythmias, followed on average for 1 year, anti-tachy-pacing efficacy was as high as 71% on atrial tachycardia and as 36% on atrial fibrillation. Shock success rate was 92% when delivered energy was adequately programmed, which means at least twice the atrial defibrillation threshold at implant. Similar results have been reported by others [<xref ref-type="bibr" rid="R32">32</xref>]. The impact of atrial prevention algorithms and atrial therapies on atrial fibrillation burden has been investigated by Friedman and coworkers [<xref ref-type="bibr" rid="R33">33</xref>]. They designed a study in which atrial fibrillation prevention and termination therapies were randomly programmed "on" or "off" for three months and then crossed over to the opposite arm for an additional 3 months. Fifty-two patients were studied. During the "on" period the arrhythmia burden (hours/month) significantly decreased: the mean burden from 58.5 to 7.8 and the median burden from 2.82 to 0.63. The mean burden reduction was 87%. The reduction in arrhythmia burden during the "on" period could be demonstrated also in the subgroup of patients (forty-one) in whom no shocks were delivered and only antitachy pacing therapies were applied.</p><p>Considering the major clinical impact of symptomatic atrial fibrillation in patients who are candidate for defibrillator implantation, a device equipped with atrial antitachycardia facilities may improve clinical outcome, by preventing acute heart failure, by decreasing inappropriate shocks, by reducing hospitalizations and by improving quality of life.</p></sec><sec id="s5"><title>Perspectives</title><p>The key point for single chamber or dual chamber ICD selection has been progressively switching from optimal tachycardia discrimination to the impact on hemodynamics and on atrial arrhythmia control. To this regard, few controlled data are available and perspectives randomized trials are strongly needed. The impact of device selection on the overall clinical outcome is the target of an ongoing trial [Dual Chamber & Atrial Tachyarrhythmias Adverse Events Study (DATAS), protocol in press [<xref ref-type="bibr" rid="R34">34</xref>]] aimed at comparing clinical benefits of dual chamber ICD with atrial antitachycardia functions with single chamber ICD. The primary end point will be the composite end-point resulting from all-causes mortality, invasive intervention, hospitalizations due to cardiovascular cause, inappropriate shocks and sustained symptomatic atrial tachyarrhythmias. The enrollment is going to be completed soon and the results will be available within the next two years.</p><p>On the other hand, new indications for ICD implantation in primary prevention of sudden death and introduction of triple chamber ICDs capable of delivering cardiac resynchronization therapy are going to change very soon the whole approach to ICD selection. The MADIT-2 trial [<xref ref-type="bibr" rid="R35">35</xref>] demonstrated that in patients with prior myocardial infarction and left ventricular ejection fraction < 30%, ICD implantion was able to reduce 2-year mortality by 31%. Rules for ICD selection in MADIT-2 patients are probably quite different from those applied for patients receiving an ICD in sudden death secondary prevention. A wider use of single chamber ICDs should be expected. First, accurate discrimination between supraventricular and ventricular tachycardia should be less meaningful in patients for whom therapy programming is focused mainly on treating fast ventricular tachycardia and ventricular fibrillation. Secondly, considering the large number of new potential candidates for ICD implantation, selecting a simpler and less costly device may improve the cost-effectiveness of ICD in primary prevention. Development of new low-cost single chamber devices, just capable of detecting ventricular fibrillation and delivering a limited number of shocks is expected for the next years. Such strategy will allow a wider protection of high risk population without an unacceptable increasing of the costs.</p><p>Cardiac resynchronization has been demonstrated to improve functional class, exercise tolerance and quality of life as well as to reduce hospitalisations due to worsening of heart failure [<xref ref-type="bibr" rid="R36">36</xref>,<xref ref-type="bibr" rid="R37">37</xref>] in patients with drug refractory heart failure with atrio-ventricular, inter-ventricular and intra-ventricular dysynchrony. Cardiac resynchronization may be combined with ICD. Considering the large number of ICD candidates with heart failure, drawing guidelines aimed to make the right choice for individual patients will be a major challenge for the next few years.</p></sec>
Nearly Fatal Torsade de Pointes with Sotalol	Could not extract abstract	<contrib contrib-type="author"><name><surname>Strohmer</surname><given-names>Bernhard</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Schernthaner</surname><given-names>Christiana</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Pichler</surname><given-names>Maximilian</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<p>A 75-year-old woman with paroxysmal atrial fibrillation (AF) experienced recurrent seizures at home. Holter monitoring (<xref ref-type="fig" rid="F1">Figure 1</xref>) (each line 50 sec) showed repetitive runs of torsade de pointes (TdP), degenerating into ventricular fibrillation and ventricular tachycardia (VT). Abrupt asytole heralded end of electrical activity and life. Amazingly, 6 min after cardiac arrest a slow ventricular escape rhythm arose spontaneously without resuscitation. At baseline, repolarisation was markedly prolonged (QTc>660 msec) and ventricular bigeminy triggered short bursts of TdP after "long-short" sequences. No hypokalemia or renal dysfunction was present. Following intensive treatment (sedation, magnesium iv, acceleration of heart rate) the patient recovered without neurological deficit. Except left ventricular hypertrophy and incomplete left bundle branch block the results of angiography, electrophysiological study and ajmaline test were normal. There was no family history of sudden death. Months ago a cardioversion attempt with ibutilide triggered polymorphic VT. Therapy with metoprolol (95 mg/day) was discontinued due to poor efficacy of rhythm control. Thus, sotalol (240 mg/day) was initiated in-hospital without signs of QT prolongation within 4 days (QT 416, QTc 432 msec). However, two weeks later the patient presented with an "idiosyncratic" proarrhythmic response to sotalol (I<sub>Kr</sub>-blocking drug) and a life-threatening arrhythmia [<xref ref-type="bibr" rid="R1">1</xref>]. There is growing evidence that drug-induced long QT syndrome (LQTS) may be due to "silent" mutations on LQT genes [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R3">3</xref>]. Although not proven by molecular analysis, our case seems to resemble a subclinical, inherited form of LQTS that makes the patient vulnerable to the QT-prolonging effects of a variety of cardiac and noncardiac drugs. The concept of "repolarization reserve" suggests that any factor that impairs the repolarizing currents renders TdP very likely when I<sub>Kr</sub>-blocking drugs are used. Avoiding torsadogenic drugs should basically prevent recurrence of TdP. However, an implantable cardioverter-defibrillator was placed for safety reasons [<xref ref-type="bibr" rid="R4">4</xref>]. During a follow-up of more than two years a few non-sustained episodes occurred, the longest, a short-coupled polymorphic VT lasted for 25 beats resulting in a diverted shock.</p>
Intra-atrial Re-entrant Tachycardia with Wenckebach Periodicity	Could not extract abstract	<contrib contrib-type="author"><name><surname>Sarubbi</surname><given-names>Berardo</given-names></name><degrees>MD, PhD</degrees></contrib><contrib contrib-type="author"><name><surname>Vergara</surname><given-names>Pasquale</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>D'Alto</surname><given-names>Michele</given-names></name><degrees>MD, PhD</degrees></contrib><contrib contrib-type="author"><name><surname>Sessa</surname><given-names>Francesco</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Calabro</surname><given-names>Raffaele</given-names></name><degrees>MD</degrees></contrib>	Indian Pacing and Electrophysiology Journal	<p>A 15-year-old girl, previously asymptomatic for palpitations, underwent a successful atrial septal defect (ASD) device closure. Twelve weeks after the procedure, the patient was admitted complaining of dyspnoea on effort and palpitations. The twelve-lead ECG showed a narrow QRS tachycardia with slight heart rate irregularity, with a mean HR of 170bpm. P waves were not clearly identified with a suspicious of negative P-waves in II, III and aVF leads. No clear relationship could be observed between the suspected P waves and QRS complexes (<xref ref-type="fig" rid="F1">Figure 1</xref>).</p><p>Echocardiographic evaluation showed normal position of ASD Device, with no residual shunts, mild dilatation of right atrium and right ventricle with a moderate biventricular systolic function impairment. Transesophageal electrophysiological study showed an atrial tachycardia with a regular A-A interval of 220ms and an A-V conduction delay with a A-V nodal Wenckebach periodicity, leading to an irregular V-V interval ranging between 375 and 300ms (<xref ref-type="fig" rid="F2">Figure 2</xref>). Overdrive atrial pacing at a cycle length of 140ms (almost 70% of the measured A-A cycle) could stop the tachycardia (<xref ref-type="fig" rid="F3">Figure 3</xref>).</p><p>The case represents a rare form of intra-atrial re-entry tachycardia with an A-V nodal Wenckebach periodicity. It has been already shown that in intra-atrial re-entry tachycardia, variables degrees of block may be present throughout the entire episode of sustained tachycardia. Neither the ventricles nor the A-V node are required for this arrhythmia. It is the appearance of AV block with maintenance of the supraventricular tachycardia that strongly suggests a supranodal origin. The presence of persistent termination of the arrhythmia through atrial pacing excluded an automatic origin.</p>
Trends in Cardiac Pacemaker Batteries	<p>Batteries used in Implantable cardiac pacemakers-present unique challenges to their developers and manufacturers in terms of high levels of safety and reliability. In addition, the batteries must have longevity to avoid frequent replacements. Technological advances in leads/electrodes have reduced energy requirements by two orders of magnitude. Micro-electronics advances sharply reduce internal current drain concurrently decreasing size and increasing functionality, reliability, and longevity. It is reported that about 600,000 pacemakers are implanted each year worldwide and the total number of people with various types of implanted pacemaker has already crossed 3 million. A cardiac pacemaker uses half of its battery power for cardiac stimulation and the other half for housekeeping tasks such as monitoring and data logging. The first implanted cardiac pacemaker used nickel-cadmium rechargeable battery, later on zinc-mercury battery was developed and used which lasted for over 2 years. Lithium iodine battery invented and used by Wilson Greatbatch and his team in 1972 made the real impact to implantable cardiac pacemakers. This battery lasts for about 10 years and even today is the power source for many manufacturers of cardiac pacemakers. This paper briefly reviews various developments of battery technologies since the inception of cardiac pacemaker and presents the alternative to lithium iodine battery for the near future.</p>	<contrib contrib-type="author"><name><surname>Mallela</surname><given-names>Venkateswara Sarma</given-names></name><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Ilankumaran</surname><given-names>V</given-names></name><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Rao</surname><given-names>N.Srinivasa</given-names></name><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><sec id="s1a"><title>Cardiac Pacemaker</title><p>The pacemaker unit delivers an electrical pulse with the proper intensity to the proper location to stimulate the heart at a desired rate. The cardiac pacemaker comprises of a pulse generator and a lead system. The pulse generator houses electrical components responsible for generating the pulse (via output circuits) at the proper time (via timing and control circuits) based on events sensed (via sensing circuits). It also contains a power supply (battery) and may include other elements such as telemetry for testability and programmability and memory (ROM or RAM) to store data for diagnostic purposes [<xref ref-type="bibr" rid="R1">1</xref>].</p><p>Impulses are transmitted to the heart by means of a lead, which is attached to the pulse generator via the connector block. A lead is either unipolar or bipolar; a unipolar lead contains one insulated coil, whereas a bipolar lead contains two coils, separated by an inner insulation. An outer insulation shields a lead from the environment. The tip of a lead, which contains an electrode, is implanted into the inner, endocardial surface of the heart, the actual location depends on the type of pacemaker. The pacemaker unit is usually implanted in the pectoral region, with the lead running through the right subclavian vein to the internal surface of the heart. A pacemaker is programmed by means of a programmer, a computer with a special user interface for data entry and display, and with special software to communicate with the pacemaker. The telemetry head is placed above the location of the pacemaker; information from the programmer to the pacemaker, and back, is transmitted by means of telemetry.</p><p>The casing of the pulse generator functions as housing for the battery and all other electronic and electrical circuits. A connector block, made of polyurethane, (glass materials were used to comprise the connector block in earlier models) is located at the top of the pacemaker. It serves to attach the pacemaker to the pacemaker lead(s). The present day pulse generator case is made of titanium, a metal that is ten times as strong as steel, but much lighter. Titanium and two of its alloys, niobium and tantalum, are biocompatible, they exhibit physical and mechanical properties superior to many other metals. The modulus of elasticity (measure of stiffness) of titanium and its alloys range between 100-120GPa. Extreme resistance to corrosion and durability make titanium and its alloys ideal materials for hermetically sealed pulse generator cases for cardiac pacemakers.</p><p>Titanium replaced ceramics and epoxy resin with silicone rubber, which were used for encapsulation of some pacemakers in the past. To assemble the pulse generator, the hybrid circuits and the battery are placed in the titanium case (ASTM Grade 1) in a specially designed clean room that has no static charge (less than 1% moisture) and no dust in it. Once the hybrid circuits and the battery are in the casing, the casing is welded shut with a high-powered laser beam. This laser beam gives the pulse generator a hermetic seal, which means that the device is airtight and liquid-tight. After welding, the top, or header of the pacemaker is attached and the entire device is covered in a thin layer of plastic (epoxy plastic). This plastic coating further seals the pacemaker.</p><p>The casing is a given a kind of elliptical shape and a typical pacemaker diagram is shown in <xref ref-type="fig" rid="F1">Figure 1</xref>. This upgrade to titanium allowed patients to safely use appliances such as microwave ovens because titanium helps to shield the internal components and reduce the external electromagnetic interference. In addition, titanium casing shields from ground level cosmic radiation.</p></sec><sec id="s1b"><title>Batteries for Cardiac Pacemakers</title><p>In 1958, Ake Senning, a thoracic surgeon at the Karolinska Hospital in Stockholm, implanted myocardial electrodes and a pulse generator with a rechargeable nickel-cadmium battery in a 40-year-old patient. Senning and his associate, Rune Elmquist, an engineer with the Swedish firm Elema Schonander, had developed and tested this pacemaker between 1956 and 1958 [<xref ref-type="bibr" rid="R2">2</xref>]. The pulse generator failed within a few hours; a successor lasted about 6 weeks. The history of the implantable cardiac pacemaker is traced from its inception in 1951, through its development and trials in 1958, to its successful implantation in 10 patients in 1960, and on to its commercial realization [<xref ref-type="bibr" rid="R3">3</xref>]. The usage of implanted pacemakers has been ever increasing since then. The battery occupies major portion of the pulse generator in terms of weight, volume, and size. The most important factor for a cardiac pacemaker battery is its reliability. Unlike many consumer products, batteries in implantable devices cannot be replaced. They are hard wired at the time of manufacture before the device is hermetically sealed. From that point on, the battery is expected to power the device during final testing at the factory, during the shelf life and throughout the useful life of the device while it is implanted. In general the power source of the implantable device is the only component, which has a known predictable service life, which in turn determines the service life of the implanted device itself.</p><p>It is indeed fascinating to see the breadth and the vision of the early investigators of implantable power sources [<xref ref-type="bibr" rid="R1">1</xref>] in the almost desperate search for a power source that would enable the pacemaker to last as long as the expected lifetime of the average patient. This paper presents a brief history and review of various types of batteries used in cardiac pacemakers since beginning. The smooth transition from zinc-mercury, nuclear batteries to the lithium-iodine batteries are presented along with product information obtained from the manufacturers. The technical advantages of lithium iodine battery in terms of its longevity, no gas generation, adaptable shapes and sizes, corrosion resistance, minimum weight, excellent current drain characteristics suitable to cardiac pacemakers are highlighted in this paper. The future of cardiac pacemaker batteries in terms of alternatives to lithium iodine battery is also presented.</p></sec><sec id="s1c"><title>Electrochemical Power sources </title><p>We need to generate electrical energy from some other source of energy. Chemical energy is the most practical source and is generally used in one of two possible ways. Fuels can be burnt in a heat engine or fuel cells can be used. Fuel cells have no moving parts and do not require the mechanical energy to generate Electrical energy. Chemical energy can also be stored in two types of electrochemical power sources, primary cells or batteries, and secondary cells or batteries. Primary cells are those used once and then discarded, whereas secondary cells can be discharged and recharged many times. In theory, many electrochemical reactions are reversible. In practice, only a few systems are worthwhile and safe. In general, electrochemical power sources have developed in an evolutionary manner.</p></sec><sec id="s1d"><title>Battery Performance Parameters</title><p>The definitions for some of the important parts of a battery and its performance parameters like voltage, duty cycle, temperature, shelf life, service life, safety and reliability, internal resistance, specific energy (watt-hours/kg), specific power (watts/kg), etc are well known [<xref ref-type="bibr" rid="R4">4</xref>]. A good battery design is a compromise between various performance parameters to meet the requirements of the specific application. Critical factors in selecting a cardiac pacemaker battery technology are: minimum and maximum voltage, initial, average, and maximum discharge current, continuous or intermittent operation (size and duration of current pulses), long shelf and service life, high specific energy and specific power, impact, and good performance in a variety of conditions (temperatures, duty cycles, etc.). Cardiac pacemaker battery design poses special challenges in development of biocompatible materials, corrosion and sealing, light weight and flat type, high reliability, accurate end of life battery predictions, etc.</p></sec><sec id="s1e"><title>Early Developments</title><p>Rechargeable (secondary batteries) nickel-cadmium batteries were used in the beginning (in 1958) of pacemaker implants in human beings. They were inductively recharged by the transmission of energy to the implanted receiver. The cell voltage was 1.25 V and the capacity was 190 mAh. The major problems were two fold, the first being very short life time and the second was to place the responsibility for recharging in the hands of patients, which is not a good medical practice. It was well known that primary or non-rechargeable batteries would give longer lifetime compared to secondary batteries. There are still some rechargeable pacemakers in use though not sold any more.</p><p>Some of the early pulse generators constructed mainly from discrete components were powered by series-wired mercury-zinc batteries [<xref ref-type="bibr" rid="R5">5</xref>]. Three to six cells in series provided 4-8 V. They were widely used at that time (around 1960s). Such mercury-zinc batteries were cast in epoxy, which was porous to the discharge of the battery released hydrogen and permitted its dissipation, which required venting and hence could not be hermetically sealed. This allowed fluid leakage into the pacemaker at times that caused electrical shorting and premature failure. The terminal voltage decay characteristic of the mercury-zinc battery is such that normal battery depletion results in little change in the terminal voltage until the end of battery’s useful life. This makes failure difficult to anticipate [<xref ref-type="bibr" rid="R6">6</xref>]. This battery was improved in its design and still the life was only about two years with an abrupt drop in voltage as they become depleted. No device of this type is currently in use [<xref ref-type="bibr" rid="R7">7</xref>].</p><p><italic>Biological batteries</italic> (which use power from within the human body) were experimented unsuccessfully1 for practical use in pacemakers.</p><p><italic>Nuclear batteries</italic> were tried successfully for some period. Practical nuclear batteries use plutonium (<sup>238</sup>Pu). It has a half-life of 87 years so the output degrades only by 11% in 10 years [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R7">7</xref>]. However it is highly toxic and 1μg in the blood stream could be fatal. Early pacemakers used metallic plutonium where as later ones used ceramic plutonium oxide. The plutonium emits alpha particles, which impact upon the container and generate heat. Thermopiles of dissimilar p- or n-doped bismuth telluride generate the electricity for the pacemaker circuits. Though these nuclear power sources had very long life, they were large and created problems when travelling between states and countries due to the presence of their radioactive fuel. They also must be removed at the time of death and returned for proper disposal. Nuclear powered pacemakers are no longer sold [<xref ref-type="bibr" rid="R7">7</xref>] but still a small number of implanted nuclear devices that remain in use. Nuclear power sources became obsolete with the development of lithium batteries.</p></sec></sec><sec sec-type="" id="s2"><title>Lithium Batteries</title><p>Lithium has the highest specific energy of all but it has only become possible since mid 1970s to manufacture practical batteries. Because lithium reacts violently with water, non-aqueous electrolytes must be used. Organic solvents such as acetonitrile and propylene carbonate, plus inorganic solvents such as thionyl chloride (SOCl2) are typical, with a compatible solute to provide conductivity. Many different materials such as sulfur di, thionyl chloride, manganese dioxide, and carbon monofluoride, are used for the active cathode material.</p><sec id="s2a"><title>Introduction</title><p>Introduction of a lithium iodine battery in 1975 greatly extended the pacemaker battery life (more than 10 years for some models) and replaced the mercury-zinc battery. Lithium Primary batteries are used in pacemakers since they meet the requirements of long life, low drain current and voltage characteristics. The shelf life of primary lithium cells is typically equivalent to a 10% loss of capacity over five years [<xref ref-type="bibr" rid="R1">1</xref>]. This compares with a similar loss for alkaline cells over only one year. The long shelf life of lithium batteries arises from the lithium metal surface becoming passivated by reaction with the electrolyte. All lithium systems are said to be thermodynamically unstable but kinetically stable. They produce no gas and hence they can be hermetically sealed. In addition, the terminal voltage decay characteristic is well behaved, falling slowly enough for battery end-of-life(EOL) to be anticipated in routine follow up.</p><p>Lithium batteries are categorized under liquid cathode cells, solid cathode cells, and solid electrolyte cells.</p><p>The <italic>liquid cathode</italic> systems, Li/SO2, Li/SOCl2 and Li/SO2Cl2, plus their derivatives, are capable of higher discharge rates than the solid cathode systems such as Li/MnO2 and Li/CFX. These are not suitable for applications in implanted cardiac pacemakers. However lithium sulfur dioxide batteries are used in automated external defibrillators (“AEDs”) that can restore a normal cardiac rhythm to victims of sudden cardiac arrest. <italic>Solid Cathode</italic> Lithium Cells use solid cathode materials such as MnO2, CuO, V2O5 and carbon monofluoride, (CF)n. They have the advantage of not being pressurized, although they cannot be discharged as rapidly as liquid cathode cells. They are available in button and cylindrical forms. About 80% (by number) of all lithium batteries in use are of the Li/MnO2 type. The energy density is similar to that of the Li/SO2 cells when discharged slowly and their slow self-discharge characteristic make them suitable for memory backup, watches, calculators, cameras, mines and munitions, etc. Voltage delay appears to be less of a problem with solid cathode cells.</p><p>The solid cathode cells do not support currents as high as the liquid cathode ones. This is because the liquid cathode undergoes a discharge at the surface of the electrode (which comprises a high surface area carbon supported on a metal mesh) where the discharge products are deposited. In contrast, discharging at a solid cathode involves diffusion of lithium ions into the bulk of the cathode, which is a slower process.</p><p>Continuous operation of liquid and solid cathode cells above 2A will lead to a significant rise in cell temperature, so this needs to be borne in mind for a particular battery application, the temperature rise being of more importance for the high pressure Li/SO2 cells. Possible hazards, like explosions associated with lithium liquid and solid cathode batteries are still a concern for absolute safety and lot of research is still going on to stipulate the rules and regulations as to how they must be disposed off towards the end of their life.</p><p><italic>Solid electrolyte lithium cells:</italic> Several solids, such as lithium iodide, are electronic insulators but reasonably good ionic conductors and can be used as the electrolyte in solid electrolyte batteries. Such batteries are characterized by extremely long service life at low drain currents, even at high temperatures. They are very much suitable for applications such as cardiac pacemakers, and for preserving volatile computer memory.</p><p>Since 1972, a variety of lithium batteries have been used. These include Li/SOCI2, Lithium-silver chromate cell [Li/Ag2CrO4], lithium copper-sulfide cell[ Li/CuS], lithium – thionyl chloride cell, Li/I2-Polyvinylpyridine (PVP), and, in more limited use, Li/LiI(Al2)3/PbI2,PbS, Pb. In addition to their widespread use in consumer products, lithium primary batteries are the power source of choice for a range of medical implants.</p><p> The lithium iodine-polyvinylpyride (PVP) is the principal cardiac pacemaker battery that has been in long use. The internal impedance (The resistance of a cell to an alternating current of a particular frequency) of the lithium iodine cell is an important factor in battery performance. The greater the impedance, the more difficult it is to pass current through the cell. Increased cell impedance corresponds to a decreased power source at the cell terminals. The beginning-of-life (BOL) impedance ranges from 50 to 100 Ohms. The impedance increases during service to values from 20,000 to 30,000 Ohms during the accumulation of discharge product [<xref ref-type="bibr" rid="R8">8</xref>].</p></sec><sec id="s2b"><title>Lithium Iodine Battery for Cardiac Pacemaker</title><p>The lithium / iodine-polyvinylpyridine battery, first implanted in 1972 has become the power source of choice for cardiac pacemaker. Since then, improvements in cell chemistry, cell design, and modeling of cell performance have been made [<xref ref-type="bibr" rid="R10">10</xref>]. Cells today exhibit an energy density over three to four times as great as cells produced in 1972. Well over 3 million pacemakers have been implanted with this chemistry, and the system has exhibited excellent reliability. The battery chemistry provides a long shelf life and high energy density. Lithium cupric sulfide was used in some pacemakers [<xref ref-type="bibr" rid="R2">2</xref>] manufactured by the Cordis Corporation due to its excellent energy density. However, due to the corrosive nature of this compound many abrupt pacemaker failures occurred when the battery chemicals ate through their containment. It is still present in some of the already implanted pacemakers but lithium cupric sulfide is no longer used.</p><p>Lithium Iodine has two characteristics that make it an excellent power source for cardiac pacemaker applications. The self-discharge rate is very low resulting in a long shelf life. It has a stable voltage through much of the useful life then tapers down in a gradual and predictable manner. This makes predicting the elective replacement time safe and easy.</p><p>The cathode is a complex of iodine and poly-2-vinyl pyridine (P2VP). Neither conducts electricity, but when mixed and heated at 149°C for 3 days, they react into a black viscous paste that conducts electricity. This is poured into the battery when molten and cools to form a solid. When this paste contacts metallic lithium, a monomolecular layer of crystalline lithium iodine forms. It is a molecular semiconductor that passes lithium ions, as required for current flow, but not iodine molecules [<xref ref-type="bibr" rid="R1">1</xref>]</p></sec><sec id="s2c"><title>Chemical Reactions</title><p>Conventional current flows through a device from anode to cathode. For a battery, the current flows from the negative anode, through the battery, to the positive cathode. Oxidation of metal occurs at the anode, <graphic xlink:href="ipej040201-02i.jpg" mimetype="image" position="float"/> and reduction of halide occurs at the cathode, <graphic xlink:href="ipej040201-03i.jpg" mimetype="image" position="float"/>. The combined reaction is, <graphic xlink:href="ipej040201-04i.jpg" mimetype="image" position="float"/> Conventional current flows from anode to cathode. The lithium reacts with iodine to form lithium-iodide, which grows in volume and increases the resistance.</p></sec><sec id="s2d"><title>Internal Resistance</title><p>The internal cell resistance (Rdc, The resistance to flow of an electric current within a cell; the sum of the ionic and electronic resistance of the cell components) as a function of capacity for PVP- coated and uncoated (see below) lithium anode are shown in <xref ref-type="fig" rid="F2">Figure 2</xref>. The open circuit voltage (OCV) and voltage at 20 μA load characteristics are shown in <xref ref-type="fig" rid="F3">Figure 3</xref>. It is seen that the voltage above 2.2V (required minimum by the pacemaker electronics) is well maintained until the 2.5 Ah rating of the battery.</p></sec><sec id="s2e"><title>Manufacturing</title><p>Lithium is easily formed into sheets that can be cut to the required sizes. It is easily pressed into specific anode shapes. The lithium anode is coated three times with a solution of PVP. The solvent is evaporated to leave a contiguous film of pure PVP on the anode surface. The precoated central lithium anode is corrugated to increase its area and lower battery impedance. To obtain lower impedance, newer designs use more concentrated active materials and larger anode surface areas. Multiple anode surfaces may be used to lower the impedance. The complex of iodine and poly-2-vinyl pyridine (P2VP) is poured into the cathode case and allowed to cool [<xref ref-type="bibr" rid="R11">11</xref>].</p></sec><sec id="s2f"><title>Testing</title><p>To maintain high reliability (of the order of 0.005 % failures per month), cells are designed conservatively. They are manufactured under stringent quality controls, as demanded by the Good Manufacturing Practices (GMP) issued by the Food and Drug Administration (FDA), USA. The qualification testing is performed under accelerated test conditions specified for Li/I2-PVP cells [<xref ref-type="bibr" rid="R11">11</xref>]. The list includes Non-destructive examinations, thermal cycling, high pressure, mechanical vibration, temperature / humidity, mechanical shock, voltage / temperature, seal terminal strength, elevated temperature discharge, destructive analysis, and solvent resistance.</p></sec><sec id="s2g"><title>Longevity and Battery life estimation</title><sec id="s2g1"><title>Longevity</title><p>The pacemaker battery provides energy required for the operation of the circuitry of a pacemaker, which includes the control, sensing and pulse-generating units. A major concern in using battery is its longevity. Longevity of a battery can be determined knowing battery capacity (Ah) and current drain (microamperes). The current drain is dependent on the type of electrode as well as the circuitry and type of pulse generation of the pacemaker.</p></sec><sec id="s2g2"><title>Life Estimation</title><p>Since the longevity of a cardiac pacemaker means its battery life, it is essential to have the circuitry to identify the remaining useful life of battery in a simple and reliable manner. Monitoring of internal resistance is a convenient tool for estimation of discharge level and for predicting the approaching end-of-service.</p><p>In many pacemaker systems, circuits are provided to measure the internal resistance of the battery to deduce the remaining life. With this circuit1, the pacemaker is first switched to “test mode“ and a resistive load is applied to the battery to measure the voltage drop. The status of the battery is indicated by generating a series of test pulses. Depending on the internal voltage drop, and thus the internal resistance, the frequency of the stimulation pulse is changed, which is measured externally. However, this circuit can only be used for batteries with increasing internal resistance as the battery discharges.</p><p>To overcome the limitations of the above technique and to measure the life expectancy of a battery with constant internal resistance, another technique was proposed [<xref ref-type="bibr" rid="R1">1</xref>]. The battery test circuit is provided with a pulse counter and input logic to measure the consumed charge from the operating parameters of the pacemaker and the number of pulses delivered over a period of time. During each test, the charge delivered since the last battery test is calculated based on the count in the pulse counter, which is then summed to the contents of the charge counter in memory. The content of the charge counter is a measure of the total charge consumed and provides information about the remaining life of the battery [<xref ref-type="bibr" rid="R12">12</xref>].</p><p>The-circuit is implemented internally in the pacemaker unit and a means is provided to report the value of the charge counter when interrogated by telemetric methods. The advantage of this method is that there is no need to alter the frequency of stimulation pulses while testing the battery.</p><p>In some models-of pacemakers, approximately twice per day the device evaluates the battery status, which is reported during follow-up; where as in some other models, battery status is automatically evaluated every 11 hours. Battery status may be displayed in the form of a gauge (showing BOL, ERT, and EOL) and longevity remaining (> 5 years to < 0.5 years in 0.5 year increments) at 100% pacing [<xref ref-type="bibr" rid="R13">13</xref>].</p></sec></sec><sec id="s2h"><title>Specifications</title><p>The battery should meet the pacemaker pulse requirements in the range of 25 μ J, a very small power (compared to 15-40 J for Implantable Cardioverter Defibrillators). The following are broad specifications. <list list-type="alpha-lower"><list-item><p>Open Circuit Voltage: 2.8 Volt</p></list-item><list-item><p>Control Circuit minimal voltage: 2.2 Volt</p></list-item><list-item><p>Control Circuit current drain: 10 μA</p></list-item><list-item><p>EOL battery resistance: 10 k Ohms</p></list-item><list-item><p>C<sub><italic>hold</italic></sub>: 10 μF</p></list-item><list-item><p>Oscillator frequency: 167 Hz</p></list-item><list-item><p>Duty Cycle; 16.7 % </p></list-item><list-item><p>Ah rating: 2 Ah (typical rating)</p></list-item><list-item><p>Reliability: 99.6% probability of survival beyond 8 years</p></list-item><list-item><p> Failure Rate: 0.005 % failures/month</p></list-item></list></p></sec><sec id="s2i"><title>Weight, Volume, Shape and Size</title><sec id="s2i1"><title>Weight</title><p>Half of the occupied space is consumed by the internal battery14 in cardiac pacemaker. Therefore the energy density (energy/volume) and specific energy (energy/mass) are important considerations for implantable batteries. Compared with lead, the same volume of lithium provides eight times as much electricity, at one-thirtieth the weight. The weight of a lithium-iodine battery varies from about 12.5 grams to 15.5 grams for different manufacturers [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R15">15</xref>,<xref ref-type="bibr" rid="R16">16</xref>] of the pacemaker unit. The variation in weight is primarily due to the longevity and current drain capabilities of the battery.</p></sec><sec id="s2i2"><title>Volume</title><p>The volume occupied by the battery in a pacemaker (pulse generator unit) is also about half the total volume. This varies from 5 to 8 cc for the units manufactured by different manufacturers [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R15">15</xref>,<xref ref-type="bibr" rid="R16">16</xref>].</p></sec><sec id="s2i3"><title>Shape</title><p>Most of the cardiac pacemakers are shaped as variations on circular or elliptical objects1 to avoid having sharp corners that might penetrate the skin or damage surrounding tissues. Therefore, the batteries in these devices are shaped to conform to the overall device geometry, and often approximate a semicircle with a radius of about 3cm and a depth of 6 to 8mm.</p></sec><sec id="s2i4"><title>Size</title><p>Typical dimensions of an implantable cardiac pacemaker are in the range of 49 mm x 46 mm x 6 mm / 47 mm x 41 mm x 7 mm / 45 mm x 52 mm x 7 mm / 44 mm x 42 mm x 8 mm / 41 mm x 50 mm x7 mm [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R15">15</xref>,<xref ref-type="bibr" rid="R16">16</xref>]. The dimensions vary from one model to an another as well as from one manufacturer to an another. The battery occupies about half of the size, and volume given in the table. Most of the companies use the lithium iodine battery developed first by Wilson Great Batch [<xref ref-type="bibr" rid="R17">17</xref>].</p></sec></sec></sec><sec sec-type="" id="s3"><title>Future Batteries</title><p>Newer designs are aimed at lowering impedance by using more concentrated active materials and increasing anode surface area9. Any increase in service life of implantable medical devices, including cardiac pacemakers is highly desirable and important. In this connection it seemed worthwhile to use power sources with higher energy densities and lower internal resistance. Indeed, batteries based on other lithium systems were also proposed; lithium-silver chromate, lithium-cupric sulfide, lithium-thionyl chloride being among them. However all these batteries were rejected.</p><p>With several features being added to the implantable cardiac pacemakers and other implanted medical devices, manufacturers are going to need to pull more energy out of the battery more quickly. Today's pacemakers typically use lithium iodine batteries and defibrillators employ lithium silver vanadium oxide, next-generation systems may slowly migrate toward a newer type of lithium battery: lithium carbon monofluoride (CFx). CFx batteries reportedly offer higher energy density and can be pulsed at currents above 20 mA, which is slightly better than today's competing batteries [<xref ref-type="bibr" rid="R18">18</xref>] [<xref ref-type="bibr" rid="R19">19</xref>]. Such innovations will be necessary, particularly if OEM visions of patient management come to fruition. Medtronic, for example, has already embarked on a decade-long program, known as Vision 2010, which calls for far-reaching use of device connectivity. Ultimately, engineers say they can foresee a day when an implanted heart monitor will detect a problem and call an ambulance; all while the patient lies sleeping.</p></sec><sec sec-type="" id="s4"><title> Lithium / carbon monofluoride (Li / CFx): a new pacemaker battery</title><p>The reduction in pacemaker size coupled with addition of more current demanding functions have motivated the development of batteries that can supply higher current densities at useful voltages than lithium / iodine batteries in use today while retaining the volumetric energy density of that system. The battery can deliver currents in the milliampere range without significant voltage drop. The system is compatible with titanium casing, allowing a 50% reduction in weight over the same size lithium / iodine battery. Cells have been designed and tested in these laboratories and have been shown to be suitable for advanced pacemaker applications [<xref ref-type="bibr" rid="R20">20</xref>].</p></sec><sec sec-type="" id="s5"><title> Lithium-polycarbon fluoride battery</title><p>This type of battery possesses very high energy density and is capable to ensure pulse discharge current as high as tens of milliamps. At the same time, in contradiction to lithium-iodine batteries, lithium-polycarbon fluoride ones use a liquid electrolyte, specifically 1 M LiBF4 in gamma- butyrolactone. This fact warrants special attention to a problem of sealing batteries for liquids and gases (due to electrolyte impurities). It is very important to check the battery leak-tightness, (meeting the reliability standards laid down for implantable cardiac pacemakers) which would qualify for medical applications. The only volatile component of lithium-polycarbon fluoride battery is gamma-butyrolactone. For detecting volatile substances, gas chromatography is viewed as the most suitable and most accurate technique. Sadly enough, gas chromatography is not able to be used for gamma-butyrolactone detection because it decomposes at a temperature below its boiling point. However, liquid chromatography is suitable for this analysis, but it is a much more sophisticated and expensive technique. Of interest, is a simple method of gamma-butyrolactone detection that was developed [<xref ref-type="bibr" rid="R21">21</xref>].</p></sec>
New Concepts in Pacemaker Syndrome	<p>After implantation of a permanent pacemaker, patients may experience severe symptoms of dyspnea, palpitations, malaise, and syncope resulting from pacemaker syndrome. Although pacemaker syndrome is most often ascribed to the loss of atrioventricular (A-V) synchrony, more recent data may also implicate left ventricular dysynchrony caused by right ventricular pacing. Previous studies have not shown reductions in mortality or stroke with rate-modulated dual-chamber (DDDR) pacing as compared to ventricular-based (VVI) pacing. The benefits in A-V sequential pacing with the DDDR mode are likely mitigated by the interventricular (V-V) dysynchrony imposed by the high percentage of ventricular pacing commonly seen in the DDDR mode. Programming DDDR pacemakers to encourage intrinsic A-V conduction and reduce right ventricular pacing will likely decrease heart failure and pacemaker syndrome. Studies are currently ongoing to address these questions.</p>	<contrib contrib-type="author"><name><surname>Farmer</surname><given-names>D. Michael</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Estes</surname><given-names>NA. Mark</given-names><suffix>III</suffix></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Link</surname><given-names>Mark S</given-names></name><degrees>MD</degrees></contrib><aff>Tufts-New England Medical Center, Tufts University School of Medicine, Boston, Massachusetts</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Pacemaker syndrome consists of the cardiovascular signs and symptoms of heart failure and hypotension induced by right ventricular (RV) pacing. The reported incidence of pacemaker syndrome likely approaches 20% in rate modulated ventricular-based (VVIR) pacing [<xref ref-type="bibr" rid="R1">1</xref>]. Over the last three decades the understanding of pacemaker syndrome has evolved. Initially described as only the sequela of A-V dysynchrony, pacemaker syndrome may be highly influenced by RV-left ventricular (LV) dysynchrony as well. The intricate interplay of these two factors along with autonomic and neurohormonal changes cause the compilation of symptoms known as the pacemaker syndrome.</p></sec><sec sec-type="" id="s2"><title>Pacemaker Syndrome</title><p>Pacemaker syndrome was first described in 1969 by Mitsui et al. as a collection of symptoms associated with right ventricular pacing [<xref ref-type="bibr" rid="R2">2</xref>]. Since its first discovery, there have been many definitions of pacemaker syndrome, and the understanding of the cause of pacemaker syndrome is still under investigation. In a general sense, pacemaker syndrome can be defined as the symptoms associated with right ventricular pacing relieved with the return of A-V and V-V synchrony.</p><p>The symptoms of pacemaker syndrome included dyspnea on exertion, paroxysmal nocturnal dyspnea, orthopnea, hypotension, pre-syncope, and even syncope [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R5">5</xref>]. Heart failure signs include elevated neck veins, rales, and pedal edema. Physical exam can often reveal cannon A-waves. This sign occurs secondary to ventricular-atrial (V-A) conduction and the contraction of the atria against closed A-V valves. Although relatively uncommon, syncope has been attributed to pacemaker syndrome. Syncope is usually associated with systolic blood pressure declines of greater than 20 mm Hg that can occur with the onset of pacing. Additional symptoms attributed to pacemaker syndrome include easy fatigability, malaise, headache, and the sensation of fullness and pulsations in the head and neck. Pacemaker syndrome is most severe when intact V-A conduction is present6. The elevated venous pressures associated with the contraction against closed A-V valves causes a vagal afferent response resulting in peripheral vasodilation and hypotension.</p></sec><sec sec-type="" id="s3"><title>Incidence</title><p>The reported incidence of pacemaker syndrome has ranged from 2% [<xref ref-type="bibr" rid="R7">7</xref>] to 83% [<xref ref-type="bibr" rid="R8">8</xref>]. The wide range of reported incidence is likely attributable to two factors. The first is the criteria used to define pacemaker syndrome. In the Pacemaker Selection in the Elderly (PASE) study, pacemaker syndrome was defined as symptoms severe enough to warrant reprogramming from ventricular to dual-chamber pacing [<xref ref-type="bibr" rid="R9">9</xref>]. The Mode Selection Trial (MOST) investigators defined pacemaker syndrome as occurring if either one of two different criteria occurred [<xref ref-type="bibr" rid="R1">1</xref>]. The first criteria was new or worsened dyspnea, orthopnea, elevated jugular venous pressure, rales, and edema with ventricular (VA) conduction during ventricular pacing. The second criteria was symptoms of dizziness, weakness, presyncope, or syncope, and a >20 mm Hg reduction of systolic blood pressure when the patient had VVIR pacing compared with atrial pacing or sinus rhythm. The second factor in the wide range of reported incidence of pacemaker syndrome is the therapy used to resolve that diagnosis. When surgical revision is required to upgrade a patient from VVIR pacing, the incidence of pacemaker syndrome has been low. In the Canadian Trial of Physiologic Pacing (CTTOP), surgical revision was required for the change from VVIR to dual chamber pacing, and the incidence of pacemaker syndrome was reported to be 2.7% at three years [<xref ref-type="bibr" rid="R10">10</xref>]. In other pacemaker mode trials, patients were implanted with dual chamber devices and then either programmed to VVIR or DDDR pacing. In these studies, patients complaining of symptoms consistent with pacemaker syndrome could be easily upgraded to DDDR mode by simple pacemaker reprogramming. In the PASE and MOST studies in which devices could be reprogrammed from VVIR to DDDR mode, the incidence of pacemaker syndrome was higher than in those studies that required an invasive intervention to change pacing mode [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R11">11</xref>].</p></sec><sec sec-type="" id="s4"><title>A-V vs. V-V dysynchrony</title><p>The majority of the symptoms of pacemaker syndrome are likely attributable to the reduction in cardiac output that is associated with right ventricular pacing [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R5">5</xref>]. Several studies have demonstrated the hemodynamic superiority and increase in cardiac output of A-V sequential pacing over ventricular pacing [<xref ref-type="bibr" rid="R12">12</xref>]. Other studies have shown that A-V and V-V synchrony are independent contributors to the hemodynamic ramifications of right ventricular pacing [<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R14">14</xref>]. Right ventricular pacing with or without A-V synchrony induces a physiologic contraction similar to that caused by left bundle branch block (LBBB). The effects of LBBB have been well studied. LBBB leads to an asynchronous ventricular contraction leading to altered diastolic filling time, increase in mitral regurgitation, as well as a reduction in left ventricular ejection fraction (LVEF) [<xref ref-type="bibr" rid="R15">15</xref>-<xref ref-type="bibr" rid="R17">17</xref>]. Thus, the reduction in cardiac output and symptoms associated with pacemaker syndrome are likely secondary to the loss of both A-V and V-V synchrony that is associated with right ventricular apical pacing [<xref ref-type="bibr" rid="R18">18</xref>].</p></sec><sec sec-type="" id="s5"><title> “Physiologic” vs. Ventricular pacing Trials</title><p>The physiologic benefits of A-V sequential have caused DDDR pacing to become common practice in most patients with sinus node dysfunction (SND). The results of completed randomized clinical trials of pacemaker mode selection have been somewhat conflicting. Overall, most trials have not shown reduction in heart failure, reduction in mortality, or improvement in quality of life with A-V sequential pacing (<xref ref-type="table" rid="T1">Table 1</xref>).</p><p>In the first of such trials, Anderson et al. compared single-chamber atrial (AAI) with VVI pacing in 225 patients with SND, normal A-V conduction, and a narrow QRS who had standard pacing indications [<xref ref-type="bibr" rid="R7">7</xref>]. The primary endpoints were frequency of atrial fibrillation (AF) and thromboembolic events. The original study follow-up was 3.3 years with a subsequent analysis at 5.5 years. Long term follow-up demonstrated persistent reduction in the primary endpoints of AF, thromboembolic events, chronic AF, and all cause mortality in the AAI paced group. At longer follow-up, the VVI group had an increased incidence of heart failure, worsening echocardiographic measurements of LV function, and increase in all-cause mortality. These long-term results have not been reproduced by any other prospective, randomized trial comparing solely atrial-based to ventricular-based pacing, although the pacing mode in subsequent trials was mainly DDDR.</p><p>In the PASE study, 407 patients older than 65 years of age, in sinus rhythm, who required a pacemaker for bradycardia, were randomized to VVIR or DDDR pacing [<xref ref-type="bibr" rid="R9">9</xref>]. The primary end point was quality-of life. Secondary endpoints included death from all causes, first nonfatal stroke or death, first hospitalization for heart failure, development of atrial fibrillation, and the development of pacemaker syndrome. At quality of life evaluations at 3, 9 and 18 months, there were no differences between the two pacing modes. There was a 26% crossover rate due to the development of pacemaker syndrome. No differences in clinical outcomes were observed. There were some trends that showed benefit in clinical outcomes in the DDDR group, especially in the subset of patients with SND, but none were of statistical significance.</p><p>As a result of the conflicting data in the Anderson et al. trial and the PASE study, two larger trials in pacemaker mode selection were performed. In the Canadian Trial of Physiologic Pacing (CTOPP) trial, 2568 patients with symptomatic bradycardia requiring pacing were randomized to atrial-based (AAI, AAIR, DDD, or DDDR) or ventricular pacing (VVI or VVIR) [<xref ref-type="bibr" rid="R10">10</xref>]. AAI and AAIR combined for only 5% of the patients assigned to the atrial-based group. The remaining 95% in the atrial-based group were either DDD or DDDR paced. The combined endpoint of stroke or death due to cardiovascular causes after a 3-year follow-up was not different between the two groups. The secondary endpoints of AF and chronic AF were observed less commonly in the atrial-based group. After an eight year follow-up, there were no significant differences in death or stroke between the two groups [<xref ref-type="bibr" rid="R19">19</xref>].</p><p>In the Mode Selection Trial in Sinus Node Dysfunction (MOST) study, 2020 patients with SND received dual chamber pacemakers and were randomized to either VVIR or DDDR pacing [<xref ref-type="bibr" rid="R11">11</xref>]. The mean follow-up was 33 months and the primary endpoints were death and nonfatal stroke. There were no differences in the primary endpoint between the two groups. The incidence of atrial fibrillation was lower with dual-chamber pacing. Subsequent analyses of the MOST data, has shown that the benefits of A-V sequential pacing are likely attenuated by the ventricular pacing that occurs in the DDDR mode [<xref ref-type="bibr" rid="R20">20</xref>]. Patients with a pre-paced QRS duration less than 120 ms were analyzed for percentage of ventricular pacing and clinical events. The percentage of ventricular pacing was determined from stored pacemaker data. The percentage of ventricular pacing was greater in the DDDR versus VVIR mode (90% vs. 58%). The percentage of ventricular pacing was a strong predictor of heart failure hospitalization in both pacing modes. The risk of AF increased linearly with percentage of ventricular pacing in both groups. The authors concluded that ventricular desynchronization imposed by ventricular pacing even when A-V synchrony is preserved increases the risk of heart failure hospitalization and AF in SND with normal baseline QRS duration. The analyses by Sweeney et al. illustrate the fact that DDDR and DDD pacing is not physiologic. The pacing mode that is most physiologic is AAI or AAIR. This fact can likely explain the discrepancies between the study by Anderson et al. and the subsequent studies in pacemaker mode selection. The benefit derived from A-V sequential pacing in the CTTOP and MOST studies was likely counterbalanced by the detrimental effects of right ventricular pacing seen in the DDDR mode.</p><p>Several ongoing trials are looking at the benefits of reduction in RV pacing in the DDDR mode. The first is the Danish Multicenter Randomized Study of Atrial Inhibited Versus Dual-Chamber Pacing in Sick Sinus Syndrome (DANPACE). This trial is comparing AAI pacing and DDDR pacing with a short A-V delay. The second is the Search A-V Extension for Promoting Atrioventricular Conduction (SAVE-PACE) study. This study will evaluate the use of search hysteresis, a pacemaker feature that allows the extension of the A-V delay to reduce right ventricular pacing. The study investigators hypothesize that the reduction in RV pacing will lead to a reduction in left ventricular remodeling and AF.</p></sec><sec sec-type="" id="s6"><title> Conclusion</title><p>Pacemaker syndrome is a common problem faced by clinicians who implant pacemakers and for those who take care of these patients. The syndrome is likely caused by both the loss of A-V and V-V synchrony imposed by right ventricular pacing. The studies on pacemaker mode selection have shown a high incidence of pacemaker syndrome. Because of the high incidence of pacemaker syndrome in VVIR-paced patients, atrial-based pacing is preferred.</p><p>These studies have also showed that DDDR pacing as compared to VVIR pacing decreases the incidence of AF, but does not affect stroke or mortality. The percentage of beats ventricular paced in these studies in patients who received dual-chamber pacemakers is likely to influence the incidence of pacemaker syndrome and heart failure. The amount of ventricular pacing in the DDDR mode is dependent on spontaneous A-V conduction and programmed A-V delay. Optimal programming should seek to reduce ventricular pacing. The reduction in the percentage of RV pacing in dual-chamber pacing modes will likely reduce the incidence of pacemaker syndrome, heart failure, and possibly mortality.</p></sec>
A Rare, Late Complication after Automated Implantable Cardioverter-Defibrillator Placement	<p>This article describes an interesting case of automated implantable cardioverter defibrillator (AICD) extrusion fifteen months after implantation. The case report is followed by a discussion of the causes and treatment of skin erosion following pacemaker/AICD insertion.</p>	<contrib contrib-type="author"><name><surname>Shapiro</surname><given-names>Michael</given-names></name><degrees>DO</degrees></contrib><contrib contrib-type="author"><name><surname>Hanon</surname><given-names>Sam</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Schweitzer</surname><given-names>Paul</given-names></name><degrees>MD</degrees></contrib><aff>Department of Cardiology, Beth Israel Medical Center, University Hospital and Manhattan Campus for the Albert Einstein College of Medicine, New York, USA</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Case report</title><p> The patient is a 65 year old man with a history of chronic renal insufficiency, dyslipidemia, hypertension, peripheral arterial disease, and coronary artery disease status post myocardial infarction in 1994. He was revascularized with 3-vessel coronary artery bypass grafting on 3/12/02 after he developed unstable angina. His post-operative course was complicated by a deep venous thrombosis. Treatment was initiated with warfarin and the patient was subsequently discharged from the hospital. He was readmitted three weeks later with mild shortness of breath. A ventilation/perfusion scan was negative for pulmonary embolism. However, while on telemetry, the patient had an episode of non-sustained ventricular tachycardia. He underwent diagnostic electrophysiologic testing at which time he was inducible for ventricular tachycardia. On 4/2/02, an AICD was implanted. He was discharged without incident and did well for over a year without any arrhythmic event requiring ICD shock. One week prior to the current admission the patient noted a skin tear over the AICD with protrusion of the device to the anterior chest wall. He visited his cardiologist to evaluate the extrusion of the device at which time he was immediately referred for hospital admission.</p><p>The patient admitted to minimal skin tenderness over the site but denied bloody or purulent drainage, fevers, chills, or malaise. His medications included metoprolol XL 50 mg daily, aspirin 81 mg daily, clopidogrel 75 mg daily, atorvastatin 10 mg daily, isosorbide mononitrate 20 mg daily, and gabapentin 300 mg daily. Social history, family history, and review of systems were non-contributory.</p><p>On physical examination, the patient had a temperature of 97.8° F, blood pressure 122/78 mm Hg, pulse 86 beats per minute and regular, and respiratory rate of 20 breaths per minute. He was a well developed, well nourished male sitting comfortably in no distress. Examination of his neck revealed no JVD at 30°, normal carotid upstrokes without bruits, and no thyromegaly. His chest was clear and heart examination revealed a non-displaced point of maximal impulse with a normal apical impulse, normal S1 and S2 without an S3 or S4. His pulse was regular with an apical 2/6 holosystolic murmur with radiation to the axilla. His anterior chest wall revealed a well healed sternotomy scar with erythema in the left supraclavicular area at the point where his AICD was extruding from his skin without drainage (see <xref ref-type="fig" rid="F1">Figure 1</xref> and <xref ref-type="fig" rid="F2">Figure 2</xref>). His abdomen was obese but soft without hepatomegaly. Neurologically, he was grossly intact. His lower extremities revealed no edema and distal pulses were intact. His initial laboratory results were normal including metabolic profile, complete blood count with differential, and coagulation profile. His chest radiograph revealed clear lungs, normal heart size, and a dual chamber pacemaker/defibrillator with leads in proper position. His electrocardiogram (ECG) revealed a normal sinus rhythm with right bundle branch block, left posterior fascicular block, and an old inferior infarct. This ECG was unchanged from prior tracings. The patient was admitted to the hospital and three sets of blood cultures were obtained. He was subsequently started on vancomycin 1 g intravenously q 12h. The following day his device was explanted and the pocket debrided. Blood cultures as well as cultures from the device and surrounding skin and subcutaneous tissue revealed no growth. He was continued on antibiotics and discharged from the hospital without complications. The discharge plan was to replace the AICD at a future date, upon healing of the operative site.</p></sec><sec sec-type="" id="s2"><title>Discussion</title><p>This case illustrates an unusually severe form of pacemaker/AICD skin erosion. This rare complication has been previously described in the literature. However, prior to the study performed by Kiviniemi et al [<xref ref-type="bibr" rid="R1">1</xref>], most of the existing data was based on information from the 1970s [<xref ref-type="bibr" rid="R2">2</xref>]. This more recent investigation retrospectively analyzed four hundred forty-six patients who received permanent pacemakers and reported the complications noted during implantation or follow-up. Pacemaker erosion was detected in 0.9% of the patients.</p><p>Two main culprits have been implicated in the pathogenesis of pacemaker extrusion. First, infection of the site can lead to skin erosion [<xref ref-type="bibr" rid="R3">3</xref>]. Obviously, prevention of and monitoring for infection is paramount to averting this process. DaCosta et al. performed a meta-analysis of antibiotic prophylaxis for pacemaker implantation [<xref ref-type="bibr" rid="R4">4</xref>]. They concluded that antibiotics administered during the peri-implant period reduced the incidence of infective complications following pacemaker implantation, including short-term pocket infection, skin erosion, or septicemia.</p><p>The other cause of skin erosion is pressure necrosis of the overlying tissue and skin [<xref ref-type="bibr" rid="R3">3</xref>]. Gross pacemaker extrusion is usually signaled by a preceding period of “ pre-erosion,” during which there is discomfort and discoloration of thinning tissue tensely stretched over a protrusion of the pacing apparatus ( <xref ref-type="fig" rid="F2">Figure 2</xref>). Griffith et al. concluded that if pacemaker erosion is not caused by infection it can be successfully managed by ipsilateral re-implantation, a financially advantageous solution [<xref ref-type="bibr" rid="R5">5</xref>]. Identification of pre-erosion allows salvage of the pacing system, as the hardware can be repositioned under the pectoralis muscle or in an abdominal location.</p><p>Risk factors for skin erosion include factors related to the device itself as well as to the implantation site. The mass and configuration of the pacemaker as well as the need for extra hardware (e.g., lead adaptor) in the pocket may lead to local tension and pressure necrosis. Additionally, precise surgical construction of the pacemaker pocket is vital. A pocket of inadequate size or a paucity of subcutaneous tissue may contribute to local complications. The pocket plane should be created on the surface of the muscle since superficial pockets lend themselves to erosion [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>If true erosion occurs, the system is considered contaminated and current opinion favors removal of the generator and leads. Extensive debridement of the pocket and prolonged irrigation and antibiotic therapy may provide an alternate option to removal in cases of both erosion and frank infection [<xref ref-type="bibr" rid="R7">7</xref>], but this approach is not generally preferred. Attempts have been made to preserve the implanted leads by debriding locally and severing the pacemaker leads while leaving them in situ. However, given the relative safety and efficacy of percutaneous lead removal, complete removal of the leads is the ideal approach if sterility is questioned.</p></sec>
Radiofrequency Catheter Ablation of Atrioventricular Nodal Reentrant Tachycardia: It Is Not Always As It Is Expected	<p>Observation of Coincident arrhythmias is not uncommon but the co-existence of idiopathic verapamil sensitive left ventricular tachycardia (ILVT) with other arrhythmias is very rare. We hereby presented a 30 year old male patient with a history of frequent episodes of palpitations and sustained narrow complex tachycardia. During electrophysiologic study two arrhythmias, one with narrow complexes which was shown to be typical atrioventricular nodal re-entrant tachycardia and the other with wide QRS complexes and right bundle branch block and left axis morphology, compatible with ILVT, were inducible. Radiofrequency catheter ablation of both arrhythmias was done at two consecutive sessions. The patient has remained asymptomatic without antiarrhythmic therapy for the past six months.</p>	<contrib contrib-type="author"><name><surname>Arya</surname><given-names>Arash</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Haghjoo</surname><given-names>Majid</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Zahra</surname><given-names>Emkanjoo</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Heydari</surname><given-names>Alireza</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Ali Sadr-Ameli</surname><given-names>Mohammad</given-names></name><degrees>MD</degrees></contrib><aff>Department of Pacemaker and Electrophysiology, Rajaie Cardiovascular Medical Center, Mellat Park, Vali-Asr Avenue, Tehran 1996911151, Iran</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Idiopathic sustained ventricular tachycardia (VT) accounts for 10-20% of patients with sustained monomorphic VT [<xref ref-type="bibr" rid="R1">1</xref>]. More than a half of idiopathic VTs originates from right ventricle especially from the right ventricular outflow tract (RVOT) and the remainder from the left ventricle (LV). Idiopathic verapamil-sensitive LV tachycardia (ILVT) is the most common type of idiopathic LV tachycardia [<xref ref-type="bibr" rid="R1">1</xref>]. Coincidence of RVOT tachycardia with atrioventricular nodal reentrant tachycardia (AVNRT) has been described [<xref ref-type="bibr" rid="R3">3</xref>], but coexistence of ILVT with other arrhythmias is much less common and rarely reported [<xref ref-type="bibr" rid="R4">4</xref>]. This case report describes the coincidence of AVNRT and ILVT in a 30 year old patient.</p></sec><sec sec-type="" id="s2"><title>Case Report</title><p>We hereby presented a 30 year old male patient with three years history of frequent episodes of palpitation. He had no structural heart disease. His previous medical records revealed several ECGs during arrhythmia all showing regular narrow complex tachycardia with a rate of 180beat/min. There was no documented wide complex tachycardia. During initial EP study a narrow complex tachycardia was reproducibly induced (<xref ref-type="fig" rid="F1">Figure 1A</xref>) which was shown to be typical AVNRT (<xref ref-type="fig" rid="F2">Figure 2A</xref>). Another regular arrhythmia with RBBB and left axis morphology (cycle length=330 ms) and negative HV (-20 ms), compatible with ILVT was also induced reproducibly by ventricular programmed electrical stimulation (<xref ref-type="fig" rid="F1">Figure 1B</xref>and<xref ref-type="fig" rid="F2">Figure 2B</xref>). RF catheter ablation of AVNRT was done successfully. After ablation, the ILVT was still inducible. The patient received verapamil and remained asymptomatic.</p><p>As the ablation procedure was prolonged the ablation of ILVT postponed to another session. After one month the patient re-admitted for ablation of ILVT, three days after discontinuation of verapamil. Three quadripolar catheters (6F, Josephson, Bard Electrophysiology) were introduced via left femoral vein and positioned at RV apex (and right ventricular outflow tract), His bundle and high right atrium. A 7F steerable decapolar catheter (2-5-2 mm) was introduced via left femoral artery (Marinr, Medtronic Inc, USA) and positioned in LV on interventricular septum for recording intracardiac signals during sinus rhythm and tachycardia. This catheter was used as a guide for localization of earliest purkinje potential (PP) recording site by ablation catheter. A 7F ablation catheter (Conductr, Medtronic Inc, USA) was introduced via right femoral artery and positioned in LV on septum for mapping and subsequent RF ablation. The ILVT was terminated by mechanical pressure at <italic>earliest</italic> PP recording site (<xref ref-type="fig" rid="F2">Figure 2C</xref>). RF energy was applied (50 W, 70°C, 60 s) from distal electrode of ablation catheter at this site during sinus rhythm. Repeated programmed electrical stimulation before and after isoprotrenol infusion (4μ/min) from atrium and ventricle failed to induce any arrhythmia. The patient remained asymptomatic without antiarrhythmics during 6 months of follow up.</p></sec><sec sec-type="" id="s3"><title>Discussion</title><p>This case describes the coincidence of AVNRT and ILVT. Occurrence of AVNRT in combination with idiopathic VT has been previously described. In patients with RVOT tachycardia the incidence of AVNRT has been reported to be as high as 15% [<xref ref-type="bibr" rid="R3">3</xref>]. Our finding also shows the possibility of such a combination in patients with ILVT. This underscores the importance of searching for dual atrioventricular node physiology and inducible AVNRT in all patients with idiopathic VT including those with ILVT.</p><p>Wagshal AB, et al. [<xref ref-type="bibr" rid="R4">4</xref>] also reported a case with ILVT and AVNRT. That report had several differences with ours. Their patient presented with ILVT, and AVNRT was discovered accidentally during EP study. Ablation of slow pathway in their case resulted in cure of AVNRT and prevention of spontaneous episodes of ILVT. They suggested that their case represented an example of tachycardia induced tachycardia and hence ablation of AVNRT resulted in cure of both arrhythmias. We can not exclude such a mechanism in their case. In our case although ILVT has never been documented before EP study, the patient remained asymptomatic after ablation of AVNRT only following administration of verapamil. In addition we were not able to exclude ILVT as arrhythmia responsible for some of his palpitation episodes. For this reasons we decided to ablate ILVT rather continuing verapamil. Also, Kautzner J, et al [<xref ref-type="bibr" rid="R5">5</xref>] in a recent study on simultaneous idiopathic VT and AVNRT, has shown that ablation of AVNRT dose not influence the induciblility of the idiopathic VT. Finally, different methods have been proposed for ablation of ILVT [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R9">9</xref>]. We used the earliest PP during ILVT as a guide for RF ablation. As Lokhandwala and his colleagues [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R11">11</xref>] has shown that PP by themselves are non-specific and are seen in normal people during sinus rhythm, if PP wanted to be used as a guide for RF ablation of ILVT, the <italic>earliest</italic> PP <italic>during</italic> tachycardia should be targeted.</p></sec><sec sec-type="" id="s4"><title>Conclusion</title><p>This case represents the co-incidence of ILVT and AVNRT and underscores the importance of searching for dual AV node physiology and AVNRT in patients with ILVT.</p></sec>
Issues in QT interval measurement	<p>The QT interval, apart from clinical implications is crucial for safety assessment of new drugs under development. A QTc prolongation of even 10 msec in a study group is a warning signal for a new drug.</p><p>There are various issues involved in the measurement of the QT interval especially regarding the ending of the T wave and different morphological pattern of T-U complex. The other issue is significant spontaneous variability in the QT interval, resulting in spurious QT prolongation and unnecessary concern.</p><p>To minimize all these confounding factors, all clinical trials for assessing QT interval prolongation should be randomized and double blinded with appropriate control groups including placebo. ECG measurements should be done by trained readers with electronic calipers at ECG core Lab. ECGs should be compared with multiple baseline values with multiple, time-matched on-treatment values.</p>	<contrib contrib-type="author"><name><surname>Lanjewar</surname><given-names>Pallavi</given-names></name><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Pathak</surname><given-names>Vaishali</given-names></name><xref ref-type="aff" rid="aff1"></xref></contrib><contrib contrib-type="author"><name><surname>Lokhandwala</surname><given-names>Yash</given-names></name><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<p> The QT interval represents the time required for completion of both ventricular depolarization and repolarisation and has been a parameter of particular interest [<xref ref-type="bibr" rid="R1">1</xref>]. Certain drugs have the ability to delay cardiac repolarization manifested as QT interval prolongation. This creates an electrophysiological environment that favors the development of cardiac arrhythmias, most important of which is torsades de pointes, but possibly other ventricular arrhythmias as well. This can result in sudden cardiac death [<xref ref-type="bibr" rid="R2">2</xref>]</p><p>Recently it has become evident that a variety of drugs induce QT interval prolongation. Besides the many antiarrhythmic drugs that are well known to widen the QT interval as their main pharmacological effects, non- antiarrhythmic drugs such as antipsychotic, antiallergic, antibiotics and gastroenterologic agents have been shown to induce QT prolongation as a side effect. A careful assessment of QT interval prolongation is often required before novel drugs are approved by regulatory authorities. For accurate evaluation of the QT interval, the fact that the QT interval varies in RR interval dependent manner must be taken into account. Clinically, several QT-correcting formulas, such as Bazett’s, Fredericia’s, etc. have been used to evaluate QT prolongation considering changes in the length of the RR interval [<xref ref-type="bibr" rid="R3">3</xref>].</p><p> The QT interval and corrected QTc are defined below: QT interval represents the duration of ventricular depolarization and subsequent repolarization, beginning at the initiation of the Q wave of the QRS complex and ending where the T wave returns to the isoelectric baseline.</p><sec sec-type="" id="s1"><title>Choice of Lead</title><p>By convention Lead II is used for QT measurements. An alternative lead is used only when lead II was not adequate.</p></sec><sec sec-type="" id="s2"><title>Measurement of QT interval</title><p> Often the end of the T wave is not clear. In such cases the end of the T-wave can be extrapolated by using the Tangent method. The start of the QT interval is defined as the first deflection of the QRS complex. The end of the QT interval is defined as the intersection of the descending part of the T-wave (positive T-wave) with the isoelectric line.</p><p> If a U-wave appears immediately after the T-wave has returned to the baseline, the QT interval is measured as the nadir between T and U-waves. If it is not clear whether a second hump occurs before the T wave has returned to baseline, it is included in the QT interval. (<xref ref-type="fig" rid="F1">Figure 1</xref>)</p><p>QT interval for research purpose is measured by vernier caliper, digitizing pad on the paper ECG as well as by on-screen calipers in digital ECGs.</p></sec><sec sec-type="" id="s3"><title>Holter vs ECGs</title><p>A continuous 12-lead Holter can be used as an alternative method for the standard digital 12-lead ECGs because of following reasons i) It provides the continuous ambulatory recording for 24 hours, thereby enabling retrieval of the ECGs based on pharmacokinetic data and (ii)It provides useful data for cardiac arrthymias by continuous beat to beat analysis [<xref ref-type="bibr" rid="R4">4</xref>]. However,12-lead Holter has many potential problems such as postural variation leading to change in ECG morphology [<xref ref-type="bibr" rid="R5">5</xref>-<xref ref-type="bibr" rid="R7">7</xref>] and multiple artifacts due to motion [<xref ref-type="bibr" rid="R8">8</xref>]. Due to the change in position of the limb electrodes in the Holter recording, there is significant change in the QT interval measurement as it is measured in Lead II and Lead II comprises of limb lead electrodes.</p><p>A study by Sarapa and colleagues [<xref ref-type="bibr" rid="R10">10</xref>] reviewed the utility of Holter-derived 12-lead ECG and standard digital 12-lead ECG in detection of sotalol-induced QT prolongation in healthy volunteers. They reported a difference between standard digital ECG and Holter-derived 12 lead ECG of 5.2 ± 19 ms, They also reported a difference of -5 ± 70 ms in measurement of the RR interval. The reason for this difference is apparently related to the technique of recording or measurement at the ECG core laboratory.</p></sec><sec sec-type="" id="s4"><title> Issues (fallacies of QT measurement)</title><sec id="s4a"><title>Wide QRS</title><p>The QT is less reliable when the QRS duration (QRSd) is ≥ 120 ms, as the increased QRSd which measures the duration of ventricular depolarization contributes to prolongation of the QT interval. The JT interval (defined as QT- QRSd) has been proposed as a more valid way to assess ventricular repolarization in such circumstances.</p><p>Zhou SH et al [<xref ref-type="bibr" rid="R11">11</xref>] examined the influence of QRS duration on the JT and QT intervals in 20,687 normal adult subjects and 2,865 subjects with various categories of intraventricular conduction delay (IVCD). Estimates for coefficients for multiple regression of QRS duration on QT and JT intervals combined with a correction term for heart rate (HR) were determined for each VCD category. QRS duration accounted for about 16% of total QT variation, but had a practically negligible effect on JT interval in complete bundle branch blocks. A single-parameter formula was derived for the JT prolongation index of the form JTI = JT (HR + 100)/518, with a JTI > or = 112 identifying repolarization prolongation in all VCD categories. They concluded that it is preferable to use JT rather than QT as a more appropriate index of duration of repolarization in IVCD.</p><p>Das G [<xref ref-type="bibr" rid="R12">12</xref>] evaluated the relative contribution of the depolarization and the repolarization time prolongation to the prolonged QT interval in patients with (IVCD). The QRS, QT, and JT intervals were measured in 72 subjects with various types of IVCD. The observed intervals in IVCD subjects were compared to similar intervals in 33 healthy individuals in whom there was no evidence for intraventricular conduction abnormalities. The QTc (QT interval corrected for heart rate) in subjects with IVCD were 445 ± 6.8 msec (mean ± SEM) in those with LAD, 470 ± 9.1 msec with RBBB, and 489 ± 6.9 msec with LBBB. All of these intervals were significantly prolonged compared to 430 ± 4.3 msec in the control group. The prolongation of QTc interval in each category of IVCD subjects was entirely secondary to a prolonged depolarization time, as the JT intervals were not significantly different from those observed in the control group (F = 0.5, p = NS). These observations may provide an explanation for the differential prognosis for subjects with prolonged QT interval with prolonged repolarization time as compared to those with prolonged QT interval with prolonged depolarization time.</p><p>Prolonged heart rate-corrected QT interval is associated with higher risk of mortality in patients with coronary heart disease (CHD). Crow RS et al [<xref ref-type="bibr" rid="R13">13</xref>] assessed JT interval in these patients with CHD and QRSd ≥ 120 ms. and concluded that the JTc is a simple measurement that is a significant independent predictor of CHD events in men with wide QRS complex.</p></sec><sec id="s4b"><title>Gender: </title><p>Bazett’s formula (QTcB) has been more frequently used in the medical literature than Fridericia’s formula, so that most reported criteria for normal and abnormal values are derived from Bazett’s formula (<xref ref-type="table" rid="T1">Table 1</xref>) [<xref ref-type="bibr" rid="R14">14</xref>]</p></sec><sec id="s4c"><title>Magnitude of QT variability</title><p>Significant spontaneous variability is seen in the QT interval, and this may result in spurious QT prolongation and unnecessary concern about cardiac safety. To avoid this spurious variability, data on QT, QTc intervals should always be presented both as analyses of central tendency (means, medians, ranges, etc.) and categorical analyses (proportion of individual subjects in each treatment group experiencing specified degrees of abnormality i.e. outlier analyses).</p></sec></sec><sec sec-type="" id="s5"><title>Analysis of central tendency</title><p>For analyses of central tendency, the emphasis should be on following factors</p><sec id="s5a"><title> Maximum change in the QT, QTc Intervals</title><p>The maximum observed difference between on-treatment and baseline QT, QTc values should be expressed both as mean and median changes in the population. This value is meaningful only as a comparison with placebo or a non-QT prolonging drug, as selection of the highest of many on- treatment values will invariably show an increase from baseline.</p></sec><sec id="s5b"><title>Time matched QT, QTc Intervals</title><p>Mean changes from baseline in the observed QT, QTc interval can be presented as time-matched control and treatment group values. Time matched ECGs counter the circadian variation in the QT interval. Although these values may show regression to the mean, they do not have the same upward bias as the maximum change.</p></sec></sec><sec sec-type="" id="s6"><title>Categorical analyses</title><p>As the absence of statistically or clinically significant differences between the test drug and comparator groups does not exclude the possibility of marked QT,QTc interval prolongation occurring in individual subjects, analyses of central tendency should always be accompanied by appropriate categorical analyses.</p><p>Categorical analyses of QT, QTc interval data are based on the number and percentage of patients meeting or exceeding some predefined upper limit value. Clinically noteworthy signals may be defined in terms of absolute (readings in excess of some specified threshold value) QT, QTc intervals or changes from baseline control Although increases from baseline in the QT, QTc interval constitute signals of interest, interpretation of these differences is complicated by the potential for changes not related to drug therapy, including regression toward the mean and choice of extreme values. Regression toward the mean refers to the tendency of subjects with high baseline values to have lower values at later time points, while subjects with low baseline values tend to experience increases. The direction of regression depends on initial selection criteria; e.g., if subjects with high baseline QT, QTc interval values are excluded from the trial, values recorded during treatment will tend to rise relative to baseline levels.</p><p>The process of choosing the highest of multiple observed values will also invariably cause an apparent change from any single baseline value, a phenomenon found in both drug and placebo-treated groups.</p><p>To minimize the effect of spontaneous variability, we put forward the following recommendations: 1) To use trained readers2) To take on-screen measurements in an ECG core Lab 3) To compare with the results in the appropriate control group(s), including placebo or a drug with no QT, QTc prolongation effect 4) To compare with multiple baseline values with multiple, time-matched on-treatment values (not just the greatest value).</p></sec>
Electrical Storms in Brugada Syndrome: Review of Pharmacologic and Ablative Therapeutic Options	<p>Electrical storm occurring in a patient with the Brugada syndrome is an exceptional but malignant and potentially lethal event. Efficient therapeutic solutions should be known and urgently applied because of the inability of usual antiarrhythmic means in preventing multiple recurrences of ventricular arrhythmias. Isoproterenol should be immediately infused while oral quinidine should be further administrated when isoproterenol is not effective. In case of failure of these therapeutic options, ablation of the triggering ventricular ectopies should be attempted.</p>	<contrib contrib-type="author"><name><surname>P</surname><given-names>Maury</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>M</surname><given-names>Hocini</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>M</surname><given-names>Haïssaguerre</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<p>The Brugada syndrome is an inherited cardiac electrical disorder occurring in the absence of obvious structural heart disease, initially defined by the association of right bundle branch block, ST elevation in right precordial leads and sudden cardiac death related to polymorphic ventricular tachycardias [<xref ref-type="bibr" rid="R1">1</xref>].</p><p>In Brugada syndrome, the prevalence of malignant ventricular arrhythmias varies from 5 % at two years follow-up in asymptomatic patients to 45 % in patients with previous cardiac arrest [<xref ref-type="bibr" rid="R2">2</xref>]. While in most patients, a single arrhythmic event occurs, some patients will suffer electrical storms. Although there is no commonly accepted definition for this event, electrical storm is generally perceived as the rapid or incessant succession of recurrent poorly tolerated ventricular arrhythmias, generally requiring repeated cardioversions, occurring during a short period of time [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R4">4</xref>] (<xref ref-type="fig" rid="F1">Figure 1</xref>). Electrical storm is fortunately a very scarce phenomenon, but which can occur twice or more in the same patient [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R6">6</xref>], or can be the first manifestation of the disease [<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R8">8</xref>].</p><p>True prevalence of electrical storm in Brugada syndrome has not been evaluated, such events having only been seldom reported in isolated case reports. To our knowledge, to date, only twelve such cases have been published [<xref ref-type="bibr" rid="R5">5</xref>-<xref ref-type="bibr" rid="R16">16</xref>]..</p><p>Although exceptional, electrical storm in Brugada syndrome is a major event with dramatic consequences [<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R10">10</xref>], leading sometimes to a fast ineluctable arrhythmic death in otherwise healthy young people despite all usual antiarrhythmic interventions [<xref ref-type="bibr" rid="R8">8</xref>]. Heart transplantation has even been once performed for such an intractable electrical storm [<xref ref-type="bibr" rid="R5">5</xref>]. That’s why some particular crucial and efficient therapeutic considerations should be known and urgently applied in order to avoid a fatal outcome.</p><sec sec-type="" id="s1"><title>Pharmacological Management</title><p>Causal factors should be searched and immediately corrected, such as fever [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R15">15</xref>,<xref ref-type="bibr" rid="R16">16</xref>], increased vagal tone during gastro-enteritis [<xref ref-type="bibr" rid="R6">6</xref>], low potassium blood levels [<xref ref-type="bibr" rid="R17">17</xref>], non febrile bronchitis [<xref ref-type="bibr" rid="R6">6</xref>] or infusion of class 1 drugs for diagnostic purposes [<xref ref-type="bibr" rid="R14">14</xref>,<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>].</p><p>Apart from class 1A agents (see further) no antiarrhythmic drug has been shown to be effective in preventing recurrence of arrhythmia in Brugada syndrome [<xref ref-type="bibr" rid="R20">20</xref>] [<xref ref-type="bibr" rid="R21">21</xref>], then antiarrhythmic drugs should be better avoided in case of electrical storm. Beta-blocker, bretylate, lidocaine, mexiletine or magnesium has been tried without any success [<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R16">16</xref>] or can even worsen the situation [<xref ref-type="bibr" rid="R22">22</xref>]. In the DEBUT study, there was an 18 % death rate in Thai patients survivors of Sudden Unexpected Death Syndrome which were treated with beta-blocker [<xref ref-type="bibr" rid="R23">23</xref>]. If amiodarone infusion has been once apparently successful (but concommitantly with a beta-adrenergic agonist, see further) [<xref ref-type="bibr" rid="R7">7</xref>], it did not seem to be beneficial in other cases [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R16">16</xref>]. Sotalol, a drug devoid of class 1 effect, seems to have been successful in one case [<xref ref-type="bibr" rid="R7">7</xref>], however, one should remind that class 3 drugs also can increase ST elevation [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R24">24</xref>] that may be deleterious.</p><p>Isoproterenol infusion at a sufficiant dosing, ranging usually from 0.1 to 1-3 μg/mn [<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>], should be immediately started. Although isoproterenol is generally proarrhythmic in most other conditions, it is an effective drug in electrical storms related to Brugada syndrome and prevents recurrences of ventricular fibrillation (VF) [<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R9">9</xref>-<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R21">21</xref>].</p><p>Both Brugada electrocardiographic pattern and ventricular excitability or vulnerability are believed to be dependent on the sympathetic imbalance and on cardiac rate. Worsening of the electrophysiological conditions have been described when vagal tonus increases and/or heart rate decreases [<xref ref-type="bibr" rid="R22">22</xref>,<xref ref-type="bibr" rid="R25">25</xref>]. Major arrhythmic events and sudden death are known to frequently occur at night, when the vagal tone is predominant [<xref ref-type="bibr" rid="R21">21</xref>]. In patients with Brugada syndrome, isoproterenol infusion normalizes the electrocardiographic pattern and avoids ventricular fibrillation induction during electrophysiological study [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R21">21</xref>].</p><p>Beta-adrenergic stimulation with isoproterenol increases I<sub>CaL</sub> and restores the dome of epicardial action potentials, reducing the degree of local and transmural heterogeneicity [<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R26">26</xref>]. This can be sufficient to decrease the degree of ST elevation [<xref ref-type="bibr" rid="R22">22</xref>] and to avoid the genesis of premature beats. Beta-adrenergic stimulation should be optimally performed using isoproterenol, since failure of dobutamine infusion had been reported [<xref ref-type="bibr" rid="R8">8</xref>], although cilostazol - e.g. an oral phosphodiesterase inhibitor - has also been shown to be efficient in preventing recurrent VF in Brugada syndrome [<xref ref-type="bibr" rid="R27">27</xref>].</p><p>Accelerating the heart rate (decreasing I<sub>to</sub>) [<xref ref-type="bibr" rid="R28">28</xref>,<xref ref-type="bibr" rid="R29">29</xref>] or decreasing the vagal tonus (decreasing I<sub>KAch</sub> and increasing I<sub>CaL</sub>) [<xref ref-type="bibr" rid="R26">26</xref>,<xref ref-type="bibr" rid="R29">29</xref>] for example by atropine infusion [<xref ref-type="bibr" rid="R28">28</xref>], can also act, particularly when associated to beta agonist [<xref ref-type="bibr" rid="R10">10</xref>]. However atropine infusion alone has been sometimes tried without any beneficial result [<xref ref-type="bibr" rid="R6">6</xref>], while success [<xref ref-type="bibr" rid="R30">30</xref>] or failure [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R9">9</xref>] of fast pacing have been reported.</p><p>In some cases, VF is incessant despite major adrenergic stimulation due to the physical and emotional stress caused by the repeated shocks. The associated alpha-adrenergic stimulation, which is leading to ST elevation in Brugada syndrome [<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R29">29</xref>], is suspected to overwhelm the beneficial effects of beta-adrenergic stimulation in those cases [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>Class 1A drugs like quinidine are another means to escape from such a critical situation and should be tested in patients with electrical storm not immediately responding to isoproterenol. Oral quinidine has been successfully used in electrical storms in Brugada syndrome [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R20">20</xref>].</p><p>Quinidine is a class 1 agent and should theoritically worsen the situation because of its sodium channel blocking properties. However quinidine is also a blocker of the transient outward current I<sub>to</sub> [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R28">28</xref>,<xref ref-type="bibr" rid="R31">31</xref>]. Blocking I<sub>to</sub> counteracts the marked abbreviation of action potentials in epicardial cells and so normalizes ECG pattern and ventricular vulnerability. Experimentally, class 1A drugs restore the action potential dome, normalize ST segment elevation and prevent arrhythmogenesis by blocking I<sub>to</sub> [<xref ref-type="bibr" rid="R28">28</xref>]. Anticholinergic properties of class 1A agents would also contribute to this beneficial effect [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R28">28</xref>].</p><p>In 1987, Belhassen and coll. first document the ability of class 1A agents for prevention of inducibility in patients with idiopathic VF [<xref ref-type="bibr" rid="R20">20</xref>], which is believed to be caused in a large part by Brugada syndrome [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R21">21</xref>]. Long term beneficial action of quinidine has then been described in a population of idiopathic VF and Brugada syndrome [<xref ref-type="bibr" rid="R32">32</xref>]. In this study, lack of inducible arrhythmia under class 1A drugs (mainly quinidine) was possible in 80 % of patients with the Brugada syndrome and in all patients with idiopathic VF, and displayed good prognostic value since no recurrent arrhythmia could be documented after a mean follow-up of 9 years when patients were treated with class 1A agents (quinidine 1 to 2 g daily) [<xref ref-type="bibr" rid="R32">32</xref>]. In a recent publication of the same group, quinidine bisulfate at a mean dose of 1.5 g daily prevented inducibility in 88 % of patients with Brugada syndrome, without any recurrence of arrhythmia with a mean follow-up of 56 months [<xref ref-type="bibr" rid="R31">31</xref>]. Normalization of the ECG pattern and suppression of the ventricular premature beats as well as the induction of VF by quinidine (1 to 1.5 g daily) has also been described by other groups in isolated cases [<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R33">33</xref>].</p><p>Hydroquinidine chlorydrate (600 to 900 mg daily) has also been recently shown to to normalize the ECG pattern in 34 % of patients with Brugada syndrome, to prevent arrhythmia inducibility in 76 % of asymptomatic patients, to prevent the occurrence of arrhythmic events in 90 % of those patients in which arrhythmia was rendered non inducible, and to avoid any recurrence of arrhythmia in implanted patients presenting with repeated shocks [<xref ref-type="bibr" rid="R34">34</xref>].</p><p>The effect of others class 1 A drugs is more controversial. Disopyramide, another blocker of Ito, is believed to be potent but in a lesser extent [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R28">28</xref>]: disopyramide can sometimes increase ST elevation [<xref ref-type="bibr" rid="R29">29</xref>] and may be proarrhythmic or antiarrhythmic according to the associated conditions [<xref ref-type="bibr" rid="R20">20</xref>]; in fact it has been shown to prevent inducibility while exagerating the ECG pattern in one case [<xref ref-type="bibr" rid="R35">35</xref>] and was inefficient in another case [<xref ref-type="bibr" rid="R11">11</xref>]. Procainamide increases ST elevation [<xref ref-type="bibr" rid="R29">29</xref>] and is proposed for the drug challenge for diagnosis of Brugada syndrome [<xref ref-type="bibr" rid="R21">21</xref>]; it failed to prevent VF induction and to normalize the ECG pattern in one case of idiopathic VF with right bundle branch, considered as a variant form of Brugada syndrome [<xref ref-type="bibr" rid="R33">33</xref>]. Other agents which block Ito without significant block of I<sub>Na</sub>, such as tedisamil, a drug currently in clinical trials for atrial fibrillation, could also be useful in this situation [<xref ref-type="bibr" rid="R20">20</xref>].</p><p>The only side effect would be an excessive QT prolongation. Because of the blocking action of class1A drugs on repolarizing K currents, these agents could unmask an associated LQT syndrome, since some congenital LQT syndrome and Brugada syndrome are both linked to mutations of the sodium channel and can coexist in the same patient [<xref ref-type="bibr" rid="R36">36</xref>]. Indeed, excessive QT prolongation in this context has been reported [<xref ref-type="bibr" rid="R12">12</xref>], although not observed by others [<xref ref-type="bibr" rid="R17">17</xref>].</p><p>Finally, in case of incessant VF recurring despite all these therapeutic options, general anesthesia has been sometimes performed with good results5, although failure has been reported [<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R8">8</xref>].</p></sec><sec sec-type="" id="s2"><title>Non Pharmacological Management</title><p>Interventional therapy has been recently developed for treating electrical storm [<xref ref-type="bibr" rid="R37">37</xref>,<xref ref-type="bibr" rid="R38">38</xref>], and appears promising in such critical situations. Even if ventricular premature beats are rather infrequent in patients with Brugada syndrome [<xref ref-type="bibr" rid="R39">39</xref>], ICD-stored electrograms and ECG monitoring have shown that ventricular premature beats (VPB’s) precede spontaneous VF in the majority of cases and that they are identical to the one initiating the arrhythmias, which are always induced by the same premature beat in a given patient [<xref ref-type="bibr" rid="R40">40</xref>].</p><p>Whether they induced VF or not, VPB’s in Brugada syndrome usually originate from the right ventricular outflow tract (RVOT) [<xref ref-type="bibr" rid="R12">12</xref>,;<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R37">37</xref>] (<xref ref-type="fig" rid="F2">Fig. 2</xref>), although right ventricular ectopics with left axis have been described [<xref ref-type="bibr" rid="R17">32</xref>,<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R37">37</xref>,<xref ref-type="bibr" rid="R41">41</xref>]. Isolated cases of monomorphic ventricular tachycardia have been also reported, with left bundle block morphology with inferior [<xref ref-type="bibr" rid="R42">42</xref>,<xref ref-type="bibr" rid="R43">43</xref>] or superior axis [<xref ref-type="bibr" rid="R41">41</xref>,<xref ref-type="bibr" rid="R44">44</xref>,<xref ref-type="bibr" rid="R45">45</xref>] or even with right bundle block morphology [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R14">14</xref>,<xref ref-type="bibr" rid="R16">16</xref>], whose relationship with Brugada syndrome could be incidental or not, but which could also be incessant [<xref ref-type="bibr" rid="R14">14</xref>] and lead to fatal electrical storm [<xref ref-type="bibr" rid="R8">8</xref>]. Arrhythmogenic preponderance of the right ventricular outflow tract is not surprising due to the local higher electrical gradient, as otherwise ilustrated by the ST elevation which is usually exclusively present in right precordial leads.</p><p>Current observations suggest an important role for VPB’s of right ventricular origin in the Brugada syndrome. Chinushi et al. [<xref ref-type="bibr" rid="R46">46</xref>] described recurrent episodes of VF in a patient with Brugada syndrome initiated by monomorphic VPB’s with left bundle block morphology. This was corroborated by Morita et al. [<xref ref-type="bibr" rid="R47">47</xref>] who observed VPB’s in 9 out of 45 patients studied. Eleven VPB morphologies were observed in these 9 patients, of which 10 were of right ventricular origin (7 lateral RVOT, 2 septal RVOT and 1 from the apex).</p><p>While the cornerstone of the management of these conditions has been the implantation of a defibrillator, these reports and our recent success with the ablation of idiopathic VF [<xref ref-type="bibr" rid="R48">48</xref>] [<xref ref-type="bibr" rid="R49">49</xref>] has led us to evaluate the role of trigger elimination in patients with Brugada syndrome [<xref ref-type="bibr" rid="R37">37</xref>,<xref ref-type="bibr" rid="R50">50</xref>]. We have studied four patients with Brugada syndrome (3 males, age 36 ± 8 years). These patients presented with documented episodes of polymorphic ventricular tachycardia or VF (1 to 21 episodes), 3 with a family history of sudden death. They had 12 ± 9 episodes of VF or syncope prior to mapping. No drug therapy had been attempted in 3 patients with Brugada syndrome while quinidine failed in 1 patient. The Brugada syndrome was diagnosed by abnormal QRST complexes in leads V1 and V2 with a coved ST segment elevation in 4 patients, one who had a familial SCN5A channelopathy. No patient had evidence of structural heart disease based on physical examination, echocardiography and right/left ventricular ejection fraction. Exercise testing and coronary angiography excluded myocardial ischemia. Brugada syndrome had been diagnosed 9 months and 3 years prior to the clinical episodes of VF in 2 patients.</p><p>All patients were studied within 2 weeks of their arrhythmic storm and had been documented to have frequent VPB’s. The triggering role of VPB’s in the initiation of VF was observed by ambulatory monitoring or stored electrograms of the defibrillator. Premature beats in the Brugada syndrome were monomorphic in all, coming from the RVOT in three patients (left bundle branch-inferior axis morphology, coupling interval of 343 ± 59 ms) and with left bundle branch block-superior axis morphology in one (coupling interval 278 ± 29 ms). The monomorphic VPB’s were first observed at the time of VF in 2 patients, whereas in 2 patients they had been documented together with a normal ECG 14 and 11 years before they triggered VF, preceding the apparition of ECG abnormalities. Exercise testing and isoproterenol infusion eliminated all premature beats, excluding catecholaminergic polymorphic ventricular tachycardia.</p><p>Mapping and ablation was performed as previously described in patients with idiopathic VF [<xref ref-type="bibr" rid="R49">49</xref>]. In the three patients with RVOT origin, VPB’s were eliminated by 7-10 minutes of radiofrequency energy applications at the earliest site of activity. In the fourth patient, the VPB’s were found to originate from the anterior right ventricular Purkinje network. Ten minutes of radiofrequency energy application eliminated all VPB’s in this patient. Noteworthy is that the inducibility of VF was modified after ablation. During a mean follow-up period of 9 ± 8 months there has been no recurrence of VF, syncope or sudden cardiac death in any patient.</p><p>While catheter ablation techniques are emerging, the initial experiences with idiopathic VF, and latterly with VF related to repolarization disorders have provided important insights into the role of focal triggers from RVOT and the Purkinje system in the initiation of VF associated with a number of clinical substrates in humans. Reducing the incidence of VF with localized ablation may reduce defibrillation requirement and replacement and improve the patients’ quality of life. The excellent long-term results after successful ablation of these triggers has been confirmed utilizing the data-logging capabilities of defibrillators in many of these patients and is being achieved in several centers.</p></sec>
Electrophysiological Mechanisms of Ventricular Fibrillation Induction	<p>Ventricular fibrillation (VF) is known as a main responsible cause of sudden cardiac death which claims thousands of lives each year. Although the mechanism of VF induction has been investigated for over a century, its definite mechanism is still unclear. In the past few decades, the development of new advance technologies has helped investigators to understand how the strong stimulus or the shock induces VF. New hypotheses have been proposed to explain the mechanism of VF induction. This article reviews most commonly proposed hypotheses that are believed to be the mechanism of VF induction.</p>	<contrib contrib-type="author"><name><surname>Chattipakorn</surname><given-names>Nipon</given-names></name><degrees>MD, PhD</degrees></contrib><contrib contrib-type="author"><name><surname>Shinlapawittayatorn</surname><given-names>Kirkwit</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Chattipakorn</surname><given-names>Siriporn</given-names></name><degrees>PhD</degrees></contrib><aff> Cardiac Electrophysiology Unit, Faculty of Medicine, Chiang Mai University, Chiang Mai, Thailand</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Ventricular fibrillation (VF) was first described by Erichsen in 1842 [<xref ref-type="bibr" rid="R1">1</xref>]. It is known as a fatal cardiac arrhythmia that can cause sudden death. This life-threatening VF has drawn the strong attention of a number of investigators for over a century. The study of VF induction can be traced back to the original Ludwig and Hoffa study in 1850 in which they used a strong faradic current to induce VF. However, it was not until 1940 that Wigger and Wegria established the fundamental work which demonstrated that VF could be induced when a strong premature stimulus was applied during a certain period of the cardiac cycle [<xref ref-type="bibr" rid="R2">2</xref>]. This period is known as the “vulnerable period”, a period which corresponded to a portion of the T-wave of the surface electrocardiogram. The finding of VF induction by a strong stimulus delivered during the vulnerable period has allowed many investigators to advance the understanding of its mechanism. Although many theories have been proposed as the fundamental mechanism of VF induction, none is universally accepted. Current debates are discussed base on whether (1) reentrant [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R5">5</xref>] or (2) focal pattern [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R8">8</xref>] that is responsible for VF induction after the shock. In this review, four commonly proposed hypotheses are presented.</p></sec><sec sec-type="" id="s2"><title>The non-uniform dispersion of refractoriness hypothesis</title><p>It is known that membrane potential differences always exist in the intact heart during systole and diastole. This is a result of the unequal levels of the resting membrane potentials as well as the depolarization potentials in myocardial cells and pacemaker cells [<xref ref-type="bibr" rid="R9">9</xref>-<xref ref-type="bibr" rid="R11">11</xref>]. This heterogeneity is known as the “dispersion of refractoriness” of the tissues [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R12">12</xref>] and already appears throughout the cardiac cycle of a normal heartbeat. It is well accepted that to induce VF, the strength of a premature stimulus must be sufficiently strong (i.e. a threshold level) and be delivered during the vulnerable period [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R13">13</xref>]. This shock strength is known as the VF threshold (VFT) [<xref ref-type="bibr" rid="R14">14</xref>,<xref ref-type="bibr" rid="R16">16</xref>]. It is believed that VF is induced when the amount of heterogeneity or dispersion of refractoriness reaches a level that allows unidirectional block to occur, leading to reentry and fibrillation [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R17">17</xref>]. This concept was supported by the discovery that VF is most likely to occur when the dispersion of refractoriness increases [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R5">5</xref>] [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>]. By preconditioning the heart in a various setting to set up a non-uniform dispersion of refractoriness such as by slowing heart rate, stimulating cardiac sympathetic nerve, or by causing ischemic myocardium, many VF induction studies have demonstrated that the VFT was decreased when the degree of dispersion of refractoriness increased [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R14">14</xref>] [<xref ref-type="bibr" rid="R20">20</xref>]. Because of the heterogeneity of refractoriness of cardiac cells in various regions on the heart, activations are generated by the stimulus in excitable areas which are blocked unidirectionally when they encountered areas of greater refractoriness, leading to reentry and eventually VF. Therefore, in this hypothesis, VF occurs by reentry caused by non-uniform dispersion of refractoriness.</p><p>It has been shown that responses of cardiac tissues to the shock can be in one of three categories, depending on the state of the myocardium at the time of the shock [<xref ref-type="bibr" rid="R21">21</xref>-<xref ref-type="bibr" rid="R23">23</xref>]. First, the action potential duration will be extended if the shock is delivered to effective or relatively refractory tissue [<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R24">24</xref>,<xref ref-type="bibr" rid="R25">25</xref>]. This is commonly known as the “graded response”. Second, the action potential will not be affected by the shock if it falls into the absolute refractory tissue. And third, a new action potential will be created if the shock falls into the completely recovered tissue. The degree of dispersion of refractoriness caused by a strong stimulus is mainly due to different responses of cardiac cells in different areas, resulting in the heterogeneity of refractory period extension in different cardiac cells throughout the heart [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R26">26</xref>,<xref ref-type="bibr" rid="R27">27</xref>].</p><p>The vulnerable period is known to have a high degree of dispersion of excitability during the cardiac cycle [<xref ref-type="bibr" rid="R9">9</xref>]. Hence, when a premature stimulus is applied to the heart during the vulnerable period, the response of the myocardium to the shock creates an even greater dispersion of refractoriness because the cardiac myocardium is irregularly excitable during that period, facilitating reentry and resulting in the initiation of VF [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R28">28</xref>,<xref ref-type="bibr" rid="R29">29</xref>]. Since a certain amount of heterogeneity exists in the normally functioning heart, the differences between this amount of preexisting heterogeneity and the heterogeneity induced by the electrical stimulus required for VF induction (i.e. at VFT) is sometimes considered the margin of safety [<xref ref-type="bibr" rid="R9">9</xref>].</p><p>Other groups of investigators, however, suggested that focal activation could initiate fibrillation due to the non-uniform dispersion of refractoriness hypothesis [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R8">8</xref>]. In vitro studies have shown that abrupt differences in repolarization of adjacent cardiac cells were found at the site where the repetitive firing occurred (i.e. focal re-excitation), [<xref ref-type="bibr" rid="R8">8</xref>] and these sudden repolarization differences have been considered to be a mechanism for VF induction [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R30">30</xref>,<xref ref-type="bibr" rid="R31">31</xref>]. The role of premature stimulus delivered during the vulnerable period in the in vivo studies, however, is still controversial. Both the increase in automaticity of the pacemaker fibers and reentry due to the unidirectional block have been demonstrated to be responsible for VF induction [<xref ref-type="bibr" rid="R32">32</xref>-<xref ref-type="bibr" rid="R34">34</xref>].</p></sec><sec sec-type="" id="s3"><title> The critical point hypothesis</title><p>It is known that the potential gradient created by the shock is very strong at sites close to the shocking electrode and is weaker at more distant sites [<xref ref-type="bibr" rid="R35">35</xref>]. This gradient distribution, therefore, creates a non-uniform gradient field. The critical point hypothesis is based on this fact and the fact that there are three possible responses of cardiac tissue when a premature stimulus is delivered to cells of different excitability: no response, graded response (i.e. refractoriness extension), and new activation. This hypothesis states that the mechanism for VF induction is due to unidirectional block and unidirectional propagation of activation caused by those three different responses of cardiac tissues, leading to reentrant activation and, finally, VF [<xref ref-type="bibr" rid="R21">21</xref>]. The difference between the critical point hypothesis and the non-uniform dispersion of refractoriness hypothesis is that the critical point hypothesis suggests the reentrant pattern as the sole mechanism for VF induction, whereas the latter could have either reentrant or focal excitation as the mechanism for VF induction [<xref ref-type="bibr" rid="R21">21</xref>].</p><p>The critical point hypothesis was first proposed mathematically by Winfree [<xref ref-type="bibr" rid="R36">36</xref>] considering the heart as an excitable media, and was later demonstrated experimentally by Frazier et al [<xref ref-type="bibr" rid="R21">21</xref>] in 1989. In this experimental study, a strong premature stimulus (S2) was delivered to myocardium after a train of basic pacing stimuli (S1). At an appropriate timing of the S2 delivery, they found the 3 types of myocardial responses to the S2 in 3 distinct regions (<xref ref-type="fig" rid="F1">Figure 1</xref>). First, the new activation created by the S2 shock arose at the recovered tissues, close to the S1 electrode, and was ready to propagate toward the excitable region (<xref ref-type="fig" rid="F1">area 1, Figure 1</xref>). (2) At the region far from the S1 electrode but close to the S2 electrode, the tissues were in their relative refractory period at the time the premature stimulus was delivered. The S2 shock was strong enough to create a graded response, prolonging the refractoriness of the tissues in that area (<xref ref-type="fig" rid="F1">area 2, Figure 1</xref>). As a result, the new activation could not propagate through it. (3) At the region farthest away from the S1 electrode (<xref ref-type="fig" rid="F1">area 3, Figure 1</xref>), the S2 shock had no effect on the tissues in this whole area because the tissues close to the S2 electrode were still in their refractory period and not excitable and the S2 shock was too weak to create any response in the tissues far from it. However, the myocardium in this region had sufficiently recovered in time to be excited by the activation front which propagates from the directly excited region. This activation front could then reenter the area 2 and return to the area 1 again, since these cardiac tissues were already excitable.</p><p>The activation front would circle around the point where the three different cellular-response regions met (<xref ref-type="fig" rid="F1">dark circle in Figure 1</xref>). This reentrant activation front could continue if the pattern of refractoriness of myocardium were maintained, or could be interrupted if the excitable pattern was changed. It is important to note that in the critical point hypothesis, the angle between the S1 and S2 stimulus must be greater than zero to create the critical point for reentry [<xref ref-type="bibr" rid="R21">21</xref>]. This reentrant activation front could continue if the pattern of refractoriness of myocardium were maintained, or could be interrupted if the excitable pattern was changed. The formation of a critical point was thought to generate fibrillation in both VF induction by a premature stimulus and failed defibrillation [<xref ref-type="bibr" rid="R37">37</xref>-<xref ref-type="bibr" rid="R40">40</xref>]. However, reentry is not always the pattern observed during VF induction or failed defibrillation. Therefore, critical point formation may not be the sole mechanism of VF induction.</p></sec><sec sec-type="" id="s4"><title>The upper limit of vulnerability hypothesis</title><p>It is known that when a premature stimulus is given during the vulnerable period, there is a minimal strength needed to generate the inhomogeneity of excitability of cardiac tissues required to induce VF. This strength is known as the VFT [<xref ref-type="bibr" rid="R14">14</xref>]. When the strength of a premature stimulus is increased up to a level that VF is no longer induced at any time during the vulnerable period, this lowest strength that cannot induce VF is known as the upper limit of vulnerability (ULV) [<xref ref-type="bibr" rid="R41">41</xref>] [<xref ref-type="bibr" rid="R42">42</xref>] The critical point hypothesis could be used to explain the existence of this ULV. Since the formation of a critical point requires the cross point between the critical potential gradient and optimal excitable tissues, if this cross point is removed from the heart, the critical point will not be formed [<xref ref-type="bibr" rid="R43">43</xref>]. It has been shown that as shock strength increases, distance of the critical potential away from the shocking electrode also increases [<xref ref-type="bibr" rid="R35">35</xref>]. When the shock reaches the ULV strength, the critical value is off the heart. Thus, no critical point is formed, and no VF is induced even when the shock strength is further increased [<xref ref-type="bibr" rid="R43">43</xref>] [<xref ref-type="bibr" rid="R44">44</xref>].</p></sec><sec sec-type="" id="s5"><title>The virtual electrode polarization hypothesis</title><p>This is the recent hypothesis proposed by Efimov to explain the induction of fibrillation [<xref ref-type="bibr" rid="R45">45</xref>]. The concept of this hypothesis is similar to that of the critical point hypothesis, except that this hypothesis is not base on the potential gradient created by the shock delivered to the myocardium. The findings that the shock can cause (1) depolarization or hyperpolarization of cardiac cells close to the shocking electrode, and (2) opposite polarization of cardiac cells in the region adjacent to (1) are the fundamental concept of this hypothesis. When the optimal transmembrane potential gradient is generated in the region near the shocking electrode, reentry can be observed as activation propagates from depolarized tissues into hyperpolarized regions (<italic>see figure 7 in reference 45</italic>). This hypothesis is proposed to explain how fibrillating activation was observed after failed defibrillation (<italic>see figure 11 in reference 45</italic>).</p><p>Although reentry has been proposed in most hypotheses as the mechanism responsible for VF induction, recent VF induction studies in pigs have demonstrated different findings. Chattipakorn et al have shown that following near ULV shocks, the first few post-shock activations arose on the epicardium in a focal manner before degenerating into VF [<xref ref-type="bibr" rid="R46">46</xref>,<xref ref-type="bibr" rid="R47">47</xref>]. No reentry was observed in these studies. It has also been shown that ablation performed in the region where the early post-shock activation occurred could significantly decrease the ULV shocks [<xref ref-type="bibr" rid="R48">48</xref>]. It is possible that focal activation observed in these studies is epicardial breakthrough resulting from transmural or endocardial reentry. Further studies are under investigation to validate this hypothesis. Other mechanisms including the vortex theory and the mother rotor theory also have been proposed to be responsible for initiation and maintenance of VF [<xref ref-type="bibr" rid="R49">49</xref>-<xref ref-type="bibr" rid="R51">51</xref>]. The definite mechanism, however, have yet to be revealed.</p></sec><sec sec-type="" id="s6"><title>Conclusion</title><p>Similar to defibrillation mechanism, the mechanism of VF induction is complicate. Although its mechanism has been investigated for so many decades, how VF is induced is still debated. Further studies of VF induction and defibrillation are essential since they will provide important information on the fundamental mechanism that can be used to improve the treatment and prevention of sudden cardiac death, which is mainly caused by VF in the future.</p></sec>
U wave: an Important Noninvasive Electrocardiographic Diagnostic Marker	<p>Study of U waves exemplifies important clinical role of noninvasive electrocardiography in modern cardiology. Present article highlights significance of U waves with a clinical case and also summarizes in brief the history of the same</p>	<contrib contrib-type="author"><name><surname>P</surname><given-names>Girish M</given-names></name></contrib><contrib contrib-type="author"><name><surname>Gupta</surname><given-names>Mohit Dayal</given-names></name></contrib><contrib contrib-type="author"><name><surname>Mukhopadhyay</surname><given-names>Saibal</given-names></name></contrib><contrib contrib-type="author"><name><surname>Yusuf</surname><given-names>Jamal</given-names></name></contrib><contrib contrib-type="author"><name><surname>N</surname><given-names>Sunil Roy T</given-names></name></contrib><contrib contrib-type="author"><name><surname>Trehan</surname><given-names>Vijay</given-names></name></contrib><aff>Department Of Cardiology, G B Pant Hospital, New Delhi-110002, India</aff>	Indian Pacing and Electrophysiology Journal	<p>Study of U waves has been more of academic importance rather than clinical application. Negative U waves when present may be of immense clinical importance. We describe here one such case where U waves were not only an early noninvasive marker for acute ischemia but they also disappeared after successful revascularization.</p><p>The U wave, named by Einthoven in 1903, is still a subject of debate with respect to its origin and clinical importance [<xref ref-type="bibr" rid="R1">1</xref>]. Three hypotheses addressing the genesis of U wave that have been put forth include late repolarization of Purkinje fibers, late repolarization of some other portions of left ventricle, and alteration in the normal action potential shape by after-potentials [<xref ref-type="bibr" rid="R2">2</xref>]. U waves have same polarity as T waves in normal subjects. Any alteration in the same with respect to T wave is of importance. A 50 year old male, chronic smoker, non hypertensive and non diabetic presented with unstable angina (Braunwald class IIIb2).Electrocardiogram showed normal sinus rhythm with upright T waves and no ST-T changes, but prominent inverted U waves in mid precordial leads (<xref ref-type="fig" rid="F1">Figure 1a</xref>). Echocardiography showed normal LV systolic and diastolic function. Coronary angiogram revealed a discrete tight stenosis (90%) in the mid left anterior descending artery (<xref ref-type="fig" rid="F2">Figure 2a</xref>). Coronary angioplasty with stenting was successfully carried out (<xref ref-type="fig" rid="F2">Figure 2b</xref>). Immediate post procedure ECG revealed remarkable disappearance of U waves (<xref ref-type="fig" rid="F1">Figure 1b</xref>).</p><p>Negative U waves have low sensitivity but high specificity for heart disease and they are recorded in approximately 1% of all electrocardiograms in general hospital [<xref ref-type="bibr" rid="R3">3</xref>]. Negative U waves at rest, may be the earliest marker of unstable angina and evolving myocardial infarction [<xref ref-type="bibr" rid="R3">3</xref>]. They may also be seen at rest in cases of hypertension, variant angina, congenital long QT syndromes, left ventricular enlargement, left anterior descending coronary artery disease and valvular heart disease (aortic and mitral valve disease) [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R4">4</xref>]. Others however showed that negative U waves during acute anterior wall myocardial infarction are useful in identifying patients with smaller infarction and better collateral circulation but they had no predictive value in localizing the diseased artery [<xref ref-type="bibr" rid="R5">5</xref>]. U waves on exercise may be seen in cases of significant left circumflex or right coronary artery disease with abundant collaterals [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>The present case exemplifies the significance of negative U waves as the important sign of acute ischemia. We believe that clinical utility of U waves remains underutilized and more clinical studies are needed to exactly define their significance in the spectrum of coronary artery disease.</p>
Pacemapping	Could not extract abstract	<contrib contrib-type="author"><name><surname>Moreno</surname><given-names>Mauricio</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Perez-Castellano</surname><given-names>Nicasio</given-names></name><degrees>MD, PhD</degrees></contrib><contrib contrib-type="author"><name><surname>Villacastin</surname><given-names>Julian</given-names></name><degrees>MD, PhD</degrees></contrib><aff>Unit of Arrhythmias, Cardiovascular Institute, San Carlos University Hospital, Madrid, Spain</aff>	Indian Pacing and Electrophysiology Journal	<p>Pacemapping (PM) is an electrophysiologic technique designed to help locating tachycardia sources by stimulating at different endocardial sites in order to reproduce the clinical tachycardia characteristics. A recorded electrocardiogram (ECG) during the clinical tachycardia has been conventionally used as reference. Yet, endocardial activation pattern during tachycardia may be utilized as well to guide the procedure. In focal tachycardia ablation, PM guide has consistently provided remarkable outcomes [<xref ref-type="bibr" rid="R1">1</xref>], while outcomes in reentrant tachycardia ablation are less favourable [<xref ref-type="bibr" rid="R2">2</xref>]</p><p>An important issue regarding PM is the electrical configuration of the stimulation catheter, bipolar or unipolar. During bipolar pacing, stimulation poles (anode and cathode) are located at the distal tip of the catheter and usually both are contacting the endocardium. Using this configuration, myocardium can be captured at the cathode, anode or both sites simultaneously decreasing mapping accuracy. On the contrary, unipolar pacing captures exclusively local myocardium next to the distal electrode [<xref ref-type="bibr" rid="R3">3</xref>]. Thus, unipolar pacing provides theoretical advantages over bipolar pacing by limiting the stimulated area. However, significant stimulation artefacts created by unipolar pacing limit its use. Bipolar pacing with minimally spaced electrodes constitutes an intermediate option capturing a limited myocardial area and inducing significantly less electrical artefacts.</p><p>Conduction properties of paced stimuli seem to depend on myocardial conditions. Pacing ventricular scar tissue, as in patients with prior myocardial infarction (MI), shows a delay between the electrical stimulus and the QRS onset, suggesting local slow conduction. These impaired conduction zones may represent potential substrates for reentrant circuits involved in ventricular tachycardia (VT).</p><p>Regarding atrial PM, it is unclear whether the conduction pattern of an atrial stimulus depend on the paced site, electrical voltage or the stimulating rate. This issue was recently addressed by Perez-Castellano el al [<xref ref-type="bibr" rid="R4">4</xref>]. Pacing at right and left atria, the authors evaluated 120 stimulation sequences determining the conduction time between a central dipole (pacing site) and two 15 mm-spaced dipoles. Conduction time intervals were significantly longer in right atrium (RA) than in left atrium (LA). The mean activation time at both atria varied 10±4 ms when the pacing site was 6 mm displaced. When pacing voltage was increased from 2 to 10 times the diastolic threshold, a non-significant reduction in conduction and activation times was observed at both atria. However, no difference was observed in the conduction and activation times when stimulating at two different cycle lengths (300 and 500 ms). According to these findings, the authors postulate that during atrial PM it is not mandatory to stimulate exactly at the same cycle length than the tachycardia to reproduce a similar atrial sequence.</p><p>A PM limitation is that pacing at two different points may induce similar surface ECG or endocavitary recordings. An optimal spatial PM resolution requires a short maximum distance between two points generating similar ECG. Usually, the spatial resolution of unipolar stimulation is shorter than 5 mm [<xref ref-type="bibr" rid="R3">3</xref>]. Spatial resolution deteriorates with wide electrodes, bipolar stimulation and pacing at pathologic areas.</p><p>The highest benefit of PM has been found in focal tachycardia mapping, especially in idiopathic VT. However, PM has been reported in a broad variety of arrhythmias. In the following sections we will review its usefulness and feasibility sorted by type of arrhythmia.</p><sec sec-type="" id="s1"><title>Atrial Tachycardia</title><p>Atrial tachycardia (AT) is defined as focal when its source is a precise point at the atria. Focal AT are due to triggered activity, increased automatism or microreentry. The precise origin of a focal AT can be readily determined using PM by pacing at different sites of the atria to emulate the P waveform of the tachycardia (<xref ref-type="fig" rid="F1">Figure 1</xref>) or the atrial activation sequence during tachycardia. However, proper interpretation of discrete changes in P wave shape is limited by its low voltage and distortion or masking by prior ventricular repolarizations.</p><p>McLean et al [<xref ref-type="bibr" rid="R5">5</xref>] reported several electrocardiographic criteria to help locating an AT origin. According to these authors, negative P waves in lead I suggest a left atrial origin, specifically an inferior pulmonary vein or coronary sinus (CS) when the P wave is negative or bimodal in lead V1. Generally, pacing close to the CS produces a negative P wave in inferior leads [<xref ref-type="bibr" rid="R2">2</xref>]. But Waldo [<xref ref-type="bibr" rid="R6">6</xref>] demonstrated that pacing at the upper site of CS generates positive or biphasic P waves in inferior leads. This could be explained by an activation of the LA through the Bachman bundle and later propagation of the excitation wavefront from the top to the bottom across the atria. Yet, Man et al [<xref ref-type="bibr" rid="R7">7</xref>] demonstrated that pacing at two different sites, up to 1.7 cm spaced at RA and 3.2 cm spaced at CS, induced similar P wave shapes. These studies confirmed a limited value of surface ECG for PM.</p><p>SippensGroenewegen et al [<xref ref-type="bibr" rid="R8">8</xref>] used a multiple lead surface ECG (62 leads) to create P wave maps during right atrial pacing at different sites. They found that the 17 evaluated sites showed different activation maps. The RA spatial resolution was 3.5 cm2 in this study; better than the RV spatial resolution (6.7 cm2) previously published [<xref ref-type="bibr" rid="R9">9</xref>].</p><p>Another tool for targeting AT ablation is PM guided by endocavitary recordings. This technique compares the atrial activation sequence during tachycardia, as shown by endocavitary recordings, and the sequence obtained during pacing with a mapping catheter at different atrial sites. Tracy et al reported excellent results using this technique in 10 focal AT ablation procedures [<xref ref-type="bibr" rid="R10">10</xref>].</p></sec><sec sec-type="" id="s2"><title>Pulmonary Veins</title><p>A focal origin has been reported in most patients with paroxysmal atrial fibrillation (AF). These foci are usually located in the pulmonary veins (PV) and its ablation is becoming a curative therapy. Ablation strategies include empirical electric isolation of the four PV or a selective approach ablating only the culprit PV. The latter can be extraordinarily demanding as finding the culprit vein is frequently challenging. Deen et al [<xref ref-type="bibr" rid="R11">11</xref>] reported the use of PM to help identify the putative PV. In 10 patients (33 PVs) the authors compared the activation sequences recorded at the crista terminalis (CT) and the CS during selective pacing from the four PVs. CT electrograms were recorded using a 20-pole catheter and a decapolar catheter was inserted in the CS.</p><p>Activation patterns at CT and CS differed at every paced PV. In order to measure activation times they considered the CS proximal electrodes (CS 9-10) as a reference, corresponding to time zero. According to their findings they concluded: 1.- Pacing of right PVs showed activation of CS was from proximal to distal: reaching CS 9-10 15.8 ms and 24 ms before CS 1-2 in right superior PV (RSPV) and right inferior PV (RIPV), respectively. Pacing of left PVs showed activation of CS from distal to proximal: reaching CS 1-2 23.8 ms and 20.1 ms before CS 9-10 in left superior PV (LSPV) and left inferior PV (LIPV), respectively. 2.- Activation of CT was earlier in CT 5-6 and CT 11-12 during pacing of RSPV and RIPV, respectively. Pacing of left PVs showed earlier activation at CT 3-4 with activation pattern from top to bottom. Total activation time of CT was longer for the LSPV (33.7 ms) than for the LIPV (19.3 ms).</p><p>An excellent correlation was observed between atrial activation sequence obtained by PM and the location of the focus at the PV under traditional electrophysiological criteria.</p></sec><sec sec-type="" id="s3"><title>Accesory pathways</title><p>Conventional strategies for accesory pathway (AcP) mapping face certain limitations, leading to prolonged procedures and the need for multiple RF applications during ablation. Perez-Castellano et al reported a new mapping approach to help localize the atrial insertion of AcP by reproducing the atrial activation sequence during orthodromic tachycardia through atrial pacing at the mapping catheter [<xref ref-type="bibr" rid="R12">12</xref>].</p><p>This method is based on the relative timing of activation between two stable reference electrograms, used to estimate the atrial activation sequence during tachycardia and during pacing (<xref ref-type="fig" rid="F2">Figure 2</xref>). The authors first estimated the location of the AcP by electrocardiographic criteria. Then, the activation interval between the two atrial references is measured during tachycardia. Atrial pacing is performed at the mapping catheter at different locations closed to the theoretical location of the AcP aiming to reproduce that interval. A difference between the activation time in tachycardia and during pacing lower than 5 ms identified sites with the maximum probability of ablation success. The highest predictive value of success was reached when the following criteria were fulfilled: difference between activation times shorter than 5 msec, V/A local index lower than 1, and a stable local electrogram (EGM).</p><p>This method is complementary with traditional mapping criteria [<xref ref-type="bibr" rid="R13">13</xref>] and its maximum usefulness is foreseen after non-intentional induction of AcP mechanical block, avoiding procedure discontinuation. A major limitation of the technique is the need for adequate stable references close to the area of interest.</p><p>In patients with preexcited tachycardia implicating Mahaim fibers underwent attempted catheter ablation of the accessory pathway is possible to locate the accessory pathway ventricular insertion site used the criteria of concordance between paced and spontaneous QRS morphologies during pace mapping [<xref ref-type="bibr" rid="R14">14</xref>]. This criteria is complementary to earliest onset of local electrogram relative to surface preexcited QRS (activation mapping) and a QS-like pattern of unfiltered unipolar electrograms with steep downstroke.</p></sec><sec sec-type="" id="s4"><title>Idiopathic Ventricular Tachycardia</title><p>The right ventricular outflow tract (RVOT) tachycardia is the most frequent idiopatic VT, it origin is usually a single and well-limited focus. In the majority of patients this focus is located in RVOT’s anterior septal wall below the pulmonary valve; occasionally, the posterior septal and free walls are involved. These patients do not show overt structural heart disease, thus ventricular excitation and conduction properties are preserved. These conditions, as opposed to scar tissue, facilitate identifying the origin of the VT using PM.</p><p>Conventionally, ventricular PM is performed in sinus rhythm, pacing at a cycle length similar to the clinical VT, trying to reproduce the QRS morphology during VT (<xref ref-type="fig" rid="F3">Figure 3</xref>). Paced and clinical VT QRS complexes must be similar regarding the pattern of left bundle-branch block, cardiac axis, R/S relationship in precordial leads and any QRS notching. These criteria should ideally be fulfilled in 11 of the 12 leads of the ECG.</p><p>Far-field capture may interfere with an adequate PM evaluation by distorting the paced QRS morphology. In order to reduce this detrimental effect, unipolar or bipolar with a distance between poles no longer than 5 mm are the recommended pacing configurations [<xref ref-type="bibr" rid="R15">15</xref>]. Regarding the PM spatial resolution in RVOT, Green et al. found similar ECGs pacing at endocardial sites separated up to 8 mm [<xref ref-type="bibr" rid="R16">16</xref>].</p><p>Activation mapping constitutes an additional tool for idiopathic VT mapping. As an example, in a recently published series of six RVOT tachycardia cases, Timmermans et al using activation mapping located the origin of the tachycardia in the pulmonary artery, followed by successful ablation [<xref ref-type="bibr" rid="R17">17</xref>]. Presystolic potentials were registered at the ablation sites. In 3 of these patients pulmonary artery pacing showed an ECG similar to the clinical VT.</p><p>Some ECG patterns may suggest the primary location of a VT within the RVOT. The presence of Q waves in DI suggests an anterior septal origin, while R waves in DI suggest a more posterior source. When the R wave transition occurs early in precordial leads suggests an origin in the upper RVOT, a later transition (from V2) points to a lower location, below the pulmonary valve. Digit et al performed PM at the RVOT in healthy individuals in order to define ECG patterns from the different stimulated points to suggest the origin of RVOT tachycardias [<xref ref-type="bibr" rid="R18">18</xref>]. The authors reported that tall and narrow R waves on DII suggest a septal origin, while a notch in the R wave on DII and/or a late QRS transition on precordial leads suggest an origin in RVOT free wall. Some limitations of these criteria are a lack of specificity and the possibility of variations according to heart position.</p><p>Other idiopathic VT is the fascicular VT, is usually verapamil-sensible and has been demonstrated to arise from the left posterior or left anterior fascicle, with a right bundle branch block configuration and left-axis deviation or right-axis deviation, respectively. The ablation of a fascicular VT can be guided by activation mapping and/or Purkinje potential preceded the QRS during VT, behind VT exit to be associated with an optimal match between the paced rhythm and the clinical VT [<xref ref-type="bibr" rid="R19">19</xref>]. Therefore, if fascicular VT is non-inducible the PM is very important for guided the ablation, but ablation at a site with an optimal PM also can be unsuccessful [<xref ref-type="bibr" rid="R20">20</xref>].</p></sec><sec sec-type="" id="s5"><title> Coronary disease-related VT</title><p>Myocardial infarction is the leading substrate for monomorphic sustained VT. In hemodinamically tolerated VT, radiofrequency ablation constitutes a major therapeutic option associated or not to implantable cardioverter defibrillators. Percutaneous VT ablation reduces the rate of recurrences, avoids potential toxic effects of antiarrhythmic drugs and it is associated to a lower risk than EP-guided surgical resection procedures. Gonska et al performing a single morphology VT ablation reported an acute success rate close to 75% [<xref ref-type="bibr" rid="R21">21</xref>]. Ablation of multiple morphology VT is associated with lower success rates.</p><p>Usefulness of PM in this subset of patients is limited. Scar tissue impairs ventricular conduction properties and favours conduction blockade. Thus, in this context, pacing in sinus rhythm to reproduce an exact VT QRS morphology is usually challenging or unsuccessful.</p><p>The electrocardiographic VT morphology is determined by the initial activation site (circuit exit), and by the ventricular activation pathway. When PM is performed at or near the exit of a reentrant circuit, the induced QRS morphology may resemble the VT QRS [<xref ref-type="bibr" rid="R22">22</xref>]. However, when pacing is performed within a wide circuit and capture involves distant zones away from its exit, the morphology of QRS may be different despite pacing inside the circuit. Therefore, PM in close zones in these patients may induce significantly different QRS morphologies [<xref ref-type="bibr" rid="R22">22</xref>].</p><p>As mentioned before, PM spatial resolution worsens with bipolar stimulation by inducing electrical capture at both electrodes. Kadish and coworkers [<xref ref-type="bibr" rid="R23">23</xref>] demonstrated that pacing at the same site may induce different QRS morphologies depending on the bipolar or unipolar configuration. Moreover, pacing at 10 mA may induce different morphologies as compared to stimulating at twice the diastolic threshold. Given that the contribution of the proximal electrode (generally anode) to depolarization is variable, Delacretaz et al do not recommend bipolar pacing with separate electrodes for PM or entrainment mapping on post-myocardial infarction VT [<xref ref-type="bibr" rid="R24">24</xref>].</p><p>PM also helps to characterize slow conduction zones usually located within scars, manifested as a delay between the stimulus and the QRS onset (S-QRS interval). In normal hearts the S-QRS interval is shorter than 40 ms, while most of myocardial areas with abnormal fractionated EGM in sinus rhythm show a S-QRS delay longer than 40 ms. Demonstrating the presence of slow conduction zones is not enough to guide an ablation procedure, as they constitute a regular finding in infarcted areas [<xref ref-type="bibr" rid="R24">24</xref>]. However, the presence of abnormal EGMs in sinus rhythm associated with a delay in the S-QRS interval and a similar ECG than the clinical VT, suggests potential key zones to be evaluated by conventional criteria during VT. In cases of fast or poorly tolerated VT, PM may be the primary strategy to guide the ablation procedure.</p></sec>
Catheter Ablation of Tachyarrhythmias in Small Children	Could not extract abstract	<contrib contrib-type="author"><name><surname>Blaufox</surname><given-names>Andrew D</given-names></name><degrees>MD</degrees></contrib><aff>Children’s Heart Program of South Carolina – MUSC</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>An estimated 80,000-100,000 radiofrequency ablation (RFA) procedures are performed in the United States each year [<xref ref-type="bibr" rid="R1">1</xref>]. Approximately 1% of these are performed on pediatric patients at centers that contribute data to the Pediatric Radiofrequency Registry [<xref ref-type="bibr" rid="R2">2</xref>]. Previous reports from this registry have demonstrated that RFA can safely and effectively be performed in pediatric patients [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R4">4</xref>]. However, patients weighing less than 15 kg have been identified as being at greater risk for complications [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R4">4</xref>]. Consequently, there has been great reluctance to perform RFA in small children such that children weighing less than 15 kg only represent approximately 6% of the pediatric RFA experience [<xref ref-type="bibr" rid="R2">2</xref>]. despite the fact that this age group carries the highest incidence of tachycardia, particularly supraventricular tachycardia (SVT) [<xref ref-type="bibr" rid="R5">5</xref>]. Factors other than the risk of complications contribute to the lower incidence of RFA in this group, including the natural history of the most common tachycardias (SVT), technical issues with RFA in small hearts, and the potential unknown long-term effects of RF applications in the maturing myocardium. Conversely, there are several reasons why ablation may be desirable in small children, including greater difficulties with medical management [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R8">8</xref>], the higher risk for hemodynamic compromise during tachycardia in infants with congenital heart disease (CHD), and the inability of these small children to effectively communicate their symptoms thereby making it more likely that their symptoms may go unnoticed until the children become more seriously ill. Before ultimately deciding that catheter ablation is indicated in small children, one must consider which tachycardias are likely to be ablated, the clinical presentation of these tachycardias, alternatives to ablation, the relative potential for success or complications, and modifications of the procedure that might reduce the risk of ablation in this group.</p></sec><sec sec-type="" id="s2"><title>Tachycardia Substrates in Small Children</title><p>It is necessary to have a clear understanding of which tachycardias that are likely to be ablated in small children prior to adequately discussing whether or how to ablate them. Atrioventricular reciprocating tachycardia (AVRT) is the most common type of SVT in small children [<xref ref-type="bibr" rid="R9">9</xref>] with a prevalence of approximately 0.1-0.15% [<xref ref-type="bibr" rid="R10">10</xref>]. AVRT, AV nodal reentrant tachycardia (AVNRT), and ectopic atrial tachycardia (EAT) respectively account for 80%, 5%, and 15% of SVT in children less than 1 year of age and approximately 65%, 25%, and 10% of SVT in children who are between 1 and 5 years [<xref ref-type="bibr" rid="R9">9</xref>]. Atrial flutter (AF) is relatively uncommon in this age group. Congenital junctional ectopic tachycardia (JET) is rare. However, JET can also be uncommonly encountered in the small group of neonates who undergo neonatal surgery for CHD. Although the true prevalence of ventricular tachycardia (VT) in small children is unknown, it is felt to be relatively uncommon. Thus far, the distribution of substrates ablated in infants less than 1.5 years of age has been shown to be similar to the relative prevalence of the tachycardias mentioned above (<xref ref-type="fig" rid="F1">Figure 1</xref>) [<xref ref-type="bibr" rid="R2">2</xref>]. Since AVRT is the most likely tachycardia to be encountered and ablated in small children, the remainder of this discussion will focus primarily, but not exclusively, on AVRT with or without associated preexcitation.</p></sec><sec sec-type="" id="s3"><title>Clinical Presentation of Tachycardias in Small Children</title><p>There are few true “natural history” studies for tachycardias that present in childhood as there has been a great propensity to treat once the problem has been identified. However, several important observations can be made with regard to the course of these tachycardias.</p><p>Approximately, 1/3-2/3 of patients who present with WPW or AVRT in infancy will not have a recurrence of tachycardia after medication is discontinues at their first birthday [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R6">6</xref>]. When tachycardia does recur, it is usually well tolerated. However, there are occasions when medical intervention is not sought for a prolonged period of time resulting in hemodynamic collapse [<xref ref-type="bibr" rid="R11">11</xref>]. Mortality rates of approximately 5% have been reported in infants with WPW and AVRT [<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R7">7</xref>]. While some of these deaths could be attributed to medication issues, it is likely that others were do to hemodynamic collapse after the development of tachycardia-induced cardiomyopathy and to cardiac arrest secondary to rapid ventricular conduction over the accessory pathway during atrial fibrillation in patients with WPW. Tachycardia-induced cardiomyopathy is a known consequence of prolonged tachycardia in infants and needs to be distinguished from myocarditis and recognized as a curable cause of cardiomyopathy [<xref ref-type="bibr" rid="R12">12</xref>] The risk of sudden death in WPW is approximately 0.1% per year overall and may be as high as 0.6% per year in “high risk” patients [<xref ref-type="bibr" rid="R13">13</xref>]. Klein et al originally reported the association between antegrade accessory pathway conduction properties and ventricular fibrillation in patients with WPW [<xref ref-type="bibr" rid="R14">14</xref>]. Although the capability for a young child’s heart to sustain atrial fibrillation has been debated, sudden death has been reported in the pediatric population and has been the presenting symptom in 2.3% [<xref ref-type="bibr" rid="R15">15</xref>]. The risk of sudden death in children with WPW has been associated a preexcited R-R interval of 190-220 msec during atrial fibrillation induced at EPS (sensitivity of 100% and a specificity of 72-74%) [<xref ref-type="bibr" rid="R16">16</xref>]. It is important to note that as the child ages, the conduction properties of the AV node and the accessory pathway will also change, so that assessing risk is somewhat of a moving target. Despite this, the vast majority of patients with WPW and AVRT will not experience severe symptoms.</p><p>There has been less published on the natural history of other tachycardia substrates. VT has a higher likelihood of causing acute hemodynamic collapse. While EAT is less likely to cause acute hemodynamic compromise, its incessant nature increases the risk of developing tachycardia-induced cardiomyopathy. Any of these tachycardias are less likely to be tolerated in children with structural heart disease.</p></sec><sec sec-type="" id="s4"><title>Alternatives to Catheter Ablation</title><p>There are three alternatives to catheter ablation. These are not treating, treating with drugs, and performing surgical ablation. Surgical ablation is a precursor to catheter ablation and is much more invasive so is rarely still performed today except in the form of an atrial maze procedure being more commonly performed for more complex arrhythmias in older patients and often with concurrent hemodynamic structural surgical intervention, thus it will not be discussed further. The other options are still practiced but are based on limited data.</p><sec id="s4a"><title>No Treatment</title><p>The decision not to treat patients with WPW carries the potential risks of SVT recurrence or sudden death. The use of transesophageal pacing studies have been shown to have a negative predictive value of 74-100% for predicting SVT recurrences [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R18">18</xref>]. This method can also be used to establish the likelihood of recurrent AVNRT or to determine the conduction properties of an accessory pathway while assessing the risk of sudden death in children with WPW. Thus, if the patient is not at risk for sudden death and is not inducible, no treatment becomes an option. However, other factors must be considered, such as access to medical care and parent comfort and abilities to handle recurrences.</p><p>No treatment strategy for other arrhythmias is less established.</p></sec><sec id="s4b"><title>Medical Treatment</title><p>Although there have been a great number of publications on the medical management of tachycardias in small children, there have been no controlled studies. Most studies report limited success of drugs to control SVT. Success rates for digoxin or beta-blockers have been reported to be approximately 50% while success for the more toxic class I and class III agents are not much better [<xref ref-type="bibr" rid="R18">18</xref>]. Although various combinations may increase success, they also increase the potential for side effects, particularly when class I and class III agents are combined. Reports of more aggressive drug combinations have been limited to a very small number of patients and thus their safety is essentially unknown. It is clear, however, that drug therapy, whether given as single agent or in combination, has the potential for adverse reactions including death [<xref ref-type="bibr" rid="R15">15</xref>]. While most drug therapies will be well tolerated, the lack of controlled data delineating their efficacy, makes balancing the risk/benefit ratio for drug therapy difficult. Another point about drug therapy is that it is unlikely to protect one against rapid conduction during atrial fibrillation in patients with WPW unless they are being treated with class I or class III drugs.</p></sec></sec><sec sec-type="" id="s5"><title>RF Ablation</title><sec id="s5a"><title>Success</title><p>Several studies on RFA in children have shown that there is no difference in success rate in small children for eliminating arrhythmias on the whole or AVRT in comparison to older children [<xref ref-type="bibr" rid="R2">2</xref>-<xref ref-type="bibr" rid="R4">4</xref>]. The comparable success rates may be partially due to the fact that ablations in smaller children are more likely to be attempted by more experienced pediatric electrophysiologists [<xref ref-type="bibr" rid="R2">2</xref>] and experience has been shown to be an important factor in successful pediatric RFA procedures [<xref ref-type="bibr" rid="R19">19</xref>]. Although pediatric ablation registry studies involving the entire pediatric age span have found lower success rates in children with structural heart disease [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R4">4</xref>], substrate elimination in infants has been shown not to be influenced by the presence of structural heart disease [<xref ref-type="bibr" rid="R2">2</xref>]. Thus, beliefs that RFA will be less successful for infants with heart disease are incorrect and should not deter attempts in those infants in whom RFA is indicated. Another interesting difference between infants and older children is that infant accessory pathway elimination may not necessarily be related to AP location [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R4">4</xref>]. Similar to reports in adult, the presence of multiple accessory pathways in infants is associated with lower success rates.</p></sec><sec id="s5b"><title>Complications</title><sec id="s5b1"><title>Overall</title><p>As stated previously, children weighing less than 15 kg have been shown to be at increased risk for complications during RFA [<xref ref-type="bibr" rid="R3">3</xref>] [<xref ref-type="bibr" rid="R4">4</xref>]. In Blaufox et al.’s pediatric ablation registry study of infants less than 1.5 years of age, a higher complication rate was found in infants in comparison to older children, but power limitations may have prevented the difference from reaching statistical significance [<xref ref-type="bibr" rid="R2">2</xref>]. When data from 231 registry patients weighing < 15 kg but being > 1.5 years old were factored in to the analysis, this study did confirm a higher complication rate in children < 15 kg. However, there was no appreciable difference in major complications for the infants less than 1.5 years of age.</p><p>Typically, RF lesions made in vivo, vary in size from non-existent to 5 or 6 mm radius. The average adult heart has a wall thickness of 3-12 mm’s. However, the size of the heart and its internal structures are proportional to body size [<xref ref-type="bibr" rid="R20">20</xref>]. Consequently, the theoretical risk of injuring cardiac structures in small children is higher and might depend specifically on the parameters that influence lesion size. In controlled animal studies, RF lesion size is directly related to catheter tip size, RF power, tip temperature and lesion duration [<xref ref-type="bibr" rid="R21">21</xref>]. Further, more RF applications is clearly more likely to increase total lesion volume. Similarly, repeated thermal injury in nearby areas can be expected to increase the chance of injury to adjacent vital structures, again with an inverse relation to patient size. Finally, the scars created by RF energy have a greater chance of expanding into vital structures when the myocardium is less mature [<xref ref-type="bibr" rid="R22">22</xref>]. As the greatest increase in heart size and maturity occurs during the neonatal and infant ages, disparities in size and myocardial maturity appear to be important even within the subgroup of small children.</p><p>The major complications in small children include pericardial effusion, pneumothorax, AV block, and death [<xref ref-type="bibr" rid="R2">2</xref>]. In addition to these complications, small children may be at particular risk for coronary artery injury.</p></sec><sec id="s5b2"><title>Death</title><p>The overall mortality associated with pediatric RFA has been reported by Schaffer et al. as 0.12% [<xref ref-type="bibr" rid="R23">23</xref>]. This study contained the report of an infant with a structurally normal heart who died 2 weeks following RFA for AVRT. Approximately 111 RFA procedures were done infants during the period covered by Schaffer et al yielding an infant mortality of approximately 0.9%. In addition, Schaffer’s study reports the death of an 18 month old child with congenital heart disease who underwent RFA and died the following day with fever and hypotension, but in whom no link between death and RFA could clearly be established. Blaufox et al reported the acute death of a separate infant yielding an infant mortality of 0.74% in that study. The discrepancy between Schaffer et al. is due to the difference in time periods covered by 2 studies and the fact that the Blaufox et al. report represented only acute results while the Schaffer et al. report included follow up. Although each report only includes one instance of death and thus the actual incidence might be somewhat inaccurate, it is evident that death can result from infant RFA.</p></sec><sec id="s5b3"><title>AV Block</title><p>Body weight less than 15 kg is an independent risk factor for AV block during RFA [<xref ref-type="bibr" rid="R4">4</xref>]. Similar to reports in older children [<xref ref-type="bibr" rid="R4">4</xref>], RFA for septal AP’s in infants is also associated with a higher incidence of heart block [<xref ref-type="bibr" rid="R2">2</xref>]. Thus, the ablation of septal substrates in small children is particularly risky. This is not surprising if one considers the relative sizes of RF lesions and the triangle of Koch in children. As stated previously, RF lesions generally have a radius of 5-6mm. Unlike in adults, the dimensions of the Triangle of Koch are proportional to body size in children (<xref ref-type="fig" rid="F2">Figure 2</xref>) [<xref ref-type="bibr" rid="R24">24</xref>]. Therefore, a lesion with a fixed size will have a greater likelihood of injuring vital structures within and around the triangle of a smaller child. So, great caution must be used when approaching these substrates.</p></sec><sec id="s5b4"><title>Coronary Injury</title><p> Coronary artery injury during RFA is a rare, but serious event [<xref ref-type="bibr" rid="R25">25</xref>] that has occasionally produced death [<xref ref-type="bibr" rid="R23">23</xref>,<xref ref-type="bibr" rid="R26">26</xref>]. Nearly all of the reports of coronary artery injury following RFA have been single case reports that have involved accessory pathway elimination. Of the 5 deaths in children with structurally normal hearts reported by the pediatric ablation registry [<xref ref-type="bibr" rid="R23">23</xref>], 1 was the result of thermal injury to the left main coronary artery and subsequent thrombosis of that vessel in a 13 year old child who underwent RFA for AVRT. In addition to these reports involving accessory pathways, there has been one report of injury to the posterior left ventricular branch artery during slow pathway ablation for AVNRT in a 15.5 kg child [<xref ref-type="bibr" rid="R27">27</xref>].</p><p>The mechanism of coronary injury is likely to be a combination of direct thermal injury and subsequent inflammatory response. The inflammatory component of tissue injury caused by RF energy has been shown to invade layers of the right coronary artery, leading to acute narrowing when RF energy is applied to the atrial side of the lateral tricuspid annulus in pigs [<xref ref-type="bibr" rid="R28">28</xref>]. Further maturation of this injury can result in significant late coronary stenosis [<xref ref-type="bibr" rid="R29">29</xref>]. Thus, with RF energy application, coronary stenosis may occur acutely or may be delayed.</p><p>In addition to the potential for coronary injury to be delayed, it may also be subtle, thus it may go unrecognized so that the incidence of sub-clinical coronary injury is likely to be underestimated. Blaufox et al. presented a patient in whom coronary injury was nearly missed because ST segment changes did not occur until 100 seconds after the last RF application and resolved spontaneously within minutes despite a significant persistent stenosis of the posterior left ventricular branch coronary artery [<xref ref-type="bibr" rid="R27">27</xref>]. In large retrospective and prospective studies where there were no coordinated attempts to investigate coronary injury after RFA, the reported incidences of injury were 0.03% in children [<xref ref-type="bibr" rid="R4">4</xref>], and 0.06-0.1% in adults [<xref ref-type="bibr" rid="R30">30</xref>]. However, in a study where coronary angiography was performed before and after RFA for accessory pathway-mediated tachycardias, Solomon et al reported a 1.3% incidence of coronary artery injury in 70 patients following RFA for accessory pathway-mediated tachycardias [<xref ref-type="bibr" rid="R31">31</xref>]. Thus, unless evidence for coronary artery injury is actively sought, it may go undiagnosed and underreported. With the exception of severe stenoses, injury may go unrecognized until premature coronary disease becomes associated with people who have undergone RFA as young children.</p></sec></sec><sec id="s5c"><title>Modifications</title><p>Modifications to the standard RFA procedure, such as, the use of smaller caliber catheters with smaller tips, the use of 5-second applications with lower temperature set points to test location accuracy, and the limitation of full applications to 20 seconds have been proposed and implemented [<xref ref-type="bibr" rid="R32">32</xref>,<xref ref-type="bibr" rid="R33">33</xref>]. Although these modifications are based upon physical and animal studies of the effects of radiofrequency energy on the maturing myocardium, aside from limiting the number of RF applications, which has been shown to decrease mortality [<xref ref-type="bibr" rid="R23">23</xref>], little clinical data exists to support these modifications in the application of RFA in small children [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R32">32</xref>,<xref ref-type="bibr" rid="R33">33</xref>]. However, for AVRT, Blaufox et al reported that the relationship between complications and application number and duration holds up for applications with durations greater than 20 seconds only when the number of these applications is indexed to body weight in kg (<xref ref-type="fig" rid="F3">Figure 3</xref>) [<xref ref-type="bibr" rid="R34">34</xref>]. In other words, the increase in risk of giving applications with a duration > 20 seconds was proportional to the patient’s weight. Inherent in the idea of limiting the number of lesions, is the abstention from giving an “insurance” lesion. In addition to limiting the length and number of applications based upon the patients size, perhaps the most important modification proposed in this study is the lowering of one’s threshold for accepting failure, for numerous studies have shown that a greater number of lesions will be given during a failed procedure in comparison to a successful one [<xref ref-type="bibr" rid="R2">2</xref>].</p></sec></sec><sec sec-type="" id="s6"><title>Alternative Energy</title><p>An alternate strategy for catheter ablation is to find an energy source that may be safer than RFA in small children. Although there have been no trials of cryoablation in small children, there are several aspects to this technology that make it a potential alternative. The results of prior animal studies suggest that some advantages of cryo-therapy may be particularly important for children. Cryo-ablation sites are histologically well delineated, discrete, and show homogeneous dense fibrous tissue without viable myocardium interspersed [<xref ref-type="bibr" rid="R35">35</xref>]. Cryo-lesions are smaller than RFA lesions [<xref ref-type="bibr" rid="R36">36</xref>], contributing to the ability to safely create cryo-lesions even adjacent to the His bundle. In addition, both cryoablation allows for reversible loss of tissue function [<xref ref-type="bibr" rid="R37">37</xref>] [<xref ref-type="bibr" rid="R38">38</xref>]. These transient effects occur for both the normal AV conduction fibers and the targeted tachycardia substrate. Friedman et al [<xref ref-type="bibr" rid="R38">38</xref>]. also reported 12 instances of transient AV block, 11 of which occurred during cryo-ablation modes, and all of them resolved completely. Because the leading edge of the ice ball during cryo-therapy is by definition near 0 °C and warmer than the temperature measured at the catheter tip, it is likely that discontinuation of cryo-therapy at the first signs of an electrophysiological effect will reverse that effect. Another potential safety feature of cryo-therapy is catheter stability at the point of tissue freeze. This lack of tip movement should both improve success when the catheter is in the correct place, and prevent lesion spread to undesirable locations through the sliding movement seen with RFA.</p><p>Because cryo-lesions are more delineated and smaller than RF lesions, cryoablation may require more precise positioning, particularly for the relatively discrete accessory pathways. Therefore it is not surprising that the acute success rate may be lower. Friedman et al reported an overall success rate of 69% for AVRT in adults [<xref ref-type="bibr" rid="R38">38</xref>]. However, success rates for septal accessory pathways and AVNRT were more comparable to those with RF ablation [<xref ref-type="bibr" rid="R38">38</xref>]. There is limited data with cryoablation in children, but our own experience with cryoablation in pediatric patients supports the data found in adults. We have experienced success rates of 96% for AVNRT and 63% for AVRT without any major complications. All instances of AV block have been transient with full recovery within a few seconds. Although our experience is limited, given the potential safety advantages for this technology, it is reasonable to consider using it prior to RF ablation in small children despite the lower expectations for success.</p></sec><sec sec-type="" id="s7"><title>Indications</title><p>Infants who have undergone RFA have done so for indications that are different than those for older children in whom RFA is done for “patient choice” 51% of the time. (<xref ref-type="fig" rid="F4">Figure 4</xref>) [<xref ref-type="bibr" rid="R2">2</xref>]. The differences in indications between infants and older children demonstrate that infants are sicker upon presentation and perceived to be at greater risk during arrhythmia. Although these perceptions are heightened for infants with structural heart disease and there is a higher incidence of structural heart disease for infants undergoing RFA, the incidence of structural heart disease does not entirely account for these perceptions because they are still true for infants with structurally normal hearts [<xref ref-type="bibr" rid="R2">2</xref>].</p><p>Because the definition of indications reported from the pediatric RFA registry, such as refractory to medical therapy, vary widely from center to center, others have sought to establish more clear cut indications. In 2002, a position statement was published by members of the Pediatric Electrophysiology Society and endorsed by the North American Society of Pacing and Electrophysiology. (Friedman RA NASPE) <italic>Class I</italic> indications, in which there is clear and consistent agreement that RFA will benefit the patient, included:1) WPW following aborted sudden death, 2) WPW and syncope with a shortest prexcited R-R < 250 msec, 3) chronic or recurrent SVT with ventricular dysfunction, 4) and recurrent VT associated with hemodynamic compromise and is amenable to RFA. <italic>Class IIA</italic> indications, in which the majority of opinion or data favor RFA, include 1) recurrent and/or symptomatic SVT refractory to medical therapy and age > 4 years, 2) impending congenital heart surgery when vascular or chamber access may be limited following surgery, 3) chronic (>6 months) or incessant tachycardia with normal ventricular function, 4) chronic or frequent recurrences of intraatrial reentrant tachycardia, and 5) palpitations with inducible SVT during EPS. <italic>Class IIB</italic> indications, in which there is a clear divergence of opinion regarding the need RFA, include: 1) asymptomatic WPW and age > 5yrs when the risk/benefits of RFA have been explained to the family, 2) SVT, age > 5 yrs, as an alternative to chronic medical therapy that has controlled the tachycardia, 3) SVT, age < 5 yrs, when medications, including sotalol and amiodarone, have not controlled the tachycardia or have resulted in intolerable side effects, 4) intraatrial reentrant tachycardia, 1-3 episodes per year requiring medical intervention, 5) AV node ablation for intratrial reentrant tachycardia, 6) one episode of VT with hemodynamic compromise and amenable to RFA. <italic>Class III</italic> indications, in which there is agreement that RFA is not indicated, include: 1) asymptomatic WPW, age < 5 yrs, 2) SVT, controlled with medication, age < 5 yrs, 3) Nonsustained and non incessant VT without ventricular dysfunction, 4) Nonsustained, asymptomatic SVT.</p></sec><sec sec-type="" id="s8"><title>Conclusion</title><p>Catheter ablation in small children should be reserved for truly life threatening or refractory arrhythmias after multiple failed attempts at medical management, which may include various combination therapies. RFA should be performed by an experienced pediatric electrophysiologist who undertakes various strategies to reduce risk, including limiting power and temperature as well as application duration and attempts based upon the patient’s size. Consideration of the use of alternate sources of energy like cryoablation prior to RFA may be helpful. Despite a high potential for success, having a lower threshold for accepting failure is essential.</p></sec>
Multiple Arrhythmogenic Substrate for Tachycardia in a Patient with Frequent Palpitations	<p>We report a 26-year-old woman with frequent episodes of palpitation and dizziness. Resting electrocardiography showed no evidence of ventricular preexcitation. During electrophysiologic study, a concealed right posteroseptal accessory pathway was detected and orthodromic atrioventricular reentrant tachycardia incorporating this pathway as a retrograde limb was reproducibly induced. After successful ablation of right posteroseptal accessory pathway, another tachycardia was induced using a concealed right posterolateral accessory pathway in tachycardia circuit. After loss of retrograde conduction of second accessory pathway with radiofrequency ablation, dual atrioventricular nodal physiology was detected and typical atrioventricular nodal reentrant tachycardia was repeatedly induced. Slow pathway ablation was done successfully. Finally sustained self-terminating atrial tachycardia was induced under isoproterenol infusion but no attempt was made for ablation. During 8-month follow-up, no recurrence of symptoms attributable to tachycardia was observed.</p>	<contrib contrib-type="author"><name><surname>Haghjoo</surname><given-names>Majid</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Arya</surname><given-names>Arash</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Dehghani</surname><given-names>Mohammadreza</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Emkanjoo</surname><given-names>Zahra</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Fazelifar</surname><given-names>Amirfarjam</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Sadr-Ameli</surname><given-names>MohammadAli</given-names></name><degrees>MD</degrees></contrib><aff> Department of Pacemaker and Electrophysiology, Shahid Rajaie Cardiovascular Center, School of Medicine, Iran University of Medical Sciences, Tehran, Iran</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Several reports demonstrated the existence of double tachycardia such as atrioventricular reentrant tachycardia (AVRT) and atrioventricular nodal reentrant tachycardia (AVNRT) [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R3">3</xref>], or AVNRT and atrial tachycardia (AT)4, but coexistence of four different type of supraventricular tachycardia (SVT) in the same patient rarely reported. We report radiofrequency (RF) catheter ablation of two different AVRTs, and typical AVNRT in a patient with four different SVT.</p></sec><sec sec-type="" id="s2"><title>Case report</title><p>A 26-year-old woman, with no evidence of structural heart disease, was referred to our center for evaluation of palpitation and dizziness. Despite frequent occurrence of palpitation, she had no chance to capture any 12-lead electrocardiogram (ECG) during palpitation. The only evidence of arrhythmia in this patient was a sustained SVT with heart rate about 185 beats/min recorded in one of her 24-hour holter monitoring. Two antiarrhythmic agents (propranolol, verapamil) were tried in our patient with no success.</p><p>The baseline standard 12-lead ECG showed no any abnormality. The physical examination and transthoracic echocardiography were unremarkable.</p><p>After obtaining of written informed consent, electrophysiological study (EPS) was done in post absorptive, and non-sedated state. All antiarrhythmic drugs were interrupted for at least five half-lives before procedure. Three 6F diagnostic catheters (Daig, St. Jude Medical Inc., Minnetonka, MN, USA) were inserted via left femoral vein and placed in the high right atrium (HRA), His bundle position, and right ventricular apex (RVA), respectively. A 7F decapolar catheter with 2/2/2 mm electrode spacing (Marinr® CS; Medtronic, Inc., Minneapolis, MN, USA) positioned retrogradely via right femoral vein into the coronary sinus (CS) for coronary sinus mapping. During programmed electrical stimulation (PES) from RVA, nondecremental retrograde conduction was seen with earliest atrial activity on the proximal pole of CS catheter (located at the ostium of CS). No ventricular preexcitation was observed on atrial incremental pacing, compatible with a concealed right posteroseptal (RPS) accessory pathway. During atrial extrastimulation, a sustained narrow complex tachycardia (cycle length=320 ms) was induced reproducibly with earliest retrograde atrial activity in the proximal CS (<xref ref-type="fig" rid="F1">Figure 1A</xref>). This tachycardia easily terminated by overdrive ventricular pacing and timed ventricular extrasystole during His refractoriness advanced atrial activity.</p><p>Application of RF energy at right posteroseptal area resulted in loss of conduction over the accessory pathway (AP) (<xref ref-type="fig" rid="F2">Figure 2A</xref>Figure 2A). Repeat ventricular PES revealed 1:1 vetriculoatrial conduction with earliest atrial activity on the HRA catheter, compatible with concealed right free wall (RFW) AP and atrial PES culminated in induction of a new narrow complex tachycardia (cycle length=340 ms) well-matched with orthodromic AVRT using RFW-AP (<xref ref-type="fig" rid="F1">Figure 1B</xref>). Atrial advancement was also seen during the second AVRT (RFW). Mapping of tricuspid annulus localized the site of second AP on the posterolateral area. RF energy delivery at this area interrupted AP conduction (<xref ref-type="fig" rid="F2">Figure 2B</xref>) and VA conduction shifted to the normal pathway. Repeat PES under isoproterenol infusion leads to induction of a new narrow complex tachycardia (cycle length=360 ms) compatible with typical AVNRT (<xref ref-type="fig" rid="F1">Figure 1C</xref>). AVNRT became noninducible after successful slow pathway (SP) ablation. During postablation arrhythmia induction under isoproterenol infusion, a sustained, self-terminating atrial tachycardia was induced but no attempt was made for ablation (<xref ref-type="fig" rid="F1">Figure 1D</xref>). Atrial origin of tachycardia was confirmed by V-A-A-V pattern after termination of ventricular pacing during tachycardia (<xref ref-type="fig" rid="F2">Figure 2C</xref>). She left the EP laboratory in good condition without any complications.</p><p>During 8-month’s follow-up, she was free of symptoms with no antiarrhythmic drugs and no recurrence of tachycardia was seen.</p></sec><sec sec-type="" id="s3"><title>Discussion</title><p>The combination of AVNRT-AVRT [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R3">3</xref>] (including multiple bypass tracts), AVNRT-AT [<xref ref-type="bibr" rid="R4">4</xref>], and AVRT-AT [<xref ref-type="bibr" rid="R5">5</xref>] was reported. To the best of our knowledge, our case is the first patient reported with a combination of two AVRT, typical AVNRT, and AT.</p><p>As atrioventricular (AV) bypass tract is a congenital abnormality due to developmental defect in the AV rings, it is not surprising that multiple bypass tracts can be present in the same patient. Incidence of multiple bypass tracts ranges from 3.1-30% [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R8">8</xref>], and the most common combination has been RPS with RFW bypass tracts [<xref ref-type="bibr" rid="R6">6</xref>]. The incidence of multiple APs is higher in: 1) patients with antidromic AVRT; 2) patients in whom atrial fibrillation results in ventricular fibrillation and; 3) patients with Ebstein anomaly [<xref ref-type="bibr" rid="R9">9</xref>].</p><p>Dual AV nodal physiology is known to occur in 8-40% of patients with AP, leading to a variety of possible reentrant circuits [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R11">11</xref>]. In the study of Csanadi Z et al [<xref ref-type="bibr" rid="R12">12</xref>]., the most common arrhythmia in these patients were AVRT without AVNRT (75%), whereas 19% of patients had both AVRT and AVNRT, and 6% had only AVNRT. They also demonstrated that absence of clinical tachycardia during follow up in those with only dual AV nodal physiology with or without single echo cycle would argue against the routine ablation of the SP in these patients. At times, one tachycardia changes into another as depicted in an interesting case report [<xref ref-type="bibr" rid="R13">13</xref>].</p><p>In the study of 176 consecutive patients who underwent SP ablation for AVNRT, Sticherling et al [<xref ref-type="bibr" rid="R4">4</xref>]. demonstrated that 15% of patients with AVNRT, without prior clinically documented AT, were found to have inducible AT. Inducible AT was non-sustained in 74% of patients and induced only under isoproterenol infusion in the most of patients (74%). The sustained episodes terminated spontaneously after 30 seconds in most of the patients. In this study, less than 10% of patients with inducible AT had recurrent AT in follow up period following SP ablation. This observation suggest that inducible AT, even when monomorphic and reproducibly inducible, often may be a non-specific finding that doesn't have clinical significance. Thus, therapy directed at the AT in patients who did not have documented prior AT should be deferred and limited to the occasional patients who later develop symptomatic AT.</p><p>Because of easy and reproducible induction of AVNRT after successful ablation of bypass tracts and absence of clinically documented tachycardia compatible with one of the specific types of SVT, we decided to ablate SP in this patient. Because the majority of patients with inducible AT without clinically sustained AT do not experience symptomatic atrial tachycardia during follow-up [<xref ref-type="bibr" rid="R4">4</xref>], ablation of AT was not attempted in this patient. The 8-month follow-up of our patient confirmed this decision.</p></sec><sec sec-type="" id="s4"><title>Conclusion</title><p>Our report demonstrated: 1) complexity of symptom origin in the patients with recurrent palpitations, 2) the importance of detailed electrophysiological assessment in the patients with AP before and after ablation, and 3) feasibility and efficacy of radiofrequency catheter ablation of multiple SVTs in single session.</p></sec>
Role of Biatrial Pacing in Prevention of Atrial Fibrillation after Coronary Artery Bypass Surgery	Could not extract abstract	<contrib contrib-type="author"><name><surname>Eslami</surname><given-names>Massoud</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Mirkhani</surname><given-names>Hamid S.</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Sanatkar</surname><given-names>Mehdi</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Bayat</surname><given-names>Homeira</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Sattarzadeh</surname><given-names>Roya</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib><contrib contrib-type="author"><name><surname>Mirhoseini</surname><given-names>Mahmood</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec sec-type="intro" id="s1"><title>Introduction</title><p>Atrial fibrillation (AF) after coronary artery bypass surgery (CABG) constitutes the most common sustained arrhythmia [<xref ref-type="bibr" rid="R1">1</xref>, <xref ref-type="bibr" rid="R2">2</xref>]. Postoperative atrial fibrillation occurs in 10% to 40% of patients undergoing open-heart surgery [<xref ref-type="bibr" rid="R3">3</xref>, <xref ref-type="bibr" rid="R4">4</xref>]. Postoperative AF is associated with stroke and a prolonged hospital stay [<xref ref-type="bibr" rid="R5">5</xref>]. Although AF rarely cause serious problem, it can cause considerable discomfort for the patient [<xref ref-type="bibr" rid="R6">6</xref>]. Similary, there is controversy regarding methods of preventing postoperative AF. Beta-adrenergic receptor blockers and Class III antiarhythmic agents such as Amiodarone confer some benefits in preventing post-CABG AF, but its incidence is still as high as 30% with these therapies [<xref ref-type="bibr" rid="R7">7</xref>, <xref ref-type="bibr" rid="R8">8</xref>]. Preoperative and postoperative digoxin was recommended as an effective preventive method by one group [<xref ref-type="bibr" rid="R9">9</xref>]but it was rejected by others [<xref ref-type="bibr" rid="R10">10</xref>]; while in one trial, only the combination of beta-blocker and digoxin was found to be effective [<xref ref-type="bibr" rid="R11">11</xref>]. The pathogenesis of postoperative AF remains unclear and is presumably multifactorial. Fuller’s study showed the close association with the patient’s age and male gender [<xref ref-type="bibr" rid="R3">3</xref>]. Multivariate analysis showed no relationship with the aortic cross-clamp time, the volume of cardioplegia, the number of grafts, the presence of postoperative infarct, or the postoperative CK_MB level [<xref ref-type="bibr" rid="R12">12</xref>], although two studies have found a relationship between postoperative AF and the length of the operation [<xref ref-type="bibr" rid="R13">13</xref>]. Previous studies demonstrated that arrhythmia was caused by operative damage to the atrial myocardium, so one would expect AF to occur immedietly after the operation. By contrast, AF develops most frequently on the second postoperative day. These observations suggest a different mechanism, such as inflamatory response with atrial edema, pericarditis, or reperfusion injury, rather than a direct ischemic insult [<xref ref-type="bibr" rid="R6">6</xref>, <xref ref-type="bibr" rid="R14">14</xref>]. The transient nature of this problem when seen after cardiac surgery suggests a reversible trigger; abnormal automaticity and atrial conduction delay are possible electrophysiological substrates. These would result in the occurrence of atrial ectopy and prolonged atrial activation, with lengthening of the P wave recorded by the ECG [<xref ref-type="bibr" rid="R7">7</xref>]. However, a signal averaged ECG of the P wave, which is a measure of regions of delayed atrial activation, is only moderately sensitive in predicting AF after CABG [<xref ref-type="bibr" rid="R15">15</xref>]. Signal-average ECG P-wave dispersion has also been recently advocated as a novel measurement of the heterogeneity of atrial depolarization [<xref ref-type="bibr" rid="R16">16</xref>].</p><p>Biatrial pacing has been shown to be effective in preventing AF after CABG [<xref ref-type="bibr" rid="R17">17</xref>, [<xref ref-type="bibr" rid="R18">18</xref>]. The purpose of this prospective study was to evaluate the efficacy of biatrial pacing as a prophylactic measure against AF after CABG when compared with no (control) or single- site atrial pacing.</p></sec><sec sec-type="materials\|methods" id="s2"><title>Materials and Methods</title><p>From July 2003 to May 2004, 120 patients that candidate for CABG were enrolled. The study design was approved by the Imam Khomeini Medical Center Committee and was a double-blind protocol in which the surgical staff and principal investigators and the patient were unaware of the assigned pacing modality. Clinical data, lead parameters, and Holter data were collected and recorded into the database by independent blinded investigators. Patients were randomly assigned in a double-blind fashion immediately after surgery to 1 of 3 pacing modes: biatrial pacing (BAP), left atrial pacing (LAP), and no pacing (control). Inclusion criteria consisted of informed consent, age > 50 years, elective heart surgery requiring cardiopulmonary bypass, and normal sinus rhythm. Exclusion criteria included participation in another investigational protocol, presence of a permanent pacemaker, or use antiarrhythmic therapy other than beta-blocker, history of supraventricular (including atrial flutter or AF) or ventricular tachyarrhythmias, redo-operation, patients with valvular heart surgery, and cardiogenic shock. All medications, including beta-blocker and digitalis, were continued until surgery. Postoperatively, these medications were used according to clinical indications. Coronary artery bypass surgery was performed in all patients. Patients underwent CABG on standard cardiopulmonary bypass with myocardial protection provided by blood cardioplegia</p><p>After completion of the surgical procedure, 2 temporary unipolar epicardial leads (model 6500, Medtronic, Inc) were attached to the anterior-superior aspect of the right atrium, and a second pair of epicardial leads was attached to the posterior-inferior aspect of the left atrium between the coronary sinus and the right inferior pulmonary vein [<xref ref-type="bibr" rid="R19">19</xref>]. The lead was a multifilament, braided, stainless steel. This electrode was attached to a curved needle with polypropylene suture, and the proximal end of the suture was coiled to reduce dislodgment. Biatrial pacing was achieved by simultaneous bipolar pacing of the right and left atria. In patients assigned to LAP group the left atrial pair of unipolar leads was connected to an external pacemaker generator (Pacesetter, Inc), and the proximal ends of the right atrial epicardial pacing electrodes were taped to the chest wall. Sensitivity of the pacing was set at 0.25 mV. The lowest rate was 90 beats per minute and the maximum pacing rate allowed was 120 beats per minute. Overdrive pacing was continued for 4 days, with continuous telemetry monitoring. The pacing and sensing thresholds were checked daily, and the output was adjusted accordingly. The 12-lead ECG was performed daily for 4 days at baseline and during pacing. The pacing wires were removed by simple transcutaneous retraction by 6 day in the absence of a clinical end point. Patients were re-evaluated in the cardiac surgery outpatient clinic 4 weeks after surgery.</p></sec><sec sec-type="results" id="s3"><title>Statistical Analysis</title><p>Continuos variables were expressed as mean ± SD. Continuous variables were compared by means of ANOVA tests and discrete variables were compared using the χ<sup>2</sup> test. p<0.05 was considered statistically significant.</p></sec><sec sec-type="results" id="s4"><title>Results</title><p>The patients in this study comprised 40 in BAP group, 40 in LAP group, and 40 patients in control group, respectively. The mean age in each group were BAP, 61.2±6.2; LAP, 62.6±7.2; and control, 60±7.5; p=0.1; 60% in BAP, 57.5% in LAP and 60% in control group were male. All clinical characteristics in each group were similar and well matched (<xref ref-type="table" rid="T1">Table 1</xref>).</p><p>Beta-blocker administration before and after operation and mean maximum sinus rate per day was not statistically different. The prevalence of postoperative atrial fibrillation was significantly less in the patients randomized to BAP group when compared with the other two remaining groups (<xref ref-type="fig" rid="F1">Figure 1</xref>).</p><p>An episode of atrial fibrillation occurred in 7 (17.5%) of 40 patients in the BAP group compared with 12 (30%) of 40 patients in the LAP group (P=0.04), and 18 (45%) of 40 patients in control group (p=0.02). The first postoperative episode of atrial fibrillation occurred 2.5±1.3 days after surgery in LAP group, 2.4±1.6 days after surgery in control group, and 2.8±0.7 days afetr surgery in BAP group (p=0.5). The mean duration of atrial fibrillation were 6.9±4.2 hours in BAP, 6.1±2.8 hours in LAP and 7.2±1.6 in control group, (p=0.3).</p><p>If AF was not converted spontaneously to sinus rate (SR) in 48 hours either pharmacological means or electrical cardioversion was used to restore SR before discharge. The mean length of hospital stay was most significantly reduced in BAP group (6.1±1.2 versus 9.0±4.1 days in the control groups; p=0.01, and 8.7±1.3 days in LAP groups; p=0.02). The mean length of stay in the intensive care unit was also significantly reduced in the BAP group (2.8±0.7 versus 4.6±4.5 days in control group; p=0.04, and 4.2±3.2 days in LAP group; p=0.01). Cerebral events (including stroke or transient ischemic attacks) occurred in 4 patients and there was no significant difference between three groups. The mean hospital charges in the BAP was 6250$, in LAP was 6820$ and in control group was 7400$, (p=0.3). Postoperative complications in each group are presented in <xref ref-type="table" rid="T2">Table 2</xref>.</p></sec><sec sec-type="results" id="s5"><title>Discussion</title><p>Previous studies demonstrated that use of beta-blockers and sotalol have beneficial effect in prevention of postopertive AF [<xref ref-type="bibr" rid="R5">5</xref>, <xref ref-type="bibr" rid="R21">21</xref>]. Daoud and associates showed that preopertive amiodarone therapy was effective in reduction of postoperive AF [<xref ref-type="bibr" rid="R7">7</xref>]. Medical therapy as a prophylactic agent against post-CABG AF, may be limited by other medical disease, such as asthma, thyroid dsyfunction, or liver function derangement [<xref ref-type="bibr" rid="R20">20</xref>, <xref ref-type="bibr" rid="R22">22</xref>]. Biatrial pacing has been shown to prevent of AF recurrence in patients with paroxysmal AF [<xref ref-type="bibr" rid="R18">18</xref> –<xref ref-type="bibr" rid="R20">20</xref>, <xref ref-type="bibr" rid="R23">23</xref> –<xref ref-type="bibr" rid="R26">26</xref>]. Another study demonstrated that single-site pacing has not been effective in patients with AF [<xref ref-type="bibr" rid="R27">27</xref>]. Our study reported the comparision of biatrial pacing with single-site atrial pacing and no pacing. Biatrial pacing prevents AF by two mechanism: a) The common cause of initiation of AF is premature atrial beat, especialy during sinus bradycardia. Biatrail pacing at a relatively high rate may result in supperession of atrial ectopy. b) Atrial conduction delay and dispresion of atrail refractoriess serve as a predictor for reentry and initiation of AF [<xref ref-type="bibr" rid="R7">7</xref>,<xref ref-type="bibr" rid="R19">19</xref>]. In our study, a triggered pacing mode was chosen to assure early activation of the atrial myocardium near the coronary sinus in response to premature atrial conduction sensed in either the right or left atria and hence reduce atrial dispresion.</p><p>Taylor et al showed that the most expensive complications post-CABG were respiratory failure and sternal wound infection, but occurred in only 3% and 0.4% of patients, respectively. However, AF was least expensive but most common complication, occurred in 20% of patients [<xref ref-type="bibr" rid="R28">28</xref>]. In our study hospital stay and hospital charges were significantly reduced by biatrial pacing, compared to single site or no pacing groups. Despite these benefits this technique was not associated with side effects. Identifying patients at risk for developing post-CABG AF and using biatrial pacing may be the optimal effective strategy.</p></sec><sec sec-type="results" id="s6"><title>Conclusion</title><p>AF is commonly encountered post-CABG and, results in an increased length of hospital stay and cost. An ideal prophylactic approach is one that is effective in a diverse patient population and that is associated with minimal expense and risk[<xref ref-type="bibr" rid="R19">19</xref>]. Biatrail pacing may be such a technique, and is more effective in preventing of post-CABG AF than single-site atrial pacing and no pacing group and it results in a shortened length of hospitalization. This technique is not associated with a risk of ventricular arrhythmia, bradycardia, or hypotension, unlike antiarrhythmic agents. Identifying patients at risk for developing post-CABG AF and using biatrial pacing may be the optimal effective strategy.</p></sec>
Cryothermal Energy Ablation Of Cardiac Arrhythmias 2005: State Of The Art	<p>At the time of antiarrhythmic surgery, cryothermal energy application by a hand-held probe was used to complement dissections and resections and permanently abolish the arrhythmogenic substrate. Over the last decade, significant engineering advances allowed percutaneous cryoablation based on catheters, apparently not very different from standard radiofrequency ablation catheters. Cryothermal energy has peculiar characteristics. In fact, it allows testing in a reversible way the effects of energy application at higher temperature, before producing a permanent lesion at -75°C. Moreover, slow formation of the lesion allows timely discontinuation of the application, as soon as inadvertent modifications of normal atrioventricular conduction are observed during ablation in the proximity of atrioventricular node and His bundle, avoiding its permanent damage. Over the last years, percutaneous cryothermal ablation has been widely used for a variety of cardiac arrhythmias. From the data gathered, it is unlikely that cryoablation will replace standard ablation in unselected cases. Nevertheless, for the above mentioned peculiarities, cryothermal ablation has proved very effective and safe for ablation of arrhythmogenic substrates close to the normal conduction pathways, becoming the first choice method to ablate anteroseptal and midseptal accessory pathways. It can be also the best treatment for ablation of the slow pathway to abolish atrioventricular node reentrant tachycardia in pediatrics or when particular anatomy of the Koch’s triangle is observed. Cryothermal ablation of the pulmonary veins for atrial fibrillation, although longer than radiofrequency ablation, is not associated with pulmonary vein stenosis and is expected to be less thrombogenic; new catheter designs for cryothermal ablation of this challenging arrhythmia are to be tested to assess their efficacy and clinical usefulness.</p>	<contrib contrib-type="author"><name><surname>De Ponti</surname><given-names>Roberto</given-names></name><degrees>MD</degrees></contrib><aff>Department of Cardiovascular Sciences, Ospedale di Circolo e Fondazione Macchi, University of Insubria, Varese, Italy</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>From surgical to catheter-based cryoablation</title><p>In the 80’s, epicardial cryoablation was introduced in antiarrhythmic surgery for ablation of accessory pathways by Klein et al [<xref ref-type="bibr" rid="R1">1</xref>]. A hand-held cryoprobe was applied at the site where intraoperative mapping localized the arrhythmogenic substrate. The probe was refrigerated to -60°C and its effectiveness was evaluated during continuous monitoring of the cardiac electrical activity. This method, largely used in the past [<xref ref-type="bibr" rid="R2">2</xref>-<xref ref-type="bibr" rid="R4">4</xref>] for antiarrhythmic surgery and still in use for the treatment of some forms of ventricular tachycardia [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R6">6</xref>], is safe and effective. During surgery, cryomapping allowed precise localization of the arrhythmogenic substrate by monitoring the effect of a cryothermal energy application with higher temperature (0 to -15°C) for a limited time (15-30s), before producing a permanent lesion at -65°C only in the most appropriate site [<xref ref-type="bibr" rid="R7">7</xref>].</p><p>In the 90’s, significant engineering advances allowed the development of systems for percutaneous cryoablation, consisting of a steerable catheter, apparently not very different from standard ablation catheters for radiofrequency energy delivery, and a dedicated console (<xref ref-type="fig" rid="F1">Figure 1</xref>). Fluid nitrous oxide is delivered under pressure to the catheter tip through a hollow injection tube, which runs internally for the whole length of the catheter. In a small chamber inside the tip electrode, nitrous oxide is made expand and a liquid to gas phase change takes place with heat extraction from the electrode-to tissue interface. The gas is constantly removed through a second coaxial lumen inside the catheter, under vacuum. The tip temperature is constantly monitored by the console, which in turn adjust the nitrous oxide flow to obtain and maintain the preset temperature. Two systems for cryoablation are currently available. The first is provided by Cryocath Technologies Inc. (Montreal, Canada) and utilizes 7 or 9 F steerable catheters with 4, 6 or 8 mm long tip electrode. The ablation catheter is connected to a dedicated console, which has two algorithms available: 1) for cryomapping with slow decrease of the temperature to - 30°C up for 80 s; 2) for cryoablation with faster decrease of the temperature to -75°C for up to 480 s. In any case, the target temperature can be manually preset on the console at any value between -30 and -75°C. The second system (CryoCor Inc., San Diego, California, USA) has 10 F steerable catheters with 6.5 or 10 mm long tip electrodes. The console has a built-in closed loop pre-cooler for the fluid nitrous oxide, whose flow at the catheter tip is adjusted during the application to maintain a temperature of -80°C.</p></sec><sec sec-type="" id="s2"><title>Lesion formation by cryothermal energy</title><p>Since cryothermal energy has been widely used in surgery, the types of cellular lesion caused by tissue freezing are well known [<xref ref-type="bibr" rid="R8">8</xref>]. The mechanisms underlying lesion formation by cryoenergy are two-fold: 1) a direct cell injury and 2) a vascular mediated tissue injury.</p><p>The direct cellular injury is due to ice formation, which has different distribution according to temperature reached during cooling. By cooling to mild temperature (0 to -20°C), ice forms only extracellularly. Consequently, extracellular environment becomes hyperosmotic and an intracellular to extracellular water shift occurs. This causes cellular shrinkage and damage to membrane. Cooling to these temperature may result in cellular death, if the application is enough prolonged. Using short applications with limited temperature, the effect produced on the cell is reversible and cellular function recovers, although minimal cellular damage may be produced. In the clinical use, the option of producing a functionally reversible lesion is quite attractive to test the effect of cryoablation without producing a permanent lesion. Conversely, by cooling down to -40°C and further, intracellular water freezes and formation of intracellular ice results in major and irreversible disruption of organelles and cell membrane with cellular death. Intracellular ice may propagate from one cell to another via intercellular channels.</p><p>The second mechanism underlying lesion formation by cryothermal energy delivery is a vascular-mediated mechanism. In fact, the initial tissue response to cooling is vasoconstriction with decreased blood flow. As tissue freezes, circulation ceases uniformly in the frozen tissue. The uniformity of cell death in a lesion produced by cryothermal energy has suggested ischemic necrosis as the main mechanism for tissue death, although it is impossible to distinguish the tissue damage caused by this mechanism from the one produced by intracellular ice formation. Upon re-warming, a hyperemic response is observed with increased vascular permeability and edema formation. Other than producing increased permeability and edema, endothelial damage results in platelet aggregation and micro-thrombus formation, with stagnation of microcirculation in about 30-45 min.</p><p>Especially in the percutaneous closed chest cryoablation, the effect produced by energy application is the result of a temperature gradient occurring at the electrode/tissue interface and possibly influenced by different factors, such as contact or blood flow. At the interface, the coldest area is the one adjacent to the catheter tip, where functional effects of energy delivery are observed earlier. Conversely, the less cooled area is the one at the periphery of the cryolesion, whose dimensions may also vary according to the duration of freezing. Due to limited (both in time and temperature) cooling of outer limit of the lesion, reversible tissue damage is more likely to occur in this area. As a consequence, the effects obtained late during cryothermal energy application are likely to revert early upon re-warming and, therefore, any expected functional modification induced by cryoenergy should occur early (usually within the first 30 s of the application) in order to obtain a successful and permanent ablation of a given arrhythmogenic substrate.</p></sec><sec sec-type="" id="s3"><title>Cryothermal vs radiofrequency energy: differences and their clinical implications</title><p>In closed chest cryoablation, the effect of cryothermal energy application greatly depends on the minimum temperature reached, the application duration and the temperature time constant [<xref ref-type="bibr" rid="R9">9</xref>]. The latter value indicates the course of the descent of temperature to the target temperature and a shorter value (expressed in seconds) identifies a more effective application. Due to intrinsic characteristics of cryothermal energy at a fixed minimum temperature, the lesion forms more slowly than the one produced by hyperthermic injury. This has two practical implications. The first is that the application duration for cryoablation is significantly longer than for radiofrequency energy and a lesion produced by cryothermal energy by a 4 mm tip 7 F cryocatheter at -75°C for 240 s has a comparable depth to the one obtained by radiofrequency energy applied in temperature control mode at 50 W, +70°C for 60 s [<xref ref-type="bibr" rid="R10">10</xref>]. Second, the longer estimated time required to create a permanent lesion may be clinically useful to better modulate the lesion formation in critical areas (i.e. close to the atrioventricular node-His bundle). In these cases, if inadvertent modifications of conduction over the normal pathways is observed during the application, immediate discontinuation of cryothermal energy application results in return to baseline conduction properties, with no permanent damage to normal conduction. Recently, it has been demonstrated in a canine model [<xref ref-type="bibr" rid="R10">10</xref>] that cryolesions are associated with significantly less endothelial disruption and overlying thrombus formation as compared to lesion produced by radiofrequency energy, regardless of the preventive use of aspirin. This characteristic could be very important, especially when multiple and prolonged energy applications are required to treat atrial arrhythmias in the left atrium. Unlike lesions produced by a hyperthermic injury, cryolesions show both in open chest [<xref ref-type="bibr" rid="R11">11</xref>] and closed chest [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R12">12</xref>] models, a well demarcated border zone and preservation of the extracellular collagen matrix with no collagen denaturation, nor contracture related to hyperthermic effects. These histologic observations combine with the clinical evidence that cryothermal energy application adjacent to coronary arteries, as well as in venous vessels (coronary sinus, middle cardiac vein and pulmonary veins) [<xref ref-type="bibr" rid="R12">12</xref>-<xref ref-type="bibr" rid="R15">15</xref>], does not results in damage nor chronic stenosis of their lumen.</p><p>In the already mentioned study [<xref ref-type="bibr" rid="R10">10</xref>], it has been pointed out that lesions by radiofrequency energy have a comparable depth of those produce by cryoablation. Nevertheless, radiofrequency energy ablation resulted in a highly significantly greater area and a nearly significantly larger volume when compared to cryolesions. Moreover, colder temperatures were associated with deeper lesions and greater area and volume were associated with use of 9 F as compared to 7 F catheters. It is not clear why cryothermal energy produces a lesion with similar depth, but smaller area and volume, when compared to the one produced by a “similar” radiofrequency energy application. One plausible explanation could be the “cryoadherence” effect, which is a tight adherence of the catheter tip to the adjacent tissue caused by cooling. Due to this effect, the lesion produced is very focal, since the “brushing” of the tip electrode, usually observed during radiofrequency ablation, is missing. On the other hand, the advantage of this effect is two-fold. First, below the critical temperature of -20°C it allows good tip electrode to tissue contact, which persists throughout the whole application and it is not dependent on the torsion/deflection manoeuvres applied to the catheter. It is well known that a fixed and stable contact during the whole application is essential, especially for ablation in proximity of critical areas, such as atrioventricular node and His bundle. Second, the cryoadherence effect allows safe continuation of the application, even when sudden changes in heart rhythm that usually displace the ablation catheter (such as tachycardia termination or pacing) occur. Moreover, cryoadherence does not compromise safety, since, upon discontinuation of cryothermal energy delivery, the defrost phase is very fast (within 3 s) and the catheter can be immediately disengaged from the ablation position.</p><p>Finally, another peculiarity of cryothermal energy is the complete absence of patient symptoms in almost every case, in spite long-lasting applications. In our experience, we have evaluated patient perceptions during cryothermal energy application in a series of non-sedated cases. In almost all cases, the absolute absence of perception was demonstrated by the fact that the patient was unable to tell when the application was started and discontinued. Only in some cases, when multiple and prolonged applications are delivered in the left heart, a light sense of cold or headache is perceived as minor discomfort. Although the full explanation of the absence of symptoms is not completely clear, this characteristic can be particularly useful in young as well as in paediatric patients.</p></sec><sec sec-type="" id="s4"><title>Clinical use in ablation of cardiac arrhythmias</title><p>Over the last five yeas, the world-wide experience in catheter ablation of cardiac arrhythmias by using cryothermal energy has increased unabated. Based on this experience, cryoablation should not be viewed as a replacement for radiofrequency energy, which will continue to be the method of choice in many clinical situations. Nevertheless, ablation by cryothermal energy should be rather considered as a useful addition to the electrophysiologist’s armamentarium. In fact, different types of arrhythmias are now successfully treated by cryoablation and in some cases, especially in proximity to normal conduction pathways, treatment by this energy source is considered the first choice therapy for its safety and efficacy. The following is a brief analysis of the experience in cryoablation for each of the considered arrhythmias.</p><sec id="s4a"><title>Atrioventricular nodal reentrant tachycardia</title><p>So far, slow pathway ablation for atrioventricular nodal reentrant tachycardia by cryothermal energy represents the numerically larger experience in the clinical application of this new technology. Unlike radiofrequency energy, accelerated junctional rhythm is not observed during cryothermal energy application on the slow atrioventricular node pathway. Therefore, the only marker of effective ablation is suppression of tachycardia inducibility during initial cooling (<xref ref-type="fig" rid="F2">Figure 2A-B</xref>). Accordingly, baseline non inducibility of the arrhythmia may be a limitation to the application of this technique. From the early report [<xref ref-type="bibr" rid="R16">16</xref>], several papers have contributed to accumulating experience in slow pathway ablation, with a satisfactory success rate and a recurrence rate varying from 6 to 9.7% [<xref ref-type="bibr" rid="R17">17</xref>-<xref ref-type="bibr" rid="R22">22</xref>]. In the “Frosty” trial [<xref ref-type="bibr" rid="R19">19</xref>], a multicentric prospective trial performed in the United States, 103 patients with atrioventricular nodal reentrant tachycardia were enrolled. On an intention-to-treat basis, the acute procedural success was 91% with no device-related complications and a recurrence rate of 6% in a 6 month follow-up. Cryomapping proved useful to predict the site of successful ablation. Nine patients had inadvertent modifications of the conduction over normal atrioventricular conduction pathways, including first to third degree atrioventricular block and right bundle branch block. These all resolved completely, usually within a minute or less and had no sequelae. A database gathering the worldwide experience and based on a combination of registry and prospective trial data reports no case of permanent atrioventricular block following cryothermal ablation of the slow pathway in more than 300 patients with atrioventricular nodal reentrant tachycardia [<xref ref-type="bibr" rid="R20">20</xref>]. Temporary first degree or higher atrioventricular block, observed in 15 cases (4.3%) during cryomapping at -30°C or during cryoablation at -75°C, was always reversible. Recently, the results of the first two prospective randomized trials on transvenous cryoablation versus radiofrequency ablation of the slow pathway for treatment of atrioventricolar nodal reentrant tachycardia have been published [<xref ref-type="bibr" rid="R21">21</xref>,<xref ref-type="bibr" rid="R22">22</xref>]. In these studies, cryoablation proved as effective and safe for the cure of atrioventricular nodal reentrant tachycardia as radiofrequency ablation. The higher recurrence rate that may be observed in the cryoablation group [<xref ref-type="bibr" rid="R21">21</xref>] suggests that, unlike radiofrequency ablation, prolonged energy applications and postablation waiting time are necessary when cryothermal ablation is used to minimize recurrence in the follow-up. In our own experience, we treat atrioventricular nodal reentrant tachycardia by slow pathway cryoablation in patients with particular anatomic characteristics, refractory to standard radiofrequency energy ablation or in pediatrics. Especially in cases with difficult anatomy, such as a small or distorted Koch’s triangle, the characteristics of cryothermal energy allow test of the ablation effect in particularly risky sites without producing irreversible damage to atrioventricular conduction, if the application is timely interrupted. In some complex cases, we found it necessary to resort to longer cryothermal energy applications (up to 480 s) and to prolong postablation observation up to 60 min. In a patient with atrioventricular nodal reentrant tachycardia, who underwent multiple unsuccessful ablation of the slow pathway, we decided to target the fast pathway by cryothermal energy [<xref ref-type="bibr" rid="R23">23</xref>]. In this particular case, selective and safe ablation of the fast pathway at the apex of the Koch’s triangle was accomplished and this resulted in permanent cure of the arrhythmia.</p><p>According to the presented data, cryothermal energy is a valuable and useful alternative to radiofrequency energy to treat patients with atrioventricular nodal reentrant tachycardia. Absence of permanent inadvertent damage of atrioventricular conduction makes this new technology particularly useful in cases with difficult anatomy, unsuccessful prior standard ablation procedure, in pediatrics and in all cases, in whom even the lower risk of atrioventricular block still possible with radiofrequency energy, is considered unacceptable.</p></sec><sec id="s4b"><title>Accessory pathways</title><p>In <xref ref-type="table" rid="T1">Table 1</xref>, published data on cryothermal ablation of anteroseptal (parahissian) and midseptal accessory pathways are reported [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R24">24</xref>-<xref ref-type="bibr" rid="R29">29</xref>]. As shown, this technique in anterospetal and midseptal areas, both at high risk of complete permanent atrioventricular block when standard radiofrequency energy in performed, is highly safe and successful. In the larger series, success rate is above 90%. Although transient modifications of the normal atrioventricular nodal conduction pathways are observed during cooling, no permanent modifications is observed with the only exception of right bundle branch block in 2 cases in a single centre. In fact, immediate discontinuation of cryothermal energy application at any temperature upon observation of modification of conduction over normal pathways results in return to baseline condition, soon after discontinuation. Resumption of accessory pathway conduction with palpitation recurrences may occur in the follow up to 20%, but, especially in young healthy individual, a recurrence is by far more acceptable than permanent complete atrioventricular block requiring pacing, which was invariably the case in many series of radiofrequency ablation of these pathways. In our experience, we have treated 18 patients with anteroseptal or midseptal accessory pathways, so far, age ranging 11-51 years. No patient was excluded from the study for proximity of the accessory pathway to the normal conduction pathways (<xref ref-type="fig" rid="F3">Figure 3</xref>). Successful ablation was obtained in all, but 1 pediatric and asymptomatic patient, in whom conduction properties over the accessory pathway indicated ablation, which was eventually postponed. Cryoadherence effect proved very useful in every case, but especially when energy delivery was performed during orthodromic atrioventricular tachycardia, to better visualize the His bundle electrogram and to monitor conduction over normal pathways. No complication or palpitation recurrences were observed during a 17±10 month follow-up. In approaching anteroseptal and midseptal accessory pathways, instead of performing “cryomapping” at -30°C in the selected site, we found it useful to test cryothermal energy applications with a step-by-step method to decrease temperature. In fact, in the most suitable site with the best contact (sometimes, a superior vena cava approach via a subclavian or a brachial vein is useful to stabilize contact), test applications are applied for 30 s, initially with a temperature of -30°C. If this test application is successful with no modification of normal conduction, then transition to ablation at -75°C up to 480 s is made. If the test application is unsuccessful, after re-warming, further 30s applications are tested, decreasing for each application the temperature by 10°C every step, up to the last application at -70°C. This is because we observed that the amount of cryothermal energy required for permanent ablation is quite individual (ranging from an application of -40°C for 40s to an application of -75°C for 480 s) and limiting test applications to only -30°C may limit the applicability of cryoablation in these patients. On the other hand, the use of cryothermal energy at temperatures lower than -30°C should be considered safer than radiofrequency energy in these critical sites.</p><p>Cryoablation can be also successfully and safely used to ablate selected cases of epicardial left-sided accessory pathways within the coronary sinus, well beyond the middle cardiac vein, once attempts by using both transseptal and transaortic approach have failed [<xref ref-type="bibr" rid="R20">20</xref>] [<xref ref-type="bibr" rid="R30">30</xref>]. Similarly, safe and successful cryothermal energy ablation of permanent junctional reciprocating tachycardia has been reported in children, in the midseptal region, at the coronary sinus os or in the middle cardiac vein [<xref ref-type="bibr" rid="R31">31</xref>].</p><p>The experience of cryoablation in unselected accessory pathways is more limited and less satisfactory [<xref ref-type="bibr" rid="R20">20</xref>]. Of 51 accessory pathways with various locations, only 69% were successfully ablated and this value is considerably lower than the one reported for radiofrequency ablation. There are many possible explanations for this including the learning curve and the smaller size of the lesion produced by cryoablation [<xref ref-type="bibr" rid="R10">10</xref>]. In any case, all the peculiarities of cryothermal energy, which are optimal for septal ablation, are less important or even useless for ablation of accessory pathways located elsewhere.</p></sec><sec id="s4c"><title>Focal atrial tachycardia and isthmus-dependent atrial flutter</title><p>Occasionally, successful cryoablation of focal atrial tachycardia has been reported and its safety has been confirmed also for ablation of atrial foci located close to the atrioventricular node [<xref ref-type="bibr" rid="R32">32</xref>].</p><p>Several papers have reported cryoablation of the cavotricuspid isthmus for typical atrial flutter with an acute and long-term success comparable to the one of radiofrequency ablation [<xref ref-type="bibr" rid="R33">33</xref>-<xref ref-type="bibr" rid="R36">36</xref>]. The use of larger catheter and longer electrode for ablation in this area is associated with a lower number of applications and a shorter procedural time. As for radiofrequency ablation, a case of transient ST segment elevation in the inferior leads was observed during cryo application at the septal isthmus, with wall irregularities in the right coronary artery without significant stenosis [<xref ref-type="bibr" rid="R35">35</xref>]. The major advantage of using cryothermal energy to produce bidirectional conduction block of the cavo-tricuspid isthmus is the absence of pain perception related to energy application. In a prospective randomized trial in which a visual analogue scale to evaluate pain was used, pain perception was by far lower if not existent in the cryothermal as compared to radiofrequency energy group [<xref ref-type="bibr" rid="R34">34</xref>].</p></sec><sec id="s4d"><title>Pulmonary vein ablation for atrial fibrillation</title><p>When radiofrequency energy is applied at the os of the pulmonary veins to prevent atrial fibrillation recurrences, a heat-induced contraction of the pulmonary vein wall can be observed early or during the follow-up, which results in a variable degree of lumen reduction and a wide spectrum of clinical presentations [<xref ref-type="bibr" rid="R37">37</xref>]. This reaction is typical of hyperthermic injury and results from a combination of edema, endothelial disruption and collagen denaturation and shrinkage [<xref ref-type="bibr" rid="R38">38</xref>]. The occurrence and the degree of stenosis correlate with the amount of energy delivered [<xref ref-type="bibr" rid="R39">39</xref>] and lesion extension [<xref ref-type="bibr" rid="R40">40</xref>]. As mentioned above, cryothermal energy ablation causes less or minimal endothelial disruption, maintenance of extracellular collagen matrix and no collagen contracture related to thermal effects. Moreover, lower incidence of thrombus formation is reported with cryoenergy as compared to radiofrequency energy ablation. For these characteristics, cryothermal energy ablation can be considered an ideal and safer energy source also for pulmonary vein ablation and the incidence of both pulmonary veins stenosis and thromboembolic events is expected to be dramatically reduced by using cryoablation. On the other hand, the presence of high blood flow in the pulmonary vein may represent a considerable heat load, which may limit the size and depth of the lesion produced by cryothermal energy at the os of the pulmonary vein. Moreover, the longer time required to produce a permanent lesion may relevantly reflect on procedure duration, limiting the clinical use of this theoretically optimal energy source. Initial experiences of electrophysiologically-guided segmental ostial ablation of the pulmonary vein by using cryothermal energy application with 10 or 7 F catheters [<xref ref-type="bibr" rid="R20">20</xref>,<xref ref-type="bibr" rid="R41">41</xref>] have been reported. These experiences show that pulmonary vein isolation is feasible with a comparable number of applications and clinical outcome with regard to radiofrequency ablation; longer procedural times, observed for both the 10 F and the 7 F catheter, correlate with longer application times required when cryothermal energy is used. Importantly, the early cryoablation experience has not evidenced, so far, development of pulmonary veins stenosis following ablation. Technologic evolution is now aimed to develop new catheter designs for circumferential ostial ablation of the pulmonary veins, with the option of deploying in the pulmonary veins an inflatable balloon to reduce the heat load related to blood flow [<xref ref-type="bibr" rid="R20">20</xref>]. These devices are to be tested in a large patient cohort to assess whether these technological improvements will lead to optimization of the use of cryothermal energy, maximizing the advantages of this new technology and limiting the drawbacks encountered in its clinical use.</p></sec><sec id="s4e"><title>Ventricular arrhythmias</title><p>Although clinical data on cryothermal ablation of ventricular arrhythmias are missing, preliminary experimental evidence shows that percutaneous cryoablation in several sites of normal ventricular myocardium is feasible with lesion deeper in the left than in the right ventricle, probably due to better contact in the former than in the latter [<xref ref-type="bibr" rid="R42">42</xref>]. In the same study, cryothermal energy has been also tested to ablate sustained ventricular tachycardias in a post-infarction sheep model. A limited number of applications was effective in suppressing the inducibility of ventricular arrhythmias, producing a transmural lesion in the majority of the cases with no acute complication.</p><p>Interestingly, cryothermal energy could be used to target ventricular tachycardias of epicardial origin, once the epicardial space has been reached by the non surgical transpericardial approach, originally described by Sosa [<xref ref-type="bibr" rid="R43">43</xref>]. As compared to radiofrequency energy, cryothermal energy seems to be safer in the epicardium, due to less probable damage to epicardial coronary arteries. The reduced heat load in the pericardial space related to the absence of blood flow could be to the advantage of cryoablation in these cases, with the possibility to produce larger transmural lesions.</p></sec></sec>
May Fever Trigger Ventricular Fibrillation?	<p>The clinical precipitants of ventricular fibrillation (VF) remain poorly understood. Clinical factors such as hypoxemia, acidosis or electrolyte imbalance, drug-related toxicity, autonomic nervous system disorders as well as viral myocarditis have been proposed to be associated with sudden cardiac death particularly in patients with structural heart disease. However, In the Brugada syndrome, concurrent febrile illness has been reported to unmask the electrocardiographic features of the Brugada syndrome and be associated with an increased propensity for VF. More recently, a febrile illnesses of infectious etiology was associated to polymorphic ventricular tachycardia or VF in patients with normal hearts and without known repolarization abnormality. In this review we detail this phenomenon and its putative mechanisms.</p>	<contrib contrib-type="author"><name><surname>Pasquié</surname><given-names>Jean Luc</given-names></name><degrees>MD, PhD</degrees></contrib><aff>Hôpital Arnaud de Villeneuve, Montpellier, France</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Ventricular fibrillation (VF) is the main mechanism of sudden cardiac death. Clinical factors such as hypoxemia, acidosis or electrolyte imbalance, drug-related toxicity, autonomic nervous system disorders as well as viral myocarditis have been proposed to be associated with sudden cardiac death particularly in patients with structural heart disease [<xref ref-type="bibr" rid="R1">1</xref>]. Emerging evidence implicates triggers dominantly originating from the distal Purkinje arborization and the right ventricular outflow tract (RVOT) in the initiation of VF in a range of clinical conditions [<xref ref-type="bibr" rid="R2">2</xref>-<xref ref-type="bibr" rid="R7">7</xref>].</p><p>Isolated reports in patients with the Brugada syndrome suggest that febrile illnesses may unmask the characteristic electrocardiographic changes that then favour the development of VF [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R15">15</xref>]. However, the clinical precipitants for VF storms in patients with idiopathic VF, without evidence of structural heart disease or known repolarization abnormality, remain poorly characterized. Recently we reported a series of patients in whom a febrile illness precipitated idiopathic VF storm, suggesting that fever may be a precipitant of sudden cardiac death in patients with normal hearts [<xref ref-type="bibr" rid="R16">16</xref>].</p></sec><sec sec-type="" id="s2"><title>Fever and Brugada Syndrome</title><p>As summarized in <xref ref-type="table" rid="T1">Table 1</xref>, isolated case reports have suggested that the phenotypic behaviour of the Brugada syndrome may be influenced by body temperature [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R15">15</xref>]. These reports have observed that not only could the electrocardiographic features of Brugada syndrome manifest but also that it may then be associated with VF during concurrent hyperthermia [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R15">15</xref>]. Such presentations with concurrent febrile illness in the literature have all been males between the ages of 50 and 70 years. The electrocardiographic features of the Brugada syndrome in each case were noted at presentation and gradually returned to normal as fever ceased. However, the Brugada electrocardiographic pattern was not reproducible by pharmacologic testing in all patients (<xref ref-type="table" rid="T1">Table 1</xref>). Very recently, ajmaline testing was reported to be a valuable test in diagnosis of SCN5A carriers with sensitivity, specificity and positive and negative predictive values of 80, 94, 93 and 83 % respectively [<xref ref-type="bibr" rid="R16">16</xref>]. However, considering risk stratification for identifying high-risk patients in Brugada syndrome, SCN5A mutation, VF inducibility by EP study and pharmacologic challenge were not predictive factors of life-threatening arrhythmic events [<xref ref-type="bibr" rid="R17">17</xref>].</p><p>Morita et al described a 69 year old patient with prior unexplained resuscitated sudden death (5 years ago), who presented with ST segment elevation, T wave alternans and premature ventricular contractions in association with a respiratory infection and mild fever (37.5°C) [<xref ref-type="bibr" rid="R10">10</xref>]. In two more patients, one with known Brugada syndrome, a febrile illness was associated to syncope due to polymorphic ventricular tachycardias or VF [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R15">15</xref>]. Dinckal and al reported the dramatic case of a 55 year old male who experienced fever due to pneumonia which unmasked typical ST changes of Brugada syndrome and lead to VF refractory VF storm resulting in the patient’s death [<xref ref-type="bibr" rid="R14">14</xref>]. Of particular interest, Mok and colleagues reported the case of a patient whose Brugada ECG pattern was unmasked by hyperthermia secondary to acute cholangitis [<xref ref-type="bibr" rid="R12">12</xref>]. In order to explain this phenomenon, they have demonstrated by genetic screening a novel mutation in the SCN5A gene related to a single amino acid substitution (H681P) confirming the diagnosis of Brugada syndrome. This mutation was associated with the loss of transmembrane current and induced a pathologic phenotype that may be elicited by hyperthermia [<xref ref-type="bibr" rid="R12">12</xref>]. In vitro expression of this amino acid substitution revealed a shift in both steady-state activation and inactivation of the sodium channel resulting in a dramatic reduction of sodium window current [<xref ref-type="bibr" rid="R12">12</xref>]. This temperature-dependency of ionic mechanisms underlying Brugada syndrome have also been described in experimental studies [<xref ref-type="bibr" rid="R18">18</xref>]. They hypothetized that at physiological temperature ranges the gating of the sodium channel may be changed by the SCN5A mutation such that the outward current is dramatically increased in the early phases of the right ventricular action potential [<xref ref-type="bibr" rid="R18">18</xref>]. Using the patch-clamp technique to study the currents at 32° C they tested a missense mutation (Thr1620Met) in SCN5A in a mammalian cell line and demonstrated a faster decay and slower reactivation of the inward sodium current with a outward shift of the current flowing during phase I. This phenomenon may be responsible for a transmural voltage gradient that may cause the repolarization abnormalities and be responsible for phase 2 reentry [<xref ref-type="bibr" rid="R18">18</xref>]. This work favours the hypothesis that some mutations may exhibit temperature-dependency properties and may explain that some patients may be at particularly high risk during febrile illnesses [<xref ref-type="bibr" rid="R18">18</xref>,<xref ref-type="bibr" rid="R19">19</xref>].</p></sec><sec sec-type="" id="s3"><title>Fever and Idiopathic Ventricular Fibrillation</title><p>In contrast, a paucity of reports (<xref ref-type="table" rid="T1">Table 1</xref>) describes association of fever and ventricular fibrillation in patients with idiopathic VF. Unlike other polymorphic ventricular tachyarrhythmias, idiopathic VF is generally not related to stress [<xref ref-type="bibr" rid="R20">20</xref>], anger or physical activity [<xref ref-type="bibr" rid="R21">21</xref>] and spontaneous VF in these patients is initiated by premature ventricular complexes with very short coupling intervals [<xref ref-type="bibr" rid="R22">22</xref>].</p><p>In patients with structural heart disease, the clinical factors that may be associated with ventricular fibrillation are transient ischemia and reperfusion, systemic factors as hypoxemia, acidosis or electrolyte imbalance, neurophysiologic interactions or toxic effects [<xref ref-type="bibr" rid="R1">1</xref>]. In patients with normal hearts, viral myocarditis is generally considered one of the main mechanism of sudden cardiac death associated with a febrile illness; although this has been poorly documented and could only be demonstrated by biopsy. We recently reported 3 patients who had no demonstrable structural heart disease or repolarization abnormalities and who presented with episodes of VF storm during a peak of a febrile illness usually occurring at night [<xref ref-type="bibr" rid="R23">23</xref>]. They demonstrated short coupled ventricular ectopies (<xref ref-type="fig" rid="F1">Figure 1</xref>). Ectopy morphology was highly suggestive of left Purkinje origin in one, and of muscular origin from RVOT in the other one. In the third patient with recurrent episodes of VF, mapping and ablation was performed confirming the origin of the triggering ectopy from the right Purkinje system (<xref ref-type="fig" rid="F2">Figure 2</xref>). None of our patients had evidence to suggest a viral myocarditis with biological screens, troponin values and viral screens being normal. In addition, one of these patients had a documented bacterial urinary sepsis. A potential role for the autonomic nervous system cannot be excluded as VF occurred at night; however, none of our patients demonstrated a change in repolarization prior to VF storm.</p><p>The distal Purkinje arborization has been implicated as the site of origin of triggers initiating VF in a variety of clinical situations. These include patients with VF associated with structural heart disease such as following myocardial infarction [<xref ref-type="bibr" rid="R7">7</xref>] abnormal repolarization syndromes [<xref ref-type="bibr" rid="R4">4</xref>] and also those with idiopathic VF [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R3">3</xref>]. The Purkinje system while being capable of sustaining spontaneous activity by automaticity, reentry or triggered activity, is known to be sensitive to a variety of clinical situations. The similarity of these observations with those made in the Brugada syndrome, indicates a potentially greater role of concurrent illnesses in the initiation of VF. Indeed, heterogeneous changes in ion-channel expression have been described in Purkinje fibers and ventricular muscle in experimental studies [<xref ref-type="bibr" rid="R24">24</xref>] Taken in conjunction with our findings [<xref ref-type="bibr" rid="R23">23</xref>], it indicates a potentially greater role of such concurrent illnesses in the initiation of VF. The morphologic features of the short coupled ectopy in the three febrile patients with normal hearts were identical to those previously demonstrated in patients with idiopathic VF [<xref ref-type="bibr" rid="R22">22</xref>]. The similar origin of triggers in this series suggests that the current observations may have resulted from increased frequency or malignancy of the triggers by the febrile illness in these patients with idiopathic VF. This series provides new information on the precipitating role of concurrent febrile illness in the initiation of VF in patients with normal hearts. A possible explanation is that some patients may exhibit temperature-dependent modifications of ion channel properties or expression that may facilitate spontaneous activity within Purkinje or RVOT as a potential mechanism for these observations.</p></sec><sec sec-type="" id="s4"><title>Conclusion</title><p>Febrile illness is a common clinical situation managed by a variety of physicians. From the data available so far, it appears that an apparently benign febrile illness may be associated with malignant ventricular arrhythmias in patients with Brugada syndrome and sometimes in the absence of cardiac disease or other factors known to precipitate sudden cardiac death. The mechanism is still unclear but may be related to temperature-dependent modifications of ion channel properties or expression that facilitate spontaneous activity within Purkinje or RVOT as triggers of ventricular fibrillation. We believe that physicians should be aware of this possible phenomenon in all cases of febrile illness associated with syncope.</p></sec>
Brugada-type Electrocardiographic Pattern Induced by Fever	<p>ST-segment elevation in Brugada syndrome is caused by a shift in the ionic current balance and the creation of a voltage gradient between the epicardium and the endocardium. This ionic mechanism have been shown to be temperature dependent. We describe a 33-year-old man who presented with fever with the dynamic electrocardiographic changes similar to the Brugada syndrome. These electrocardiographic anomalies disappeared when the temperature returned to normal.</p>	<contrib contrib-type="author"><name><surname>Ozeke</surname><given-names>Ozcan</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Aras</surname><given-names>Dursun</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Geyik</surname><given-names>Bilal</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Deveci</surname><given-names>Bulent</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Selcuk</surname><given-names>Timur</given-names></name><degrees>MD</degrees></contrib><aff>Yuksek Ihtisas Hospital, Department of Cardiology, Ankara, Turkey</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Case Report</title><p>A 33-year-old man with fever due to common cold was referred to our hospital with suspicion of having an acute coronary syndrome. Three hours before admission, the patient with complaints of fever, chills, sore throat, was admitted to a local hospital, where the temperature was 38.7°C and his electrocardiogram showed characteristic Brugada pattern with coved type ST-segment elevation in leads V1-V3 and first-degree atrioventricular block (<xref ref-type="fig" rid="F1">Figure 1</xref>). He was diagnosed as having common cold and was referred to our hospital because of his electrocardiographic changes resembling acute coronary syndrome. He had no cardiac complaints and did not have a history of syncope or a family history of sudden cardiac death. On physical examination, the patient was dehydrated but alert. In the emergency room, his temperature was 38.3°C, the blood pressure was 120/70 mm Hg and the pulse was 88 per min. There was an inflammation of the pharynx. No other abnormalities were found. Hematological and biochemical tests including cardiac enzymes and electrolytes were within normal limits and the chest X-ray was unremarkable.The white blood cell count was 9600/cu mm. There was no evidence of structural heart disease on echocardiographic examination. Later that day, the patient’s temperature was 36.9°C with antipyretics and electrocardiogram showed saddle-back type ST-segment elevation in leads V1-V2 (<xref ref-type="fig" rid="F2">Figure 2A</xref>). Subsequently, drug challenge test with propafenone was performed; the ST-segment elevation changed from the saddle-back type (<xref ref-type="fig" rid="F2">Figure 2A</xref>) to the coved type (<xref ref-type="fig" rid="F2">Figure 2B</xref>). The next day, the patient was afebrile and the incomplete right bundle branch block disappeared on electrocardiogram, which showed only minimal ST-segment elevation in V2 (<xref ref-type="fig" rid="F1">Figure 1B</xref>). An electrophysilogical study was not performed because neither the patient nor any of his family members had experienced any arrhytmic symptom or sudden cardiac death, and medical follow up was decided. The patient was recommended to came hospital for urgent therapy of fever when a fever developed.</p></sec><sec sec-type="" id="s2"><title>Discussion</title><p>Brugada syndrome, first described as a new clinical entity by Pedro and Josep Brugada in 1992, is an inherited cardiac disease causing life-threatening ventricular tachyarrhythmias in patients with structurally normal heart and a characteristic electrocardiogram showing a pattern of right bundle branch block and ST segment elevation in right precordial leads V1 to V3 [<xref ref-type="bibr" rid="R1">1</xref>]. Wilde et al [<xref ref-type="bibr" rid="R2">2</xref>] reported 3 types of repolarization patterns as the ECG criteria for diagnosis of Brugada Syndrome; type 1 ECG, a coved-type ST segment elevation of ≥ 2 mm; type 2 ECG, saddle-back type a J wave amplitude of ≥ 2 mm and gradually descending ST segment elevation remaining ≥ 1 mm above baseline; and type 3 ECG, a coved-type or saddle-back type with ST segment elevation of ≤ 2 mm.</p><p>In spite of all the knowledge acquired over the last decade on this syndrome, diagnosis and risk stratification can be difficult sometimes. The ECG pattern may be dynamic over time and may include transient normalization [<xref ref-type="bibr" rid="R2">2</xref>]. The pathophysiology of Brugada syndrome is still under investigation. Insight from cellular electrophysiology suggests that the ST-segment elevation is caused by a shift in the ionic current balance and the creation of a voltage gradient, with predominance of transient outward current in the epicardium over the endocardium [<xref ref-type="bibr" rid="R3">3</xref>]. The ionic mechanism responsible for the Brugada syndrome have been shown to be temperature dependent [<xref ref-type="bibr" rid="R4">4</xref>-<xref ref-type="bibr" rid="R8">8</xref>]. In the present case, the electrocardiographic anomalies were evident when the fever was present and disappeared once the temperature returned to normal. It is possible that an increase in temperature may be the most important factor in the appearance of these electrocardiographic changes. In consideration of its variability, physicians should be aware of its often transient electrocardiographic features and fever should be aggressively treated. Our present report supports the effect of the body temperature in diagnosis and of on the underlying ion channelopathy of Brugada syndrome.</p></sec>
Naxos Disease<xref ref-type="fn" rid="fn1">1</xref>	<p>Since 1995, according to the World Health Organisation’s classification of cardiomyopathies, Naxos disease has been considered as the recessive form of arrhythmogenic right ventricular dysplasia/cardiomyopathy (ARVD/C). It is a stereotype association of ARVD/C with a cutaneous phenotype, characterised by woolly hair and palmoplantar keratoderma.</p>	<contrib contrib-type="author"><name><surname>Protonotarios</surname><given-names>Nikos</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Tsatsopoulou</surname><given-names>Adalena</given-names></name><degrees>MD</degrees></contrib><aff>Yannis Protonotarios Medical Center, Hora Naxos, Naxos 84300, Greece</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Definition</title><p>Since 1995, according to the World Health Organisation’s classification of cardiomyopathies, Naxos disease has been considered as the recessive form of arrhythmogenic right ventricular dysplasia/cardiomyopathy (ARVD/C) [<xref ref-type="bibr" rid="R1">1</xref>]. It is a stereotype association of ARVD/C with a cutaneous phenotype, characterised by woolly hair and palmoplantar keratoderma [<xref ref-type="bibr" rid="R2">2</xref>].</p></sec><sec sec-type="" id="s2"><title>Epidemiology</title><p>Naxos disease was first reported in 1986 by Protonotarios et al in patients originating from the Hellenic island of Naxos [<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R3">3</xref>]. Apart from Naxos, cases have also been reported from other Hellenic islands, as well as from Turkey, Israel and Saudi Arabia [<xref ref-type="bibr" rid="R4">4</xref>]. The prevalence of the disease in Hellenic islands reaches 1:1000. A variety of Naxos disease presenting at a younger age with more pronounced left ventricular involvement has been described in families from India and Ecuador (Carvajal syndrome) [<xref ref-type="bibr" rid="R5">5</xref>,<xref ref-type="bibr" rid="R6">6</xref>].</p></sec><sec sec-type="" id="s3"><title>Molecular Genetics</title><p>Genetic studies have located two causative genes, encoding for the proteins plakoglobin and desmoplakin. Plakoglobin and desmoplakin are proteins of cell-cell adhesion.</p><p>A 2-base-pair deletion mutation of the plakoglobin gene (Pk2157del2TG) truncating the C-terminal of the protein causes Naxos disease [<xref ref-type="bibr" rid="R7">7</xref>]. This mutation was identified in 13 families from Greece and in one family from Turkey [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R8">8</xref>]. The prevalence of heterozygous carriers is up to 5% of the Naxos population (20,000 inhabitants). Heterozygotes present normal phenotype except in a small minority who show woolly hair as well as a few electrocardiographic or echocardiographic abnormalities not fulfilling the criteria for ARVC [<xref ref-type="bibr" rid="R9">9</xref>]. Two different mutations of the desmoplakin gene (Dsp7901del1G and DspG2375R), truncating the C-terminal of the protein, have been found to underlie a similar cardiocutaneous syndrome in families from Ecuador and Israel [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R11">11</xref>].</p></sec><sec sec-type="" id="s4"><title>Clinical presentation and natural history</title><p>In patients with Naxos disease woolly hair was apparent from birth (<xref ref-type="fig" rid="F1">Figure 1A</xref>), whereas palmoplantar keratoderma developed during the first year of life as soon as the infant started to use the hands and feet (<xref ref-type="fig" rid="F1">Figure 1B and C</xref>) [<xref ref-type="bibr" rid="R9">9</xref>,;<xref ref-type="bibr" rid="R12">12</xref>]. In childhood the patients were asymptomatic with no diagnostic cardiac findings. During adolescence all affected members presented electrocardiographic and/or echocardiographic abnormalities fulfilling the criteria for ARVC [<xref ref-type="bibr" rid="R13">13</xref>]. Resting 12-lead electrocardiogram was abnormal in more than 90% of patients [<xref ref-type="bibr" rid="R9">9</xref>]. Electrocardiographic abnormalities included inverted T waves in leads V1 to V3 or across the precordial leads (77%), QRS complex prolongation in leads V1 to V3 (73%), epsilon waves (42%) (<xref ref-type="fig" rid="F1">Figure 1E</xref>) and complete or incomplete right bundle branch block (35%) [<xref ref-type="bibr" rid="R9">9</xref>]. Low voltage and/or flat T waves in left precordial leads were mostly observed in severe right or biventricular involvement. Ventricular extrasystoles of left bundle branch block configuration were recorded in the majority of patients. Ventricular extrasystoles of right bundle branch block configuration were less common. All patients presented right ventricular structural/functional alterations; minor alterations, consisting of mild dilatation or regional hypokinesia, were detected in 27% while major, consisting of severe dilatation, diffuse hypokinesia and aneurysms mostly of the outflow tract, apex or inferior wall, were detected in 73% of patients [<xref ref-type="bibr" rid="R9">9</xref>]. In one fourth of patients left ventricular abnormalities ranging from regional hypokinesia particularly of posterior wall or apex to diffuse dilatation and global hypokinesia were detected [<xref ref-type="bibr" rid="R9">9</xref>].</p><p>The symptomatic presentation was usually with syncope and/or sustained ventricular tachycardia during adolescence with a peak in young adulthood [<xref ref-type="bibr" rid="R14">14</xref>]. During follow-up arrhythmic events occurred in half of the patients [<xref ref-type="bibr" rid="R9">9</xref>]. The ECG during ventricular tachycardia always showed a left bundle branch block pattern (<xref ref-type="fig" rid="F1">Figure 1F</xref>) with a frontal plane axis ranging from -30 to -60 degrees or +60 to +165 degrees [<xref ref-type="bibr" rid="R14">14</xref>]. The inducibility of sustained monomorphic ventricular tachycardia in the electrophysiology laboratory was high among those presenting with a clinical episode. Heart disease progressed during time to the right or both ventricles. Congestive heart failure developed in one fourth of patients at the end stage of severe right or biventricular involvement.</p><p>The annual total cardiac mortality rate was 3% and the annual sudden death mortality rate reached 2.3% [<xref ref-type="bibr" rid="R9">9</xref>]. Syncope, left ventricular involvement and the appearance of symptoms and/or structural progression before the age of 35 years were risk factors for sudden death [<xref ref-type="bibr" rid="R9">9</xref>].</p><p>Cardiac histology revealed the typical pattern of ARVD/C15 with fibrofatty replacement of right ventricular myocardium mainly in subepicardial and mediomural layers being regionally transmural with formation of aneurysms [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R16">16</xref>]. Surviving myocytes surrounded by fibrosis were embedded within fatty tissue (<xref ref-type="fig" rid="F1">Figure 1D</xref>). Involved left ventricular myocardium showed fibrofatty or fibrous replacement [<xref ref-type="bibr" rid="R4">4</xref>]. Lymphocyte infiltrates were observed particularly when the biopsy was performed at the time of clinical progression. Cardiac histology in one patient from Ecuador with the Carvajal variant of Naxos disease differed only in that the fatty component of replacement process was absent [<xref ref-type="bibr" rid="R17">17</xref>]. Immunohistochemical studies, showed that the signal of the mutated protein (plakoglobin or desmoplakin) and the signal of connexin43 were diminished at intercellular junctions [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R18">18</xref>].</p></sec><sec sec-type="" id="s5"><title>Pathogenesis of ARVD/C in Naxos disease</title><p>Myocardial cells are differentiated bipolar cells coupled at intercalated discs where adherence junctions, desmosomes and gap junctions are located [<xref ref-type="bibr" rid="R19">19</xref>]. Adherence junctions and desmosomes secure mechanical coupling while gap junctions serve electrical coupling. Plakoglobin (γ-catenin) is the only known common component of both adherence junctions and desmosomes functioning also as a signaling molecule apart from its structural role in securing the intercellular adhesion [<xref ref-type="bibr" rid="R20">20</xref>]. At the adherence junctions it is connected to the actin cytoskeleton and at desmosomes to the intermediate filaments of desmin. Desmoplakin is also a cytoplasmic protein of the desmosomes that interlinks plakoglobin or plakophilin with desmin intermediate filaments. Defects in linking sites of these proteins can interrupt the contiguous chain of cell adhesion, particularly under conditions of increased mechanical stress or stretch, leading to cell isolation and death, progressive loss of myocardium and fibro-fatty replacement [<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R7">7</xref>]. The degree of participation of fat in the repair process may be related to the rate of disease progression or may be mutation specific [<xref ref-type="bibr" rid="R4">4</xref>]. Surviving myocardial fibers within fibro-fatty tissue provide a slow conduction substrate inducing re-entrant ventricular arrhythmias [<xref ref-type="bibr" rid="R21">21</xref>]. Recent studies on Naxos disease revealed that the genetically determined defect in cell adhesion results in early gap junctions remodeling and altered electrical coupling which may act synergically with the progressive pathologic changes in myocardium contributing to a highly arrhythmogenic substrate [<xref ref-type="bibr" rid="R18">18</xref>].</p></sec>
Role of Catheter Ablation in Arrhythmogenic Right Ventricular Dysplasia<xref ref-type="fn" rid="fn1">1</xref>	<p>Arrhythmogenic right ventricular dysplasia/cardiomyopathy is a disorder characterized by frequent ventricular tachycardia originating from the right ventricle and fibro-fatty replacement of right ventricular myocardium. Though the disorder was originally described during surgical ablation of refractory ventricular tachycardia, catheter ablation of tachycardia is one of the options for patients not responding to anti arrhythmic agents. Direct current fulguration was used in the initial phase followed by radiofrequency catheter ablation. In the present day scenario, all patients with risk for sudden cardiac death should receive an implantable cardioverter defibrillator. Radiofrequency catheter ablation remarkably reduces the frequency of defibrillator therapies. Direct current fulguration can still be considered in cases when radiofrequency ablation fails, though it requires higher expertise, general anesthesia and carries a higher morbidity. Newer mapping techniques have helped in identification of the site of ablation. In general, the success rate of ablation in arrhythmogenic right ventricular dysplasia is less than in other forms of right ventricular tachycardias like right ventricular outflow tract tachycardia.</p>	<contrib contrib-type="author"><name><surname>Francis</surname><given-names>Johnson</given-names></name><degrees>MBBS, MD, DM</degrees><xref ref-type="aff" rid="aff1">*</xref></contrib><contrib contrib-type="author"><name><surname>Fontaine</surname><given-names>Guy</given-names></name><degrees>MD, PhD, FACC, FAHA, FESC</degrees><xref ref-type="aff" rid="aff2">†</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Arrhythmogenic right ventricular dysplasia/cardiomyopathy (ARVD) is a disorder characterized by fibro-fatty replacement of the right ventricular myocardium, frequent ventricular tachycardia originating from the right ventricle and right heart failure. It was originally described by Fontaine et al during surgical ablation of refractory ventricular tachycardia [<xref ref-type="bibr" rid="R1">1</xref>]. The first case of ARVD underwent surgical ablation in October 1973 with a simple incision made at the site of origin of ventricular tachycardia (VT). This was successful to prevent recurrence of the arrhythmia and contributed to the identification of this disease [<xref ref-type="bibr" rid="R2">2</xref>]. The first clinical series of 24 adult cases was published by Marcus et al in1982 [<xref ref-type="bibr" rid="R3">3</xref>]. The refractory ventricular arrhythmias of ARVD has always been a challenge for the clinician. Catheter ablation has been tried in spite of the justifiable fear of perforation of the dysplastic ventricle. This short review aims at bringing together the available literature on catheter ablation in ARVD.</p></sec><sec sec-type="" id="s2"><title>Therapeutic Options in ARVD</title><p>Though historically the original description of ARVD was during surgical ablation, pharmacological therapy was the initial mode of treatment in most cases. Surgical ablation by right ventricular disconnection was resorted to in resistant cases [<xref ref-type="bibr" rid="R4">4</xref>]. Peroperative cryoablation of 8 cases were reported by Isobe from Japan [<xref ref-type="bibr" rid="R5">5</xref>]. None of the eight patients died during a mean follow up of 3.25 year. VT recurred in two patients and a new VT was seen in another patient. Endocavitary electrode catheter ablation using direct current shocks of 100 to 320 J was another mode of therapy which was tried in that period [<xref ref-type="bibr" rid="R6">6</xref>]. Nowadays, implantable cardioverter defibrillators (ICD) are being recommended more often to cover the risk of sudden cardiac death (SCD) in ARVD. Even then, patients may need pharmacological therapy to reduce the number of shocks. Patients having recurrent sustained VT while on optimal medical therapy are candidates for catheter ablation.</p></sec><sec sec-type="" id="s3"><title>Catheter Ablation in ARVD</title><p>The initial reports on catheter ablation in ARVD were using direct current fulguration [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R8">8</xref>]. One of the earliest series of fulguration in ARVD was that of 13 patients who were treated with shocks ranging from 160 to 280 J [<xref ref-type="bibr" rid="R9">9</xref>]. Single or multiple shocks were required in up to three sessions. There were two deaths and four of the 11 survivors required antiarrhythmic treatment following the fulguration therapy. The mean follow up was 45 months. With the advent of radiofrequency catheter ablation, direct current fulguration went out of vogue. Fontaine et al has suggested that fulguration should still be tried if radiofrequency ablation is not successful [<xref ref-type="bibr" rid="R10">10</xref>]. In their 16 year experience of ablation in ARVD, the effectiveness of radiofrequency was less than 40% in the first session. At same time, fulguration is effective in the same session after ineffective radiofrequency ablation. Complications have disappeared since the use of soft and steerable ablation catheters. This work also classifies thoroughly the results of VT ablation alone and in combination with antiarrhythmic drugs, definition of relapses, etc. But the disadvantage of fulguration is that it requires expertise, general anesthesia and more than one session in half the patients.</p></sec><sec sec-type="" id="s4"><title>Radiofrequency Catheter Ablation</title><p>Radiofrequency catheter ablation for ARVD has been in use since early nineties [<xref ref-type="bibr" rid="R11">11</xref>]. It has been suggested that only patients with focal dysplasia are potential candidates for ablation [<xref ref-type="bibr" rid="R12">12</xref>].</p><p>Entrainment mapping can be used to characterize reentry circuits in ARVD to guide ablation [<xref ref-type="bibr" rid="R13">13</xref>-<xref ref-type="bibr" rid="R15">15</xref>]. The concept of concealed entrainment was first reported by Fontaine et al in 1989 [<xref ref-type="bibr" rid="R13">13</xref>]. This paper stresses the identification of the zone of slow conduction. The concept was originally found in a patient with ARVD. It was later extended to other forms of chronic VT (post-myocardial infarction), and more recently is used as a marker of the reentry pathway in re-entrant supraventricular tachycardias. Ellison et al mapped 19 VTs in 5 patients with ARVD. Radiofrequency current was applied to the 58 sites where pacing entrained the VT to assess acute termination, with only 22% success. Eight of the 19 VTs were rendered noninducible and three were modified to a longer cycle length. In two patients ablation at a single site abolished two VTs [<xref ref-type="bibr" rid="R14">14</xref>]. Harada et al did entrainment mapping in 8 VTs in 7 patients with ARVD. Radiofrequency applications were done at 31 sites identified by mapping and terminated 7 of them [<xref ref-type="bibr" rid="R15">15</xref>].</p><p>Endocardial mapping can detect abnormal fragmented electrograms with delayed potentials. Pacemapping confirms the ablation site by producing a QRS morphology identical to the clinical VT [<xref ref-type="bibr" rid="R16">16</xref>]. Recently non-contact mapping has been used to guide catheter ablation in ARVD [<xref ref-type="bibr" rid="R17">17</xref>]. The endocardial exit point was defined in all three ARVD patients and the diastolic pathway (earliest endocardial diastolic activity) was identified in one of them. Catheter ablation was completely effective in only one of the three. Reithmann et al used electroanatomic mapping of right ventricular endocardial activation as a guide for catheter ablation in patients with ARVD [<xref ref-type="bibr" rid="R18">18</xref>]. Both electroanatomic mapping and entrainment procedures were performed in 5 patients. Endocardial mapping during tachycardia demonstrated a focal activation pattern with radial spreading of activation from the site of earliest activation. The sites of earliest activation were in an aneurysmal outflow tract in two patients, at the border of aneurysms near the tricuspid annulus in two patients and at the apex of the right ventricle in one. Entrainment mapping showed that these were the exit sites of the reentrant circuits. The clinical VTs were noninducible in 4 of the 5 patients after catheter ablation. During a mean follow up of 7 months, the frequency of ICD therapies came down from 49 ± 61 episodes per month to 0.3 ± 0.5 episodes per month.</p><p>Three dimensional Real-time Positioning Management System (RPM) has also been used for guiding ablation in ARVD [<xref ref-type="bibr" rid="R19">19</xref>]. RPM uses sonomicrometry to determine the spatial location of the ablation catheter relative to two reference catheters positioned in the right atrium and right ventricle.</p><p>O'Donnell et al have highlighted the electrophysiological differences between patients with ARVD and right ventricular outflow tract tachycardia (RVOT VT) [<xref ref-type="bibr" rid="R20">20</xref>]. Though radiofrequency ablation is the first line treatment for symptomatic RVOT VT, the role is limited in ARVD. In their study they compared 33 patients with RVOT VT and 17 patients with ARVD. Re-entry was the mechanism of tachycardia in 80% of the ARVD group while 97% of RVOT VT had features of triggered automaticity. Partial or complete success was obtained only in 71% of patients with ARVD while complete success was obtained in 97% of RVOT VT. The recurrence rate was 48% in ARVD and 6% in RVOT VT.</p><p>Ablation of ventricular tachycardias in ARVD still remains a clinical challenge, though more and more cases are being reported in the literature [<xref ref-type="bibr" rid="R21">21</xref>-<xref ref-type="bibr" rid="R23">23</xref>].</p></sec>
Idiopathic Epicardial Ventricular Arrhythmias: Diagnosis and Ablation Technique from the Aortic Sinus of Valsalva	<p>Idiopathic outflow tract arrhythmias (ventricular tachycardias or symptomatic premature ventricular contractions; OT-VT/PVCs) can originate from the left ventricular (LV) epicardium (Epi-VT/PVCs), and radiofrequency (RF) energy applications from the aortic sinus of Valsalva can eliminate Epi-VT/PVCs in selected patients. Among the various ECG findings, the R-wave duration index and R/S amplitude index in leads V1 or V2 are useful for identifying Epi-VT/PVCs, and the Q-wave ratio of leads aVL to aVR and S-wave amplitude in lead V1 are useful for differentiating between an Epi-VT/PVC originating from the LV epicardium remote from the left sinus of Valsalva (LSV) and that from the LSV. Tissue tracking imaging is a promising modality for identifying the origin of OT-VT/PVCs and for differentiating between an Epi-VT/PVC originating from the LV epicardium remote from the LSV and that from the LSV.</p><p>If the origin of the Epi-VT/PVC is identified within the LSV, coronary and aortic angiography should be performed to assess the anatomic relationships between the Epi-VT/PVC origin and coronary arteries and aortic valve before the RF energy delivery. To avoid potential complications, RF ablation should be performed at the LSV using a maximum power of 35 watts and maximum temperature of 55°C. Epicardial mapping through the coronary venous system and the presence of potentials recorded from the ablation site within the LSV and their changes before and after the RF energy applications may be useful for diagnosing Epi-VT/PVCs or predicting a successful catheter ablation from the LSV.</p>	<contrib contrib-type="author"><name><surname>Tada</surname><given-names>Hiroshi</given-names></name><degrees>MD</degrees></contrib><aff>Division of Cardiology, Gunma Prefectural Cardiovascular Center, Maebashi, Gunma Japan</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Radiofrequency (RF) catheter ablation has been established as an effective and curative therapy for ventricular tachycardias (VTs) or symptomatic premature ventricular contractions (PVCs) originating from the outflow tract (OT-VT/PVCs) in structurally normal hearts [<xref ref-type="bibr" rid="R1">1</xref>-<xref ref-type="bibr" rid="R6">6</xref>]. Recenty, idiopathic OT-VT/PVCs originating from the left ventricular (LV) epicardium (Epi-VT/PVCs) have been reported [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R13">13</xref>]. These OT-VT/PVCs are thought to originate from the LV epicardium around the transitional area from the great cardiac vein to the anterior interventricular vein [<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R14">14</xref>]. Some can be ablated through the coronary venous system [<xref ref-type="bibr" rid="R15">15</xref>] or by percutaneous epicardial instrumentation [<xref ref-type="bibr" rid="R16">16</xref>], but others can be ablated from the left sinus of Valsalva (LSV) [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R13">13</xref>]. The correct identification of the latter type of OT-VT/PVC before the RF energy applications is important in order to avoid futile RF applications from the LSV and the ensuing complications [<xref ref-type="bibr" rid="R17">17</xref>,<xref ref-type="bibr" rid="R18">18</xref>]. In this article, we focused on the diagnosis and RF catheter ablation of Epi-VT/PVCs which can be ablated from the LSV.</p></sec><sec sec-type="" id="s2"><title>Diagnosis</title><p>Characteristic ECG findings originating from each portion of the outflow tract and several ECG findings useful for differentiating an OT-VT/PVC from other OT-VT/PVCs originating from different portions of the outflow tract have been reported. The precordial R-wave transition [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R19">19</xref>], QRS morphology in lead I [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R6">6</xref>], R-wave duration in lead V1 or V2, and R/S-wave amplitude ratio in leads V1 or V2 have been reported as useful indices for differentiating OT-VT/PVCs originating from the right side from those arising from the left side (LVOT, LSV or LV epicardium remote from the LSV) [<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R13">13</xref>]. Furthermore, the presence of an S wave in lead V5 or V6 is considered a characteristic, useful ECG finding in OT-VT/PVCs of LVOT origin [<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R9">9</xref>]. Among those findings, the usefulness of both the R-wave duration index and R/S amplitude index in leads V1 or V2 for identifying Epi-VT/PVCs has been reported (<xref ref-type="fig" rid="F1">Figure 1</xref>) [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R8">8</xref>]. The R-wave duration index is obtained by dividing the QRS complex duration by the longer R-wave duration in lead V1 or V2, and the R/S-wave amplitude index is calculated as the greater value for the R/S-wave amplitude ratio in lead V1 or V2 [<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R8">8</xref>]. A recent study demonstrated that when the R-wave duration index is ≥0.5 or the R/S-wave amplitude index is ≥0.3, there is a fair possibility that the OT-VT/PVC originates from the LSV (<xref ref-type="fig" rid="F1">Figure 1A, left panel</xref>) or LV epicardium remote from the LSV which cannot be ablated from the LSV (<xref ref-type="fig" rid="F1">Fig. 1A, right panel</xref>) [<xref ref-type="bibr" rid="R6">6</xref>]. However, both indices do not significantly differ when differentiating between Epi-VT/PVCs originating from the LSV and those originating from the LV epicardium remote from the LSV [<xref ref-type="bibr" rid="R6">6</xref>]. Instead, a Q-wave ratio of leads aVL to aVR >1.4 or an S-wave amplitude ≥1.2 mV in lead V1 was useful for differentiating between an Epi-VT/PVC originating from the LV epicardium remote from the LSV and that from the LSV (<xref ref-type="fig" rid="F1">Figure 1</xref>) [<xref ref-type="bibr" rid="R6">6</xref>]. In a previous report using a pace mapping techinique [<xref ref-type="bibr" rid="R13">13</xref>], the QRS morphology during pacing from the proximal site of the anterior interventricular vein (AIV) exhibited a deep S-wave in V1, and pacing from the distal portion of the great cardiac vein (GCV) revealed a high Q-wave ratio of aVL to aVR. Therefore, a high Q-wave ratio for aVL to aVR or a deep S-wave in lead V1 may indicate that the Epi-VT/PVC originates from the LV epicardium remote from the LSV around the transitional area from the GCV to the AIV [<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R14">14</xref>]. For determination of the location of OT-VT/PVCs, including Epi-VT/PVCs, a recently developed ECG algorithm is useful [<xref ref-type="bibr" rid="R6">6</xref>].</p><p>Although rare (3%), some OT-VT/PVCs have demonstrated dynamic changes in the QRS morphology following the RF catheter ablation, requiring additional RF ablation applications at a different portion of the outflow tract to cure the OT-VT/PVCs [<xref ref-type="bibr" rid="R20">20</xref>]. These OT-VT/PVCs consistently showed a change in the R wave amplitude in the inferior leads between the 1<sup>st</sup> OT-VT/PVC and 2nd OT-VT/PVC, while the earliest ventricular activation during either the 1<sup>st</sup> or 2<sup>nd</sup> OT-VT/PVC was recorded from the LSV [<xref ref-type="bibr" rid="R20">20</xref>]. When an increase in the R-wave amplitude in the inferior leads follows the RF catheter ablation of the 1st OT-VT/PVC from an endocardial site of the RVOT or LVOT, the earliest ventricular activation during the 2<sup>nd</sup> OT-VT/PVC may be recorded at the LSV. On the other hand, when a decrease in the R-wave amplitude in the inferior leads follows the RF catheter ablation from the LSV, the earliest ventricular activation during the 2<sup>nd</sup> OT-VT may be recorded from an endocardial site of the RVOT or LVOT. Therefore, detailed continuous observation of the QRS morphology of OT-VT/PVCs, especially the R wave amplitude in the inferior leads, is important for identifying the changes in the QRS morphology during the RF catheter ablation of OT-VT/PVCs.</p><p>Recently, tissue tracking imaging (TTI) has been demonstrated as a novel non-invasive modality for identifying the origin of OT-VT/PVCs [<xref ref-type="bibr" rid="R21">21</xref>]. TTI is an ultrasonographic technique that measures the myocardial motion amplitude toward the transducer in each region during systole, identifying regional myocardial displacement on the basis of myocardial velocities using color Doppler myocardial imaging principles [<xref ref-type="bibr" rid="R22">22</xref>,<xref ref-type="bibr" rid="R23">23</xref>]. It allows rapid semiquantitative visual assessment of the systolic distance of the tissue motion along the Doppler axis using a graded color display [<xref ref-type="bibr" rid="R21">21</xref>-<xref ref-type="bibr" rid="R23">23</xref>]. In this technique, the origin of the OT-VT/PVC could be recognized as the site where the earliest color-coded signal (ECCS) appeared on the myocardium at the onset of the OT-VT/PVC (<xref ref-type="fig" rid="F2">Figure 2</xref>) [<xref ref-type="bibr" rid="R21">21</xref>]. The OT-VT/PVCs in which the earliest ventricular activation was recorded from the LSV had the ECCS in the myocardium above the pulmonary valve, and most of them in which the ECCS appeared close to the pulmonary valve could be ablated from the LSV (<xref ref-type="fig" rid="F2">Figure 2B</xref>) [<xref ref-type="bibr" rid="R21">21</xref>]. The distance between the attachment of the pulmonary valve to the septum and the center of the ECCS (8±4 mm) in these OT-VT/PVCs with a successful ablation from the LSV was significantly shorter than that in those with a failed ablation (18±6 mm, p<0.05). On the other hand, the ECCS was always found below or at the level of the pulmonary valve in all arrhythmias which could be ablated from the RVOT(<xref ref-type="fig" rid="F2">Figure 2A</xref>). These results indicate that TTI can provide detailed and accurate information on the arrhythmia origin of OT-VT/PVCs and may be useful for differentiating between an OT-VT/PVC originating from the LV epicardium remote from the LSV and that from the LSV.</p></sec><sec sec-type="" id="s3"><title>Mapping and Ablation Technique</title><p>Under guidance with fluoroscopy, catheters were introduced into the RV apex, RVOT and/or His bundle region via the right femoral vein. If the clinical arrhythmia did not occur spontaneously, programmed ventricular stimulation from the RV apex and RVOT or incremental burst pacing were performed. If the clinical arrhythmia was not induced in the baseline state, intravenous isoproterenol (0.5 to 2.0 μg/min) was administered to induce the clinical arrhythmia. Activation mapping and pace mapping were performed during the clinical arrhythmia. Mapping of the OT-VT/PVC was initially started in the RVOT region. If suitable ablation sites were not found in the RVOT, the endocardium of the LVOT and aortic sinus of Valsalva were mapped. If the origin of the OT-VT/PVC was mapped to a location above the aortic valve, coronary and aortic angiography were performed to assess the anatomic relationship between the OT-VT/PVC origin and coronary arteries and aortic valve before the RF energy delivery (<xref ref-type="fig" rid="F3">Figure 3</xref>) [<xref ref-type="bibr" rid="R6">6</xref>-<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R13">13</xref>].</p><p>It is believed that an RF application from the LSV does not ablate the valve or aortic wall itself, but destroys the epicardium above the septum [<xref ref-type="bibr" rid="R9">9</xref>]. In our laboratory, a 7 Fr quadripolar catheter with a 4-mm distal electrode, embedded thermistor, interelectrode spacing of 2-5-2 mm, and deflectable tip was usually used for mapping and ablation (<xref ref-type="fig" rid="F3">Figure 3</xref>). To avoid potential complications, the RF ablation was performed at the aortic sinus of Valsalva using a maximum power of 35 watts, maximum electrode-tissue interface temperature of 55°C and maximum of 6 RF applications [<xref ref-type="bibr" rid="R9">9</xref>]. The applications of RF energy were 60 to 90 sec in duration. If an application of RF energy was unsuccessful, the ablation site was changed under fluoroscopic guidance. Pace mapping from the LSV was possible (<xref ref-type="fig" rid="F1">Figure 1B</xref>), and several studies reported that a perfect pace mapping could be obtained from the successful ablation site [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R13">13</xref>]. However, the pacing threshold was usually too high (>8 V) to capture the myocardium through the LSV [<xref ref-type="bibr" rid="R9">9</xref>], and no studies have sufficiently examined whether or not an RF energy application at a site where a perfect pace map is obtained can always result in successful ablation, or whether or not an RF energy application at a site where a perfect pace map cannot be obtained results in a failed ablation.</p><sec id="s3a"><title>Utility of Epicardial Mapping through the Coronary Venous System</title><p>It is well known that even if the earliest ventricular activation of the Epi-VT/PVC is at the LSV, RF catheter ablation at that site is not always effective. Because the tip temperature should be maintained at <55°C to avoid potential complications [<xref ref-type="bibr" rid="R9">9</xref>], the distance between the ablation site within the LSV and origin of the Epi-VT/PVC will mainly affect the results of the RF ablation from the LSV. Therefore, some Epi-VT/PVCs with the earliest ventricular activation in the LSV cannot be ablated from the LSV because of the distance from the LSV. No criterion based on the electrograms for predicting a successful catheter ablation from the LSV has been established. However, when the earliest ventricular activation during the clinical arrhythmia was found within the LSV, the placement of a 2-Fr octapolar electrode catheter from the GCV to the AIV (GCV-AIV) and obtaining 7 bipolar electrogram recordings from the adjacent electrodes of the catheter placed along the GCV-AIV may be helpful in diagnosing Epi-VT/PVCs and predicting the successful RF catheter ablation from the LSV [<xref ref-type="bibr" rid="R24">24</xref>]. In Epi-VT/PVCs, the earliest ventricular activation during the arrhythmias is usually found at the LSV or GCV-AIV [<xref ref-type="bibr" rid="R13">13</xref>,<xref ref-type="bibr" rid="R14">14</xref>]. Therefore, several studies have attempted to evaluate the utility of epicardial mapping through the coronary venous system for diagnosising and treating Epi-VT/PVCs [<xref ref-type="bibr" rid="R6">6</xref>,<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R13">13</xref>]. However, their results differ. Kanagaratnam et al. reported simultaneous endocardial and percutaneous epicardial mapping in patients with Epi-VTs eliminated by an RF application from the LSV [<xref ref-type="bibr" rid="R7">7</xref>]. They found that the Epi-VT/PVC had an earlier activation on the epicardial surface than at the endocardial site [<xref ref-type="bibr" rid="R7">7</xref>]. However, Hachiya et al. reported that the ventricular activation recorded from the LSV must be earlier than the earliest ventricular activation recorded in the GCV-AIV in Epi-VT/PVCs which can be ablated from the LSV [<xref ref-type="bibr" rid="R9">9</xref>]. A recent study reported the usefulness of comparing the earliness of the ventricular activation recorded from the LSV and earliest ventricular activation of 7 bipolar recordings from the GCV-AIV during the Epi-VT/PVCs for predicting a successful catheter ablation from the LSV [<xref ref-type="bibr" rid="R24">24</xref>]. In patients in whom the earliest ventricular activation was found at the LSV or GCV-AIV, the earliest ventricular activation in the GCV-AIV preceding the ventricular activation from the LSV by less than 10 ms identified a successful RF catheter ablation from the LSV with a sensitivity of 88 %, specificity of 100%, positive predictive value of 100 % and negative predictive value of 75% [<xref ref-type="bibr" rid="R24">24</xref>].</p></sec><sec id="s3b"><title>Significance of Potentials recorded at the Ablation Site within the Left Sinus of Valsalva</title><p>It is well known that a potential preceding the QRS complex is often recorded at the ablation site within the LSV (<xref ref-type="fig" rid="F4">Figure 4A</xref>) [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R13">13</xref>,;<xref ref-type="bibr" rid="R25">25</xref>]. However, it is unclear if complete elimination of Epi-VT/PVCs can always be achieved with an application of RF energy at the site where this potential is recorded. In a recent study in which a quantitative analysis of the electrograms was performed in 23 patients with symptomatic Epi-VT/PVCs that underwent RF applications from the LSV, this potential was found at 19 (90%) of 21 successful ablation sites [<xref ref-type="bibr" rid="R26">26</xref>]. However, it was also recorded at 19 (79%) of 24 unsuccessful ablation sites, and there was no significant difference in the incidence of the potentials between the successful and unsuccessful ablation sites (p=0.5). In addition, there were no significant differences in the amplitude (p=0.12) or interval from the potential to the QRS complex (p=0.10) between the successful and unsuccessful ablation sites. Therefore, the presence of this potential may not help in identifying successful ablation sites within the LSV.</p><p>Another potential recorded in the late phase of the QRS complex during sinus rhythm in patients with Epi-VT/PVCs has been reported (<xref ref-type="fig" rid="F4">Figure 4B</xref>)[<xref ref-type="bibr" rid="R8">8</xref>]. A recent study demonstrated that this potential was more often observed at the successful ablation sites than at the unsuccessful ablation sites before the ablation (p<0.05), and the difference between the two groups became greater after the ablation (p<0.001) [<xref ref-type="bibr" rid="R26">26</xref>]: In the unsuccessful ablation sites, this potential was absent (Type 1) or its location did not change before and after the ablation (Type 4). However, this potential appeared at 38% of the successful ablation sites (Type 2), and pre-existing potentials were more delayed after the ablation at 53% of the successful ablation sites (Type 3), demonstrating significant differences in the pattern of this potential between the successful and unsuccessful ablation sites (p<0.001). As a result, type 2 or type 3 patterns for this potential identified a successful ablation with a sensitivity of 100%, specificity of 92%, positive predictive value of 90% and negative predictive value of 100% [<xref ref-type="bibr" rid="R26">26</xref>]. Therefore, this potential is also characteristic of Epi-VT/PVCs, and may be more useful in identifying successful ablation sites than the former potential recorded during the Epi-VT/PVCs. However, the precise mechanism responsible for these two potentials recorded during the Epi-VT/PVCs (the former) or during sinus rhythm (the latter) is still unclear. There is a possibility that the analysis of the electrograms recorded at the successful and unsuccessful ablation sites in the same patient might be affected by the effects of the RF energy applications. Therefore, the significance and utility of these 2 potentials recorded within the LSV have not been sufficiently clarified.</p></sec></sec><sec sec-type="" id="s4"><title>Conclusions</title><p>With an RF energy application from the LSV, some Epi-VT/PVCs can be abolished. However, RF catheter ablation within the LSV runs the risk of potential complications. Thus, attempting to diagnose these types of VT/PVCs non-invasively before the ablation procedure, performing detailed mapping and assessment of the electrograms obtained during the procedure, and obtaining a precise assessment of the anatomic relationships between the arrhythmia origin and coronary arteries and aortic valve, are crucial for avoiding futile RF applications from the LSV and the ensuing complications, and for successfully eliminating this type of Epi-VT/PVC.</p></sec>
Coexistence of Atrioventricular Nodal Reentrant Tachycardia and Idiopathic Left Ventricular Outflow-Tract Tachycardia	<p>Double tachycardia is a relatively rare condition. We describe a 21 year old woman with history of frequent palpitations. In one of these episodes, she had wide complex tachycardia with right bundle branch and inferior axis morphology. A typical atrioventricular nodal tachycardia was induced during electrophysiologic study, aimed at induction of clinically documented tachycardia. Initially no ventricular tachycardia was inducible. After successful ablation of slow pathway, a wide complex tachycardia was induced by programmed stimulation from right ventricular outflow tract. Mapping localized the focus of tachycardia in left ventricular outflow tract and successfully ablated via retrograde aortic approach. During 7 month's follow-up, she has been symptom free with no recurrence. This work describes successful ablation of rare combination of typical atrioventricular nodal tachycardia and left ventricular outflow tract tachycardia in the same patient during one session.</p>	<contrib contrib-type="author"><name><surname>Haghjoo</surname><given-names>Majid</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Arya</surname><given-names>Arash</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Dehghani</surname><given-names>Mohammadreza</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Emkanjoo</surname><given-names>Zahra</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Fazelifar</surname><given-names>Amirfarjam</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Heidari</surname><given-names>Alireza</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Sadr-Ameli</surname><given-names>MohammadAli</given-names></name><degrees>MD</degrees></contrib><aff>Department of Pacemaker and Electrophysiology, Shahid Rajaie Cardiovascular Medical and Research Center, Iran University of Medical Sciences, Tehran, Iran</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Double tachycardia, defined as the simultaneous occurrence of atrial and ventricular [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>] or junctional and ventricular tachycardia (VT) [<xref ref-type="bibr" rid="R3">3</xref>], has been rarely reported and usually occurs in patients with poor left ventricular function or in association with digitalis intoxication [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>,<xref ref-type="bibr" rid="R4">4</xref>,<xref ref-type="bibr" rid="R5">5</xref>]. The coexistence of atrioventricular reentrant tachycardia (AVRT) and idiopathic right ventricular outflow tract (RVOT) tachycardia [<xref ref-type="bibr" rid="R3">3</xref>], or atrioventricular nodal reentrant tachycardia (AVNRT) and RVOT-tachycardia [<xref ref-type="bibr" rid="R6">6</xref>] has also been reported, but coincidence of AVNRT and idiopathic left ventricular outflow tract (LVOT) tachycardia has rarely been reported. To the best of our knowledge, this case is the first report of successful ablation of rare combination of AVNRT and LVOT-tachycardia in the same patient.</p></sec><sec sec-type="" id="s2"><title>Case report</title><p>A 21 year old woman with no evidence of structural heart disease referred to our center for evaluation of palpitation and dizziness. The structural heart disease was excluded by physical examination and transthoracic echocardiography. Transthoracic echocardiography showed normal cardiac chambers (including right ventricle), normal valvular function and ejection fraction (EF) without any wall motion abnormalities. During an episode of palpitation, the standard 12-lead electrocardiogram (ECG) showed documented wide complex tachycardia with a heart rate of 125 beats /min. The tachycardia was refractory to two intravenous antiarrhythmics (amiodarone, procainamide). The wide complex tachycardia had inferior axis and right bundle branch block morphology compatible with LVOT-tachycardia (<xref ref-type="fig" rid="F1">Figure 1</xref>). The baseline ECG showed no abnormality.</p><p>After obtaining written informed consent, electrophysiologic study was done in the postabsorptive and nonsedated state. During programmed electrical stimulation from atrium and ventricle, dual AV nodal physiology with nonsustained AVNRT was induced. Then programmed ventricular stimulation was performed with standard protocol at three cycle length (600,500,400ms) and three extrastimuli up to coupling interval of 200 ms from two sites (RV apex, RVOT). No ventricular tachycardia was induced with and without isoproterenol infusion. Repeat programmed atrial stimulation resulted in induction of sustained AVNRT under isoproterenol infusion (<xref ref-type="fig" rid="F2">Figure 2</xref>). Radiofrequency catheter ablation of slow pathway was done at right posteroseptal area. Postablation programmed stimulation failed to induce any supraventricular tachycardia with and without isoproterenol infusion but a wide complex tachycardia (cycle length=480 ms) identical to clinically documented arrhythmia was induced by overdrive pacing from RVOT. Mapping of RVOT failed to show any early ventricular activation site, thus LVOT was mapped and tachycardia focus was localized in this area with 53 ms early ventricular activation relative to surface electrocardiogram (<xref ref-type="fig" rid="F3">Figure 3</xref>) Radiofrequency energy delivery (50 W, 60°C) at this site resulted in termination of tachycardia (<xref ref-type="fig" rid="F4">Figure 4</xref>). Thirty minutes after ablation, no tachycardia was induced with and without isoproterenol infusion. During 12 month follow-up, she has been symptom free with no antiarrhythmic drugs.</p></sec><sec sec-type="" id="s3"><title>Discussion</title><p>Double tachycardia was a relatively uncommon type of tachycardia in previous reports [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R3">3</xref>,<xref ref-type="bibr" rid="R5">5</xref>]. In recent study by Kautzner et al [<xref ref-type="bibr" rid="R7">7</xref>], this combination (RVOT-tachycardia and AVNRT) was not as uncommon (15% of RVOT-tachycardia patients had AVNRT). In this study, three of seven patients with coexistent idiopathic ventricular outflow tachycardia and AVNRT had the arrhythmogenic focus localized in the uppermost part of the septum or more epicardially near the great cardiac vein as documented by detailed mapping but no early site was reported in LVOT area or aortic cusps in any patients. In this study, no attempt for ablation was made in the patients with epicardial variant.</p><p>Idiopathic VT most commonly arises from RV than LV (70% versus 30%) [<xref ref-type="bibr" rid="R8">8</xref>]. Idiopathic LVOT-tachycardia is one of the three subtypes of idiopathic left ventricular tachycardia that analogous to adenosine sensitive RVOT-tachycardia originate from deep within the septum and exit from left side of septum, and result from cAMP-mediated triggered activity [<xref ref-type="bibr" rid="R9">9</xref>]. AVNRT is a typical reentrant tachycardia originating from the AV nodal and perinodal tissues [<xref ref-type="bibr" rid="R10">10</xref>].</p><p>In this patient, clinically documented arrhythmia was ventricular tachycardia arising from the LVOT area whereas AVNRT was the first tachycardia induced by programmed stimulation in the electrophysiologic laboratory, although this had not been documented clinically. Slow pathway ablation was done because of patient request and report of future recurrence of AVNRT in such patients [<xref ref-type="bibr" rid="R7">7</xref>]. Then LVOT-tachycardia was induced and ablated successfully because catheter ablation of one arrhythmia substrate did not prevent inducibility or clinical recurrence of the other [<xref ref-type="bibr" rid="R7">7</xref>].</p><p>Co existence of AVNRT and LVOT-tachycardia may be a no more than chance association. This suggestion appears to be supported by presence of different mechanisms for both each type of tachycardia. On the other hand, some debate was made for this hypothesis in Kautzner study explaining this combination by presence of common trigger in patients with combination of AVNRT and RVOT-tachycardia [<xref ref-type="bibr" rid="R7">7</xref>].</p></sec><sec sec-type="" id="s4"><title>Conclusion</title><p>Our case demonstrated: 1) Presence of rare coexistence of AVNRT and LVOT-tachycardia 2) Feasibility of successful ablation of combination of AVNRT and LVOT-tachycardia in the same patient during one session.</p></sec>
Bundle Branch Reentrant Ventricular Tachycardia	<p>Bundle branch reentrant (BBR) tachycardia is an uncommon form of ventricular tachycardia (VT) incorporating both bundle branches into the reentry circuit. The arrhythmia is usually seen in patients with an acquired heart disease and significant conduction system impairment, although patients with structurally normal heart have been described. Surface ECG in sinus rhythm (SR) characteristically shows intraventricular conduction defects. Patients typically present with presyncope, syncope or sudden death because of VT with fast rates frequently above 200 beats per minute. The QRS morphology during VT is a typical bundle branch block pattern, usually left bundle branch block, and may be identical to that in SR. Prolonged His-ventricular (H-V) interval in SR is found in the majority of patients with BBR VT, although some patients may have the H-V interval within normal limits. The diagnosis of BBR VT is based on electrophysiological findings and pacing maneuvers that prove participation of the His- Purkinje system in the tachycardia mechanism. Radiofrequency catheter ablation of a bundle branch can cure BBR VT and is currently regarded as the first line therapy. The technique of choice is ablation of the right bundle. The reported incidence of clinically significant conduction system impairment requiring implantation of a permanent pacemaker varies from 0% to 30%. Long-term outcome depends on the underlying cardiac disease. Patients with poor systolic left ventricular function are at risk of sudden death or death from progressive heart failure despite successful BBR VT ablation and should be considered for an implantable cardiovertor-defibrillator.</p>	<contrib contrib-type="author"><name><surname>Mazur</surname><given-names>Alexander</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Kusniec</surname><given-names>Jairo</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Strasberg</surname><given-names>Boris</given-names></name><degrees>MD</degrees></contrib><aff>Cardiology Department, Rabin Medical Center, Beilinson Campus, Petah-Tikva and Sackler School of Medicine, Tel Aviv University, Israel</aff>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>Bundle branch reentrant (BBR) tachycardia is a form of ventricular tachycardia (VT) incorporating both bundle branches into the reentry circuit. Another variant of the His-Purkinje macro-reentry utilizing ramifications of the left bundle branch is referred to as interfascicular reentrant tachycardia.</p><p>Reentry within the His-Purkinje system (HPS) in humans was first documented by Akhtar et al [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>] in studies involving isolated ventricular beats commonly produced by programmed ventricular stimulation (also known as “V3 phenomenon”). These isolated reentrant beats within the HPS represent a normal response to stimulation. It is well recognized that sustained bundle branch reentry can not be induced in patients with normal HPS. The latter may be explained by electrophysiological properties of normal HPS characterized by the combination of very fast conduction velocity and a relatively long refractory period which precludes formation of a stable reentry circuit. Persistent bundle branch reentry as a mechanism of sustained VT has been demonstrated in patients with conduction system impairment [<xref ref-type="bibr" rid="R3">3</xref>-<xref ref-type="bibr" rid="R21">21</xref>]. In series of consecutive patients undergoing electrophysiological studies, this mechanism has been responsible for up to 6% of induced sustained monomorphic VT [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R10">10</xref>]. However, there is significant variability in the reported incidence of inducible BBR VT that could be related to the patient population studied as well as to the stimulation protocol used [<xref ref-type="bibr" rid="R22">22</xref>].</p></sec><sec sec-type="" id="s2"><title>Clinical characteristics of patients</title><p>BBR VT is usually seen in patients with an acquired structural heart disease and significant conduction system impairment [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R11">11</xref>]. Dilated cardiomyopathy, both ischemic and non-ischemic, is an important underlying cardiac pathology accounting for the majority of cases in the largest reported series [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R10">10</xref>]. The relative contribution of BBR mechanism to inducible sustained monomorphic VT is significantly higher in non-ischemic (up to 40%) than in ischemic etiology (up to 6%) [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R10">10</xref>]. Whether this relates to specific cardiac pathology or to significantly higher total incidence of inducible sustained monomorphic VT in patients with ischemic cardiomyopathy due to much more readily inducible intra-myocardial reentry is not clear. Valvular disease is another relatively common cause of bundle branch reentry. Aortic or mitral valve surgery can facilitate the development of the arrhythmia [<xref ref-type="bibr" rid="R11">11</xref>]. Close anatomical proximity of the HPS to the valvular annuli makes it vulnerable to surgical manipulations in this area. In contrast to postoperative intra-myocardial reentrant VT, BBR VT usually occurs in the early postoperative period (within first 2 weeks) and can be associated with preserved systolic left ventricular function. Bundle branch reentry is an important mechanism of arrhythmia in patients with myotonic myocardial dystrophy [<xref ref-type="bibr" rid="R12">12</xref>]. The disease is characterized by relatively selective and significant conduction system impairment. In the majority of these patients there is either minor myocardial involvement or none at all. Conduction abnormalities due to sodium channel blockade with flecainide have been implicated in the development of bundle branch reentry [<xref ref-type="bibr" rid="R13">13</xref>-<xref ref-type="bibr" rid="R14">14</xref>]. Isolated cases of this arrhythmia mechanism have been described in other diseases associated with conduction impairment [<xref ref-type="bibr" rid="R15">15</xref>-<xref ref-type="bibr" rid="R17">17</xref>]. The arrhythmia has also been reported in patients with idiopathic isolated conduction system disease and no apparent structural heart abnormalities [<xref ref-type="bibr" rid="R18">18</xref>-<xref ref-type="bibr" rid="R21">21</xref>]</p></sec><sec sec-type="" id="s3"><title>Baseline ECG</title><p>Surface ECG in sinus rhythm characteristically shows intraventricular conduction defects with or without PR interval prolongation. The conduction defects are presented by non-specific or typical bundle branch block patterns with prolonged QRS duration[<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R12">12</xref>]. Although total interruption of conduction in one of the bundle branches would theoretically prevent occurrence of bundle branch reentry, an ECG pattern of “complete” bundle branch block may not be an accurate marker of complete conduction block [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R23">23</xref>]. Rarely, BBR VT can also occur in patients with relatively narrow baseline QRS complex suggesting a role of functional conduction delay in the genesis of bundle branch reentry [<xref ref-type="bibr" rid="R24">24</xref>].</p></sec><sec sec-type="" id="s4"><title>Clinical features</title><p>Patients usually present with presyncope, syncope or sudden death because of VT with fast rates frequently above 200 beats per minute [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R12">12</xref>]. The surface ECG QRS morphology during VT is a typical bundle branch block pattern and may be identical to that in sinus rhythm [<xref ref-type="bibr" rid="R25">25</xref>]. A left bundle branch block (LBBB) pattern is the most commonly reported VT morphology [<xref ref-type="bibr" rid="R7">7</xref>].[<xref ref-type="bibr" rid="R12">12</xref>]. In contrast to VT of myocardial origin, bundle branch reentry with a LBBB pattern characteristically shows rapid intrinsicoid deflection in the right precordial leads indicating that initial ventricular activation occurs through the HPS. During BBR VT with a LBBB pattern the activation propagates in the antegrade direction down the right bundle (RB) and in the retrograde direction up the left bundle (LB). During BBR VT with a right bundle branch block (RBBB) pattern the direction of activation is reversed (<xref ref-type="fig" rid="F1">Figure 1A</xref>). The orientation of frontal plane QRS axis is usually to the left. Less frequently it may be to the right or normal. This may relate to the activation pattern of the left bundle branch fascicles.</p></sec><sec sec-type="" id="s5"><title>Electrophysiological data</title><p>Prolonged His-ventricular (H-V) interval in sinus rhythm is found in the majority of patients with BBR VT [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R12">12</xref>]. Although some patients may have the H-V interval in sinus rhythm within normal limits, functional HPS impairment in these patients manifested as H-V interval prolongation or split His-bundle potentials, commonly become evident during atrial programmed stimulation or burst pacing [<xref ref-type="bibr" rid="R24">24</xref>].</p><p>BBR VT is usually inducible with conventional pacing protocols. Both atrial and ventricular programmed stimulation or burst pacing can be useful. In some patients, the arrhythmia may be inducible only with atrial pacing stressing the importance of using atrial stimulation for evaluation of patients with VT or syncope [<xref ref-type="bibr" rid="R19">19</xref>]. Left ventricular stimulation, introduction of short-long-short sequences, isoproterenol infusion, or sodium channel blockade can aid in the induction of bundle branch reentry [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R9">9</xref>]. Atrial fibrillation or flutter may also facilitate the initiation of BBR VT [<xref ref-type="bibr" rid="R26">26</xref>]. Most induced BBR VTs demonstrate LBBB morphology. The initiation of bundle branch reentry with RBBB morphology may require left ventricular pacing [<xref ref-type="bibr" rid="R18">18</xref>].</p><p>The following electrophysiological features are consistent with the bundle branch reentry mechanism (<xref ref-type="fig" rid="F2">Figure 2</xref>): (1) reproducible initiation of tachycardia with critical V-H interval prolongation suggesting that induction of the tachycardia depends on conduction delay within the HPS; (2) a stable His or bundle branch potential preceding each ventricular activation. The H-V interval is usually longer than that recorded in sinus rhythm. However, in rare cases, it can be equal to or slightly shorter (by less than 15 ms) than the HV interval in sinus rhythm. During bundle branch reentry the His bundle is activated in the retrograde direction simultaneously with the proximal part of the bundle branch serving as the antegrade limb of the reentry circuit (<xref ref-type="fig" rid="F1">Figure 1A</xref>). The relative duration of the H-V interval recorded during VT as compared to sinus rhythm would depend on 2 factors: (a) the balance between antegrade and retrograde conduction times from the upper turnaround point of the reentry circuit; and (b) the site of His bundle recording relative to the upper turnaround point [<xref ref-type="bibr" rid="R27">27</xref>]. Conduction delay in the bundle branch used as the antegrade limb of the circuit would tend to prolong H-V interval during VT, while retrograde conduction delay to the His bundle recording site as well as the use of relatively proximal His bundle recording site (far from the turnaround point) would tend to shorten it. The RB (LB)-V interval must always be = than that recorded in sinus rhythm emphasizing the importance of recording the RB (LB) potential during VT [<xref ref-type="bibr" rid="R28">28</xref>]; (3) H-RB (LB)-V activation sequence consistent with ventricular activation through an appropriate bundle branch with regard to VT QRS morphology; (4) changes in H-H (RB-RB or LB-LB) interval during VT precede changes in the VV interval. In other words, the tachycardia cycle length is affected by variation in the V-H (V-RB or V-LB) interval. This would strongly argue against passive retrograde HPS activation. However, the opposite relation may not rule out the bundle branch reentry mechanism. V-V oscillations preceding H-H oscillations during BBR VT suggesting conduction variation in the antegrade rather than in the retrograde limb of the reentry circuit have been reported [<xref ref-type="bibr" rid="R29">29</xref>,<xref ref-type="bibr" rid="R30">30</xref>]. On the other hand, changes in the relationship between V and H (RB or LB) that do not affect the tachycardia cycle length would support lack of participation of the HPS in the tachycardia mechanism; (5) reproducible termination of VT with block in the HPS. This is another finding suggesting that the HPS is not a passive bystander; (6) inability to induce VT after ablation of the right or left bundle branch.</p><p>Recording from both sides of the septum may help in the identification of the bundle branch reentry mechanism. Documentation of typical H-RB-V-LB (during VT with LBBB morphology) or H-LB-V-RB (during VT with RBBB morphology) activation sequence would further support BBR VT diagnosis. In addition, during VT with LBBB morphology right ventricular excitation must precede the left ventricular excitation. The opposite is true for the VT with RBBB morphology [<xref ref-type="bibr" rid="R22">22</xref>].</p><p>Pacing maneuvers, if feasible, can be extremely helpful. Ability to dissociate His or, particularly, RB (LB) potential would strongly argue against bundle branch reentry mechanism. The combination of concealed entrainment (concealed QRS fusion) by atrial pacing and manifest entrainment (manifest QRS fusion) by ventricular pacing has been recently proposed as a useful diagnostic criterion for BBR VT with LBBB QRS morphology [<xref ref-type="bibr" rid="R30">30</xref>]. Analysis of the difference between the first postpacing interval after VT entrainment from the right ventricular apex and the tachycardia cycle length can be used to rapidly screen for VT mechanism. Bundle branch reentry is unlikely if the difference is longer than 30 ms [<xref ref-type="bibr" rid="R31">31</xref>]. The latter two findings can be potentially helpful in identifying the bundle branch reentry mechanism when stable recording of the His (RB or LB) potential during VT is not possible, although their ultimate utility needs further validation. Application of the pacing maneuvers during bundle branch reentry is often hampered by fast VT rates commonly associated with hemodynamic compromise. Furthermore, entrainment of BBR VT by atrial pacing has a limited success and usually requires isoproterenol infusion to improve atrio-ventricular (AV) nodal conduction.</p></sec><sec sec-type="" id="s6"><title>Interfascicular tachycardia</title><p>Interfascicular tachycardia has been less commonly reported [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R28">28</xref>,<xref ref-type="bibr" rid="R32">32</xref>-<xref ref-type="bibr" rid="R34">34</xref>]. BBR and interfascicular tachycardia may be present in the same patient [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R33">33</xref>,<xref ref-type="bibr" rid="R34">34</xref>]. The tachycardia usually has RBBB morphology. The orientation of the frontal plain axis is variable and may depend on the direction of the reentrant circuit. Antegrade activation over the left anterior fascicle and retrograde through the posterior fascicle would be associated with right axis deviation while the reversed activation sequence with left axis deviation. In contrast to BBR VT, the H-V interval during interfascicular tachycardia is usually shorter by more than 40 ms than that recorded in sinus rhythm [<xref ref-type="bibr" rid="R22">22</xref>]. This is because the upper turnaround point of the circuit (the left bundle branching point) is relatively far from the retrogradely activated His bundle (<xref ref-type="fig" rid="F1">Figure 1B</xref>). During interfascicular tachycardia the LB potential should be inscribed before the His potential (<xref ref-type="fig" rid="F3">Figure 3</xref>) [<xref ref-type="bibr" rid="R22">22</xref>]. During BBR VT with RBBB morphology the His potential usually precedes the LB potential, although the reverse is theoretically possible if retrograde conduction time to the His bundle recording point is significantly prolonged.</p></sec><sec sec-type="" id="s7"><title>Differential diagnosis</title><p>The differential diagnosis of BBR VT includes other mechanisms of VT and different types of supraventricular tachycardia with aberrant conduction. All electrophysiological findings and pacing maneuvers described above that prove participation of the HPS in the tachycardia mechanism and exclude passive retrograde activation of the HPS help to differentiate between bundle branch reentry and other mechanisms of VT. The exclusion of supraventricular tachycardia is particularly important because QRS morphology during BBR VT is a typical bundle branch block pattern and also may be similar to that in sinus rhythm. The differential diagnosis should be based on the complimentary use of the diagnostic criteria of bundle branch reentry as well as supraventricular tachycardia [<xref ref-type="bibr" rid="R22">22</xref>]. Since AV dissociation is typically present during BBR VT, the differential diagnosis is usually narrowed to AV nodal reentrant tachycardia, junctional tachycardia and hypothetical mechanisms such as intrahisian reentry and orthodromic tachycardia using a retrograde nodo-fascicular (ventricular) pathway. Orthodromic AV reciprocal tachycardia, tachycardia using an antegrade atrio-fascicular (ventricular) accessory pathway, and atrial tachycardia need to be also considered when 1:1 ventriculo-atrial conduction is present. Entrainment with manifest QRS fusion during ventricular pacing and ability to terminate or reset the tachycardia with ventricular extrastimulus introduced when His bundle is refractory will rule out atrial tachycardia, AV nodal reentrant tachycardia, junctional tachycardia and intrahisian reentry. In all types of supraventricular tachycardia, the difference between the first postpacing interval after entrainment from the right ventricular apex and the tachycardia cycle length should be much longer than 30 ms [<xref ref-type="bibr" rid="R31">31</xref>].</p></sec><sec sec-type="" id="s8"><title>Treatment</title><p>Pharmacologic antiarrhythmic therapy, both empiric and electrophysiologically guided, is usually ineffective. Radiofrequency catheter ablation of a bundle branch can cure BBR VT and is currently regarded as the first line therapy. The technique of choice is ablation of the RB [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R10">10</xref>]. Using this technique, the catheter is initially placed at the His bundle area and then gradually advanced toward anterior-superior ventricular septum with clockwise torque. The RB potential is identified by the following characteristics: (1) a sharp deflection inscribed at least 20 ms later than the His potential; and (2) absence of the atrial electrogram on the same recording. The RB-V interval value of < 30 ms may not be a reliable marker of the RB potential in these patients because of the HPS disease that can cause prolongation of RB-V conduction time [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R27">27</xref>]. The reported incidence of clinically significant conduction system impairment requiring implantation of a permanent pacemaker varies from 0% to 30% [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R12">12</xref>,<xref ref-type="bibr" rid="R24">24</xref>,<xref ref-type="bibr" rid="R34">34</xref>]. Ablation of the LB has been proposed in patients with preexisting LBBB pattern on the baseline ECG to reduce the need for pacing [<xref ref-type="bibr" rid="R35">35</xref>]. However, the technique is difficult to perform because of non-discrete anatomy of the LB that usually requires application of multiple lesions. Moreover, the RB ablation seems to be safe in the majority of patients with baseline LBBB pattern [<xref ref-type="bibr" rid="R8">8</xref>,<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R24">24</xref>].</p><p>Ablation of interfascicular tachycardia is guided by fascicular potentials. Successful ablation of the arrhythmia can be performed by targeting either the left anterior or posterior fascicle [<xref ref-type="bibr" rid="R9">9</xref>,<xref ref-type="bibr" rid="R19">19</xref>,<xref ref-type="bibr" rid="R32">32</xref>-<xref ref-type="bibr" rid="R34">34</xref>].</p></sec><sec sec-type="" id="s9"><title>Long-term outcome and management</title><p>The available data on long-term outcome of patients with BBR VT treated by catheter ablation come from small retrospective series including predominantly patients with left ventricular dysfunction [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R24">24</xref>,<xref ref-type="bibr" rid="R27">27</xref>]. These data suggest that the prognosis depends on the underlying cardiac disease. Patients with dilated cardiomyopathy and poor systolic LV function, especially those who have inducible VT other than BBR, are at high risk of non-BBR VT recurrence and sudden death despite successful abolition of bundle branch reentry. Progressive heart failure is a common cause of death in this population of patients [<xref ref-type="bibr" rid="R7">7</xref>-<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R24">24</xref>,<xref ref-type="bibr" rid="R27">27</xref>]. Therefore, most of these patients should be considered for an implantable cardiovertor-defibrillator (ICD) with or without cardiac resynchronization capabilities. Because BBR VT has a limited response to antiarrhythmic drugs and can be an important cause of repetitive ICD therapies, catheter ablation of the arrhythmia should always be considered as an important adjunct to the device therapy. The survival and management strategy of patients with bundle branch reentry and preserved left ventricular function who have no other risk markers for sudden death is less clear. Very limited data suggest that these patients may have a favorable long-term prognosis after successful ablation of bundle branch reentry [<xref ref-type="bibr" rid="R10">10</xref>,<xref ref-type="bibr" rid="R11">11</xref>,<xref ref-type="bibr" rid="R18">18</xref>]. Patients with myotonic dystrophy may need prophylactic permanent pacemaker implantation because of the progressive nature of the conduction system disease [<xref ref-type="bibr" rid="R12">12</xref>]. The recurrence of BBR VT after successful ablation would be extremely unlikely [<xref ref-type="bibr" rid="R8">8</xref>-<xref ref-type="bibr" rid="R10">10</xref>]. Complete RBBB pattern on the ECG appears to be a good marker of long-term success after ablation of the RB. However, it still remains unclear whether or not a routine follow-up electrophysiological study should be undertaken in patients who do not have ICD back-up because of potentially life threatening consequences of recurrent BBR VT.</p></sec>
Ventricular Tachycardia in the Absence of Structural Heart Disease	<p>In up to 10% of patients who present with ventricular tachycardia (VT), obvious structural heart disease is not identified. In such patients, causes of ventricular arrhythmia include right ventricular outflow tract (RVOT) VT, extrasystoles, idiopathic left ventricular tachycardia (ILVT), idiopathic propranolol-sensitive VT (IPVT), catecholaminergic polymorphic VT (CPVT), Brugada syndrome, and long QT syndrome (LQTS). RVOT VT, ILVT, and IPVT are referred to as idiopathic VT and generally do not have a familial basis. RVOT VT and ILVT are monomorphic, whereas IPVT may be monomorphic or polymorphic. The idiopathic VTs are classified by the ventricle of origin, the response to pharmacologic agents, catecholamine dependence, and the specific morphologic features of the arrhythmia. CPVT, Brugada syndrome, and LQTS are inherited ion channelopathies. CPVT may present as bidirectional VT, polymorphic VT, or catecholaminergic ventricular fibrillation. Syncope and sudden death in Brugada syndrome are usually due to polymorphic VT. The characteristic arrhythmia of LQTS is torsades de pointes. Overall, patients with idiopathic VT have a better prognosis than do patients with ventricular arrhythmias and structural heart disease. Initial treatment approach is pharmacologic and radiofrequency ablation is curative in most patients. However, radiofrequency ablation is not useful in the management of inherited ion channelopathies. Prognosis for patients with VT secondary to ion channelopathies is variable. High-risk patients (recurrent syncope and sudden cardiac death survivors) with inherited ion channelopathies benefit from implantable cardioverter-defibrillator placement. This paper reviews the mechanism, clinical presentation, and management of VT in the absence of structural heart disease.</p>	<contrib contrib-type="author"><name><surname>Srivathsan</surname><given-names>Komandoor</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="A1">1</xref></contrib><contrib contrib-type="author"><name><surname>Lester</surname><given-names>Steven J</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="A2">2</xref></contrib><contrib contrib-type="author"><name><surname>Appleton</surname><given-names>Christopher P</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="A2">2</xref></contrib><contrib contrib-type="author"><name><surname>Scott</surname><given-names>Luis RP</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="A2">2</xref></contrib><contrib contrib-type="author"><name><surname>Munger</surname><given-names>Thomas M</given-names></name><degrees>MD</degrees><xref ref-type="aff" rid="A1">1</xref></contrib>	Indian Pacing and Electrophysiology Journal	<sec sec-type="" id="s1"><title>Introduction</title><p>It is estimated that 10% of patients who present with ventricular tachycardia (VT) have no obvious structural heart disease [<xref ref-type="bibr" rid="R1">1</xref>]. An absence of structural heart disease is usually suggested if an electrocardiogram (ECG) (except in Brugada syndrome and long QT syndrome [LQTS]), echocardiogram, and coronary arteriogram collectively are normal [<xref ref-type="bibr" rid="R2">2</xref>]. However, structural abnormalities may be identified by magnetic resonance imaging (MRI) even if all other test results are normal [<xref ref-type="bibr" rid="R3">3</xref>]. In addition, focal dysautonomia in the form of localized sympathetic denervation has been reported in patients with VT and no other obvious structural heart disease [<xref ref-type="bibr" rid="R4">4</xref>].</p><p>Types of VT that occur in the absence of structural heart disease include right ventricular (RV) monomorphic extrasystoles, RV outflow tract (RVOT) VT, left ventricular (LV) outflow tract (LVOT) VT, idiopathic LV tachycardia (ILVT), idiopathic propranolol-sensitive (automatic) VT (IPVT), catecholaminergic polymorphic VT (CPVT), Brugada syndrome, and LQTS. RV monomorphic extrasystoles, RVOT VT, LVOT VT, ILVT, and IPVT are referred to as idiopathic VT. Idiopathic VT from the RVOT and LV are monomorphic and generally not familial. Idiopathic VTs are classified with respect to the ventricle of origin, the response to pharmacologic agents, evidence of catecholamine dependence, and the specific morphologic features (QRS morphology, axis, pattern, and whether tachycardia is repetitive, nonsustained, or sustained) (<xref ref-type="table" rid="T1">Table 1</xref>). CPVT, Brugada syndrome, and LQTS are inherited ion channelopathies.</p><p>In this review of VT in the absence of structural heart disease, we discuss the clinical recognition and management of idiopathic VT and inherited ion channelopathies. The articles were selected for review from a search of PubMed using search terms “idiopathic VT,” “LQTS,” “Brugada syndrome,” and “CPVT.” For each topic, articles focusing on diagnosis and management were preferentially selected.</p></sec><sec sec-type="" id="s2"><title>RV Monomorphic Extrasystoles and RVOT VT</title><p>RV monomorphic extrasystoles and RVOT VT appear to be on a continuum of the same process. RV monomorphic extrasystoles are characterized by ventricular ectopy with left bundle branch block (LBBB) morphology, and, on ECG, the QRS axis is directed inferiorly. Ventricular ectopy of this type was defined in 1969 and considered “typical for normal subjects” [<xref ref-type="bibr" rid="R5">5</xref>]. Resting ECG in these patients has no identifiable abnormalities, and the prognosis is generally benign. These extrasystoles occur more often during the day than at night and are transiently suppressed by sinus tachycardia [<xref ref-type="bibr" rid="R6">6</xref>]. The extrasystoles may diminish or disappear with exercise during stress testing. The site of origin is most often the RVOT and, to a lesser extent, the interventricular septum in the region of the RVOT [<xref ref-type="bibr" rid="R7">7</xref>]. An echocardiogram is normal in most of these patients [<xref ref-type="bibr" rid="R8">8</xref>], although anatomic changes, such as focal thinning and fatty replacement of the RVOT, have been demonstrated with MRI [<xref ref-type="bibr" rid="R3">3</xref>].</p><p>A potential relationship between these seemingly benign ventricular extrasystoles of RVOT origin and structural cardiac disease (arrhythmogenic RV dysplasia [ARVD]) has been investigated.9 Sixty-one patients with RVOT ventricular extrasystoles were contacted 15 years after their initial visit; no patient had died of sudden death nor developed ARVD in this study. Two-thirds of the patients were asymptomatic, and, in half, the ventricular ectopy had disappeared. Focal fatty replacement of the RV was present on MRI in most patients, in contrast to the diffuse pattern of fatty replacement observed in patients with ARVD.</p><p>In North America, 70% of cases of idiopathic VT arise from the RV, chiefly the RVOT just inferior to the pulmonic valve [<xref ref-type="bibr" rid="R2">2</xref>]. The characteristic morphology of RVOT VT is a wide QRS complex tachycardia with LBBB pattern and an inferior axis [<xref ref-type="bibr" rid="R10">10</xref>]. Among the outflow tract tachycardias, 90% originate from the RVOT and 10% from the LVOT. Either or both forms may be found in the same patient. RVOT VT is usually diagnosed in the third to fifth decade of life, although cases at the extremes of age have been reported. Most patients (80%) present with palpitations or presyncope (50%) but rarely present with frank syncope. Exercise or emotional stress usually precipitates the tachycardia. Sudden death is rare.</p><p>Two phenotypic forms of RVOT VT occur: nonsustained, repetitive, monomorphic VT (<xref ref-type="fig" rid="F1">Figure 1</xref>) [<xref ref-type="bibr" rid="R11">11</xref>] and paroxysmal, exercise-induced, sustained VT (<xref ref-type="fig" rid="F2">Figure 2</xref>). Both are terminated by the administration of adenosine. Specific ECG characteristics have been described to differentiate RVOT VT from LVOT VT [<xref ref-type="bibr" rid="R12">12</xref>]. LVOT VT may originate from either the supravalvular region of a coronary cusp or the infravalvular endocardial region of a coronary cusp of the aortic valve. The distinction between supravalvular and infravalvular location of the tachycardia has important therapeutic implications, particularly if radiofrequency (RF) ablation is performed. LVOT VT is suggested if the ECG during VT shows an S wave in lead I and an R-wave transition in lead V1 or V2. The absence of an S wave in V5 or V6 suggests a supravalvular location, whereas an S wave in leads V5 and V6 indicates an infravalvular location (<xref ref-type="fig" rid="F3">Figure 3</xref>). In addition, in leads V1 and V2, an R:S amplitude ratio of 30% or more or an R:QRS duration ratio of 50% or more suggests an LV (aortic sinus cusp) origin of the tachycardia [<xref ref-type="bibr" rid="R13">13</xref>].</p><p>Exercise stress testing is used frequently to initiate and evaluate RVOT VT (unlike RV monomorphic extrasystoles, which are suppressed by sinus tachycardia) but is not clinically helpful in most cases. Initiation of the tachycardia depends on a critical heart rate that differs in each patient. The VT may be initiated during exercise or recovery [<xref ref-type="bibr" rid="R14">14</xref>]. ECG and echocardiogram in sinus rhythm are usually normal, as is coronary angiography. MRI may show abnormalities of the RV in up to 70% of patients, including focal thinning, diminished systolic wall thickening, and abnormal wall motion [<xref ref-type="bibr" rid="R15">15</xref>].</p><p>RVOT VT should be distinguished from ARVD, a disorder with a more serious clinical outcome. The VT in ARVD may have morphologic features similar to RVOT VT (LBBB with inferior axis) but does not terminate with adenosine. In ARVD, the resting 12-lead ECG typically shows inverted T waves in right precordial leads. When present, RV conduction delay with an epsilon wave (<xref ref-type="fig" rid="F4">Figure 4</xref>), best seen in leads V1-V2, is helpful in the diagnosis of ARVD. Measurement of serum brain natriuretic peptide may help distinguish ARVD from RVOT VT [<xref ref-type="bibr" rid="R16">16</xref>]. The level of brain natriuretic peptide is increased in ARVD, most likely due to increased expression by the surviving myocytes surrounded by atrophic tissue, which is indicative of the severity of RV dysfunction. Mechanisms of ARVD-related VT include both reentry facilitated by slow conduction through areas of fatty infiltration and increased automaticity [<xref ref-type="bibr" rid="R17">17</xref>]. In ARVD, the areas typically affected on echocardiography or MRI include the apex, interventricular septum below the tricuspid septal leaflet, and the RVOT. In some cases, fatty infiltration of the LV occurs. An RV biopsy and histopathologic characterization may help determine the correct diagnosis.</p><p>The differential diagnosis of RVOT VT also includes tachycardias associated with atriofascicular fibers (Mahaim fibers), atrioventricular reentrant tachycardia using a right-sided accessory pathway, and VT occurring in patients after repair of tetralogy of Fallot.</p><sec id="s2a"><title>Mechanism of RVOT VT</title><p>Intracellular calcium overload appears to be the principal underlying mechanism of RVOT VT. Cytosolic calcium overload enhances the function of the Na+/Ca2+ exchanger, which leads to increasing inward current and delayed afterdepolarization. When the inward current is of sufficient threshold, the delayed afterdepolarization may cause another action potential and initiate tachycardia. Cyclic adenosine monophosphate (cAMP) has a substantial role in regulating intracellular calcium. When the concentration of cAMP is increased, intracellular calcium levels are high. Adenosine is effective in terminating RVOT VT because of its ability to lower cAMP concentration [<xref ref-type="bibr" rid="R18">18</xref>]. Beta-blockers are often effective because of their inhibition of adenylate cyclase, which leads to a decrease of cAMP. Verapamil inhibits L-type calcium channels, which decreases the concentration of intracellular calcium and thereby has salutary effects.</p><p>Triggered activity, rather than reentry or enhanced automaticity, as the cause of RVOT VT is evidenced by termination with administration of adenosine and inability to entrain. The tachycardia may be inducible by programmed extrastimuli or by burst pacing the ventricle or atrium or by infusion of isoproterenol. Somatic mutation involving the G-protein signaling cascade could give rise to RVOT VT by disrupting adenosine signaling. Of interest, mutation of the G protein subunit Alphai2 has been identified on myocardial biopsy in only the RVOT (the site of origin) and not in myocardium remote from the site of VT [<xref ref-type="bibr" rid="R19">19</xref>].</p></sec><sec id="s2b"><title>Treatment of RVOT VT</title><p>Acute termination of RVOT VT can be achieved by vagal maneuver or intravenous administration of adenosine, 6 mg, which can be titrated up to 24 mg as needed. Intravenous verapamil, 10 mg, given over 1 minute is an alternative, provided the patient has adequate blood pressure and has a previously established diagnosis of verapamil-sensitive VT. Lidocaine also may be effective in some cases. Hemodynamic instability warrants emergent cardioversion.</p><p>Long-term treatment options for RVOT VT include medical therapy or RF ablation. Medications, including beta-blockers or verapamil (diltiazem is equally effective), have a 25% to 50% rate of efficacy [<xref ref-type="bibr" rid="R20">20</xref>]. Alternative therapy includes class IA, class IC, and class III agents including amiodarone [<xref ref-type="bibr" rid="R20">20</xref>]. RF ablation now has cure rates of 90%, [<xref ref-type="bibr" rid="R10">10</xref>] which makes it a preferable option, given the young age of patients with RVOT VT. Ablation of sites at the aortic sinus cusp has been successful for treatment of LVOT VT [<xref ref-type="bibr" rid="R13">13</xref>], but serious complications may occur, including left main coronary artery occlusion. Coronary arteriography before and during ablation is recommended to keep the tip of the ablation catheter 1 cm away from the ostia of the coronary arteries. After ablation, arteriography should be repeated to assess the patency of coronary arteries. Epicardial foci of the LVOT remain a challenging ablation target.</p></sec></sec><sec sec-type="" id="s3"><title>ILVT</title><p>Most VTs of LV origin are verapamil-sensitive intrafascicular tachycardias. Intrafascicular tachycardia has a right bundle branch block (RBBB) left-axis configuration in 90% to 95% of cases (exit site, left posterior fascicle) and the rest have RBBB with a right-axis pattern (exit site, left anterior fascicle). This form of VT is seen in the second to fourth decade of life and occurs more often in men (60%-80%) [<xref ref-type="bibr" rid="R21">21</xref>]. Symptoms during tachycardia include palpitations, dizziness, presyncope, and syncope. Sudden death is usually not seen, but one possible case has been reported [<xref ref-type="bibr" rid="R22">22</xref>].</p><p>A proposed diagnostic triad of ILVT includes: 1) induction with atrial pacing, 2) RBBB with left axis configuration, and 3) no evidence of structural heart disease [<xref ref-type="bibr" rid="R23">23</xref>]. A fourth feature, verapamil sensitivity, has since been described [<xref ref-type="bibr" rid="R24">24</xref>].</p><sec id="s3a"><title>Mechanism of ILVT</title><p>Focal reentry appears to be the principal mechanism of ILVT. The tachycardia cycle length can be increased with the administration of verapamil [<xref ref-type="bibr" rid="R25">25</xref>]. Some evidence implicates the Purkinje fibers of the fascicle as the area of slow conduction because of the presence of high-frequency potentials (Purkinje potentials) [<xref ref-type="bibr" rid="R26">26</xref>]. Others have found late diastolic potentials near the main trunk of the left bundle branch. Another hypothesis implicates a false tendon extending from the inferoposterior aspect of the LV to the basal septum as directly or indirectly having a role in causing this arrhythmia [<xref ref-type="bibr" rid="R27">27</xref>].</p></sec><sec id="s3b"><title>Treatment of ILVT</title><p>In the acute setting, this tachycardia responds to intravenous verapamil. Termination with adenosine is rare, except for cases in which isoproterenol is used for induction of the tachycardia. Long-term therapy with verapamil is useful in mild cases and RF ablation is highly effective (85%-90%) in those with severe symptoms [<xref ref-type="bibr" rid="R21">21</xref>]. Identifying the focus of ablation may involve recognition of Purkinje potential, late diastolic potential, or earliest ventricular activation. Electroanatomic mapping may help localize the area of slow conduction [<xref ref-type="bibr" rid="R28">28</xref>]. In about 10% of cases of both ILVT and RVOT VT, a tachycardia with a different morphology may be inducible after successful ablation of clinical VT. This second tachycardia may be a cause for recurrence and should preferably be ablated during the initial attempt [<xref ref-type="bibr" rid="R29">29</xref>].</p></sec></sec><sec sec-type="" id="s4"><title>IPVT</title><p>This form of idiopathic VT usually occurs by the fifth decade of life and can arise from the LV or RV [<xref ref-type="bibr" rid="R21">21</xref>]. The morphology of the tachycardia may be monomorphic or polymorphic. IPVT is not inducible with programmed stimulation. Isoproterenol infusion usually induces this VT. Beta-blockers are effective in terminating the tachycardia.</p><sec id="s4a"><title>Treatment of IPVT</title><p>Beta-blockers are used to treat this form of VT because they are effective in acute situations. There is insufficient information available regarding long-term management of IPVT. Survivors of sudden cardiac death may receive an implantable cardioverter-defibrillator (ICD).</p></sec></sec><sec sec-type="" id="s5"><title>Inherited Channelopathies</title><sec id="s5a"><title>CPVT</title><p>CPVT is characterized by a uniform pattern of bidirectional polymorphic VT that can be easily and reproducibly induced during exercise or catecholamine infusion. A third of patients with CPVT have a family history of premature sudden death or stress-related syncope [<xref ref-type="bibr" rid="R30">30</xref>]. Exercise or acute emotion usually triggers syncope. Symptoms typically manifest in childhood; onset in adulthood has been reported but is uncommon.</p><p>The ryanodine receptor 2 (RyR2) is important for the regulation of intracellular calcium fluxes [<xref ref-type="bibr" rid="R31">31</xref>]. In patients with CPVT, the RyR2 gene is mutated, with autosomal dominant inheritance suggested [<xref ref-type="bibr" rid="R32">32</xref>]. One family with recessive CPVT has been reported, and the gene responsible produces the protein calsequestrin, which is functionally related to RyR2 [<xref ref-type="bibr" rid="R33">33</xref>].</p><p>CPVT and IPVT can be distinguished by means of family history, morphology (CPVT is usually bidirectional), age of onset (childhood vs the fifth decade), and in some cases genetic testing (genetic defect vs idiopathic).</p><sec id="s5a1"><title>Treatment</title><p>Beta-blockers are the preferred therapy for CPVT [<xref ref-type="bibr" rid="R30">30</xref>]. Beta-blockers may prevent syncope and sudden death because adrenergic activation is the main mechanism of delayed after depolarization-dependent triggered activity in these patients [<xref ref-type="bibr" rid="R34">34</xref>]. An ICD is required in 30% of patients because of symptomatic recurrence of life-threatening arrhythmia in spite of beta-blocker therapy [<xref ref-type="bibr" rid="R35">35</xref>].</p></sec></sec><sec id="s5b"><title>Brugada Syndrome</title><p>Brugada syndrome is characterized by apparent RBBB with ST elevation in V1 to V3 (V2 always present) (<xref ref-type="fig" rid="F5">Figure 5</xref>), life-threatening cardiac arrhythmia (polymorphic VT) with no demonstrable structural cardiac disease, and familial occurrence [<xref ref-type="bibr" rid="R36">36</xref>,<xref ref-type="bibr" rid="R37">37</xref>]. The ECG changes may mimic acute myocardial infarction. The ECG findings may not be evident on resting 12-lead ECG but may be unmasked by flecainide or procainamide [<xref ref-type="bibr" rid="R38">38</xref>,<xref ref-type="bibr" rid="R39">39</xref>]. Two different types of ST elevation have been described: coved and saddleback. The coved type is more relevant to the syndrome than is the saddleback type [<xref ref-type="bibr" rid="R40">40</xref>]. Genetic analysis indicates that Brugada syndrome is due to mutation of the SCN5A protein [<xref ref-type="bibr" rid="R41">41</xref>]. The incidence of the disease is about 5 per 10,000 persons. The Brugada-type ECG (“Brugada sign”) may be much more common than is the clinical syndrome [<xref ref-type="bibr" rid="R42">42</xref>]. Sudden death is usually due to polymorphic VT or ventricular fibrillation. The disease predominantly affects young males.</p><p>The risk of sudden cardiac death with Brugada syndrome is substantial. In a study of 334 patients with typical Brugada-type ECG findings, which included symptomatic (cardiac arrest and syncope) and asymptomatic patients, the risk of recurrent events during 4 years of follow-up was 62% for those with cardiac arrest and 19% for those with syncope [<xref ref-type="bibr" rid="R37">37</xref>]. The asymptomatic group had an 8% event rate during 2 years of follow-up.</p><sec id="s5b1"><title>Management</title><p>ICD placement is the treatment of choice in symptomatic patients. Asymptomatic patients with Brugada-type ECG results should undergo electrophysiologic testing. If ventricular arrhythmia is inducible (two-thirds of patients are noninducible) the patient should receive an ICD. Asymptomatic patients with normal baseline ECG do not require further testing.</p></sec></sec></sec><sec sec-type="" id="s6"><title>LQTS</title><p>LQTS is an uncommon disorder in the general population. It is an inherited disorder, and mutations in 7 genes for LQTS have been identified to date (<xref ref-type="table" rid="T2">Table 2</xref>) [<xref ref-type="bibr" rid="R43">43</xref>]. This syndrome was initially identified in a family in which several children had syncope and sudden death. A recessive inheritance pattern was identified, and the syndrome was associated with deafness (Jervell and Lange-Nielsen syndrome). A similar and more common disorder without deafness, inherited in an autosomal dominant pattern, was subsequently identified (Romano-Ward syndrome).</p><sec id="s6a"><title>Clinical Diagnosis</title><p>Syncope, sudden cardiac death, or family screening of an affected individual is the reason that physicians evaluate patients for LQTS. Prolonged QT interval on ECG makes a diagnosis of LQTS likely. Medications that prolong QT interval must be carefully excluded from the patient’s medication list. Family history may be helpful for diagnosis of LQTS. An incidental finding of prolonged QTc (not due to medications) in an asymptomatic person is rare. Syncope and sudden death with LQTS occur with higher frequency during adolescence.</p></sec><sec id="s6b"><title>Triggers of Clinical Events</title><p>In patients with LQT1 subtype, exercise seems to precipitate clinical events [<xref ref-type="bibr" rid="R44">44</xref>]. In those with LQT2, acute arousal, such as a sudden loud noise, tends to be a precipitating factor. In patients with LQT3, clinical events occur at rest or during sleep.</p></sec><sec id="s6c"><title>Clinical Course</title><p>The risk of cardiac events is higher with certain genotypes; patients with LQT1 and LQT2 have higher risk of events than do those with LQT3 [<xref ref-type="bibr" rid="R45">45</xref>]. The risk of events also is higher during adulthood in females and during adolescence (before puberty) in males. The length of the QTc interval and the number of mutations also increase the risk. Once a clinical event occurs (syncope or survival after sudden cardiac death), recurrence is frequent.</p></sec><sec id="s6d"><title>ECG Findings</title><p>Eighty percent of LQT1 and LQT2 carriers and 65% of LQT3 carriers have typical ECGs. In LQT1, the T wave is broad-based with an indistinct onset. In LQT2, bifid T waves may be seen in all 12 leads, and the ECG in LQT3 may have a long isoelectric ST segment [<xref ref-type="bibr" rid="R46">46</xref>].</p></sec><sec id="s6e"><title>Treatment</title><p>Adrenergic modulation with beta-blockers is the most useful therapy in both symptomatic and asymptomatic patients, even though beta-blockers do not alter QTc interval [<xref ref-type="bibr" rid="R47">47</xref>]. However, the benefits of beta-blockers have not been proven in a randomized trial. Surgical sympathectomy is an adjuvant treatment and has been done rarely since the introduction of beta-blockers. Oral potassium may be useful in certain genotypes [<xref ref-type="bibr" rid="R48">48</xref>]. ICD placement, along with beta-blocker therapy, offers the best protection in high-risk patients (survivors of sudden death and those with recurrent syncope) [<xref ref-type="bibr" rid="R49">49</xref>].</p></sec></sec><sec sec-type="" id="s7"><title>Acquired LQTS and Torsades de Pointes</title><p>Acquired prolongation of QT interval and pause-dependent, early afterdepolarization-mediated torsades de pointes most often is caused by medication and occasionally is caused by metabolic derangement (hypokalemia and hypomagnesemia) [<xref ref-type="bibr" rid="R50">50</xref>]. Correction of electrolyte abnormalities and discontinuation of precipitating drugs usually lead to amelioration of the arrhythmia. Intravenous magnesium, although it may not have an effect on QT interval, is highly effective at suppressing torsades de pointes. Occasionally, a temporary transvenous pacemaker or isoprenaline may be needed for effective management.</p></sec><sec sec-type="" id="s8"><title>Diagnostic Approach to Ventricular Arrhythmia in the Absence of Structural Heart Disease</title><p>The approach to diagnosis of the subtypes of VT in the absence of heart disease depends on the morphology of the tachycardia precipitating the clinical event. If the presentation is monomorphic VT, RVOT VT, LVOT VT, ILVT, and, rarely, IPVT are in the differential diagnosis. Specific ECG criteria should help clarify the diagnosis (<xref ref-type="fig" rid="F6">Figure 6</xref>). If the inciting clinical event is precipitated by polymorphic VT, torsades de pointes, or ventricular fibrillation, the diagnosis may be approached by evaluating the baseline or postresuscitation QTc interval (without antiarrhythmic medications) (<xref ref-type="fig" rid="F7">Figure 7</xref>). If the QTc is prolonged, LQTS and its subtypes are the predominant diagnoses, provided drug-induced QTc prolongation can be reasonably excluded. If the QTc is normal, Brugada syndrome, CPVT, and IPVT are the conditions in the differential diagnosis. The Brugada ECG, either at baseline or on induction with antiarrhythmic medication, may help identify Brugada syndrome. Bidirectional VT and presentation during childhood identify most patients with CPVT. IPVT remains a diagnosis when all other causes are unlikely and the arrhythmia is propranolol sensitive. Genetic testing may be helpful in the long-term evaluation of polymorphic VT but acutely is not helpful.</p></sec><sec sec-type="" id="s9"><title>Conclusion</title><p>Ventricular arrhythmia in the absence of structural heart disease is a small subset in the clinical spectrum of patients with VT. Overall, the prognosis is better in patients with idiopathic ventricular arrhythmia than in patients with structural heart disease and VT. Prognosis in hereditary channelopathies is variable; CPVT, in particular, has a malignant course when untreated. Understanding the different characteristics of these tachycardias, their diagnostic features, and physiologic substrates is essential for successful therapy and management. Stress testing and response to antiarrhythmics each have an important role in identifying the specific arrhythmia. RF ablation and placement of an ICD are important in the overall management of specific arrhythmia.</p></sec>
Risk Stratification for Sudden Cardiac Death In Patients With Non-ischemic Dilated Cardiomyopathy	<p>Non ischemic dilated cardiomyopathy (NIDCM) is a disorder of myocardium. It has varying etiologies. Albeit the varying etiologies of this heart muscle disorder, it presents with symptoms of heart failure, and rarely as sudden cardiac death (SCD). Manifestations of this disorder are in many ways similar to its counterpart, ischemic dilated cardiomyopathy (IDCM). A proportion of patients with NIDCM carries a grave prognosis and is prone to sudden cardiac death from sustained ventricular arrhythmias. Identification of this subgroup of patients who carry the risk of sudden cardiac death despite adequate medical management is a challenge .Yet another method is a blanket treatment of patients with this disorder with anti arrhythmic medications or anti tachyarrhythmia devices like implantable cardioverter defibrillators (ICD). However this modality of treatment could be a costly exercise even for affluent economies. In this review we try to analyze the existing data of risk stratification of NIDCM and its clinical implications in practice.</p>	<contrib contrib-type="author"><name><surname>Shekha</surname><given-names>Karthik</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Ghosh</surname><given-names>Joydeep</given-names></name><degrees>MD</degrees></contrib><contrib contrib-type="author"><name><surname>Thekkoott</surname><given-names>Deepak</given-names></name><degrees>MD, MRCP (UK)</degrees></contrib><contrib contrib-type="author"><name><surname>Greenberg</surname><given-names>Yisachar</given-names></name><degrees>MD</degrees></contrib><aff> Department of Cardiology and Clinical Electrophysiology, Maimonides Medical Center (Mount Sinai Health System), Brooklyn, New York.</aff>	Indian Pacing and Electrophysiology Journal	<p>Non-ischemic dilated cardiomyopathy (NIDCM) is a primary disease of the myocardium, characterized by dilatation of all four chambers of the heart, but primarily the left ventricle, with associated systolic dysfunction. The incidence of dilated cardiomyopathy is 5 - 8 / 100,000 / year [<xref ref-type="bibr" rid="R1">1</xref>,<xref ref-type="bibr" rid="R2">2</xref>]. In an international heart failure study it was found that 18% of symptomatic patients with ejection fraction less than 30% were diagnosed with NIDCM[<xref ref-type="bibr" rid="R3">3</xref>]. The age-adjusted mortality in patients with NIDCM ranges from 0.10 per 10,000 person-years among men aged 35-39 up to 1.16 per 10,000 person-years among men aged 55-57 [<xref ref-type="bibr" rid="R2">2</xref>]. Currently, mortality in NIDCM is 12-13% at 3 years [<xref ref-type="bibr" rid="R4">4</xref>]. Independent risk factors for death include smoking, diabetes mellitus, and high diastolic blood pressure [<xref ref-type="bibr" rid="R2">2</xref>]. Patients with NIDCM suffer from heart failure mortality and sudden cardiac death in near equal numbers [<xref ref-type="bibr" rid="R5">5</xref>]. Sudden cardiac death (SCD) may well be the first manifestation of NIDCM, and idiopathic NIDCM is responsible for 10% of all sudden cardiac deaths in adults [<xref ref-type="bibr" rid="R6">6</xref>]. Survivors of SCD with NIDCM often have recurrent ventricular fibrillation as their clinical arrhythmia [<xref ref-type="bibr" rid="R7">7</xref>].</p><p>The majority of SCDs occur among patients that are classically defined as low risk, including those with NYHA functional class II or I. In the MERIT-HF trial (NYHA heart failure class II-IV, EF < 40%), the most common cause of mortality in patients with NYHA functional classes II and III was sudden cardiac death[<xref ref-type="bibr" rid="R8">8</xref>]. The incidence of SCD in NYHA class IV heart failure is also high but the competing risk of pump failure makes it the second cause of death in this very sick category of patients.</p><p>With the advent of the implantable cardioverter defibrillator (ICD) and their near-perfect termination of lethal arrhythmias, a determination needs to be made of which patients would benefit most from ICD insertion. In order to balance the potential risks of device implantation and the associated cost, many investigators have sought to establish risk factors to select patients with NIDCM who would benefit most from ICD implantation.</p><sec sec-type="" id="s1"><title>Secondary Prevention Trials</title><p>Patients with symptomatic NIDCM and documented ventricular tachycardia or ventricular fibrillation (VT/VF) have a mortality rate of up to 22% in the first two years after diagnosis. This is comparable to a similar group of patients in the AVID registry with CAD[<xref ref-type="bibr" rid="R9">9</xref>]. It has been clearly established that patients with symptomatic sustained ventricular tachycardia and underlying structural heart disease benefit from ICD implantation irrespective of the etiology of their heart disease [<xref ref-type="bibr" rid="R10">10</xref>-<xref ref-type="bibr" rid="R21">21</xref>]. Data from the AVID registry reveal a similar mortality rate in patient’s surviving symptomatic VT or VF regardless of the underlying nature of cardiac disease. The AVID trial, on the other hand, randomized patients resuscitated from ventricular arrhythmias to either ICD therapy or treatment with class III antiarrhythmic agents (primarily amiodarone) [<xref ref-type="bibr" rid="R22">22</xref>]. Patients were randomized (n=1016) and 3117 patients were followed in the AVID registry. The registry comprised 73% of the patients with CAD, while the remaining 27% were diagnosed as NIDCM. 54% of NIDCM patients received ICDs, while 48% of CAD patients received ICDs. Analysis of the registry revealed that the overall 2-year survival of patients with NIDCM was 78.2%, which was similar to those with CAD.</p></sec><sec sec-type="" id="s2"><title>Primary Prevention</title><p>Primary prevention of sudden cardiac death is an important goal in the NIDCM population. Two trials using amiodarone in a randomized fashion against placebo have been performed in the NIDCM population. The GESICA trial randomized 516 patients with predominantly non-ischemic cardiomyopathies and ejection fraction <35% on optimal medical therapy to amiodarone or placebo [<xref ref-type="bibr" rid="R23">23</xref>]. The trial demonstrated a substantial reduction in all cause mortality in patients treated with amiodarone (33.5% vs. 41.4% RR reduction 28%, 95% CI 4-45%) independent of the presence of ventricular arrhythmias. Although the trial was randomized, it was criticized for its non-blinded enrollment. Also many patients enrolled in this trial had cardiomyopathy from Chaga’s disease, which is a very different group of patients compared to NIDCM patients in the US or other parts of the world.</p><p>Similarly, CHF-STAT evaluated the use of amiodarone in a predominantly though not exclusively ischemic cardiomyopathy population with ≥ 10 PVCs /hr. Although the overall study showed no mortality benefit of amiodarone therapy, subgroup analysis of the 193 non-ischemic participants demonstrated a trend towards reduction in all cause mortality [<xref ref-type="bibr" rid="R24">24</xref>], with a P value of 0.07. One of the intriguing findings of this randomized, double blind trial was that arrhythmia suppression with amiodarone had no effect on survival.</p><p>The Cardiomyopathy Trial (CAT) randomized 104 patients with NIDCM to ICD placement versus optimal medical management and followed them for greater than five years (5.5 ± 2.2 years). ICD therapy had no significant benefit in comparison to medical therapy [<xref ref-type="bibr" rid="R25">25</xref>]. The trial may have been underpowered to detect a survival benefit, as the overall mortality did not reach the anticipated 30% in the control group.</p><p>The Amiodarone Versus Implantable Cardioverter-Defibrillator in patients with non-ischemic dilated cardiomyopathy and asymptomatic nonsustained ventricular tachycardia (AMIOVIRT trial) evaluated empiric amiodarone therapy vs. prophylactic ICD implantation in 103 patients with NIDCM, EF ≤35%, and asymptomatic NSVT. The three-year survival rate was 89% in both arms of the study with no significant difference between amiodarone and ICD placement in either survival or quality of life endpoints (the trial was stopped at the first interim analysis). Patients treated with amiodarone, however, showed a trend toward improved arrhythmia-free survival compared to the ICD arm.</p><p>These two smaller trials, evaluating optimal medical therapy [<xref ref-type="bibr" rid="R25">25</xref>] and amiodarone [<xref ref-type="bibr" rid="R4">4</xref>] against prophylactic ICD placement, have suggested medical therapy to be as effective as ICD implantation in preventing all cause mortality in asymptomatic patients with NIDCM and non-sustained ventricular tachycardia (NSVT). A more definitive assessment of the benefit of prophylactic ICD implantation in the NIDCM population was, however, performed in the Defibrillators in Non-ischemic Cardiomyopathy Treatment Evaluation (DEFINITE) trial. DEFINITE compared the use of the ICD with standard oral medical therapy vs. medical therapy only of patients with NIDCM and NSVT. Inclusion criteria for the trial included EF ≤35%, symptomatic heart failure, and documented ventricular ectopy (> 10 PVCs/hour) or NSVT on Holter monitor or telemetry within the last 6 months [<xref ref-type="bibr" rid="R26">26</xref>]. The trial randomized 458 patients. Arrhythmia mortality accounted for 33% of deaths. At 2 years, mortality in the medical therapy arm was 13.8%, and was reduced to 8.1 % in the ICD arm (p=0.06). The ICD group had a 74% reduction in arrhythmic deaths (p ≤ 0.05). The trial’s primary endpoint (total mortality) failed to reach statistical significance; however, ICD’s were associated with a significantly lower rate of arrhythmic death, the study’s secondary endpoint. Subgroup analysis found that male patients, EF < 20%, QRS duration > 120 msec, and NYHA Class III received the greatest benefit from ICD implantation. NYHA class III was associated with a 67% risk reduction in all cause mortality (p=0.009). The DEFINITE trial has also been criticized for being underpowered.</p><p>The Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) is a recently published primary prevention trial. It enrolled patients with EF < 36% and NYHA class II or III heart failure [<xref ref-type="bibr" rid="R27">27</xref>] into three arms, medical therapy vs. medical therapy plus amiodarone vs. medical therapy plus ICD. The endpoint of the study was all-cause mortality. About half of the 2521 patients enrolled had an underlying diagnosis of NIDCM. The five-year all-cause mortality in the ICD group was 28.9%, compared to 34.1% in the amiodarone group and 35.8% in the placebo group. This resulted in a reduction in five-year all-cause mortality by 23% with ICD use compared with the control group. Amiodarone had no effect on survival. The conclusion is that ICD implantation saves lives (but amiodarone doesn't) in patients with class 2 or 3 heart failure, irrespective of the etiology of heart failure. It is interesting to note that in patients with class III heart failure, irrespective of etiology, ICD implantation was associated with a significant survival advantage when compared to amiodarone but not when compared to placebo. As a result of this large randomized, double-blind trial incorporating a placebo arm, prophylactic ICD implantation in NIDCM patients with a LVEF < 36% and NYHA functional class II or III seems justified and risk stratification in this group of patients may no longer be required. Other patient groups (eg. LVEF 36-40% or NYHA class I functional status) will, however, still need risk stratification.</p></sec><sec sec-type="" id="s3"><title>Risk Stratification</title><p>Based on the primary prevention trials there seems to be still some controversy regarding improved survival with ICD therapy in patients with NIDCM. It is believed that benefit would be more readily achievable if a higher risk subgroup could be selected instead of allocating this expensive therapy to the broad population of patients with NIDCM. A number of diagnostic methods have been used to risk stratify patients with cardiomyopathy for elevated risk for sudden death. History of syncope [<xref ref-type="bibr" rid="R28">28</xref>-<xref ref-type="bibr" rid="R30">30</xref>], Electrocardiographic monitoring [<xref ref-type="bibr" rid="R31">31</xref>-<xref ref-type="bibr" rid="R36">36</xref>], programmed electrical stimulation (PES) [<xref ref-type="bibr" rid="R37">37</xref>-<xref ref-type="bibr" rid="R56">56</xref>], signal-averaged ECG (SAECG) [<xref ref-type="bibr" rid="R56">56</xref>-<xref ref-type="bibr" rid="R58">58</xref>], heart rate variability [<xref ref-type="bibr" rid="R48">48</xref>-<xref ref-type="bibr" rid="R59">59</xref>], QT dispersion [<xref ref-type="bibr" rid="R46">46</xref>], and baroreceptor sensitivity testing [<xref ref-type="bibr" rid="R60">60</xref>], heart rate turbulence [<xref ref-type="bibr" rid="R63">63</xref>-<xref ref-type="bibr" rid="R65">65</xref>] have all been evaluated.</p></sec><sec sec-type="" id="s4"><title>Syncope</title><p>Syncope may be a harbinger of SCD. Comparisons of patients without documented VT/VF but with a history of unexplained syncope and patients who have survived documented cardiac arrest reveal similar mortality [<xref ref-type="bibr" rid="R29">29</xref>]. In one particular trial 23 patients with NIDCM and syncope were compared to a similar group of 201 patients that did not have syncope [<xref ref-type="bibr" rid="R28">28</xref>]. This trial was non-randomized, with more frequent use of amiodarone in the syncope group (p < 0.04). During the follow-up, there was no statistical difference in mortality between the two groups. However, 83% of the mortality in the syncope group was due to sudden death, compared to 32% in the control group (P < 0.025).</p><p>Knight et al. prospectively followed 14 patients who presented with unexplained syncope and who were diagnosed with NIDCM, and had negative electrophysiology study [<xref ref-type="bibr" rid="R29">29</xref>] along with a control group of 19 patients with NIDCM who sustained a cardiac arrest. All patients in both arms of the study received an ICD. During the 24 ± 13 months of follow-up, 50% of patients with unexplained syncope received appropriate ICD shocks in comparison to 42% of the patients who received an ICD for cardiac arrest (P=0.1). Time to first appropriate ICD shock actually occurred earlier in the syncope patients as compared to those with cardiac arrest (32±7 vs. 72±12 months, p =0.01). Due to the small number of patients in this study, the results were not statistically significant. However, the trend suggests that syncope is a marker for arrhythmic death in patients with NIDCM.</p><p>Fonarow et al. evaluated 147 patients with NIDCM, syncope, and severely decreased left ventricular function [<xref ref-type="bibr" rid="R30">30</xref>]. Twenty-five patients underwent ICD implantation and were compared to 122 controls that were treated medically. The 2-year survival was 84.9% in the ICD group compared to 66.9% in the medically treated group (P=0.04). While this trial was non-randomized, the ICD group had significantly improved 2-year survival although they had more frequent ventricular arrhythmias during monitoring (56% vs. 15% P=0.0001).</p><p>These studies suggest that unexplained syncope in patients with NIDCM is a grave prognostic indicator with a significant proportion benefiting from ICD implantation; however that is not to imply that absence of syncope confers any protection from sudden death.</p></sec><sec sec-type="" id="s5"><title>Electrocardiographic Monitoring</title><p>The prognostic influence of electrocardiographic monitoring for the presence of ventricular arrhythmias has been debated for nearly two decades. The incidence of NSVT in patients with NIDCM varies from 33 to 79% [<xref ref-type="bibr" rid="R31">31</xref>-<xref ref-type="bibr" rid="R32">32</xref>]. Survival in such patients at one year is 92%, and decreases to 88% at two years [<xref ref-type="bibr" rid="R4">4</xref>]. Huang et al. followed 35 patients with recently diagnosed NIDCM that had ambulatory electrocardiographic monitoring to evaluate for ventricular arrhythmias [<xref ref-type="bibr" rid="R33">33</xref>]. At baseline, 83% had frequent PVCs, 93% had complex PVCs and 60% of patients had NSVT. During follow-up of 34 ± 17 months there were 4 deaths, two of which were sudden. Of the two patients who suffered sudden death during follow-up, one had no evidence of ventricular arrhythmias at baseline. While this study showed a high incidence of NSVT in the NIDCM population, it was not powered to show any correlation between NSVT and sudden death.</p><p>Olshausen et al. followed a group of 73 patients with NIDCM for a minimum of 3 years [<xref ref-type="bibr" rid="R34">34</xref>]. All patients received ambulatory holter monitoring at baseline. During follow-up, 38% of patients expired. The deaths were evenly divided between those that died of pump failure and those that died due to sudden death. NSVT was a risk factor for death due to pump failure, but was not for sudden death. On the other hand, Becker et al. followed 256 patients with NIDC [<xref ref-type="bibr" rid="R35">35</xref>]. Of those, 99 patients had documented asymptomatic NSVT, while 157 controls with NIDCM were free of documented ventricular arrhythmias. During the 22 ± 14 months of follow-up, both overall mortality and mortality due to sudden death were higher in patients with NSVT than those without arrhythmias (34.2 vs. 9.8%, P=0.0001 and 15.8 vs. 3.7%, P=0.0037).</p><p>Grimm et al. recently assessed a cohort of 343 patients with NIDCM using multiple diagnostic tests to assess risk for sudden death [<xref ref-type="bibr" rid="R36">36</xref>]. During the 52 ± 21 months of follow-up, 13% of patients had a major arrhythmic event (sustained VT, VF, or sudden death). Ejection fraction was found to be the only significant predictor of a major arrhythmic event, with a relative risk reduction of 2.3 for every 10% increase in ejection fraction (95% CI, 1.5 to 3.3, p=0.0001). The finding of NSVT on holter monitoring was associated with a trend towards an increased risk of major arrhythmic events (RR 1.7, 95% CI 0.9 to 3.3, p=0.11).</p><p>The prognostic value of NSVT or frequent ventricular arrhythmias remains unclear. The weight of evidence suggest that it is predominantly a marker of risk for all cause mortality but is not very effective in selecting a high risk group for sudden arrhythmic death.</p></sec><sec sec-type="" id="s6"><title>Electrophysiologic Testing</title><p>Programmed electrical stimulation (PES) has been evaluated in NIDCM. The studies performed were often non-randomized and had small numbers of patients (<xref ref-type="table" rid="T1">Table 1</xref>). In addition, the variability of stimulation protocols makes assessment of predictive values difficult. Sustained ventricular arrhythmias (monomorphic and polymorphic VT, or VF) may be induced in up to 38% of patients with NIDCM undergoing PES, but this may have limited clinical significance [<xref ref-type="bibr" rid="R37">37</xref>].</p><p>The reproducibility of programmed stimulation is greater in patient’s who presented with clinical VT as opposed to those with ventricular fibrillation. Milner et al. performed programmed stimulation in 19 patients with NIDCM and symptomatic ventricular arrhythmias; 10 had survived an out-of-hospital cardiac arrest, 8 had monomorphic VT (MMVT), and 1 had NSVT [<xref ref-type="bibr" rid="R38">38</xref>]. The mean ejection fraction in this group was 26 ± 9%. 13 (68%) of these patients had their clinical arrhythmia reproduced in the EP lab. The most frequently reproduced rhythm was MMVT (92%), while VF was reproducible in only one patient (8%). Monomorphic VT induced with programmed stimulation does appear to predict future occurrence of spontaneous monomorphic VT of a similar rate and configuration [<xref ref-type="bibr" rid="R38">38</xref>-<xref ref-type="bibr" rid="R39">39</xref>].</p><p>Nevertheless, the diagnostic yield of PES in the NIDCM population has been poor. Stamato et al. performed PES in 15 patients with NIDCM and heart failure [<xref ref-type="bibr" rid="R40">40</xref>]. The mean EF in this group was 17%, and they all had NSVT on cardiac monitoring. PES in this group failed to produce a sustained ventricular arrhythmia in any patients. Similarly, Das et al. performed PES in 24 patients with NIDCM and a mean ejection fraction of 25 ± 12%. While ventricular arrhythmias were induced in 42% of patients, MMVT was seen in only 8% of the group.</p><p>Furthermore the specificity of induced arrhythmias has been called into question. The most common finding with PES in patients with NIDCM is inducible polymorphic VT or VF [<xref ref-type="bibr" rid="R37">37</xref>,<xref ref-type="bibr" rid="R39">39</xref>,<xref ref-type="bibr" rid="R41">41</xref>]. Meinertz et al. evaluated 42 patients with NIDCM and performed PES in all patients [<xref ref-type="bibr" rid="R41">41</xref>]. 86% of the study group had polymorphic ventricular arrhythmias, with three or more beats of PMVT induced in 26% of the study group. During the 16 ± 7 months of follow-up, 5% of patients died of sudden death. None of the patients who suffered sudden death had an induced arrhythmia during PES. This suggests that the positive predictive value of induced polymorphic ventricular arrhythmias is low in the NIDCM population.</p><p>The use of PES for risk stratification in the NIDCM population has not provided satisfying results. Aside from the rare occasion of inducible MMVT, PES may frequently result in the induction of non-specific arrhythmias. The negative predictive value of a normal study is poor as well.</p></sec><sec sec-type="" id="s7"><title>Signal Averaged ECG</title><p>Several groups have evaluated the performance of the signal averaged ECG (SAECG) to stratify risk for SCD. Many of these studies did not differentiate between CAD and non-ischemic causes of cardiomyopathy, and were not powered for mortality in the NIDCM population. Galinier et al. prospectively followed 151 patients with CHF, 48% of which were diagnosed with NIDCM [<xref ref-type="bibr" rid="R42">42</xref>]. At baseline, late potentials were detected in 32.5% of the total patient population. 34% of the NIDCM patients were found to have late potentials (similar to the incidence of late potentials in patient’s with CAD). Late potentials were not found to be predictive of overall mortality or sudden cardiac death but did improve risk stratification for sustained ventricular tachycardia.</p><p>In a group of NIDCM patients, Fauchier et al. performed SAECG in 131 patients with mean ejection fraction of 33 ± 12% and followed them for 54 ± 41 months[<xref ref-type="bibr" rid="R43">43</xref>]. Late potentials were present in 27% of the patients. During follow-up, 15% of the patients had major arrhythmic events. Patients with late potentials were found to be at increased risk of all-cause cardiac death (RR 3.3, 95% CI 1.5 to 7.5, P = 0.004) and of arrhythmic events (RR 7.2, 95% CI 2.6 to 19.4, P = 0.0001), but not sudden death.</p><p>The studies of SAECG in the NIDCM population are complicated by the inclusion of individuals with bundle branch blocks (BBB), which compromises the utility of the SAECG as an adequate method for risk stratification in this population. Mancini et al. evaluated 114 patients with NIDCM with SAECG [<xref ref-type="bibr" rid="R44">44</xref>]. In the first year of follow-up, survival was 95% in patients with a normal SAECG, 88% in patients with a BBB, and only 39% in patients with an abnormal SAECG (P < 0.001). During follow-up, of the 20 patients with an abnormal SAECG, 5 (20%) suffered a sudden death, and three others either died from progressive heart failure or required urgent cardiac transplantation. This suggests that an abnormal SAECG confers an increased risk for an arrhythmic episode within a relatively short follow-up period. This study was flawed, however, because of the non-uniform implantation of ICDs in the study group, with patients with BBB having significantly more ICDs placed at baseline (P < 0.05).</p><p>An abnormal result on SAECG may be a marker of increased risk of sustained ventricular tachycardia or death [<xref ref-type="bibr" rid="R43">43</xref>,<xref ref-type="bibr" rid="R44">44</xref>]. As such, the SAECG appears to have an excellent negative predictive value, with the caveat that the presence of bundle branch block may significantly decrease the specificity of SAECG [<xref ref-type="bibr" rid="R45">45</xref>]. However, the poor positive predictive value for arrhythmic events and decreased specificity in the significant number of patients with bundle branch block lessens the value of this test.</p></sec><sec sec-type="" id="s8"><title>QT Dispersion</title><p>QT dispersion (QTd) has been studied in the risk stratification of SCD in patients with NIDCM. Few studies have been done, and the results have been unfavorable. Grim et al. performed QT dispersion and related measures in 107 patients with NIDCM and 100 controls without structural heart disease and followed them for 13 ± 7 months [<xref ref-type="bibr" rid="R46">46</xref>]. During follow-up, 11% of patients with NIDCM had an arrhythmic event, defined as sustained VT, VF, or sudden death. QT dispersion was noted to be increased in patients with arrhythmic events compared to patients without an arrhythmia during follow-up (76 ± 17 vs 60 ± 26 ms, P=0.03). On the other hand, corrected QT dispersion (QTc) and adjusted QTc were not statistically different between those with and those without arrhythmias during follow-up (80 ± 21 vs 75 ± 35 ms, and 27 ± 6 vs 24 ± 10 ms). This modality is limited due to the overlap between groups with positive and negative results. The Marburg Cardiomyopathy Study [<xref ref-type="bibr" rid="R36">36</xref>], confirmed that QT dispersion does not significantly risk stratify patients for sudden death.</p></sec><sec sec-type="" id="s9"><title>HR Variability</title><p>Measures of autonomic function have been evaluated to assess risk for SCD. Heart rate variability is an indirect measure of autonomic tone. Previous studies have shown that heart rate variability (HRV) is predictive of arrhythmic events following myocardial infarction (<xref ref-type="table" rid="T2">Table 2</xref>). The ATRAMI (Autonomic Tone and Reflexes After Myocardial Infarction) study showed that both decreased HRV and baro-receptor sensitivity, another measure of autonomic tone, are associated with worse outcomes in patients after myocardial infarction [<xref ref-type="bibr" rid="R47">47</xref>].</p><p>HRV was evaluated in 70 patients with NIDCM and prophylactic ICD implantation. Patients with sustained ventricular arrhythmias as recorded by their defibrillator had similar HRV to those who had no arrhythmic events [<xref ref-type="bibr" rid="R48">48</xref>]. Other trials have shown that in patients with NIDCM, HRV was predictive of cardiac death but not ventricular arrhythmias [<xref ref-type="bibr" rid="R49">49</xref>,<xref ref-type="bibr" rid="R50">50</xref>]. Patients with NIDCM have decreased heart rate variability. However, this decrease is associated with systolic ventricular function, and does not correlate with increased risk of ventricular arrhythmias [<xref ref-type="bibr" rid="R49">49</xref>]. The weight of the evidence suggests that due to the poor positive predictive value of HRV, this test has no role in the risk stratification for sudden death of patients with NIDCM [<xref ref-type="bibr" rid="R48">48</xref>-<xref ref-type="bibr" rid="R50">50</xref>].</p></sec><sec sec-type="" id="s10"><title>Baroreceptor Sensitivity Testing</title><p>Studies of baroreceptor sensitivity (BRS) in the NIDCM population have not focused on risk stratification for sudden death until recently [<xref ref-type="bibr" rid="R36">36</xref>]. Menz et al. compared BRS against heart rate variability in a series of 179 patients including those with CAD (27%) and NIDCM (73%) [<xref ref-type="bibr" rid="R51">51</xref>]. This trial was designed to determine if there were any differences in these two diagnostic tests in a baseline group of patients with decreased EF. It was not designed for assessment of prognostic value. Baroreceptor sensitivity was comparable in patients with CAD and NIDCM (6.1 ± 3 vs 6.9 ± 5 ms/mmHg, P=NS).</p><p>Grimm, et al. performed an exhaustive trial of risk stratification techniques in patients with NIDCM, including BRS [<xref ref-type="bibr" rid="R36">36</xref>]. 343 patients with NIDCM underwent BRS testing, with 26 patients (10%) being excluded due to insufficient response to phenylephrine during BRS testing. No correlation was noted between BRS results and sudden death during the 52 ± 21 months of follow-up. The evidence to date suggests that BRS is not a reliable risk stratifier for sudden death in NIDCM, and may not have additional benefit over other markers of autonomic tone, such as heart rate variability.</p></sec><sec sec-type="" id="s11"><title>Microvolt T Wave Alternans</title><p>Microvolt T Wave Alternans (TWA) is due to beat-to-beat alterations in cellular repolarization due to changes in action potential duration. Kitamura et al described an increased risk of SCD or sustained VT/VF in patients with a history of NIDCM with occurrence of microvolt level TWA at sufficiently low heart rate [<xref ref-type="bibr" rid="R52">52</xref>]. Onset of TWA with heart rates below 100/minute conferred increased risk of SCD or sustained VT/VF, with a predictive accuracy of 78% [<xref ref-type="bibr" rid="R52">52</xref>]. The onset HR of patients with TWA is independent of the more standard risk factors, including LV ejection fraction, gender, systolic or diastolic BP, or NYHA functional class. This makes it attractive in risk stratifying patients with NIDCM. Unfortunately TWA can be indeterminate in as many as 20% of patients with NIDCM [<xref ref-type="bibr" rid="R52">52</xref>,<xref ref-type="bibr" rid="R53">53</xref>]. Combining it with the presence of late potentials on SAECG [<xref ref-type="bibr" rid="R54">54</xref>] may increase the sensitivity of TWA.</p><p>A recent study by Hohnloser et al looked at 137 patients with NIDCM referred to a single center for diagnosis or management of heart failure or evaluation of symptomatic arrhythmias [<xref ref-type="bibr" rid="R55">55</xref>]. The follow-up was 14 ± 6 months, with endpoints defined as sudden death and hemodynamically unstable VT or VF. 37 patients in the study had prior ICD implantation. TWA at heart rates ≤110/minutes had a positive predictive value of 22% ± 5% and a negative predicative value of 94% ± 4%. The strong negative predictive value of sustained TWA demonstrated in this study at heart rates <110 bpm suggests a role of TWA testing in risk stratification of patients with NIDCM (<xref ref-type="table" rid="T3">Table 3</xref>). Further studies to look at larger, more diverse populations, long-term follow-up, and the benefits of serial TWA measurements will clarify the role of this test in risk stratification of patients with NIDCM.</p></sec><sec sec-type="" id="s12"><title>Heart Rate Turbulence</title><p>Heart rate turbulence is an emerging risk profiling method in patients with heart disease [<xref ref-type="bibr" rid="R63">63</xref>-<xref ref-type="bibr" rid="R65">65</xref>]. It is a method which is based on the electrical property of the heart after a ventricular ectopic beat and can be measured from a 12 lead electrocardiogram or holter monitoring. Immediately post ventricular ectopic beat the native ventricular response accelerates (due to baroreflex activity because of decreased cardiac output) and followed by a slow deceleration back to normalcy. Measurement of these variables when studied along with other risk factors has shown superior risk prediction for sudden cardiac death. This method however will need to be validated with prospective clinical studies.</p></sec><sec sec-type="" id="s13"><title>Conclusion</title><p>As the treatment of heart failure from Non-ischemic dilated cardiomyopathy CHF becomes more refined, it is expected that death from pump failure will be delayed, and risk stratification of these patients for arrhythmic death will become more important. Proper risk stratification based on symptoms, signs, and judicious use of noninvasive and invasive testing is important to determine the subgroup of patients that would benefit most from ICD implantation.</p><p>The most effective treatment and prophylaxis for SCD is implantation of an ICD. Due to the considerable cost considerations of these devices it would probably be uneconomical to place these devices in all patients with NIDCM. While it may be expected that the price of these devices individually will decrease over time, due to the increasing number of expected implantations the overall cost to the healthcare system is not likely to diminish. More over majority of the health systems cannot afford these expensive devices on a “blanket” basis. When compared to similar patients with ischemic cardiomyopathy, it appears that the non-ischemic dilated cardiomyopathy patient has a less certain clinical course in terms of sudden cardiac death. Most of the current trials of risk stratification for patients with NIDCM, with the exception of the SCD-HeFT trial, are underpowered and non-randomized. Clearly further studies that define the precise role of various risk stratification modalities need to be performed. Current indications for ICD implantation miss a fairly large number of patients at risk for sudden death. Appropriate primary prevention and secondary prevention of cardiac arrest in this population will require effective risk stratification to allow delivery of life saving therapy in an economically reasonable fashion.</p></sec>
A statistical method for predicting splice variants between two groups of samples using GeneChip<sup>® </sup>expression array data	<sec><title>Background</title><p>Alternative splicing of pre-messenger RNA results in RNA variants with combinations of selected exons. It is one of the essential biological functions and regulatory components in higher eukaryotic cells. Some of these variants are detectable with the Affymetrix GeneChip<sup>® </sup>that uses multiple oligonucleotide probes (i.e. probe set), since the target sequences for the multiple probes are adjacent within each gene. Hybridization intensity from a probe correlates with abundance of the corresponding transcript. Although the multiple-probe feature in the current GeneChip<sup>® </sup>was designed to assess expression values of individual genes, it also measures transcriptional abundance for a sub-region of a gene sequence. This additional capacity motivated us to develop a method to predict alternative splicing, taking advance of extensive repositories of GeneChip<sup>® </sup>gene expression array data.</p></sec><sec><title>Results</title><p>We developed a two-step approach to predict alternative splicing from GeneChip<sup>® </sup>data. First, we clustered the probes from a probe set into pseudo-exons based on similarity of probe intensities and physical adjacency. A pseudo-exon is defined as a sequence in the gene within which multiple probes have comparable probe intensity values. Second, for each pseudo-exon, we assessed the statistical significance of the difference in probe intensity between two groups of samples. Differentially expressed pseudo-exons are predicted to be alternatively spliced. We applied our method to empirical data generated from GeneChip<sup>® </sup>Hu6800 arrays, which include 7129 probe sets and twenty probes per probe set. The dataset consists of sixty-nine medulloblastoma (27 metastatic and 42 non-metastatic) samples and four cerebellum samples as normal controls. We predicted that 577 genes would be alternatively spliced when we compared normal cerebellum samples to medulloblastomas, and predicted that thirteen genes would be alternatively spliced when we compared metastatic medulloblastomas to non-metastatic ones. We checked the consistency of some of our findings with information in UCSC Human Genome Browser.</p></sec><sec><title>Conclusion</title><p>The two-step approach described in this paper is capable of predicting some alternative splicing from multiple oligonucleotide-based gene expression array data with GeneChip<sup>® </sup>technology. Our method employs the extensive repositories of gene expression array data available and generates alternative splicing hypotheses, which can be further validated by experimental studies.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>Fan</surname><given-names>Wenhong</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Khalid</surname><given-names>Najma</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" contrib-type="author"><name><surname>Hallahan</surname><given-names>Andrew R</given-names></name><xref ref-type="aff" rid="I2">2</xref><xref ref-type="aff" rid="I3">3</xref><email>[email protected]</email></contrib><contrib id="A4" contrib-type="author"><name><surname>Olson</surname><given-names>James M</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A5" corresp="yes" contrib-type="author"><name><surname>Zhao</surname><given-names>Lue Ping</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	Theoretical Biology & Medical Modelling	<sec><title>Background</title><p>Alternative splicing of pre-messenger RNA is an essential biological functional and regulatory component in higher eukaryotic cells. It increases the complexity of biological processes and gives the cells enhanced capability to respond to various factors, such as developmental changes and environmental stimuli. Some splice variants have been associated with diseases, such as mammary tumorigenesis [<xref ref-type="bibr" rid="B1">1</xref>] and ovarian cancer [<xref ref-type="bibr" rid="B2">2</xref>]. About 15% of single nucleotide mutations in the exon recognition process are associated with human genetic diseases [<xref ref-type="bibr" rid="B3">3</xref>]. Understanding the alternative splicing mechanism may also lead to finding potential treatments for related diseases [<xref ref-type="bibr" rid="B4">4</xref>].</p><p>In this paper, we describe a method for detecting alternative splicing variants using the GeneChip<sup>® </sup>gene expression array data. Affymetrix GeneChip<sup>® </sup>technology employs multiple probes per gene to measure gene expression. These multiple probes are short sequences located in different positions within each gene. Even though distributions of these probe sequences are not optimized for detecting alternative splicing, the probe sequence data obtained by the current GeneChip<sup>® </sup>technology can be used to assess alternative splicing. In our method, we infer "pseudo-exons" from hybridization intensities of multiple probes that are spread over a probe set. A pseudo-exon is defined as a range of expressed sequence on the genome that we infer to be an exon based on probe intensities and physical adjacency.</p><p>Figure <xref ref-type="fig" rid="F1">1</xref> illustrates how GeneChip<sup>® </sup>expression array data can be used to detect alternative splicing. We show the probe locations for a hypothetical gene in Figure <xref ref-type="fig" rid="F1">1A</xref> and their corresponding hybridization intensities in Figure <xref ref-type="fig" rid="F1">1B</xref>. From the probe intensities, we infer that three clusters of probes represent three pseudo-exons (Figure 1C). For each of the pseudo-exons, we test whether the difference in probe intensities between tissue 1 and tissue 2 is significant. If the difference is statistically significant, we infer that there is alternative splicing between the two tissues for the region corresponding to the selected pseudo-exon. In our illustration, the region between probe #7 and probe #14, i.e. pseudo-exon 2 is predicted to be alternatively spliced between tissue 1 and tissue 2.</p><fig position="float" id="F1"><label>Figure 1</label><caption><p><bold>A </bold>Multiple probes are used to quantify the expression value for a gene in GeneChip<sup>® </sup>technology. Currently the probe design has a 3' bias, i.e. probes are selected from the sequence at the 3'end of the gene. In the Hu6800 array, twenty probes are used for a single gene. 1 <bold>B </bold>Intensities of the twenty probes are plotted for both tissues 1 and 2. 1 <bold>C </bold>The twenty probes are clustered into three groups based on the similarity of probe intensity and probe adjacency. Each cluster, called a pseudo-exon in this paper, represents a sub-region of the gene.</p></caption><graphic xlink:href="1742-4682-3-19-1"/></fig><p>Previously, Hu et al reported a method, based on fold changes, to predict alternative splicing from GeneChip<sup>® </sup>expression array data on ten tissue types [<xref ref-type="bibr" rid="B5">5</xref>]. For each probe, they calculated the difference in the fold change between each tissue type and the average of the remaining tissue types for the corresponding probe. If the fold change was greater than an empirically-determined threshold value R, they selected the gene sequence corresponding to that selected probe as an alternative splicing site for that tissue type. However, there are some problems with Hu's approach. First, the fold-change approach does not take into account sample variation and thus is less reliable when sample-to-sample variations are large. Second, their method is designed to predict splice variants in a dataset with multiple tissue types. Hu et al reported that prediction power decreased for a dataset that contained only three tissue types compared to a dataset that consisted of ten tissue types. The robustness of their method depended on the number of the tissue types in the dataset. Thus, their method is not suitable for the comparison of two tissue types such as detection of splice variants between two phenotypes, or two disease status, or two experimental stimuli.</p><p>In this paper, we propose an approach to predict splice variants between two groups of samples from GeneChip<sup>® </sup>expression array data, taking into consideration sample variation. Our t-test based approach is more statistically vigorous and reliable than fold-change based methods. Furthermore, our method does not rely on a large number of tissue types. We implemented the method from Hu et al and compared the splice variants predicted from the two approaches. Our dataset consists of normal cerebellum, non-metastatic medulloblastomas, and metastatic medulloblastomas. The comparisons were made between normal cerebellum versus medulloblastomas, and non-metastatic medulloblastomas versus metastatic medulloblastomas.</p></sec><sec><title>Results</title><sec><title>The computational algorithms</title><p>Our approach has two steps. In STEP 1, we infer pseudo-exons using multiple probe intensities. In STEP 2, we identify pseudo-exons that are differentially expressed between two groups of samples. In STEP 1, for each probe, we first compute the average of the difference in probe intensities between the two groups of samples. Then, based on the similarity of probe intensities and probe adjacency on the gene sequence, we merge probes into clusters that represent one pseudo-exon. In STEP 2, we test if the pseudo-exons are differentially expressed between the two groups of samples. The expression value from a pseudo-exon is treated as an entity in the current analysis, comparable to the gene expression from a complete probe set in customary analyses of gene expression data. The selected pseudo-exons are interpreted as an indication of alternative splicing at this region of the gene between the two comparison groups.</p></sec><sec><title>Predicting splice variants between normal cerebellum and medulloblastomas</title><p>For illustrative purposes, we applied the above method to predict splice variants between the normal cerebellum and medulloblastoma tumor samples, which included both non-metastatic and metastatic tumors. In STEP 1, using a significance level of 0.05 in the t-test, we identified 10,838 pseudo-exons out of a total of 142,580 (7129 × 20) probes that represent the 7,129 probe sets on the Hu6800 GeneChip<sup>®</sup>. In STEP 2, we compared the difference in expression values between the two groups for each pseudo-exon. The histogram of Z-scores from these tests is shown in Figure <xref ref-type="fig" rid="F2">2</xref>. With the significance threshold of the Z-score set to 4.8 (equivalent to one false positive error in the discovery), we discovered 811 pseudo-exons, derived from 577 genes, were significantly different between normal cerebellum and medulloblastoma tumor samples. Note that for some genes more than one pseudo-exon was selected.</p></sec><sec><title>Predicting splice variants between non-metastatic medulloblastomas and metastatic medulloblastomas</title><p>Following the same procedure, we predicted splice variants between the non-metastatic and the metastatic medulloblastomas. We identified 8,319 pseudo-exons, thirteen of which were significantly different between non-metastatic and metastatic medulloblastomas (Table <xref ref-type="table" rid="T1">1</xref>). Instead of conducting validation in a biological experiment, we searched two genome browsers for supportive evidence for our prediction. We input the thirteen genes in Table <xref ref-type="table" rid="T1">1</xref> into the Integrated Genome Browser (IGB) from Affymetrix [<xref ref-type="bibr" rid="B7">7</xref>] to see whether the probes in the identified pseudo-exons were positioned on separate exons within the same gene, which is a pre-requisite for alternative splicing. For further consistency, we checked whether the predicted pseudo-exons were reported as splice variants in the UCSC Human Genome Browser [<xref ref-type="bibr" rid="B8">8</xref>] under the track named "mRNA sequences from GenBank". In the IGB, we found four out of thirteen genes with predicted alternatively spliced pseudo-exons resided on different exons. These four genes were glutaredoxin (GLRX), carboxypeptidase N polypeptide 1 (CPN1), Keratin 7 (KRT7) and killer cell lectin-like receptor subfamily C member 3 (KLRC3). For instance, we predicted the last three probes for GLRX were within one pseudo-exon. In IGB, based on RefSeq information, these three probes are on a different exon. We searched alternatively transcribed variants deposited in GenBank in the "mRNA sequences from GenBank" track in UCSC Human Genome Browser for the genes confirmed by IGB. All of them except for CPN1 have at least two transcript sequences in the GenBank database. At least one of these sequences lack the region that we predicted to be alternatively spliced, and at least one of these sequences contain the predicted region. We also searched PubMed for reported splice variants for the thirteen identified genes. Five of out of the thirteen genes were reported in the literature to have splice variants. They are nitric oxide synthase 1 (NOS1) [<xref ref-type="bibr" rid="B9">9</xref>], low density lipoprotein receptor (LDLR) [<xref ref-type="bibr" rid="B10">10</xref>], thrombopoietin (THPO) [<xref ref-type="bibr" rid="B11">11</xref>], Down syndrome critical region gene 1 (DSCR1) [<xref ref-type="bibr" rid="B12">12</xref>], paired box gene 2 (PAX2) [<xref ref-type="bibr" rid="B13">13</xref>].</p><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>Alternative spliced genes selected by our method: Comparison of non-metastatic medulloblastomas with metastatic medulloblastomas</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center">Affymetrix Probe Set ID</td><td align="center">Gene Symbol</td><td align="center">Number of Affymetrix Probes in the Predicted Pseudo-exon</td><td align="center">Nucleotide Positions of Predicted Pseudo-exon in the Gene</td><td align="center">Mean Difference</td><td align="center">Standard Error</td><td align="center">Z-score</td><td align="left">Description of the Genes</td></tr></thead><tbody><tr><td align="center">M81882_at</td><td align="center">GAD2</td><td align="center">4</td><td align="center">(2135–2285)</td><td align="center">-1.28</td><td align="center">0.20</td><td align="center">-6.45</td><td align="left">glutamate decarboxylase 2 (pancreatic islets and brain, 65 kDa)</td></tr><tr><td align="center">M13955_at</td><td align="center">KRT7</td><td align="center">5</td><td align="center">(1402–1474)</td><td align="center">-0.63</td><td align="center">0.12</td><td align="center">-5.23</td><td align="left">keratin 7</td></tr><tr><td align="center">U17327_at</td><td align="center">NOS1</td><td align="center">7</td><td align="center">(6805–7003)</td><td align="center">-0.66</td><td align="center">0.13</td><td align="center">-5.19</td><td align="left">nitric oxide synthase 1 (neuronal)</td></tr><tr><td align="center">X14329_at</td><td align="center">CPN1</td><td align="center">4</td><td align="center">(1569–1665)</td><td align="center">-0.62</td><td align="center">0.12</td><td align="center">-5.18</td><td align="left">carboxypeptidase N, polypeptide 1, 50 kD</td></tr><tr><td align="center">M89470_s_at</td><td align="center">PAX2</td><td align="center">6</td><td align="center">(2855–2972)</td><td align="center">-0.92</td><td align="center">0.19</td><td align="center">-4.91</td><td align="left">paired box gene 2</td></tr><tr><td align="center">L14542_at</td><td align="center">KLRC3</td><td align="center">5</td><td align="center">(916–1006)</td><td align="center">-1.18</td><td align="center">0.24</td><td align="center">-4.91</td><td align="left">killer cell lectin-like receptor subfamily C, member 3</td></tr><tr><td align="center">X76648_at</td><td align="center">GLRX</td><td align="center">3</td><td align="center">(704–776)</td><td align="center">-1.35</td><td align="center">0.28</td><td align="center">-4.86</td><td align="left">glutaredoxin (thioltransferase)</td></tr><tr><td colspan="8"><hr></hr></td></tr><tr><td align="center">U82987_at</td><td align="center">BBC3</td><td align="center">3</td><td align="center">(1578–1638)</td><td align="center">2.25</td><td align="center">0.32</td><td align="center">6.98</td><td align="left">BCL2 binding component 3</td></tr><tr><td align="center">U01102_at</td><td align="center">SCGB1A1</td><td align="center">2</td><td align="center">(409–439)</td><td align="center">1.42</td><td align="center">0.25</td><td align="center">5.62</td><td align="left">secretoglobin, family 1A, member 1 (uteroglobin)</td></tr><tr><td align="center">M28219_at</td><td align="center">LDLR</td><td align="center">15</td><td align="center">(67–277)</td><td align="center">0.77</td><td align="center">0.14</td><td align="center">5.42</td><td align="left">low density lipoprotein receptor (familial hypercholesterolemia)</td></tr><tr><td align="center">X68194_at</td><td align="center">SYPL</td><td align="center">5</td><td align="center">(1915–2089)</td><td align="center">1.67</td><td align="center">0.31</td><td align="center">5.42</td><td align="left">synaptophysin-like protein</td></tr><tr><td align="center">U85267_at</td><td align="center">DSCR1</td><td align="center">10</td><td align="center">(64–169)</td><td align="center">1.20</td><td align="center">0.24</td><td align="center">5.08</td><td align="left">Down syndrome critical region gene 1</td></tr><tr><td align="center">L36051_at</td><td align="center">THPO</td><td align="center">6</td><td align="center">(1647–1809)</td><td align="center">1.05</td><td align="center">0.21</td><td align="center">4.96</td><td align="left">thrombopoietin (myeloproliferative leukemia virus oncogene ligand, megakaryocyte growth and development factor)</td></tr></tbody></table><table-wrap-foot><p>Number of Affymetrix Probes in the Predicted Pseudo-exon: number of probes that are contained in a predicted alternatively spliced pseudo-exon. Nucleotide Positions of Predicted Pseudo-exon in the Gene: nucleotide positions of the pseudo-exon from the beginning of the gene it resides. Mean difference: Mean difference of the expression values between the two tissue types being compared for each predicted pseudo-exon in the t-test in STEP 2. Standard Error: the standard error calculated in the same t-test. Z-score: the ratio of mean difference over standard error (noise), a measure of significance of the difference between the two tissues being compared. The sign of the Z-scores indicate direction of the difference. A negative Z-score means a lower expression in metastatic medulloblastomas than in non-metastatic medulloblastomas, and vice-versa for a positive Z-score.</p></table-wrap-foot></table-wrap></sec><sec><title>Comparison with Hu et al's approach</title><p>To compare our method with the Hu et al's, we implemented their method and applied it to our dataset. When comparing normal cerebellum and medulloblastomas samples using Hu et al's method, we inferred 31 alternatively spliced genes with the selection criterion used by Hu et al in their paper (Table <xref ref-type="table" rid="T2">2</xref>). Among these 31 genes, seven overlapped with the findings from our approach (Table <xref ref-type="table" rid="T3">3</xref>). For four of them, D87119_at, U14971_at, U29953_rna1_at, X04828_at, the locations of the alternative splicing were consistent between the two methods. In the comparison between non-metastatic and metastatic medulloblastoma samples, we did not find any gene that was alternatively spliced by Hu et al's method. We also investigated the effect of different selection criteria in Hu et al's method (i.e. the R threshold, which is the ratio of the probe intensity in a tissue over the mean of the probe intensities in the remaining nine tissue types for the same probe). Table <xref ref-type="table" rid="T4">4</xref> shows the relation between the 577 genes predicted by our approach and the genes selected with different R thresholds in Hu's approach. Numbers of predicted alternatively spliced genes increase as smaller R values (less stringent) are used.</p><table-wrap position="float" id="T2"><label>Table 2</label><caption><p>Alternative spliced genes inferred by applying Hu's method to our dataset: Comparison of normal cerebellum with medulloblastoma samples</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center">Affy Probe Set ID</td><td align="left">Gene Symbol</td><td align="center">Number of Affymetrix Probes in the Predicted Pseudo-exon</td><td align="center">Nucleotide Positions of Predicted Pseudo-exon in the Gene</td><td align="left">Description of the Genes</td></tr></thead><tbody><tr><td align="center">X51362_s_at</td><td align="left">DRD2</td><td align="center">2</td><td align="center">(2541–2574)</td><td align="left">dopamine receptor D2</td></tr><tr><td align="center">M15517_cds5_at</td><td align="left">TTR</td><td align="center">3</td><td align="center">(155–197)</td><td align="left">transthyretin (prealbumin, amyloidosis type I)</td></tr><tr><td align="center">Y10141_s_at</td><td align="left">SLC6A3</td><td align="center">2</td><td align="center">(96–125)</td><td align="left">solute carrier family 6 (neurotransmitter transporter, dopamine), member 3</td></tr><tr><td align="center">Z14982_rna1_at</td><td align="left">PSMB8</td><td align="center">2</td><td align="center">(820–850)</td><td align="left">proteasome (prosome, macropain) subunit, beta type, 8 (large multifunctional protease 7)</td></tr><tr><td align="center">X69654_at</td><td align="left">RPS26</td><td align="center">2</td><td align="center">(9–35)</td><td align="left">ribosomal protein S26</td></tr><tr><td align="center">U63842_at</td><td align="left">NEUROG1</td><td align="center">2</td><td align="center">(834–891)</td><td align="left">neurogenin 1</td></tr><tr><td align="center">M97815_at</td><td align="left">CRABP2</td><td align="center">2</td><td align="center">(524–554)</td><td align="left">cellular retinoic acid binding protein 2</td></tr><tr><td align="center">D00017_at</td><td align="left">ANXA2</td><td align="center">2</td><td align="center">(1229–1265)</td><td align="left">annexin A2</td></tr><tr><td align="center">U13021_s_at</td><td align="left">CASP2</td><td align="center">3</td><td align="center">(844–913)</td><td align="left">caspase 2, apoptosis-related cysteine protease (neural precursor cell expressed, developmentally down-regulated 2)</td></tr><tr><td align="center">U30999_at</td><td align="left">ALCAM</td><td align="center">2</td><td align="center">(373–403)</td><td align="left">activated leukocyte cell adhesion molecule</td></tr><tr><td align="center">X04828_at</td><td align="left">GNAI2</td><td align="center">3</td><td align="center">(1668–1701)</td><td align="left">guanine nucleotide binding protein (G protein), alpha inhibiting activity polypeptide 2</td></tr><tr><td align="center">U14971_at</td><td align="left">RPS9</td><td align="center">2</td><td align="center">(319–373)</td><td align="left">ribosomal protein S9</td></tr><tr><td align="center">U79299_at</td><td align="left">OLFM1</td><td align="center">2</td><td align="center">(1342–1372)</td><td align="left">olfactomedin 1</td></tr><tr><td align="center">L20298_at</td><td align="left">CBFB</td><td align="center">2</td><td align="center">(2298–2334)</td><td align="left">core-binding factor, beta subunit</td></tr><tr><td align="center">X93017_at</td><td align="left">SLC8A3</td><td align="center">2</td><td align="center">(1725–1821)</td><td align="left">solute carrier family 8 (sodium-calcium exchanger), member 3</td></tr><tr><td align="center">M17886_at</td><td align="left">RPLP1</td><td align="center">2</td><td align="center">(127–163)</td><td align="left">ribosomal protein, large, P1</td></tr><tr><td align="center">D16480_at</td><td align="left">HADHA</td><td align="center">2</td><td align="center">(2335–2365)</td><td align="left">hydroxyacyl-Coenzyme A dehydrogenase/3-ketoacyl-Coenzyme A thiolase/enoyl-Coenzyme A hydratase (trifunctional protein), alpha subunit</td></tr><tr><td align="center">D38305_at</td><td align="left">TOB1</td><td align="center">2</td><td align="center">(707–749)</td><td align="left">transducer of ERBB2, 1</td></tr><tr><td align="center">U32519_at</td><td align="left">G3BP</td><td align="center">2</td><td align="center">(1534–1564)</td><td align="left">Ras-GTPase-activating protein SH3-domain-binding protein</td></tr><tr><td align="center">U07919_at</td><td align="left">ALDH1A3</td><td align="center">3</td><td align="center">(3363–3411)</td><td align="left">aldehyde dehydrogenase 1 family, member A3</td></tr><tr><td align="center">U29953_rna1_at</td><td align="left">SERPINF1</td><td align="center">2</td><td align="center">(1288–1324)</td><td align="left">serine (or cysteine) proteinase inhibitor, clade F (alpha-2 antiplasmin, pigment epithelium derived factor), member 1</td></tr><tr><td align="center">D55716_at</td><td align="left">MCM7</td><td align="center">2</td><td align="center">(2288–2396)</td><td align="left">MCM7 minichromosome maintenance deficient 7 (S. cerevisiae)</td></tr><tr><td align="center">J05448_at</td><td align="left">POLR2C</td><td align="center">2</td><td align="center">(1575–1605)</td><td align="left">polymerase (RNA) II (DNA directed) polypeptide C, 33 kDa</td></tr><tr><td align="center">U46570_at</td><td align="left">TTC1</td><td align="center">2</td><td align="center">(1226–1262)</td><td align="left">tetratricopeptide repeat domain 1</td></tr><tr><td align="center">D87119_at</td><td align="left">TRB2</td><td align="center">2</td><td align="center">(4022–4136)</td><td align="left">tribbles homolog 2</td></tr><tr><td align="center">X69910_at</td><td align="left">CKAP4</td><td align="center">2</td><td align="center">(2543–2573)</td><td align="left">cytoskeleton-associated protein 4</td></tr><tr><td align="center">U50078_at</td><td align="left">HERC1</td><td align="center">2</td><td align="center">(14885–14915)</td><td align="left">hect (homologous to the E6-AP (UBE3A) carboxyl terminus) domain and RCC1 (CHC1)-like domain (RLD) 1</td></tr><tr><td align="center">J04164_at</td><td align="left">IFITM1</td><td align="center">2</td><td align="center">(798–828)</td><td align="left">interferon induced transmembrane protein 1 (9–27)</td></tr><tr><td align="center">AFFX-HUMRGE/M10098_3_at</td><td align="left">N/A</td><td align="center">2</td><td align="center">(1562–1613)</td><td align="left">N/A</td></tr><tr><td align="center">HG2788-HT2896_at</td><td align="left">N/A</td><td align="center">2</td><td align="center">(N/A-N/A)</td><td align="left">N/A</td></tr><tr><td align="center">HG2994-HT4850_s_at</td><td align="left">N/A</td><td align="center">2</td><td align="center">(N/A-N/A)</td><td align="left">N/A</td></tr></tbody></table><table-wrap-foot><p>Number of Affymetrix Probes in the Predicted Pseudo-exon: number of probes that are contained in a predicted alternatively spliced pseudo-exon. Nucleotide Positions of Predicted Pseudo-exon in the Gene: nucleotide positions of the pseudo-exon from the beginning of the gene it resides. Mean difference: Mean difference of the expression values between the two tissue types being compared for each predicted pseudo-exon in the t-test in STEP 2. Standard Error: the standard error calculated in the same t-test. Z-score: the ratio of mean difference over standard error (noise), a measure of significance of the difference between the two tissues being compared. The sign of the Z-scores indicate direction of the difference. A negative Z-score means a lower expression in metastatic medulloblastomas than in non-metastatic medulloblastomas, and vice-versa for a positive Z-score.</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T3"><label>Table 3</label><caption><p>Overlapping of the predicted gene from our method and Hu's method for the comparison of normal cerebellum and medulloblastoma samples</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left">Affy Probe Set ID</td><td align="left">Gene Symbol</td><td align="left" colspan="2">Number of Affymetrix Probes in the Predicted Pseudo-exon</td><td align="left" colspan="2">Nucleotide Positions of Predicted Pseudo-exon in the Gene</td><td align="left">Descriptions of the Genes</td></tr></thead><tbody><tr><td></td><td></td><td align="left">Ours</td><td align="left">Hu's</td><td align="left">Ours</td><td align="left">Hu's</td><td></td></tr><tr><td align="left">X04828_at</td><td align="left">GNAI2</td><td align="left">3</td><td align="left">3</td><td align="left">(1668–1701)</td><td align="left">(1668–1701)</td><td align="left">guanine nucleotide binding protein (G protein), alpha inhibiting activity polypeptide 2</td></tr><tr><td align="left">U14971_at</td><td align="left">RPS9</td><td align="left">19</td><td align="left">2</td><td align="left">(103–685)</td><td align="left">(319–373)</td><td align="left">ribosomal protein S9</td></tr><tr><td align="left">U29953_rna1_at</td><td align="left">SERPINF1</td><td align="left">13</td><td align="left">2</td><td align="left">(1288–1492)</td><td align="left">(1288–1324)</td><td align="left">serine (or cysteine) proteinase inhibitor, clade F (alpha-2 antiplasmin, pigment epithelium derived factor), member 1</td></tr><tr><td align="left">D87119_at</td><td align="left">TRB2</td><td align="left">13</td><td align="left">2</td><td align="left">(3824–4184)</td><td align="left">(4022–4136)</td><td align="left">tribbles homolog 2</td></tr><tr><td align="left">X69910_at</td><td align="left">CKAP4</td><td align="left">5</td><td align="left">2</td><td align="left">(2789–2891)</td><td align="left">(2543–2573)</td><td align="left">cytoskeleton-associated protein 4</td></tr><tr><td align="left">U30999_at</td><td align="left">ALCAM</td><td align="left">16</td><td align="left">2</td><td align="left">(25–337)</td><td align="left">(373–403)</td><td align="left">activated leukocyte cell adhesion molecule</td></tr><tr><td align="left">D55716_at</td><td align="left">MCM7</td><td align="left">8</td><td align="left">2</td><td align="left">(1952–2096)</td><td align="left">(2288–2396)</td><td align="left">MCM7 minichromosome maintenance deficient 7 (S. cerevisiae)</td></tr></tbody></table><table-wrap-foot><p>* Consistent alternative splice sites between two methods.</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T4"><label>Table 4</label><caption><p>Comparison of the results from our approach and those from Hu's using different R thresholds when normal cerebellum samples are compared with medulloblastomas</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left">R used</td><td align="left">Number of Genes Found in Hu's Approach</td><td align="left">Number of Overlap Between Hu's and Our 577 Genes</td><td align="left">Percentage of the overlapping genes based on number of genes found in Hu's method</td><td align="left">Percentage of the overlapping genes based on our 577 selected genes</td></tr></thead><tbody><tr><td align="left">4</td><td align="left">324</td><td align="left">69</td><td align="left">21%</td><td align="left">11.9%</td></tr><tr><td align="left">6</td><td align="left">103</td><td align="left">28</td><td align="left">27%</td><td align="left">4.9%</td></tr><tr><td align="left">8</td><td align="left">53</td><td align="left">14</td><td align="left">26%</td><td align="left">2.4%</td></tr><tr><td align="left">10</td><td align="left">31</td><td align="left">7</td><td align="left">23%</td><td align="left">1.2%</td></tr></tbody></table><table-wrap-foot><p>Genes found in Hu's methods using different R thresholds are compared to each other. Larger R value represents more stringent selection criterion. Genes found using smaller R values always include those found using larger R values, i.e. gene list of 324 genes contains gene list of 103 genes, etc. Genes obtained from Hu's method are also compared with 577 genes from our approach. Numbers of overlapping genes are presented in the third column for different R values. Similarly, overlapping genes for the smaller R values contains those for the larger R values, i.e. gene list of 69 genes contains gene list of 28 genes, etc.</p></table-wrap-foot></table-wrap><p>We checked both IGB and UCSC Human Genome Browsers for supportive evidence for the seven predicted alternatively spliced variants in Table <xref ref-type="table" rid="T3">3</xref>. We found four genes that had predicted pseudo-exons located on separate exons according to IGB and alternative spliced mRNA from GenBank in UCSC Human Genome Browser. They are guanine nucleotide binding protein alpha inhibiting activity polypeptide 2 (GNAI2), ribosomal protein S9 (RPS9), activated leukocyte cell adhesion molecule (ALCAM), and minichromosome maintenance deficient 7 (MCM7). There are splicing variants reported in PubMed literature for ALCAM [<xref ref-type="bibr" rid="B14">14</xref>].</p></sec></sec><sec><title>Discussion</title><p>We have developed a two-step approach to predict splice variants between two groups of samples using GeneChip<sup>® </sup>gene expression array data. We illustrated the method using empirical data from normal cerebellum, metastatic medulloblastoma and non-metastatic medulloblastoma samples. We predicted a total of 577 alternatively spliced genes when we compared normal cerebellum with medulloblastomas tumor samples and thirteen alternatively spliced genes when we compared non-metastatic medulloblastomas with metastatic medulloblastomas. A comparison of the results from our approach and the method described by Hu et al on the same dataset revealed some overlapping alternatively spliced genes.</p><p>Our proposed method can be used to predict splice variants and takes advantage of the extensive repositories of gene expression array data. Inferred splice variants can be used to generate alternative splicing hypotheses for subsequent experimental validation. Higher signal quality in the newer generation GeneChip<sup>®</sup>, such as U133 Plus 2.0 array, should make our predictions more robust. Recently, a genome-wide human exon array became available from Affymetrix [<xref ref-type="bibr" rid="B7">7</xref>] to detect known alternative splicing in a biological sample. Bypassing the need for defining "pseudo-exons" in the STEP 1 of our approach, one can directly use STEP 2 of our method to predict splice variants. As expected, such an exon array coupled with our rigorous statistical method may improve the power to predict more splice variants.</p><p>There are some limitations associated with using GeneChip<sup>® </sup>gene expression array data to detect alternatively spliced variants. Currently, GeneChip<sup>® </sup>probes cover 600 base pairs in sequence from the 3' end. Thus we can only detect splice variants at the 3' end. Furthermore, some 3' end splice variants could be due to alternative polyadenylation sites, and our method does not differentiate between these in the analysis. The splice variants resulting from the 3' non-translational region could be removed by checking whether the predicted pseudo-exons on the 3' end are located in translational regions.</p><p>Since our approach depends on probe intensities to cluster probes into pseudo-exons within a single gene, non-specific hybridization in an expression array could complicate this step (STEP 1), thus result in both false positive and false negative findings. Cross-hybridization can be partially addressed by excluding lower grade probe sets, such as probe sets with the suffix _s or _x, which could hybridize to multiple genes either before analysis or from the gene list after analysis.</p></sec><sec><title>Conclusion</title><p>In this paper we describe a method that can generate hypotheses of alternative splicing for further investigation. Our approach overcomes two limitations of a previously proposed method [<xref ref-type="bibr" rid="B5">5</xref>]: 1) we use t-tests instead of fold changes, 2) we can predict splicing variants between two groups of samples. These differences make our inference more robust and not dependent on multiple tissue types to stabilize the inference.</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Dataset</title><p>Our empirical dataset consists of GeneChip<sup>® </sup>Hu6800 expression array data from sixty-nine medulloblastoma samples and four cerebellum samples as normal controls. Among the medulloblastoma samples, forty-two are from non-metastatic tumors and twenty-seven are from metastatic tumors. There are 7,129 probe sets in the Hu6800 expression array, and twenty probes in each probe set.</p></sec><sec><title>Inferring pseudo-exons within a gene (STEP 1)</title><p>In this step, we merge probes within a gene into clusters that represent pseudo-exons. First, we compute the difference in probe hybridization intensity between two groups of samples for each probe. Then, for each gene, we merge probes into clusters based on the similarity of the differences in probe intensity (between the two groups of samples) and the probe adjacency on the genome sequence. For a gene, let <bold><italic>Y</italic></bold><sub>(<italic>i</italic>, 1)</sub>, <inline-graphic xlink:href="1742-4682-3-19-i1.gif"/> and <italic>n</italic><sub>1 </sub>be the probe intensity for the <italic>i</italic>th probe in sample group 1, variance, and sample size, respectively. Similarly, <bold><italic>Y</italic></bold><sub>(<italic>i</italic>, 2)</sub>, <inline-graphic xlink:href="1742-4682-3-19-i2.gif"/> and <italic>n</italic><sub>2 </sub>are for the sample group 2. Within the gene, the index <italic>i </italic>increases from the direction of the 5' end to the 3' end. We start with the first probe from the 5' end and compute:</p><p><inline-graphic xlink:href="1742-4682-3-19-i3.gif"/></p><p><inline-graphic xlink:href="1742-4682-3-19-i4.gif"/></p><p>where <inline-graphic xlink:href="1742-4682-3-19-i5.gif"/> is the mean of probe intensities. If the absolute value of <bold><italic>t</italic></bold><sub>i </sub>does not exceed the threshold value at the significance level <italic>α </italic>= 0.05, we merge the <italic>i</italic>th probe with the (<italic>i</italic>+1)th probe to generate a pseudo-exon. The resulting pseudo-exon becomes the new <italic>i</italic>th probe in the next iteration of the t-test. The pseudo-exon extends with each iteration until the results of the t-test become significant or reach the last probe within a probe set. If <bold><italic>t</italic></bold><sub><italic>i </italic></sub>exceeds the significance threshold value, we do not merge the <italic>i</italic>th probe with the (<italic>i</italic>+1)th probe, but start generating a new pseudo-exon from this (<italic>i</italic>+1)th probe, using the same iteration procedure. After we finish the last probe at the 3' end, we may either have several pseudo-exons or only one pseudo-exon (i.e. the entire probe set) if every t-statistic within a probe set is not significant.</p></sec><sec><title>Testing for statistical significance (STEP 2)</title><p>For each pseudo-exon, we determine whether there is a difference in hybridization intensity between the two groups x<sub>1 </sub>and x<sub>2</sub>. Our null hypothesis is that, for any pseudo-exon, the difference in probe intensity between the two groups is zero. If we reject the null hypothesis for a pseudo-exon, meaning that the hybridization intensities between the two groups are significant different for that pseudo-exon, we then infer that there is a splice variant between the two groups of samples for the corresponding region within the gene.</p><p>In the same vein as Li and Wong's model to analyze gene expression at the probe level [<xref ref-type="bibr" rid="B15">15</xref>], we propose a multiplicative heterogeneity factor model to associate the probe intensities of a pseudo-exon directly with the covariate, i.e. group indictor x<sub><italic>k</italic></sub>:</p><p><inline-graphic xlink:href="1742-4682-3-19-i6.gif"/></p><p>where <italic>Y</italic><sub><italic>jik </italic></sub>is the hybridization intensity for the <italic>i</italic>th probe in the <italic>j</italic>th pseudo-exon in the <italic>k</italic>th sample, N is the number of probes in the <italic>j</italic>th pseudo-exon, <italic>δ</italic><sub><italic>k </italic></sub>and <italic>λ</italic><sub><italic>k </italic></sub>are heterogeneity factors for normalization,<italic>x</italic><sub><italic>k </italic></sub>is the group indicator for the <italic>k</italic>th sample, <italic>β</italic><sub><italic>j </italic></sub>is the coefficient for <italic>j</italic>th pseudo-exon, <italic>φ</italic><sub><italic>ji </italic></sub>is the multiplicative probe-specific parameter for <italic>i</italic>th probe in <italic>j</italic>th pseudo-exon, and <italic>ξ </italic>is random variation term. To avoid making any distributional assumptions, we applied estimating equation techniques to estimate the coefficients and their standard errors for making statistical inferences [<xref ref-type="bibr" rid="B16">16</xref>-<xref ref-type="bibr" rid="B19">19</xref>].</p></sec></sec><sec><title>Abbreviations</title><p>IGB: Integrated Genome Browser; UCSC: University of California, Santa Cruz</p></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>WF performed the data analysis, drafted the manuscript and developed method jointly with LPZ. NK revised the manuscript. ARH and JMO conceived the study. LPZ conceived the study and developed the method jointly with WF. All authors read and approved the final manuscript.</p><fig position="float" id="F2"><label>Figure 2</label><caption><p>Histogram of the Z-scores for all 10,838 pseudo-exons obtained in the comparison of normal cerebellum samples with medulloblastomas.</p></caption><graphic xlink:href="1742-4682-3-19-2"/></fig></sec><sec sec-type="supplementary-material"><title>Supplementary Material</title><supplementary-material content-type="local-data" id="S1"><caption><title>Additional File 1</title><p>Alternative spliced pseudo-exons selected by our method: Comparison of normal cerebellum with medulloblastomas. Complete results for the 811 pseudo-exons predicted to be alternatively spliced between normal cerebellum and medulloblastomas.</p></caption><media xlink:href="1742-4682-3-19-S1.xls" mimetype="application" mime-subtype="vnd.ms-excel"><caption><p>Click here for file</p></caption></media></supplementary-material></sec>
Premature ovarian failure	<p>Premature ovarian failure (POF) is a primary ovarian defect characterized by absent menarche (primary amenorrhea) or premature depletion of ovarian follicles before the age of 40 years (secondary amenorrhea). It is a heterogeneous disorder affecting approximately 1% of women <40 years, 1:10,000 women by age 20 and 1:1,000 women by age 30. The most severe forms present with absent pubertal development and primary amenorrhea (50% of these cases due to ovarian dysgenesis), whereas forms with post-pubertal onset are characterized by disappearance of menstrual cycles (secondary amenorrhea) associated with premature follicular depletion. As in the case of physiological menopause, POF presents by typical manifestations of climacterium: infertility associated with palpitations, heat intolerance, flushes, anxiety, depression, fatigue. POF is biochemically characterized by low levels of gonadal hormones (estrogens and inhibins) and high levels of gonadotropins (LH and FSH) (hypergonadotropic amenorrhea). Beyond infertility, hormone defects may cause severe neurological, metabolic or cardiovascular consequences and lead to the early onset of osteoporosis. Heterogeneity of POF is also reflected by the variety of possible causes, including autoimmunity, toxics, drugs, as well as genetic defects. POF has a strong genetic component. X chromosome abnormalities (<italic>e.g. </italic>Turner syndrome) represent the major cause of primary amenorrhea associated with ovarian dysgenesis. Despite the description of several candidate genes, the cause of POF remains undetermined in the vast majority of the cases. Management includes substitution of the hormone defect by estrogen/progestin preparations. The only solution presently available for the fertility defect in women with absent follicular reserve is ovum donation.</p>	<contrib id="A1" corresp="yes" contrib-type="author"><name><surname>Beck-Peccoz</surname><given-names>Paolo</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Persani</surname><given-names>Luca</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib>	Orphanet Journal of Rare Diseases	<sec><title>Disease name and synonyms</title><p>Premature ovarian failure (POF; POF1: OMIM 311360); Hypergonadotropic ovarian failure; Menopausa precoce.</p><sec><title>Included diseases</title><p>POF2 (OMIM #300511); POF3 (OMIM #608996)</p></sec></sec><sec><title>Definition</title><p>Premature ovarian failure is defined as a primary ovarian defect characterized by absent menarche (primary amenorrhea) or premature depletion of ovarian follicles/arrested folliculogenisis before the age of 40 years (secondary amenorrhea) [<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B2">2</xref>].</p></sec><sec><title>Epidemiology</title><p>POF affects approximately: one in 10,000 women by age 20; one in 1,000 women by age 30; one in 100 women by age 40 [<xref ref-type="bibr" rid="B3">3</xref>]. The familial form of POF is rare, representing 4 to 31% of all cases of POF [<xref ref-type="bibr" rid="B4">4</xref>-<xref ref-type="bibr" rid="B6">6</xref>].</p></sec><sec><title>Etiology</title><p>The causes of POF are extremely heterogeneous. Acquired forms such as those occurring after treatments for neoplastic diseases or autoimmune diseases account for many cases [<xref ref-type="bibr" rid="B1">1</xref>]. POF has a strong genetic component with X chromosome abnormalities playing a primary role mainly in the cases with ovarian dysgenesis [<xref ref-type="bibr" rid="B7">7</xref>-<xref ref-type="bibr" rid="B10">10</xref>]. A gene (or genes) for POF (POF1) was localised to Xq21.3–Xq27 on the basis of deletions in various patients and families. A second gene (or genes) for POF (POF2) implicated by the analysis of balanced X/autosomal translocations has been localised to Xq13.3–q21.1. Despite the description of several candidate genes [<xref ref-type="bibr" rid="B11">11</xref>-<xref ref-type="bibr" rid="B16">16</xref>], the cause of POF still remains undetermined in the majority of the cases (idiopathic). This idiopathic form of POF can show sporadic and familial forms.</p><p>The different causes of POF are illustrated as follows:</p><p>• Iatrogenic origin (surgery, chemotherapy, radiations);</p><p>• Autoimmune, including polyglandular autoimmune syndrome, as well as autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy (APECED) due to mutations in <italic>AIRE </italic>gene);</p><p>• Infections (<italic>e.g. </italic>herpes zoster, cytomegalovirus);</p><p>• Chromosome X defects:</p><p>■ Turner syndrome</p><p>■ Fragile X syndrome (<italic>FMR1 </italic>gene premutation)</p><p>• Monogenic defects</p><p>■ Syndromic defects:</p><p>○ Congenital disorders of glycosylation (CDG, formerly named carbohydrate-deficient glycoprotein syndromes) (recessive)</p><p>○ Galactosemia (recessive)</p><p>○ Blepharophimosis-ptosis-epicanthus inversus syndrome (BPES) (female-limited, dominant)</p><p>○ Pseudohypoparathyroidism (PHP) type Ia (parental imprinting : maternal inheritance)</p><p>■ Isolated defects:</p><p>○ Follicle stimulating hormone (FSH) receptor mutations (<italic>FSHR</italic>), (recessive)</p><p>○ Luteinizing hormone (LH) receptor mutations (<italic>LHR</italic>), (recessive)</p><p>○ <italic>FOXL2 </italic>(transcription factor involved in BPES) mutations (female-limited defect, dominant)</p><p>○ Bone morphogenetic protein 15 (<italic>BMP15</italic>) mutations (female-limited defect, heterozygous mutation)</p><p>• Idiopathic</p><p>Defects in some of these candidate genes may present with different phenotypes. <italic>FOXL2 </italic>defects may present either with BPES type 1 (without POF) or with BPES type 2 (with POF), condition designated as POF3 [<xref ref-type="bibr" rid="B14">14</xref>]. Rarely, <italic>FOXL2 </italic>mutations may be associated with POF in the absence of eyelid/palpebral alterations (isolated POF) [<xref ref-type="bibr" rid="B17">17</xref>,<xref ref-type="bibr" rid="B18">18</xref>]. Depending on the degree of FSH resistance, <italic>FSHR </italic>defects are associated with primary [<xref ref-type="bibr" rid="B12">12</xref>] or secondary amenorrhea [<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B19">19</xref>]. Mutations in <italic>LHR </italic>have been described in women with secondary amenorrhea (characterized by elevated serum LH/FSH ratio and cystic follicles at ultrasound) belonging to pedigrees of male patients with Leydig hypoplasia [<xref ref-type="bibr" rid="B20">20</xref>]. Two of the candidate genes are located on the X chromosome. <italic>FMR1 </italic>gene (Xq27.3) mutations or pre-mutations are typically associated with secondary amenorrhea in female relatives of male patients with mental retardation [<xref ref-type="bibr" rid="B8">8</xref>]. <italic>BMP15 </italic>gene (Xp11.2) defect has so far been described in two sisters with primary amenorrhea and heterozygous for the mutation. This defect represents an unusual example of a X-linked disease in which affected females inherit the mutation from their unaffected father [<xref ref-type="bibr" rid="B15">15</xref>].</p></sec><sec><title>Clinical description</title><p>The symptoms can vary considerably from patient to patient and the disorder may occur abruptly or spontaneously or it may develop gradually over several years. The most severe forms of hypergonadotropic ovarian failure present with absent pubertal development and primary amenorrhea [<xref ref-type="bibr" rid="B2">2</xref>,<xref ref-type="bibr" rid="B21">21</xref>]. The clinical picture is characterized by absent menarche and pubertal delay results in absent sexual maturation and reduced growth velocity. In the female, pubertal delay is defined as the absence of mammary and pubic hair development and menarche at 13 years. Moderate hirsutism may be seen due to the action of androgens originating from adrenals.</p><p>About half of the cases of primary amenorrhea are due to ovarian dysgenesis, which is revealed by the finding of streak ovaries accompanied by uterus hypoplasia at ultrasound. In the other patients, follicles (<10 mm) may be found at histological evaluation such as in the case of <italic>FSHR </italic>mutations [<xref ref-type="bibr" rid="B22">22</xref>]. In these cases, almost normal pubertal development may be seen.</p><p>Post-pubertal onset of ovarian failure represents the large majority of the cases [<xref ref-type="bibr" rid="B1">1</xref>]. This is characterized by secondary amenorrhea associated with premature follicular depletion or arrested folliculogenisis. As in the case of physiological menopause, POF is clinically characterized by typical manifestations of climacterium such as palpitations, heat intolerance, flushes, night sweats, irritability, anxiety, depression, sleep disturbance, decreased libido, hair coarseness, vaginal dryness, fatigue.</p><p>Female infertility is an obvious and presently irreversible consequence of POF. Importantly, POF and prolonged lack of estrogen treatment may lead to the early onset of osteopenia and osteoporosis. Moreover, sexual hormone defects represent an important risk factor for frequent and severe neurological, metabolic or cardiovascular disorders such as Alzheimer's disease, hypercholesterolemia or ischemic diseases.</p><p>Hypergonadotropic ovarian failure may be part of other syndromic features (see the causes of POF): Autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy, Blepharophimosis-ptosis-epicanthus inversus syndrome, Carbohydrate-deficient glycoprotein syndromes, Galactosemia, Turner) and PHP I.</p><p>PHP I is associated with mutation in the <italic>GNAS </italic>gene encoding Gs protein alpha [<xref ref-type="bibr" rid="B23">23</xref>]. The diagnosis of PHP I is based on the findings of resistance to several peptide hormones acting through the adenylyl cyclase/cAMP pathway. The key findings are elevated parathyroid hormone (PTH) with low/normal calcemia, high thyrotropin (TSH) with normal thyroid hormone levels, growth hormone deficiency and high gonadotropins in patient with delayed puberty and skeletal abnormalities (Albright osteodystrophy). PHP I syndrome occurs when the mutant allele is inherited from the mother, due to the imprinting of the paternal <italic>GNAS </italic>allele in the affected tissues.</p><p>The early diagnosis of familial POF will provide the opportunity to predict the likelihood of early menopause, and allow other reproductive choices to be made, such as freezing embryos or having children earlier. As POF has cumulative negative effects over time, it is important for clinicians to make a timely diagnosis and begin appropriate strategies for symptom management, emotional support, and risk reduction.</p></sec><sec><title>Diagnostic methods</title><p>Both primary and secondary forms of ovarian failure are biochemically characterized by low levels of gonadal hormones (estrogens and inhibins) and high gonadotropins (LH and FSH) (hypergonadotropic amenorrhea). The elevation of FSH is usually more marked than that of LH and an FSH value >30 U/L is indicative of ovarian failure.</p><p>Ultrasound frequently reveals small ovaries without evidence of growing follicles. In the cases with primary amenorrhea, gonadal dysgenesis is documented by the finding of streak ovaries. Histological examination of biopsies performed during pelvic laparoscopy in the case of hypoplastic ovaries (0.20–0.30 ml on ultrasound) may reveal the presence of primary follicles. Forms of POF linked to the finding of ovarian cysts may be due to LH resistance (<italic>LHR </italic>mutations) which presents with secondary amenorrhea. In contrast to what is generally found in POF, defects in LH receptor are typically associated with a serum LH elevation (> 10 U/L) more pronounced than that of serum FSH. The evaluation of other peptide factors of ovarian origin, such as inhibin B and anti-mullerian hormone (AMH), may be useful to determine the follicular reserve when POF is suspected. Low levels of inhibin B may predict follicular depletion before the large FSH rise.</p><p>Karyotype evaluation and other cytogenetic investigations are useful to identify major X chromosome abnormalities.</p></sec><sec><title>Differential diagnosis</title><p>The differential diagnosis is based on the exclusion of other causes of primary and secondary amenorrhea (absence of menstruation for more than 6 months). Parameters useful for the exclusion of each of the following conditions are illustrated:</p><p>• Pregnancy: high chorionic gonadotropin (CG) levels.</p><p>• Iatrogenic causes (surgery, anti-neoplastic treatments, radiations, antidopaminergic drugs): complete anamnestic investigation.</p><p>• Hypothalamic-pituitary disease (pituitary tumors, hyperprolactinemia, Kallmann syndrome, ....): high prolactin (PRL) and low/normal gonadotropin levels, alterations at imaging of brain/sella region.</p><p>• Hypothalamic amenorrhea (induced by stress, intensive exercise, anorexia, weight loss, fasting, severe diseases,): low/normal gonadotropin levels.</p><p>• Polycystic ovaries: alterations at ovarian ultrasound, normal gonadotropin and high androgen levels.</p><p>• Enzymatic defects of steroidogenesis (e.g. 21-hydroxylase deficiency): alterations at physical and adrenal ultrasound, normal gonadotropin, high androgen and adrenocorticotropic hormone (ACTH) levels.</p><p>• Endocrine disorders, such as hyperthyroidism, hypothyroidism, Cushing syndrome: complete clinical/biochemical evaluations, normal gonadotropin levels.</p><p>• Only in patients with primary amenorrhea:</p><p>- vaginal/uterus anatomical abnormalities, such as Rokitanski syndrome or Asherman syndrome: alterations at physical examination/pelvic ultrasound, normal gonadotropin levels.</p><p>- disorders of sexual differentiation (<italic>e.g. </italic>resistance to androgens): alterations at physical/ultrasound examination, evaluation of karyotype, measure androgen/anti-mullerian hormone levels.</p></sec><sec><title>Genetic counseling</title><p>Genetic counseling is nowadays recommended for several reasons, when a genetic form of POF is suspected or identified.</p><p>Counseling is of particular importance in POF cases from families with X-linked mental retardation (Fragile X syndrome). Fragile X syndrome is due to CGG expansion (>55 repeats) at the 5'UTR of <italic>FMR1 </italic>gene (Xq27.3). The expansion of CGG repeats is associated with gene silencing resulting in male mental retardation and in POF with secondary amenorrhea in female carriers [<xref ref-type="bibr" rid="B8">8</xref>].</p><p>Genetic investigations may be useful for the early diagnosis of genetic defects underlying POF, when a female is born from a family with other female members affected with POF. Pedigree studies on affected families showed a mode of inheritance suggestive of autosomal dominant sex-limited transmission or X-linked inheritance with incomplete penetrance. In families with POF, the risk of other females developing POF will depend on the mode of inheritance and the mode of transmission. With autosomal dominant inheritance, the risk of POF will be 50% with either maternal or paternal transmission. However, with X-linked inheritance and paternal transmission this risk may be as high as 100%. These risks will be smaller with incomplete penetrance. If a POF patient appears to be a sporadic case, the risk of other female relatives developing POF will probably be equal to the risk in the general population.</p><p>All women who experience POF before the age of 30 years should perform a blood test for chromosomal assessment. Older women should discuss the option of chromosomal studies, as identification of abnormality may influence other family members, sisters or daughters, who carry the same defect in term of planning pregnancies. Carriers of the genetic defect may be advised for early pregnancy or oocyte collection and preservation.</p></sec><sec><title>Antenatal diagnosis</title><p>Not relevant at present.</p></sec><sec><title>Management</title><p>Patients with POF have infertility and hormone deficits. At present, fertility cannot be restored if the diagnosis is made after complete follicular depletion. In some cases, early diagnosis by genetic investigation may instead lead to advice for early conception or oocyte harvesting and preservation. Hormone defect may be substituted by estrogen/progestin preparations. The only solution presently available for the fertility defect in women with absent follicular reserve is represented by ovum donation.</p></sec><sec><title>Unresolved questions</title><p>In most of the isolated defects the cause is still unknown. Several candidate genes have been identified, but causative mutations have been found in a strict minority of patients [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B11">11</xref>,<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B16">16</xref>]. The prevalence of some genetic defects remains to be determined (<italic>e.g. BMP15 </italic>mutations). Though one paper described auto-antibodies against FSHR in a series of women with POF [<xref ref-type="bibr" rid="B24">24</xref>], auto-antigens and specific auto-antibodies for the diagnosis of autoimmune forms of isolated POF remain to be determined.</p></sec>
Effect of Ebola virus proteins GP, NP and VP35 on VP40 VLP morphology	<p>Recently we described a role for Ebola virus proteins, NP, GP, and VP35 in enhancement of VP40 VLP budding. To explore the possibility that VLP structure was altered by co-expression of EBOV proteins leading to the observed enhancement of VP40 VLP budding, we performed density gradient analysis as well as electron microscopy studies. Our data suggest that VP40 is the major determinant of VLP morphology, as co-expression of NP, GP and VP35 did not significantly change VLP density, length, and diameter. Ultra-structural changes were noted in the core of the VLPs when NP was co-expressed with VP40. Overall, these findings indicate that major changes in morphology of VP40 VLPs were likely not responsible for enhanced budding of VP40 VLPs in the presence of GP, NP and/or VP35.</p>	<contrib id="A1" contrib-type="author"><name><surname>Johnson</surname><given-names>Reed F</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Bell</surname><given-names>Peter</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A3" corresp="yes" contrib-type="author"><name><surname>Harty</surname><given-names>Ronald N</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	Virology Journal	<sec><title>Introduction</title><p>Ebola and Marburg viruses are members of the <italic>Filoviridae </italic>family of the order <italic>Mononegavirales</italic>. Both viruses are associated with recurrent outbreaks of deadly hemorrhagic fevers with mortality rates as high as 90% [<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B2">2</xref>]. Currently, there are no approved vaccines, nor treatments for Ebola virus (EBOV) infection. A better understanding of the molecular aspects of EBOV replication will be necessary for successful development of specific treatments for EBOV infection.</p><p>Ebola virus matrix protein, VP40, is the major virion protein and plays an essential role in virus assembly and budding [<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B4">4</xref>]. VP40 buds from the cell surface forming virus-like particles (VLPs). VLP budding is mediated by viral L-domains present in the N-terminus of the protein, which interact with host factors such as Nedd4 and TSG101, leading to VLP release [<xref ref-type="bibr" rid="B3">3</xref>-<xref ref-type="bibr" rid="B7">7</xref>]. It is hoped that investigations into the mechanisms of VP40 VLP budding will lead to possible vaccines and therapeutics that will block late stages of the virus life-cycle.</p><p>Recent evidence suggests that co-expression of other EBOV proteins will enhance VP40 VLP budding [<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B9">9</xref>]. For example, co-expression of VP40+GP+NP enhanced VP40 release approximately 40-fold over that observed for VP40 alone [<xref ref-type="bibr" rid="B9">9</xref>]. We have also demonstrated that VP35 interacts with VP40, is enclosed within VP40 VLPs, and functions to specifically package the EBOV 3E-5E minigenome into VLPs [<xref ref-type="bibr" rid="B10">10</xref>]. Currently, the mechanism by which EBOV proteins enhance VP40 budding is unclear, as is their affect on VLP morphology. Thus, we are interested in examining VLPs that contain combinations of VP40, GP, NP, and VP35 to determine whether co-expression of different EBOV proteins affects density, length, diameter, and overall morphology. Investigating the morphology of EBOV VLPs may give us insight into the mechanism by which EBOV proteins contribute to the observed enhancement of VLP budding.</p><p>Early EBOV reports suggest the virus particle is 970 nm in length and 80 nm in diameter with a density of 1.14 g/mL [<xref ref-type="bibr" rid="B11">11</xref>-<xref ref-type="bibr" rid="B13">13</xref>]. Since EBOV is a bio-safety level 4 pathogen, alternate means to study its properties have been developed. The most commonly used method to study EBOV proteins is transfection and co-expression of plasmids coding for individual viral proteins. Using this approach, Bavari et al. have demonstrated that co-expression of VP40 and GP yielded VLP particles 50–70 nm in diameter and 1–2 μm in length [<xref ref-type="bibr" rid="B13">13</xref>], while Jasenosky et al. determined the VP40 VLP particle density to be 1.11–1.13 g/ml [<xref ref-type="bibr" rid="B4">4</xref>]. In addition, Noda et al. demonstrated that GP formed 10 nm long spikes on the surface of VP40 VLPs, and VLPs were found to be 10 μm in length.</p><p>In this report, we performed sucrose density gradient sedimentation, electron microscopy (EM), and protease protection assays on VLPs from cells transfected with combinations of VP40, GP, NP, and/or VP35. We demonstrate that there are minimal changes in VLP density, diameter, and wall thickness with co-expression of other viral proteins. Statistically significant differences were found in measurements of wall thickness between VP40 VLPs and VP40+VP35 VLPs. Lastly, NP was packaged within VP40+NP VLPs, and VLP morphology was altered when NP was co-expressed with VP40.</p></sec><sec><title>Results</title><sec><title>NP is packaged within VP40 VLPS</title><p>We have demonstrated previously that NP enhances VP40 VLP budding 3.5 fold over VP40 alone, but did not demonstrate that NP was packaged within VP40 VLPs [<xref ref-type="bibr" rid="B9">9</xref>]. To prove that NP is packaged within VP40 VLPs, protease protection assays were performed. Similar experiments have been performed with VP35 to demonstrate that VP35 is also packaged within VP40 VLPs [<xref ref-type="bibr" rid="B10">10</xref>]. Human 293T cells were transfected with pCAGGS vector alone, VP40, NP, or VP40+NP. Purified VLPs were divided into six equal fractions. As reported previously, VP40 was only digested in the presence of both Triton X-100 and trypsin (Fig <xref ref-type="fig" rid="F1">1A</xref>, Lane 5) [<xref ref-type="bibr" rid="B6">6</xref>]. Similarly, we found that NP was degraded completely only in the presence of both Triton X-100 and trypsin (Fig. <xref ref-type="fig" rid="F1">1B</xref>, lane 5). Treatment with trypsin alone was insufficient to digest NP (Fig <xref ref-type="fig" rid="F1">1B</xref>, lane 4), indicating that NP is packaged within VP40 VLPs. It should be noted that NP was unable to bud from cells as a VLP when expressed alone in mammalian cells [<xref ref-type="bibr" rid="B9">9</xref>].</p><fig position="float" id="F1"><label>Figure 1</label><caption><p>Protease Protection Assay of VLPs containing NP and VP40. VLP pellets were split into six fractions and treated with STE buffer (lane 1), soybean trypsin inhibitor (lane 2), Triton X-100 (lane 3), Trypsin (lane 4), Trypsin and Triton X-100 (lane 5), Trypsin and soybean trypsin inhibitor (lane 6). The fractions were divided equally, and viral proteins were immunoprecipitated with α-VP40 (Panel A), or α-NP (Panel B). Samples were resolved by SDS-PAGE and detected by autoradiography.</p></caption><graphic xlink:href="1743-422X-3-31-1"/></fig></sec><sec><title>VP40 is the major determinant of EBOV VLP density</title><p>VLP density was determined by centrifugation of VLPs purified from transfected 293T cells over 20% sucrose in STE, followed by centrifugation on a 20–50% sucrose gradient for 14 hours. Five hundred microliter fractions were collected and analyzed for density using a refractometer. Fractions positive for VP40 as determined by Western blot were converted to density. Table <xref ref-type="table" rid="T1">1</xref> lists the densities for each combination of EBOV proteins expressed, and these data represent an average of 7 independent experiments.</p><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>VLP Density on 20–50% sucrose in STE.</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="center"><bold>Density g/mL</bold></td><td align="center"><bold>Standard Deviation</bold></td><td align="center"><bold>pNUM </bold>(t-test compared to VP40)</td></tr></thead><tbody><tr><td align="left">VP40 (n = 7)</td><td align="center">1.102</td><td align="center">0.0262</td><td></td></tr><tr><td align="left">VP40+VP35 (n = 7)</td><td align="center">1.125</td><td align="center">0.003</td><td align="center">0.5</td></tr><tr><td align="left">VP40+VP35+NP (n = 7)</td><td align="center">1.118</td><td align="center">0.006</td><td align="center">0.15</td></tr><tr><td align="left">VP40+NP (n = 7)</td><td align="center">1.102</td><td align="center">0.035</td><td align="center">0.34</td></tr><tr><td align="left">VP40+GP (n = 7)</td><td align="center">1.091</td><td align="center">0.038</td><td align="center">0.45</td></tr><tr><td align="left">VP40+NP+GP (n = 7)</td><td align="center">1.104</td><td align="center">0.031</td><td align="center">0.27</td></tr></tbody></table></table-wrap><p>The average density of VP40 VLPs was 1.102 g/mL. Co-expression of VP35 and VP40 increased VLP density to that of 1.125 g/mL. Co-expression of VP40+VP35+NP yielded VLPs with a density of 1.118 g/mL, whereas VP40+NP VLPs had a density of 1.102 g/mL (Table <xref ref-type="table" rid="T1">1</xref>). These data indicate a slight decrease in VLP density with the inclusion of GP to VP40 VLPs yielding a density of 1.091 g/mL. When NP was included with VP40+GP, the density increased to 1.104 g/mL; closer to that obtained with VP40 alone. It should be noted that no statistically significant differences were measured, and each VLP density measurement was well within the standard deviation of VP40 VLPs of 0.026 (Table <xref ref-type="table" rid="T1">1</xref>). The density of the VLPs suggests that VP40 is the major determinant of density of the VLP particle. Jasenosky et al reported VP40 VLP density of 1.11 to 1.13 g/mL, which is close to EBOV observations (1.14 g/mL) [<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B11">11</xref>,<xref ref-type="bibr" rid="B12">12</xref>]. Our findings suggest that VP40 determines density of the VLP particle, as co-expression of EBOV proteins did not significantly change VLP density.</p></sec><sec><title>Electron microscopic analysis of EBOV VLPs</title><p>Combinations of EBOV proteins (VP40, VP40+VP35, VP40+VP35+NP, VP40+NP+GP) were expressed in 293T cells, and samples were prepared for analysis by electron microscopy. Measurements of length, diameter, and wall thickness were taken for each set of VLPs, and these values were compared to those obtained for VP40 alone. Overall length of VLPs was measured (Fig. <xref ref-type="fig" rid="F2">2</xref> and Fig. <xref ref-type="fig" rid="F3">3</xref>) resulting in an average length of 540.70 nm for VP40 VLPs, 575.83 nm for VP40+VP35, 686.15 nm for VP40+VP35+NP, and 609.36 nm for VP40+NP+GP. It should be noted that length measurements likely reflect a mixture of full and partial-length VLPs. We identified several particles that correlated with full-length measurements reported previously by Bavari et al. (1.0–2.0 μm) [<xref ref-type="bibr" rid="B13">13</xref>]. The largest overall length of 1277 nm was observed for VP40+NP+GP VLPs, while the smallest overall length of 204 nm was observed for VP40 alone.</p><fig position="float" id="F2"><label>Figure 2</label><caption><p>A. Electron micrographs of individual VLP particles demonstrating the dimensions measured. B. Summary measurements of VLP Length, Diameter, and Wall Thickness. The data demonstrate that measured differences between VLPs are not statistically significant, except for VP40+VP35 VLPs. The (*) indicates statistically significant differences.</p></caption><graphic xlink:href="1743-422X-3-31-2"/></fig><fig position="float" id="F3"><label>Figure 3</label><caption><p>Electron micrographs for A) VP40 VLPs, B) VP40+VP35 VLPs, C) VP40+VP35+NP VLPs, and D) VP40+NP+GP VLPs.</p></caption><graphic xlink:href="1743-422X-3-31-3"/></fig><p>VP40 VLPs had an average diameter of 57.58 nm, with slight increases for VP40+VP35 at 63.64 nm, VP40+VP35+NP at 60.35 nm, and VP40+NP+GP at 66.57 nm (Fig. <xref ref-type="fig" rid="F2">2</xref>). These data are consistent with diameters reported by Bavari et al of 50–70 nm and narrower than that reported for EBOV of 80 nm [<xref ref-type="bibr" rid="B13">13</xref>]. These data support a role for VP40 as the major component that determines VLP length and diameter. Co-expression of NP, GP, and/or VP35 did not significantly alter VLP diameter. (Fig. <xref ref-type="fig" rid="F2">2</xref>.)</p><p>A morphological difference was observed following examination of cross-sections of VLPs expressing VP40 alone vs. those expressing additional EBOV proteins. For example, VP40 and VP40+VP35 VLPs had disorganized centers, while VLPs which included NP possessed a "bull's-eye" appearance (Fig. <xref ref-type="fig" rid="F4">4</xref>). In addition, the region around the membrane appeared more uniform in VLPs that included NP (Fig. <xref ref-type="fig" rid="F4">4</xref>). Next, we measured the wall thickness of VLPs in cross section (see Figure <xref ref-type="fig" rid="F2">2</xref> for detailed description of what we defined as diameter and "wall thickness"). The average measurement of wall thickness was 17.48 nm for VP40 VLPs, 21.15 nm for VP40+VP35, 17.67 nm for VP40+VP35+NP, and 18.55 nm for VP40+NP+GP (Figs. <xref ref-type="fig" rid="F2">2</xref> and <xref ref-type="fig" rid="F4">4</xref>). A statistically significant difference was found between VP40 and VP40+VP35 with a pNum value of 0.008.</p><fig position="float" id="F4"><label>Figure 4</label><caption><p>Representative examples of VLPs in cross section. Note the appearance of a bull's-eyes structure within VLPs containing NP.</p></caption><graphic xlink:href="1743-422X-3-31-4"/></fig></sec></sec><sec><title>Conclusion</title><p>Overall, expression of GP, NP and/or VP35 with VP40 did not drastically affect VLP density or diameter. However, wall thickness increased slightly with co-expression of VP35 and VP40. We also observed packaging of NP within VP40+NP VLPs and altered VLP morphology when NP was co-expressed with VP40+GP or VP40+VP35. These data suggest that VP40 is the major determinant of VLP morphology. These data support prior findings by Noda, Bavari, and Jasenosky [<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B13">13</xref>], and expand upon these findings by incorporation of NP and VP35 into VLPs.</p><p>Our data demonstrate that VLP density increased with the inclusion of the nucleocapsid proteins NP and VP35 (NP+VP35 – 1.118 g/mL, VP35 – 1.125 g/mL) suggesting that these VLPs may be more organized and more closely resemble authentic virus particles. In contrast, when GP was co-expressed with VP40, a slight decrease in density was observed (1.102 vs. 1.091). This may be due to a change in morphology of the particle and a resulting retardation in the sucrose gradient by extension of GP from the VLP surface [<xref ref-type="bibr" rid="B5">5</xref>]. VP40+NP+GP and VP40+NP VLPs had densities roughly equal to that of VP40 alone, with VP40+NP particles being slightly denser. These findings suggest that the contribution of NP to density of the particle offsets the decrease in density by GP. Statistical analysis suggests that the densities are within one standard deviation of the density of VP40 alone. It is also important to point out that densities observed here and by Jasenosky et al. are slightly less than that observed for mature virions (1.14 g/mL). This difference could be due to complete incorporation of nucleocapsids and less variability in particles produced due to infection vs. transfection.</p><p>Morphometric analyses suggest that there are differences in the lengths of the VLP particles (Fig. <xref ref-type="fig" rid="F2">2</xref>); however, length measurements likely vary due to the possibility that complete VLP particles were not consistently measured. Indeed it is likely that only a portion of a VLP was measured due to the VLP exiting the plane of section being inspected. Analysis by EM also suggests that there were slight differences in VLP diameter with co-expression of the nucleocapsid components VP35 and NP (Fig. <xref ref-type="fig" rid="F2">2</xref>). These differences were not found to be statistically significant. Overall particle diameter was as follows: 57.58 nm for VP40 VLPs, 63.64 nm for VP40+VP35 VLPs, 60.35 nm for VP40+VP35+NP VLPs, and 66.57 nm for VP40+NP+GP VLPs. The increase in diameter of particles containing VP35 correlates with cross-sectional observations of wider particles with no organized centers (Fig. <xref ref-type="fig" rid="F4">4</xref>). Additionally, morphological differences were evident in VLPs expressing NP (bull's-eye, Fig. <xref ref-type="fig" rid="F4">4</xref>). The formation of the bull's-eye within the VP40+VP35+NP VLPs and VP40+NP+GP VLPs may indicate that NP self-aggregated (as is seen with Marburg NP [<xref ref-type="bibr" rid="B14">14</xref>]) resulting in electron-dense material within VLPs. VP40+VP35+NP VLPs were also slightly more narrow than those containing VP40+VP35, suggesting an organization of the core of the virus imparted by NP+VP35 nucleocapsid formations [<xref ref-type="bibr" rid="B15">15</xref>]. The "bull's-eye" present in VP40+GP+NP VLPs along with the results of the protease protection assay suggest that NP is packaged within VLPs.</p><p>The wall thickness of the VLPs containing EBOV proteins was found to vary. Indeed, statistical analysis indicated a significant difference between VLPs containing VP40 alone vs. those containing VP40+VP35. Gross inspection of the particles indicated a less uniform appearance and thickening of the VLP wall in VP40+VP35 VLPs. When NP was co-expressed, the wall appeared more uniform around the circumference of the VLP, suggesting that NP has an organizing effect during the formation of VLPs.</p><p>In sum, our data suggest that VP40 is the major determinant of VLP density, and co-expression of NP, GP, and/or VP35 had minor affects on VLP morphology. It will be of interest to use immuno-EM to more precisely map the location of EBOV proteins within VLPs. Thus, the enhancement of VP40 release following co-expression of additional EBOV proteins [<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B9">9</xref>] is unlikely to be due to gross changes in VLP morphology. One possibility is that enhancement of VP40 release may be due to modifications of the cellular architecture, or effects on host protein function as a result of EBOV protein expression [<xref ref-type="bibr" rid="B16">16</xref>].</p></sec><sec sec-type="materials\|methods"><title>Materials and methods</title><sec><title>Cells, plasmids, and antibodies</title><p>293T cells were maintained in Dulbecco Modified Eagles Media (DMEM) 10% FBS, 5% CO<sub>2</sub>. pCAGGS, pCAGGS VP40, pCAGGS NP, and pCAGGS GP have been described previously [<xref ref-type="bibr" rid="B9">9</xref>]. The pCAGGS VP35 construct was provided by Chris Basler (Mount Sinai, New York) and contains an influenza hemagglutanin (HA) epitope tag at the amino terminus of VP35. Antibodies to VP40 were provided by Roland Grunow (Marburg, Germany). Antibodies to NP were provided by Jason Paragas (Fort Detrick, USAMRIID). Antibodies to GP were provided by Paul Bates (Univ. of Penn). HA antibody was obtained from Roche.</p></sec><sec><title>Protease protection assay</title><p>VLP budding assays were performed as described previously [<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B17">17</xref>] with minor modifications as follows. 293-T cells on 100 mm tissue culture dishes (Corning) were transfected with 10 μg of pCAGGS, pCAGGS VP40, or pCAGGS NP+ pCAGGS VP40 plasmids using Lipofectamine (Invitrogen). 1600 μCi of [<sup>35</sup>S]Met/Cys (Perkin Elmer) was added to each 100 mm dish 24 hours post-transfection. At thirty hours post transfection, culture medium was clarified at 1500 rpm for 10 minutes, layered over a 20% sucrose in STE buffer (0.01 M Tris-HCl [pH 7.5], 0.01 M NaCl, 0.001 M EDTA [pH 8.0]) cushion, and VLPs were purified at 36,000 rpm for 2 hours at 4°C. VLPs were suspended in 400 μL of STE buffer and divided into six, 60 μL fractions. The fractions were treated with either 6 μL of STE buffer, 3 mg/mL soybean trypsin inhibitor (Roche Biochemicals), 1% Triton X-100 (Fisher), 0.1 mg/mL trypsin (Promega), 1% Triton X-100 and 0.1 mg/ml trypsin, or 0.1 mg/ml trypsin and 3 mg/mL trypsin inhibitor. All fractions were incubated at room temperature for 30 minutes followed by the addition of 25 μL of soybean trypsin inhibitor at a concentration of 100 mg/mL to quench the reactions. Each fraction was lysed in RIPA buffer (50 mM Tris [pH 8.0], 150 mM NaCl, 1.0% NP-40, 0.5% deoxycholate, 0.1% sodium dodecyl sulfate [SDS]) in a final volume of 500 μL. Proteins were immunoprecipitated with appropriate antibodies, and resolved by SDS-Polyacrylamide gel electrophoresis (PAGE). As a control for protein expression, the transfected cells were lysed in RIPA buffer and split into equal fractions. Immunoprecipitated proteins were resolved by SDS-PAGE, and visualized by autoradiography.</p></sec><sec><title>VLP density analysis</title><p>VLP budding assays were performed as described previously [<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B17">17</xref>] with minor modifications as follows. 293T cells on 100 mm tissue culture dishes (Corning) were transfected with 10 μg of pCAGGS, pCAGGS VP40, pCAGGS VP35+pCAGGS VP40, pCAGGS NP+pCAGGS VP40, or pCAGGS VP40+pCAGGS GP, or pCAGGS VP40+pCAGGS NP+pCAGGS VP35, or pCAGGS VP40+pCAGGS NP+pCAGGS GP plasmids using Lipofectamine (Invitrogen). VLPs were purified over 20% sucrose as previously described [<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B9">9</xref>]. VLPs were resuspended in 500 μL STE buffer and overlayed onto a 20–50% sucrose gradient prepared in STE buffer. VLPs were centrifuged at 36,000 × G for 14 hours and 0.5 mL fractions were collected for each sample. An equal portion of each fraction was run on a 10% SDS-PAGE gel, transferred and probed with anti-VP40 antibody. Transfected cells were lysed in RIPA buffer and electrophoresed over a 10% SDS-PAGE gel, transferred and probed with anti-VP40, anti-GP, anti-NP, and anti-HA to ensure that all viral proteins were expressed.</p></sec><sec><title>Electron microscopy and morphometry</title><p>Human 293T cells on 6-well tissue culture dishes (Corning) were transfected with 1 μg of pCAGGS, pCAGGS VP40, pCAGGS VP40+pCAGGS VP35, pCAGGS VP40+pCAGGS NP, or pCAGGS VP40+pCAGGS GP, or pCAGGS VP40+pCAGGs NP+pCAGGS VP35, or pCAGGS VP40+pCAGGS NP+pCAGGS GP plasmids using Lipofectamine (Invitrogen). Transfected cells were harvested and centrifuged at 3000 rpm in a microcentrifuge for 5 min to form a loose pellet. Cell pellets were then fixed overnight in 2.5% glutaraldehyde, 2% paraformaldehyde in 0.1 M sodium cacodylate buffer (pH 7.3), washed in cacodylate buffer, and incubated in 2% OsO<sub>4 </sub>for 2 hours. After washing with water, samples were stained overnight in 0.5% uranyl acetate, washed again, dehydrated through a graded ethanol series and propylene oxide, and finally embedded in resin (LX-112, Ladd Research Industries). Ultrathin sections (80 nm) were stained with uranyl acetate and lead citrate according to standard protocols and examined with a Philips CM-100 transmission electron microscope equipped with a KeenView digital camera system. Morphometric measurements were performed on digital images using AnalySIS software (Soft Imaging System, Lakewood, CO).</p></sec></sec>
Violent crime exposure classification and adverse birth outcomes: a geographically-defined cohort study	<sec><title>Background</title><p>Area-level socioeconomic disparities have long been associated with adverse pregnancy outcomes. Crime is an important element of the neighborhood environment inadequately investigated in the reproductive and public health literature. When crime has been used in research, it has been variably defined, resulting in non-comparable associations across studies.</p></sec><sec sec-type="methods"><title>Methods</title><p>Using geocoded linked birth record, crime and census data in multilevel models, this paper explored the relevance of four spatial violent crime exposures: two proximal violent crime categorizations (count of violent crime within a one-half mile radius of maternal residence and distance from maternal residence to nearest violent crime) and two area-level crime categorizations (count of violent crimes within a block group and block group rate of violent crimes) for adverse birth events among women in living in the city of Raleigh NC crime report area in 1999–2001. Models were adjusted for maternal age and education and area-level deprivation.</p></sec><sec><title>Results</title><p>In black and white non-Hispanic race-stratified models, crime characterized as a proximal exposure was not able to distinguish between women experiencing adverse and women experiencing normal birth outcomes. Violent crime characterized as a neighborhood attribute was positively associated with preterm birth and low birth weight among non-Hispanic white and black women. No statistically significant interaction between area-deprivation and violent crime category was observed.</p></sec><sec><title>Conclusion</title><p>Crime is variably categorized in the literature, with little rationale provided for crime type or categorization employed. This research represents the first time multiple crime categorizations have been directly compared in association with health outcomes. Finding an effect of area-level violent crime suggests crime may best be characterized as a neighborhood attribute with important implication for adverse birth outcomes.</p></sec>	<contrib id="A1" corresp="yes" contrib-type="author"><name><surname>Messer</surname><given-names>Lynne C</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" equal-contrib="yes" contrib-type="author"><name><surname>Kaufman</surname><given-names>Jay S</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A3" equal-contrib="yes" contrib-type="author"><name><surname>Dole</surname><given-names>Nancy</given-names></name><xref ref-type="aff" rid="I3">3</xref><email>[email protected]</email></contrib><contrib id="A4" equal-contrib="yes" contrib-type="author"><name><surname>Herring</surname><given-names>Amy</given-names></name><xref ref-type="aff" rid="I4">4</xref><email>[email protected]</email></contrib><contrib id="A5" equal-contrib="yes" contrib-type="author"><name><surname>Laraia</surname><given-names>Barbara A</given-names></name><xref ref-type="aff" rid="I3">3</xref><email>[email protected]</email></contrib>	International Journal of Health Geographics	<sec><title>Background</title><p>The literature related to area-level effects, particularly socioeconomic disadvantage, on public health outcomes has grown substantially in recent years [<xref ref-type="bibr" rid="B1">1</xref>-<xref ref-type="bibr" rid="B18">18</xref>]. Research in perinatal health has demonstrated modest but consistent effects of neighborhood-level socioeconomic disparities in key pregnancy outcomes using census variables [<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B19">19</xref>,<xref ref-type="bibr" rid="B20">20</xref>]. Low birth weights (LBW) have been associated with a variety of neighborhood level socioeconomic variables including poverty [<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B11">11</xref>,<xref ref-type="bibr" rid="B12">12</xref>], unemployment [<xref ref-type="bibr" rid="B11">11</xref>], education and income [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B11">11</xref>] and median rent [<xref ref-type="bibr" rid="B8">8</xref>], as well as socio-demographic indices of economic disadvantage [<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B17">17</xref>]. Area-level research has a long-standing tradition in the United Kingdom (U.K.) [<xref ref-type="bibr" rid="B21">21</xref>-<xref ref-type="bibr" rid="B24">24</xref>]. Established area-level indices such as the Townsend Material Deprivation Score and the Carstairs Deprivation Index have been widely utilized in the U.K., which have allowed for the comparison of deprivation effects across a variety of geographic regions. While the last decade has seen advancements in the field of neighborhood research in the United States (U.S.), including innovative data collection approaches [<xref ref-type="bibr" rid="B25">25</xref>-<xref ref-type="bibr" rid="B28">28</xref>] and increased access to relevant methodology such as multilevel modeling and geographic information systems (GIS) applications, the range of exposures used to represent relevant area-level effects has been largely limited to socio-demographic variables available from the census.</p><p>Census data, critical for identifying important associations between socioeconomic disadvantage and a variety of adverse health outcomes, are limited in their utility for public health research several reasons. First, census data are available only at decennial intervals in the US, whereas neighborhood condition and its subsequent effect can change within the span of a few years. Second, the exclusive use of census variables, which are produced by aggregating individual responses to census questions, implies that the important features of 'neighborhoods' can be captured by aggregating individual measures, ignoring the important role of other neighborhood features, such as the presence neighborhood of parks, the quality of area resources or the quantity of local disamenities such as land fills or strip clubs [<xref ref-type="bibr" rid="B29">29</xref>,<xref ref-type="bibr" rid="B30">30</xref>]. Third, census variables are limited to socio-demographic features. Economic and demographic features, while clearly important, are not the only neighborhood characteristics likely to affect health intermediates and outcomes. Fourth, census data are restricted to census geography, such as block groups and tracts, which may bear no resemblance to the salient features of 'neighborhoods' for most individuals. Therefore, while census variables continue to function as crude surrogates for neighborhood attributes, other aspects of the neighborhood must be assessed to elucidate more clearly the pathways through which neighborhoods might influence health.</p><p>Crime is a potentially important neighborhood characteristic inadequately examined in public health research despite documented relevance to birth outcomes, asthma and health behaviors, such as physical activity. Crime, and fear of crime, are both considered in the literature. Fear of crime is thought to contribute to an underlying mechanism explaining area differences in health [<xref ref-type="bibr" rid="B31">31</xref>-<xref ref-type="bibr" rid="B33">33</xref>] and has been directly associated with poor health outcomes in several studies [<xref ref-type="bibr" rid="B34">34</xref>-<xref ref-type="bibr" rid="B36">36</xref>], even after adjusting for health behaviors and a number of individual and household level socioeconomic factors [<xref ref-type="bibr" rid="B31">31</xref>]. Neighborhood-level crime is inversely associated with physical activity [<xref ref-type="bibr" rid="B37">37</xref>], especially among adolescent females [<xref ref-type="bibr" rid="B38">38</xref>], with important implications for youth overweight. Exposure to violence, including neighborhood violence, also predicts higher numbers of asthma symptom days and more nights of lost sleep for caretakers, after adjustment for socioeconomic status, housing deterioration and negative life events [<xref ref-type="bibr" rid="B39">39</xref>].</p><p>Four previous studies have reported an association between crime and poor pregnancy outcomes. In Santiago, Chile, Zapata and colleagues assigned ratings for sociopolitical violence and found exposure to violent environments associated with poor pregnancy outcomes, adjusting for individual-level risk factors [<xref ref-type="bibr" rid="B40">40</xref>]. In their research on impoverished women in Chicago (those living in census tracts with family median incomes <$10,000), Collins and David (1997) found more small-for-gestational-age and LBW deliveries among women living in high (16% LBW), compared with low crime rate neighborhoods (12% LBW). These relationships remained significant after controlling for individual risk factors [<xref ref-type="bibr" rid="B41">41</xref>]. Similarly, using the violent crime rate in Chicago, Morenoff (2003) found violent crime to be a robust neighborhood predictor of LBW after controlling for individual covariates [<xref ref-type="bibr" rid="B18">18</xref>]. Most recently, we found living in block groups in the highest and second highest quartiles of violent crime rate, compared with the lowest quartile of violent crime rate, was associated with increased odds of preterm birth among black non-Hispanic women (OR = 1.5, 95%CI: 0.9, 2.6, OR = 1.4, 95%CI: 1.0, 2.1, respectively) in adjusted models [<xref ref-type="bibr" rid="B42">42</xref>]. The relationship between neighborhood violence and birth outcomes appears suggestive and may help explain the disparity in birth outcomes between white and black women.</p><p>Crime exposure has been variably categorized in the literature and have included episodes of political violence [<xref ref-type="bibr" rid="B40">40</xref>,<xref ref-type="bibr" rid="B43">43</xref>,<xref ref-type="bibr" rid="B44">44</xref>], one-half mile violent crime density [<xref ref-type="bibr" rid="B38">38</xref>], violent crime incidents [<xref ref-type="bibr" rid="B45">45</xref>], and area crime rates [<xref ref-type="bibr" rid="B18">18</xref>,<xref ref-type="bibr" rid="B41">41</xref>,<xref ref-type="bibr" rid="B42">42</xref>,<xref ref-type="bibr" rid="B46">46</xref>]. Increased accessibility to geocoded data has enabled more sophisticated crime modeling techniques [<xref ref-type="bibr" rid="B47">47</xref>,<xref ref-type="bibr" rid="B48">48</xref>]. Crime is often reported as a rate or count for a specific geographic space, such as city, county or state. Geocoding, and the creation of distance-based measures to area resources or disamenities, has been an important innovation in health-related research [<xref ref-type="bibr" rid="B49">49</xref>-<xref ref-type="bibr" rid="B51">51</xref>]. Geocoding allows one to observe the spatial distribution of crime over multiple geographies to more clearly see the areas within a given locality (city, census tract) where crimes are most likely to be reported. It further allows the creation of different crime exposures, like distance to nearest crimes or count of crimes within a certain spatial area. The utility of different crime modeling techniques and the most relevant exposure for health outcomes has not been previously explored, despite probable implications for health.</p><p>Using the health example of adverse birth outcomes, this paper will explore the relevance of various violent crime categorizations by assessing their respective associations with two adverse birth outcomes, preterm birth and low birth weight. <italic>A priori</italic>, we hypothesized that proximal crime measures would be more predictive of adverse birth outcomes than area-level measures. Because no diagnostic criteria are available to formally compare the different exposure models, the predictive utility of each crime categorization, or that crime exposure categorization that proves most useful in differentiating women who experience an adverse birth outcome, will be considered most relevant. Previous published work on Wake County NC indicates black non-Hispanic and white non-Hispanic women live in different socioeconomic and demographic environments. Because of these known heterogeneities in neighborhood structures, which are associated with crime exposure, separate analyses were conducted for black non-Hispanic and white non-Hispanic women using multilevel logistic regression models.</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Data sources</title><p>Exposure data from 1999–2001 City of Raleigh NC crime reports contain event locations, of which 99% were geocoded to latitude and longitude using Geographic Data Technology, Inc. (GDT) and assigned a crime category based on methods described in the literature [<xref ref-type="bibr" rid="B52">52</xref>]. These analyses are limited to violent crimes, including homicides, assaults, sexual assaults and kidnappings, of which 21,037 (22.5% of total crimes) occurred during the study period in Raleigh. Violent crime events were assigned to 2000 U.S. Census block groups to produce violent crime counts.</p><p>The birth outcome and maternal characteristics data are from 1999–2001 Wake County NC birth records (N = 30,481). Each maternal address identified on the birth certificate was geocoded to latitude and longitude and assigned to year 2000 U.S. census block groups. Of the 98.6% of birth records with complete addresses sent to GDT for geocoding, 93.2% achieved an exact census tract match using GDT's methods. The North Carolina birth records contain birth outcome, personal characteristics and health behavior information on each woman.</p><p>The third data source is year 2000 U.S. census data, from which the deprivation score, a neighborhood control variable, was derived. The deprivation score is a single summary representing four socioeconomic domains including poverty, housing, employment and education [<xref ref-type="bibr" rid="B53">53</xref>].</p></sec><sec><title>Neighborhood definition</title><p>Neighborhood is a term loosely used to refer to a person's immediate residential environment, which is hypothesized to have both material and social characteristics related to health [<xref ref-type="bibr" rid="B54">54</xref>,<xref ref-type="bibr" rid="B55">55</xref>]. Census block groups were used to approximate the neighborhood environment. This level of aggregation is large enough to contain women who delivered during the study years, but small enough to approximate the immediate physical neighborhood for study subjects. Previous research has advocated using the smallest possible level of aggregation due to the considerable crime variability within larger ecological units [<xref ref-type="bibr" rid="B25">25</xref>,<xref ref-type="bibr" rid="B56">56</xref>].</p></sec><sec><title>Violent crime exposures</title><p>The study was limited to women who resided in the Raleigh crime reporting area and were therefore subject to influence by reported area-level crimes (N = 13,960). Wake County comprises 263 block groups of which 114 represent the Raleigh crime report area. Crime exposure was conceptualized in two distinct ways: as an attribute of the neighborhood and as an attribute of the individual.</p><sec><title>Neighborhood-level crime variables</title><p>At the neighborhood level, block group measures include a) block group violent crime count and b) block group violent crime rate per thousand population ([block group violent crime count / block group population] × 1000). Neighborhood crime tertiles were created based on block group violent crime distributions with 38 block groups per tertile, then merged with maternal data, resulting in common neighborhood categories for black non-Hispanic (hereafter referred to as "black") and white non-Hispanic (hereafter referred to as "white") women.</p></sec><sec><title>Proximal crime variables</title><p>Two variables conceptualize violent exposure to crime as an individual-level attribute, subject to the geographic distribution of the women in the sample. Proximal violent crime exposure was defined as: a) the count of violent crimes within a one-half mile radius of each woman's address and b) the distance in feet from maternal residence to the nearest violent crime. Previous work [<xref ref-type="bibr" rid="B42">42</xref>] and preliminary analyses indicated white and black women were exposed to different quantities of violent crime. Creating proximal crime cutpoints based on the combined distribution of black and white women resulted in the majority of white women falling in the lowest crime categories and the majority of black women falling in the highest crime categories, with virtually no representation of the other racial group in the extremes of the violent crime distribution. Using race-specific cutpoints allowed for a more equal distribution of women of both races across the continuum of individually defined crime. For these reasons, race-specific tertiles were created for the two variables that defined violent crime as an individual attribute. Furthermore, the geographic area represented by these proximal crime exposures differed substantially from the block group-level crime variables. Raleigh NC block groups are quite large, with a mean size of 1.26 square miles (range: 0.10, 15.64). These proximal crime variables represented a much smaller unit of geographic space around the study women.</p></sec></sec><sec><title>Outcome definition</title><p>This research explored the association between crime and two adverse birth outcomes, preterm birth and low birth weight (LBW). Preterm birth is defined as birth at gestational age <37 weeks and weighing less than 3,888 grams [<xref ref-type="bibr" rid="B57">57</xref>]. Clinically estimated gestational age was obtained from the birth record. Comparisons between clinically estimated gestational age and last menstrual period calculated gestational age found that for Wake County births, the clinical estimate better approximated the expected preterm, term and post-term proportions that were obtained from a clinical sample with ultrasound dated gestational age. For this reason, the clinical estimate of gestational age, combined with the weight restriction for preterm births, was used to calculate preterm birth. Less than one percent of the records were missing gestational age information. Low birth weight is defined as birth at less than 2500 grams. The vital records were missing no birth weight data. The birth cohort for this analysis was limited to singleton births.</p></sec><sec><title>Covariates</title><p>Individual covariates considered include maternal age, education and marital status. These individual-level variables are established risk factors for preterm birth and possible confounders to the neighborhood crime-preterm birth relationship. The neighborhood-level covariate considered for this analysis was neighborhood deprivation and was controlled using a neighborhood deprivation score. The deprivation score is the weighted sum of nine standardized census variables including block group percents of households below 1999 poverty level, female headed households with dependent children, earning < $30,000 per year, on public assistance, with no car, selected owner and renter costs in excess of 50% of income, unemployment, individuals over age 25 with less than a high school education and median household value. The deprivation index has a median value of -0.4, a mean of 0.3 and standard deviation of 2.3 and a range of -2.3 to +12.5. The low end of the deprivation score indicates lack of deprivation (i.e., affluence) whereas a high end of the range suggests a large amount of deprivation. Continuous and categorical forms of the covariates were considered and the categorical forms used in the models. Location of police substation, while associated with crime reporting, is not associated with birth outcomes [<xref ref-type="bibr" rid="B42">42</xref>] and was not included as a neighborhood covariate in these analyses.</p></sec><sec><title>Data analysis</title><p>Distributions and prevalence ratios of each exposure variable and individual and neighborhood-level covariates were examined. Analyses of variance (ANOVA) tested if mean neighborhood-level characteristics differed by block-group violent crime rate tertile. Race-specific multilevel logistic regression analyses were conducted to explore the contribution of the neighborhood environment (second level variables) over that of the individual-level predictors and to account for any clustering of the birth outcomes. The authors estimated random effects logistic models with a fixed slope value for each predictor variable but with block group-specific intercepts, adjusting the models for individual and neighborhood covariates. Adjustment for confounders was made when the crude odds ratio differed from the adjusted odds ratio by 10% or more [<xref ref-type="bibr" rid="B58">58</xref>]. Terms for the interaction between neighborhood deprivation and crime were introduced to assess if these terms improved the model fit. Analyses were conducted in Stata 8.2.</p></sec></sec><sec><title>Results</title><p>Of the 11,256 non-Hispanic women delivering singleton live births in Raleigh NC during the study years, 471 (6.7% of 7036) white and 539 (12.8% of 4220) black women delivered preterm and 308 (4.4%) white and 485 (11.5%) black women delivered a LBW infant (Table <xref ref-type="table" rid="T1">1</xref>). White women in this sample were generally older with more years of education than black women. Black women were exposed on average to four times as many violent crimes within a half-mile radius: mean number = 106.9 (standard deviation = 139.1) and lived closer to the nearest violent crime: mean = 377.8 (745.9) feet compared with white women: who were exposed to, on average, 24.7 (43.2) crimes and lived on average 1173 (1414) feet from the nearest violent crime (Table <xref ref-type="table" rid="T2">2</xref>).</p><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>Individual-level characteristics of study sample, Raleigh crime report area, 1999–2001</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td></td><td align="center"><bold>NH White N (%)</bold></td><td align="center"><bold>NH Black N (%)</bold></td><td align="center"><bold>% PTB</bold></td><td align="center"><bold>% LBW</bold></td><td align="center"><bold>PR* (95% CI)</bold></td></tr></thead><tbody><tr><td align="left"><bold>Preterm birth</bold></td><td></td><td align="center">471 (6.7)</td><td align="center">539 (12.8)</td><td align="center">8.7</td><td align="center">70.6</td><td align="center">1.91 (1.69, 2.15)</td></tr><tr><td align="left"><bold>Low birth weight</bold></td><td></td><td align="center">308 (4.4)</td><td align="center">485 (11.5)</td><td align="center">70.6</td><td align="center">6.9</td><td align="center">2.49 (2.17, 2.87)</td></tr><tr><td align="left" colspan="7">(column percent) <bold>Individual-level maternal characteristics</bold></td></tr><tr><td align="left"><bold>Marital status</bold></td><td></td><td></td><td></td><td></td><td></td><td></td></tr><tr><td></td><td align="left"><bold>Married</bold></td><td align="center">6611 (89.7)</td><td align="center">1815 (41.5)</td><td align="center">7.3</td><td align="center">5.2</td><td align="center">0.46 (0.44, 0.48)</td></tr><tr><td></td><td align="left"><bold>Not married</bold></td><td align="center">761 (10.3)</td><td align="center">2562 (58.5)</td><td align="center">11.8</td><td align="center">10.6</td><td align="center">5.70 (5.30, 6.12)</td></tr><tr><td align="left"><bold>Maternal age</bold></td><td></td><td></td><td></td><td></td><td></td><td></td></tr><tr><td></td><td align="left"><bold>< 20 years</bold></td><td align="center">210 (2.9)</td><td align="center">555 (12.7)</td><td align="center">9.7</td><td align="center">9.9</td><td align="center">4.45 (3.81, 5.19)</td></tr><tr><td></td><td align="left"><bold>20–24 years</bold></td><td align="center">772 (10.5)</td><td align="center">1344 (30.7)</td><td align="center">9.4</td><td align="center">8.2</td><td align="center">3.11 (2.87, 3.38)</td></tr><tr><td></td><td align="left"><bold>25–29 years</bold></td><td align="center">2000 (27.1)</td><td align="center">1155 (27.4)</td><td align="center">8.7</td><td align="center">6.1</td><td align="center">0.97 (0.91, 1.03)</td></tr><tr><td></td><td align="left"><bold>30–34 years</bold></td><td align="center">2798 (38.0)</td><td align="center">823 (18.8)</td><td align="center">7.5</td><td align="center">5.8</td><td align="center">0.50 (0.46, 0.53)</td></tr><tr><td></td><td align="left"><bold>35+ years</bold></td><td align="center">1592 (21.6)</td><td align="center">500 (11.4)</td><td align="center">9.3</td><td align="center">6.8</td><td align="center">0.53 (0.48, 0.58)</td></tr><tr><td align="left"><bold>Maternal education</bold></td><td></td><td></td><td></td><td></td><td></td><td></td></tr><tr><td></td><td align="left"><bold>< 12 years</bold></td><td align="center">309 (4.2)</td><td align="center">794 (18.2)</td><td align="center">10.3</td><td align="center">8.6</td><td align="center">4.33 (3.82, 4.91)</td></tr><tr><td></td><td align="left"><bold>12 years</bold></td><td align="center">936 (12.7)</td><td align="center">1350 (30.9)</td><td align="center">10.6</td><td align="center">9.2</td><td align="center">2.43 (2.26, 2.62)</td></tr><tr><td/><td align="left"><bold>> 12 years</bold></td><td align="center">6110 (83.1)</td><td align="center">2219 (50.9)</td><td align="center">7.6</td><td align="center">5.6</td><td align="center">0.61 (0.59, 0.63)</td></tr></tbody></table><table-wrap-foot><p>* Distribution, percent preterm birth, percent low birth weight and prevalence ratio [PR] (95% confidence intervals [95% CI]) for individual-level attributes of women living in Raleigh crime report area, 1999–2001</p><p>*PR = prevalence in blacks/prevalence in whites</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T2"><label>Table 2</label><caption><p>Proximal violent crime characteristics of study sample, Raleigh crime report area, 1999–2001</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left">(column percent)</td><td align="center"><bold>NH White N (%)</bold></td><td align="center"><bold>NH Black N (%)</bold></td><td align="center"><bold>% PTB</bold></td><td align="center"><bold>% LBW</bold></td><td align="center"><bold>PR** (95% CI)</bold></td></tr></thead><tbody><tr><td align="left" colspan="6"><bold>Count of violent crimes within one-half mile of all maternal address</bold></td></tr><tr><td align="left"><bold>Low (0 – 12)</bold></td><td align="center">3049 (52.6)</td><td align="center">735 (19.0)</td><td align="center">7.1</td><td align="center">5.3</td><td align="center">0.36 (0.34, 0.39)</td></tr><tr><td align="left"><bold>Medium (13 – 49)</bold></td><td align="center">2014 (34.7)</td><td align="center">1098 (28.3)</td><td align="center">9.0</td><td align="center">7.3</td><td align="center">0.82 (0.77, 0.87)</td></tr><tr><td align="left"><bold>High (50 – 633)</bold></td><td align="center">736 (12.7)</td><td align="center">2043 (52.7)</td><td align="center">10.2</td><td align="center">8.8</td><td align="center">4.15 (3.86, 4.47)</td></tr><tr><td align="left" colspan="6"><bold>Distance in feet from maternal residence to nearest violent crime</bold></td></tr><tr><td align="left"><bold>Close (0 – 89)</bold></td><td align="center">829 (12.7)</td><td align="center">2120 (51.9)</td><td align="center">10.2</td><td align="center">8.9</td><td align="center">4.09 (3.82, 4.39)</td></tr><tr><td align="left"><bold>Med (89 – 771)</bold></td><td align="center">2279 (34.9)</td><td align="center">1387 (34.0)</td><td align="center">8.7</td><td align="center">7.0</td><td align="center">0.97 (0.92, 1.03)</td></tr><tr><td align="left"><bold>Far (772 – 15,616)</bold></td><td align="center">3430 (52.5)</td><td align="center">577 (14.1)</td><td align="center">6.8</td><td align="center">4.8</td><td align="center">0.27 (0.25, 0.29)</td></tr></tbody></table><table-wrap-foot><p>* Distribution, percent preterm, percent low birth weight and prevalence ratios [PR] (95% confidence intervals [95% CI]) for proximal crime variables for women living in Raleigh crime report area, 1999–2001</p><p>*PR = prevalence in blacks/prevalence in whites</p></table-wrap-foot></table-wrap><p>Levels of neighborhood deprivation also differed by race (Table <xref ref-type="table" rid="T3">3</xref>), with 54.2% of white women living in the block groups in the lowest tertile of deprivation (most affluent) compared with a roughly equal proportion of black women, 52.1%, living in the most deprived neighborhoods, or those block groups in the highest tertile of deprivation. Consistent with this pattern of exposure to neighborhood deprivation, black women lived in block groups with more violent crimes (68.6 [53.1]) and a higher violent crime rate (31.4 [37.2]) than white women (27.4 [36.1] and 8.5 [16.5], respectively) in this study.</p><table-wrap position="float" id="T3"><label>Table 3</label><caption><p>Area-level deprivation and crime characteristics of study sample, Raleigh crime report area, 1999–2001</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left">(column percent)</td><td align="center"><bold>NH White N (%)</bold></td><td align="center"><bold>NH Black N (%)</bold></td><td align="center"><bold>% PTB</bold></td><td align="center"><bold>% LBW</bold></td><td align="center"><bold>PR** (95% CI)</bold></td></tr></thead><tbody><tr><td align="center" colspan="6"><bold>Neighborhood-level deprivation</bold></td></tr><tr><td align="left" colspan="3"><bold>Neighborhood deprivation</bold></td><td></td><td></td><td></td></tr><tr><td align="left"><bold>Low ([-2.8]–[-0.7])</bold></td><td align="center">3922 (54.2)</td><td align="center">659 (15.1)</td><td align="center">6.8</td><td align="center">4.8</td><td align="center">0.28 (0.26, 0.30)</td></tr><tr><td align="left"><bold>Med ([-0.6]–0.8)</bold></td><td align="center">2667 (36.2)</td><td align="center">1436 (32.8)</td><td align="center">8.7</td><td align="center">6.8</td><td align="center">0.91 (0.86, 0.96)</td></tr><tr><td align="left"><bold>High (0.82 – 12.5)</bold></td><td align="center">713 (9.7)</td><td align="center">2282 (52.1)</td><td align="center">11.0</td><td align="center">9.7</td><td align="center">5.39 (5.00, 5.81)</td></tr><tr><td align="center" colspan="6"><bold>Neighborhood-level violent crime</bold></td></tr><tr><td align="left" colspan="3"><bold>Neighborhood violent crime count</bold></td><td></td><td></td><td></td></tr><tr><td align="left"><bold>Low (0 – 12)</bold></td><td align="center">3176 (43.1)</td><td align="center">583 (13.3)</td><td align="center">6.8</td><td align="center">4.7</td><td align="center">0.31 (0.29, 0.33)</td></tr><tr><td align="left"><bold>Medium (13 – 50)</bold></td><td align="center">3019 (41.0)</td><td align="center">1348 (30.8)</td><td align="center">8.6</td><td align="center">6.7</td><td align="center">0.75 (0.71, 0.79)</td></tr><tr><td align="left"><bold>High (52 – 378)</bold></td><td align="center">1177 (16.0)</td><td align="center">2446 (55.9)</td><td align="center">10.4</td><td align="center">8.8</td><td align="center">3.50 (3.30, 3.71)</td></tr><tr><td align="left" colspan="3"><bold>Neighborhood violent crime rate</bold></td><td></td><td></td><td></td></tr><tr><td align="left"><bold>Low (0 – 6)</bold></td><td align="center">4553 (61.8)</td><td align="center">920 (21.0)</td><td align="center">7.0</td><td align="center">4.9</td><td align="center">0.34 (0.32, 0.36)</td></tr><tr><td align="left"><bold>Medium (6 – 16)</bold></td><td align="center">2138 (29.0)</td><td align="center">1404 (32.1)</td><td align="center">9.1</td><td align="center">7.2</td><td align="center">1.11 (1.05, 1.17)</td></tr><tr><td align="left"><bold>High (17 – 205)</bold></td><td align="center">681 (9.2)</td><td align="center">2053 (46.9)</td><td align="center">11.0</td><td align="center">9.6</td><td align="center">5.07 (4.70, 5.49)</td></tr></tbody></table><table-wrap-foot><p>* Distribution, percent preterm, percent low birth weight and prevalence ratios [PR] (95% confidence intervals [95% CI]) for neighborhood-level deprivation and violent crime variables for women living in Raleigh crime report area, 1999–2001</p><p>*PR = prevalence in blacks/prevalence in whites</p></table-wrap-foot></table-wrap><p>Neighborhood-level features were associated with block group crime tertiles in Raleigh NC (Table <xref ref-type="table" rid="T4">4</xref>). Block groups with low rates of violent crime (tertile 1) generally had less poverty (6.1%) and unemployment (2.3%) and fewer households with the following characteristics: female headed with dependent children (7.1%), earning < $30,000/year (19.5%), on public assistance (1.2%), with no car (1.4%) and with low education (6.4%) compared with block groups characterized by high rates of violent crime (tertile 3). Block group level median home value decreased as block groups experienced more violent crimes from approximately $22,000 to $12,000. Interestingly, one neighborhood feature that did not differ across crime tertiles was high housing costs; housing costs appear generally high across Raleigh block groups, regardless of crime rate. The correlation between the neighborhood deprivation index and the continuous violent crime rate was 0.55.</p><table-wrap position="float" id="T4"><label>Table 4</label><caption><p>Association between area-level characteristics and tertiles of block group crime rate 1999–2001</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left">Mean percent (standard deviation)</td><td align="center"><bold>Tertile 1: (0.0 – 0.005)</bold></td><td align="center"><bold>Tertile 2: (0.006 – 0.019)</bold></td><td align="center"><bold>Tertile 3: (0.022 – 0.21)</bold></td><td align="center"><bold>F-test P-value</bold></td></tr></thead><tbody><tr><td align="left"><bold>Black Non-Hispanic</bold></td><td align="center">9.7 (11.5)</td><td align="center">24.3 (21.5)</td><td align="center">52.0 (32.2)</td><td align="center">P < 0.001</td></tr><tr><td align="left"><bold>Poverty</bold></td><td align="center">6.1 (5.4)</td><td align="center">11.7 (9.9)</td><td align="center">22.1 (14.8)</td><td align="center">P < 0.001</td></tr><tr><td align="left"><bold>Female headed HH</bold></td><td align="center">7.1 (4.9)</td><td align="center">12.2 (7.9)</td><td align="center">21.9 (17.5)</td><td align="center">P < 0.001</td></tr><tr><td align="left"><bold>HH income <$30,000</bold></td><td align="center">19.5 (10.3)</td><td align="center">30.1 (15.3)</td><td align="center">47.4 (18.1)</td><td align="center">P < 0.001</td></tr><tr><td align="left"><bold>Public assistance</bold></td><td align="center">1.2 (1.4)</td><td align="center">1.5 (2.8)</td><td align="center">5.3 (5.5)</td><td align="center">P < 0.001</td></tr><tr><td align="left"><bold>No car</bold></td><td align="center">1.4 (1.4)</td><td align="center">1.9 (2.4)</td><td align="center">8.0 (12.4)</td><td align="center">P < 0.001</td></tr><tr><td align="left"><bold>High housing costs*</bold></td><td align="center">23.0 (12.7)</td><td align="center">24.5 (10.2)</td><td align="center">27.0 (10.1)</td><td align="center">P = 0.29</td></tr><tr><td align="left"><bold>Unemployment</bold></td><td align="center">2.3 (2.3)</td><td align="center">4.2 (5.2)</td><td align="center">9.8 (11.7)</td><td align="center">P < 0.001</td></tr><tr><td align="left"><bold>Low education*</bold></td><td align="center">6.4 (7.7)</td><td align="center">11.5 (9.1)</td><td align="center">25.5 (14.6)</td><td align="center">P < 0.001</td></tr><tr><td align="left"><bold>Median home value***</bold></td><td align="center">22.0 (88.3)</td><td align="center">17.4 (11.1)</td><td align="center">12.0 (60.7)</td><td align="center">P < 0.001</td></tr></tbody></table><table-wrap-foot><p> Percent of block group characteristics and F-statistic P-value comparing mean values for block group crime rate tertiles in Raleigh crime report area, 1999–2001.</p><p> Percent households reporting living under the 1999 federal poverty line</p><p>* Percent households headed by females with dependent children</p><p>** Specified owner or renter costs in excess of 50% of income</p><p>* Less than 12 years of education among adults 25 years or older</p><p>**** Median household value(× $10000.00)</p></table-wrap-foot></table-wrap><p>White women's odds of adverse birth outcomes appear modestly associated with violent crime in this sample. The proximal exposure categorizations of violent crime, such as the count of violent crimes within a half-mile and distance to nearest violent crimes (Table <xref ref-type="table" rid="T5">5</xref>), do not distinguish women at increased odds of delivering preterm; the odds ratios in adjusted models remain close to the null. Exposure to high counts of violent crimes within a half-mile radius of maternal residence suggests a small association with both preterm birth and LBW in unadjusted models, but following adjustment, the association is attenuated. White women living in neighborhoods with high, compared with low rates of violent crime (Table <xref ref-type="table" rid="T6">6</xref>) appear to be at increased odds of preterm birth (OR= 1.4; 95% CI: 1.1, 1.9) and LBW (OR = 1.7; 95% CI: 1.2, 2.4). The relationship between living in a block group with a high violent crime rate (between 17 and 205 crimes per 1000 population) and adverse birth outcomes is attenuated following adjustment for individual and neighborhood covariates, though still suggestive of increased odds for LBW (OR = 1.5; 95% CI: 1.0, 2.3) but less so for preterm birth (OR = 1.3; 95% CI: 0.9, 1.9).</p><table-wrap position="float" id="T5"><label>Table 5</label><caption><p>Associations between proximal crime exposures and adverse birth outcomes among non-Hispanic white women</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="center" colspan="4"><bold>Preterm birth (<37 weeks gestation & <3888 g)</bold></td><td align="center" colspan="4"><bold>Low birth weight (<2500 g)</bold></td></tr></thead><tbody><tr><td></td><td align="center"><bold>Not adjusted</bold></td><td align="center"><bold>Individual covariates</bold></td><td align="center"><bold>Area-level covariates</bold></td><td align="center"><bold>Fully adjusted</bold></td><td align="center"><bold>Not adjusted</bold></td><td align="center"><bold>Individual covariates</bold></td><td align="center"><bold>Area-level covariates</bold></td><td align="center"><bold>Fully adjusted</bold></td></tr><tr><td colspan="9"><hr></hr></td></tr><tr><td align="left" colspan="9"><bold>Count of violent crimes within 1/2 mile of maternal address (White non-Hispanic cutpoints used in model)</bold></td></tr><tr><td align="left"><bold>Low (0 – 6)</bold></td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td></tr><tr><td align="left"><bold>Medium (7 – 21)</bold></td><td align="center">1.0 (0.8, 1.4)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.1 (0.8, 1.5)</td><td align="center">1.0 (0.7, 1.4)</td><td align="center">1.0 (0.7, 1.4)</td><td align="center">1.0 (0.7, 1.4)</td></tr><tr><td align="left"><bold>High (22 – 562)</bold></td><td align="center">1.2 (1.0, 1.6)</td><td align="center">1.1 (0.9, 1.5)</td><td align="center">1.1 (0.8, 1.6)</td><td align="center">1.1 (0.8, 1.5)</td><td align="center">1.2 (0.9, 1.7)</td><td align="center">1.1 (0.8, 1.5)</td><td align="center">1.1 (0.8, 1.7)</td><td align="center">1.0 (0.7, 1.5)</td></tr><tr><td align="left" colspan="9"><bold>Distance (in feet) from maternal address to nearest violent crime (White non-Hispanic cutpoints used in model)</bold></td></tr><tr><td align="left"><bold>Close (0–512)</bold></td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td></tr><tr><td align="left"><bold>Medium (513–1239)</bold></td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.1 (0.8, 1.4)</td><td align="center">1.1 (0.8, 1.4)</td><td align="center">1.1 (0.9, 1.5)</td><td align="center">0.8 (0.6, 1.1)</td><td align="center">0.9 (0.7, 1.2)</td><td align="center">0.8 (0.6, 1.1)</td><td align="center">0.9 (0.7, 1.2)</td></tr><tr><td align="left"><bold>Far (<15,617)</bold></td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.1 (0.9, 1.4)</td><td align="center">1.1 (0.8, 1.4)</td><td align="center">1.2 (0.9, 1.5)</td><td align="center">0.8 (0.6, 1.1)</td><td align="center">0.9 (0.7, 1.3)</td><td align="center">0.9 (0.6, 1.2)</td><td align="center">1.0 (0.7, 1.3)</td></tr></tbody></table><table-wrap-foot><p> Unadjusted, individual (maternal age, education), neighborhood (area-level deprivation), and fully-adjusted (individual and area-level covarites) odds ratios [OR] (95% Confidence Intervals [95% CI]) of preterm birth and low birth weight for 1/2 mile violent crime count and distance to nearest violent crime among non-Hispanic white women living in Raleigh NC.</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T6"><label>Table 6</label><caption><p>Associations between area-level violent crime and adverse birth outcomes among non-Hispanic white women</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="center" colspan="4"><bold>Preterm birth (<37 weeks gestation & <3888 g)</bold></td><td align="center" colspan="4"><bold>Low birth weight (<2500 g)</bold></td></tr></thead><tbody><tr><td></td><td align="center"><bold>Not adjusted</bold></td><td align="center"><bold>Individual covariates</bold></td><td align="center"><bold>Area-level covariates</bold></td><td align="center"><bold>Fully adjusted</bold></td><td align="center"><bold>Not adjusted</bold></td><td align="center"><bold>Individual covariates</bold></td><td align="center"><bold>Area-level covariates</bold></td><td align="center"><bold>Fully adjusted</bold></td></tr><tr><td colspan="9"><hr></hr></td></tr><tr><td align="left" colspan="9"><bold>Block group violent crime count</bold></td></tr><tr><td align="left"><bold>Low (0 – 12)</bold></td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td></tr><tr><td align="left"><bold>Med (13 – 50)</bold></td><td align="center">1.0 (0.8, 1.2)</td><td align="center">1.0 (0.8, 1.2)</td><td align="center">1.0 (0.8, 1.2)</td><td align="center">0.9 (0.8, 1.2)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.0 (0.8, 1.3)</td></tr><tr><td align="left"><bold>High (52 – 378)</bold></td><td align="center">1.2 (0.9, 1.6)</td><td align="center">1.1 (0.8, 1.4)</td><td align="center">1.1 (0.8, 1.5)</td><td align="center">1.1 (0.8, 1.5)</td><td align="center">1.2 (0.9, 1.6)</td><td align="center">1.0 (0.8, 1.5)</td><td align="center">1.0 (0.7, 1.5)</td><td align="center">1.0 (0.7, 1.4)</td></tr><tr><td align="left" colspan="9"><bold>Block group violent crime rate (count/population 1000)</bold></td></tr><tr><td align="left"><bold>Low (0 – 5.7)</bold></td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td></tr><tr><td align="left"><bold>Medium (6.1–16.3)</bold></td><td align="center">1.0 (0.8, 1.2)</td><td align="center">0.9 (0.7, 1.2)</td><td align="center">1.0 (0.8, 1.2)</td><td align="center">1.0 (0.7, 1.2)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.0 (0.7, 1.3)</td><td align="center">1.0 (0.7, 1.4)</td><td align="center">1.0 (0.7, 1.3)</td></tr><tr><td align="left"><bold>High (<204.7)</bold></td><td align="center">1.4 (1.1, 1.9)</td><td align="center">1.3 (1.0, 1.8)</td><td align="center">1.3 (0.9, 1.9)</td><td align="center">1.3 (0.9, 1.9)</td><td align="center">1.7 (1.2, 2.4)</td><td align="center">1.5 (1.1, 2.2)</td><td align="center">1.6 (1.0, 2.4)</td><td align="center">1.5 (1.0, 2.3)</td></tr></tbody></table><table-wrap-foot><p>* Unadjusted, individual (maternal age, education), neighborhood (area-level deprivation), and fully-adjusted (individual and area-level covarites) odds ratios [OR] (95% Confidence Intervals [95% CI]) of preterm birth and low birth weight for area-level violent crime count and violent crime rate among non-Hispanic white women living in Raleigh NC.</p></table-wrap-foot></table-wrap><p>Similar to the results observed for white women, crime categorized as a proximal exposure shows little association with adverse birth outcomes for black women. In unadjusted models (Table <xref ref-type="table" rid="T7">7</xref>), living far from the nearest violent crime suggests protection against LBW (OR = 0.8, 95% CI: 0.6, 1.0), which is the direction of the relationship one might anticipate, but this association is attenuated following adjustment for individual covariates (OR = 0.9, 95% CI: 0.7, 1.1). Violent crime defined as a neighborhood attribute, however, appears to be modestly associated with adverse birth outcomes among black women. In unadjusted models (Table <xref ref-type="table" rid="T8">8</xref>), living in a block group with medium and high counts of violent crime confers 80% and 60% increased odds of preterm birth (OR = 1.8, 95% CI: 1.3, 2.5, and OR = 1.6, 95% CI: 1.2, 2.3, respectively). Following adjustment, odds of preterm birth are reduced for the highest tertile (OR = 1.2, 95% CI: 0.9, 1.8), but remain associated for the middle tertile (OR = 1.7, 95% CI: 1.2, 2.4). A similar pattern is apparent for LBW; medium levels of violent crime count are associated with LBW after adjustment for individual covariates (OR = 1.6, 95% CI: 1.1, 2.4) while high levels of violent crime count appear to confer reduced odds of low birth following adjustment (OR = 1.2, 95% CI: 0.8, 1.7). In the models employing violent crime rates, odds for adverse birth outcomes show a similar pattern. Unadjusted results suggest living in block groups characterized by medium (OR = 1.6; 95% CI: 1.2, 2.0) or high (OR = 1.5; 95% CI: 1.1, 1.9) rates of violent crime are associated with increased odds of preterm birth, compared with living in low violent crime rate block groups. Following adjustment for covariates, the associations are attenuated (OR = 1.3; 95% CI: 1.0, 1.8 and OR = 1.1; 95% CI: 0.8, 1.6) but remain modestly associated for medium violent crime rate block groups. Similarly, the odds of LBW are associated with medium (OR = 1.3; 95% CI: 1.0, 1.8) and high rates of violent crime (OR = 1.5; 95% CI: 1.2, 2.0) in unadjusted models, but these relationships are reduced following adjustment (OR = 1.2, 95% CI: 0.9, 1.7 and OR = 1.2, 95% CI: 0.8, 1.7, respectively).</p><table-wrap position="float" id="T7"><label>Table 7</label><caption><p>Associations between proximal violent crime exposures and adverse birth outcomes among non-Hispanic black women</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="center" colspan="4"><bold>Preterm birth (<37 weeks gestation & <3888 g)</bold></td><td align="center" colspan="4"><bold>Low birth weight (<2500 g)</bold></td></tr></thead><tbody><tr><td></td><td align="center"><bold>Not adjusted</bold></td><td align="center"><bold>Individual covariates</bold></td><td align="center"><bold>Area-level covariates</bold></td><td align="center"><bold>Fully adjusted</bold></td><td align="center"><bold>Not adjusted</bold></td><td align="center"><bold>Individual covariates</bold></td><td align="center"><bold>Area-level covariates</bold></td><td align="center"><bold>Fully adjusted</bold></td></tr><tr><td colspan="9"><hr></hr></td></tr><tr><td align="left" colspan="9"><bold>Count of violent crimes within 1/2 mile of maternal address (Black non-Hispanic cutpoints used in this model)</bold></td></tr><tr><td align="left"><bold>Low (0 – 29)</bold></td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td></tr><tr><td align="left"><bold>Medium(30 – 81)</bold></td><td align="center">1.0 (0.8, 1.3)</td><td align="center">0.9 (0.7, 1.2)</td><td align="center">0.9 (0.7, 1.2)</td><td align="center">0.9 (0.7, 1.2)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">0.9 (0.7, 1.2)</td><td align="center">0.9 (0.7, 1.2)</td><td align="center">0.9 (0.6, 1.1)</td></tr><tr><td align="left"><bold>High (82 – 633)</bold></td><td align="center">1.1 (0.9, 1.4)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.0 (0.7, 1.4)</td><td align="center">0.9 (0.7, 1.3)</td><td align="center">1.3 (1.0, 1.6)</td><td align="center">1.1 (0.8, 1.4)</td><td align="center">1.1 (0.8, 1.5)</td><td align="center">1.0 (0.7, 1.3)</td></tr><tr><td align="left" colspan="9"><bold>Distance (in feet) from maternal address to nearest violent crime (Black non-Hispanic cutpoints used in model)</bold></td></tr><tr><td align="left"><bold>Close (0 – 21)</bold></td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td></tr><tr><td align="left"><bold>Medium (22 – 287)</bold></td><td align="center">1.1 (0.9, 1.4)</td><td align="center">1.1 (0.9, 1.4)</td><td align="center">1.1 (0.9,1.4)</td><td align="center">1.1 (0.9, 1.4)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.1 (0.8, 1.4)</td><td align="center">1.0 (0.8, 1.3)</td><td align="center">1.1 (0.8, 1.4)</td></tr><tr><td align="left"><bold>Far (<12857)</bold></td><td align="center">0.8 (0.7, 1.1)</td><td align="center">0.9 (0.7, 1.2)</td><td align="center">0.9 (0.7, 1.1)</td><td align="center">0.9 (0.7,1.2)</td><td align="center">0.8 (0.6, 1.0)</td><td align="center">0.8 (0.6, 1.1)</td><td align="center">0.8 (0.6, 1.0)</td><td align="center">0.9 (0.7, 1.1)</td></tr></tbody></table><table-wrap-foot><p>* Unadjusted, individual (maternal age, education), neighborhood (area-level deprivation), and fully-adjusted (individual and area-level covarites) odds ratios [OR] (95% Confidence Intervals [95% CI]) of preterm birth and low birth weight for 1/2 mile violent crime count and distance to nearest violent crime among non-Hispanic black women living in Raleigh NC.</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T8"><label>Table 8</label><caption><p>Associations between area-level violent crime and adverse birth outcomes among non-Hispanic black women</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="center" colspan="4"><bold>Preterm birth (<37 weeks gestation & <3888 g)</bold></td><td align="center" colspan="4"><bold>Low birth weight (<2500 g)</bold></td></tr></thead><tbody><tr><td></td><td align="center"><bold>Not adjusted</bold></td><td align="center"><bold>Individual covariates</bold></td><td align="center"><bold>Area-level covariates</bold></td><td align="center"><bold>Fully adjusted</bold></td><td align="center"><bold>Not adjusted</bold></td><td align="center"><bold>Individual covariates</bold></td><td align="center"><bold>Area-level covariates</bold></td><td align="center"><bold>Fully adjusted</bold></td></tr><tr><td colspan="9"><hr></hr></td></tr><tr><td align="left" colspan="9"><bold>Block group violent crime count</bold></td></tr><tr><td align="left"><bold>Low (0 – 12)</bold></td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td></tr><tr><td align="left"><bold>Med (13 – 50)</bold></td><td align="center">1.8 (1.3, 2.5)</td><td align="center">1.8 (1.3, 2.5)</td><td align="center">1.7 (1.2, 2.4)</td><td align="center">1.7 (1.2, 2.4)</td><td align="center">1.6 (1.1, 2.3)</td><td align="center">1.7 (1.2, 2.4)</td><td align="center">1.6 (1.1, 2.3)</td><td align="center">1.6 (1.1, 2.4)</td></tr><tr><td align="left"><bold>High (52 – 378)</bold></td><td align="center">1.6 (1.2, 2.3)</td><td align="center">1.5 (1.1, 2.1)</td><td align="center">1.3 (0.9, 1.9)</td><td align="center">1.2 (0.9, 1.8)</td><td align="center">1.6 (1.2, 2.2)</td><td align="center">1.4 (1.0, 2.0)</td><td align="center">1.3 (0.9, 1.9)</td><td align="center">1.2 (0.8, 1.7)</td></tr><tr><td align="left" colspan="9"><bold>Block group violent crime rate (count/population 1000)</bold></td></tr><tr><td align="left"><bold>Low (0 – 5.7)</bold></td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td><td align="center">Referent</td></tr><tr><td align="left"><bold>Medium (6.1–16.3)</bold></td><td align="center">1.6 (1.2, 2.0)</td><td align="center">1.5 (1.2, 2.0)</td><td align="center">1.4 (1.0, 1.8)</td><td align="center">1.3 (1.0, 1.8)</td><td align="center">1.3 (1.0, 1.8)</td><td align="center">1.3 (1.0, 1.8)</td><td align="center">1.2 (0.9, 1.7)</td><td align="center">1.2 (0.9, 1.7)</td></tr><tr><td align="left"><bold>High (<204.7)</bold></td><td align="center">1.5 (1.1, 1.9)</td><td align="center">1.3 (1.0, 1.8)</td><td align="center">1.2 (0.9, 1.7)</td><td align="center">1.1 (0.8, 1.6)</td><td align="center">1.5 (1.2, 2.0)</td><td align="center">1.3 (1.0, 1.8)</td><td align="center">1.3 (0.9, 1.8)</td><td align="center">1.2 (0.8, 1.7)</td></tr></tbody></table><table-wrap-foot><p>* Unadjusted, individual (maternal age, education), neighborhood (area-level deprivation), and fully-adjusted (individual and area-level covarites) odds ratios [OR] (95% Confidence Intervals [95% CI]) of preterm birth and low birth weight for area-level violent crime count and violent crime rate among non-Hispanic black women living in Raleigh NC.</p></table-wrap-foot></table-wrap><p>To examine if the interaction between neighborhood deprivation and crime influenced birth outcomes, we developed 8 models interacting the dichotomized deprivation index with each category of violent crime. Based on likelihood ratio tests, none of the interaction models was an improvement over the main effect models (p-values ranging from 0.3 – 0.8). In each interaction model, the adjusted odds ratios were consistent with the main effect models.</p></sec><sec><title>Discussion</title><p>The measurement of neighborhood effects on health has generally been imprecise, in part due to conceptual and methodological limitations [<xref ref-type="bibr" rid="B59">59</xref>]. Researchers often have to make use of administrative or other types of data not explicitly collected for research on health outcomes that offer few options for variable creation and exposure categorization. Using geocoded crime report data, this research sought to contribute to the neighborhood effects literature by testing various spatial and area-level violent crime exposure categorizations and assessing their association with adverse birth outcomes.</p><p>Crime was grouped into two main categories: attributes of individuals, or proximal measures, and attributes of neighborhoods. Crimes characterized as proximal, including the count of violent crimes within a half mile of maternal address and distance from maternal address to nearest violent crime were less predictive than those estimating the neighborhood environment in associations with adverse birth outcomes.</p><p>Among white women, the rate of neighborhood violent crime showed the strongest association with low birth weight, and a modest association with preterm birth. Several possible explanations exist for this modest effect. First, it is possible that violent crime has little effect on birth outcomes. This potential explanation is somewhat refuted by the violent crime effect observed in other studies. And while possible, this explanation seems unlikely given the research showing that women, and particularly white women, are more fearful of crime than others, despite their decreased risk of victimization compared with other racial and gender groups [<xref ref-type="bibr" rid="B60">60</xref>]. The second explanation for the modest violent crime effect is that the white women in this area are minimally exposed. The research reported here supports this explanation. Well over half the white women in this study had low counts of violent crimes within a one-half mile, lived in the farthest tertile from the nearest violent crime, and in block groups with the lowest counts and rate of violent crimes. A related explanation is that the relatively crime free and affluent neighborhoods in which the majority of white women in Raleigh reside offer some protection against, or buffer against, the possibly harmful effects of crime exposure. This explanation is at least partially supported by the study findings, as well. Observing an effect of neighborhood crime rate for white women living in block groups characterized by the highest tertile of violent crime is suggestive and supports the need for further investigation of crime as a neighborhood phenomenon.</p><p>Finding the strongest violent crime effect among black women living in the middle, compared with the lowest tertile of violent crime count was unanticipated. One possible explanation for this finding lies in the fact these middle tertile block groups are more racially heterogeneous than the neighborhoods characterized by the upper tertiles of violent crime and previous research has found racial homogeneity to be of some health benefit [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B61">61</xref>]. Neighborhoods in the middle tertile of violent crime house more white women; black women living in these more heterogeneous neighborhoods may be exposed to other, unmeasured stressors that women at the extremes do not face. One candidate stressor is interpersonal or institutional racism, which may affect birth outcomes. Additionally, black women living in the highest tertile of violent crime are simultaneously exposed to multiple economic stressors. While one might anticipate economic deprivation would lead to increased odds of adverse birth outcomes, as multiple studies have found, these women may have psychologically and physiologically accommodated this increased level of environmental stress and become relatively inured to its effects. In the absence of better individual-level exposure assessment and information regarding women's coping with and perceptions of crime, it is difficult to know what unmeasured individual or neighborhood confounders may influence the association between crime and birth outcomes and put the black women in this middle crime tertile at a relative disadvantage.</p><p>Finding an effect of neighborhood-level violent crime is consistent with previous research. In their research on impoverished women in Chicago (those living in census tracts with family median incomes <$10,000), Collins and David (1997) found more small-for-gestational-age and LBW deliveries among women living in high, compared with low crime rate neighborhoods. Similarly, using the violent crime rate in Chicago, Morenoff found violent crime to be a robust neighborhood predictor of LBW after controlling for individual covariates [<xref ref-type="bibr" rid="B18">18</xref>]. The research reported here confirms previous low birth weight studies and expands birth outcomes consideration to the etiologically significant outcome of preterm birth. This work builds on earlier findings, which focus largely on violent crime rates, by categorizing crime exposure in multiple ways. It was interesting to note that for white women, relative quantities of violent crime, or the violent crime rate, appeared associated with adverse birth outcomes whereas for black women, absolute violent crime (the violent crime count) appeared more influential.</p><p>Finding no effect on birth outcomes for living in close proximity to violent crime was unexpected. We anticipated living farther from violent crime would be protective against preterm birth while living within a half mile of a large number of violent crime episodes would increase a woman's risk of preterm birth. These expectations were not borne out. It is possible that the measurement error associated with geocoded crime events and maternal addresses is sufficiently large to preclude capturing salient distances or densities of violent crime or that systematic crime reporting errors may result in crime misclassification. Additionally, crime report data does not represent actual crime experiences and in the absence of daily diaries or other individual data collection techniques, it is impossible to assess actual crime exposure.</p><p>The mechanisms through which crime may influence health, and birth outcomes in particular, are uncertain. Exposure to violent environments has been associated with mental health in children [<xref ref-type="bibr" rid="B62">62</xref>] and sleep disturbances, nightmares and other anxiety manifestations among adults [<xref ref-type="bibr" rid="B63">63</xref>]. Another possible mechanism by which crime can influence health is through chronic stress. Research in this area is new, but some work has found violence exposure to be an important constituent of chronic environmental stress, suggested to play a role in developing essential hypertension through elevated sympathetic nervous system activity [<xref ref-type="bibr" rid="B45">45</xref>]. Other work finds evidence for an association between maternal stress or stressful life events and adverse pregnancy outcomes [<xref ref-type="bibr" rid="B64">64</xref>-<xref ref-type="bibr" rid="B68">68</xref>], including PTB [<xref ref-type="bibr" rid="B67">67</xref>-<xref ref-type="bibr" rid="B69">69</xref>], while other research does not [<xref ref-type="bibr" rid="B70">70</xref>,<xref ref-type="bibr" rid="B71">71</xref>]. Crime environments may influence health outcomes through behavior. Crime has been previously associated with health risk behaviors in young men [<xref ref-type="bibr" rid="B72">72</xref>] and women [<xref ref-type="bibr" rid="B73">73</xref>]. Health behaviors, especially those used to reduce stress such as cigarette smoking and alcohol consumption, can be particularly harmful during pregnancy.</p><p>This research is limited in several ways. The choice of control variables was limited to those reliably collected on the birth record. Other covariates, including tobacco and alcohol use, may partially explain the study results, but because of data quality issues were not explored in these analyses. Further, recognizing the presence of a crime event may differentially impact pregnancy health depending on perceptions of overall neighborhood quality, threat, pre-existing anxieties and a host of other psychosocial factors not assessed in this research. This study is further limited by its reliance on administratively defined boundaries to approximate the 'neighborhood'. The census block group is a relatively small unit of aggregation, but may bear no resemblance to the salient neighborhood-level exposure. Further, each woman's definition of neighborhood may differ, adding another level of complexity to determining relevant exposure types and levels. Despite the potential misattribution of "neighborhood" influence to an administrative unit, other authors have found using the census block group as the unit of analysis useful in studies of birth outcomes [<xref ref-type="bibr" rid="B17">17</xref>]. The study relied on birth certificate data for all individual-level data; the quality of birth record data is variable. Among North Carolina birth certificates, research indicates reporting is very accurate for birth weight and fair to good for most other variables, but poor for medical history and alcohol use [<xref ref-type="bibr" rid="B74">74</xref>]. The greatest concern with using birth record data in this research involves the construction of gestational age, from which one outcome measure is obtained. This limitation has been addressed through use of the clinical gestational age estimate, as discussed in the methods section.</p><p>One issue that frequently perplexes epidemiologic research is that of the modifiable areal unit problem [MAUP]. The MAUP arises from the imposition of artificial units of spatial reporting on continuous geographical phenomenon resulting in the generation of artificial spatial patterns [<xref ref-type="bibr" rid="B75">75</xref>]. The MAUP describes two effects that influence statistical and epidemiological results: scale and aggregation effects. The scale effect produces different statistical results by altering the denominator within the same dataset [<xref ref-type="bibr" rid="B76">76</xref>]. The aggregation or zoning effect arises from variability in the way units can be grouped at a given scale [<xref ref-type="bibr" rid="B76">76</xref>]. By assessing both census tract (data not shown) and census block group data, and finding similar effects at both levels of aggregation, this research can be considered robust to the scale effects of the MAUP. While this research did not address zoning effects directly, its use of different spatial categorizations of crime was an attempt to forward the field of area-level effects on health outcomes.</p><p>This research represents an important step in refining neighborhood exposures for health outcomes. The use of geocoded data allowed for multiple violent crime categorization forms. To our knowledge, this paper is the first to employ multiple categorizations of crime exposure and assess the relationship of each to a health outcome. Additionally, this paper considered two adverse birth outcomes, thereby providing a broader array of information. The study benefited from a large number of women, births and crime events; these numbers enabled the investigators to observe modest effects on relatively rare outcomes.</p></sec><sec><title>Conclusion</title><p>Living in close proximity to crime, estimated by the count of violent crimes within a half-mile radius of residence and the distance from residence to nearest violent crime, was not associated with adverse birth outcomes in this research. Area-level crime, whether measured as the count of violent crimes within a given block group or the block group rate of violent crime, was more useful for differentiating areas where women would and would not be at increased odds of an adverse birth outcome, even after adjusting for neighborhood-level deprivation. This is the first work to consider multiple categorizations of crime exposure in association with health outcomes. The results from this paper suggest that, at least for birth outcomes, crime appears best considered as part of the general neighborhood environment, rather than as a proximal exposure.</p><p>Preterm birth and LBW are important public health outcomes, but pregnancy is a resilient time for many women. Most pregnancies can endure a variety of 'insults' and still result in healthy, normal weight, term infants. Other health outcomes may prove more sensitive to the 'individual exposures' approach to crime characterization and this possibility should be considered in further research.</p></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>LCM conceived of and drafted the manuscript and conducted the study analyses. JSK co-conceived of the manuscript and participated in the study design and analysis plan. NDR participated in the study conception, design and coordination. AH contributed to the statistical analyses and modifications of the manuscript. BAL helped conceived of the study, and participated in its design and coordination. All authors read and approved the final manuscript.</p></sec>
Apolipoprotein E-specific innate immune response in astrocytes from targeted replacement mice	<sec><title>Background</title><p>Inheritance of the three different alleles of the human apolipoprotein (apo) E gene (<italic>APOE</italic>) are associated with varying risk or clinical outcome from a variety of neurologic diseases. ApoE isoform-specific modulation of several pathogenic processes, in addition to amyloid β metabolism in Alzheimer's disease, have been proposed: one of these is innate immune response by glia. Previously we have shown that primary microglia cultures from targeted replacement (TR) APOE mice have apoE isoform-dependent innate immune activation and paracrine damage to neurons that is greatest with TR by the ε4 allele (TR APOE4) and that derives from p38 mitogen-activated protein kinase (p38MAPK) activity.</p></sec><sec sec-type="methods"><title>Methods</title><p>Primary cultures of TR APOE2, TR APOE3 and TR APOE4 astrocytes were stimulated with lipopolysaccharide (LPS). ApoE secretion, cytokine production, and nuclear factor-kappa B (NF-κB) subunit activity were measured and compared.</p></sec><sec><title>Results</title><p>Here we showed that activation of primary astrocytes from TR APOE mice with LPS led to TR APOE-dependent differences in cytokine secretion that were greatest in TR APOE2 and that were associated with differences in NF-κB subunit activity.</p></sec><sec><title>Conclusion</title><p>Our results suggest that LPS activation of innate immune response in TR APOE glia results in opposing outcomes from microglia and astrocytes as a result of TR APOE-dependent activation of p38MAPK or NF-κB signaling in these two cell types.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>Maezawa</surname><given-names>Izumi</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Maeda</surname><given-names>Nobuyo</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A3" contrib-type="author"><name><surname>Montine</surname><given-names>Thomas J</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A4" corresp="yes" contrib-type="author"><name><surname>Montine</surname><given-names>Kathleen S</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	Journal of Neuroinflammation	<sec><title>Background</title><p>Humans are different from other mammals in that we have 3 common alleles of the apolipoprotein E gene (<italic>APOE</italic>): the ε2 (<italic>APOE2</italic>), ε3 (<italic>APOE3</italic>), and ε4 (<italic>APOE4</italic>) alleles [<xref ref-type="bibr" rid="B1">1</xref>]. Numerous genetic studies have associated inheritance of <italic>APOE4 </italic>with increased risk, earlier onset, or poorer clinical outcome for a number of neurodegenerative diseases, including Alzheimer's disease (AD), Parkinson's disease (PD), amyotrophic lateral sclerosis (ALS), traumatic brain injury, and HIV-encephalitis [<xref ref-type="bibr" rid="B2">2</xref>-<xref ref-type="bibr" rid="B10">10</xref>]. At least for AD, inheritance of <italic>APOE2 </italic>is associated with apparent neuroprotection, perhaps related to delayed onset of illness by many years [<xref ref-type="bibr" rid="B11">11</xref>]. While apoE isoforms play a role in the metabolism of beta amyloid (Aβ) peptides and thereby may modulate the risk of developing AD [<xref ref-type="bibr" rid="B12">12</xref>], the influence of inheriting different <italic>APOE </italic>alleles extends well beyond diseases thought to involve Aβ peptide-mediated neurotoxicity, as noted above. For this reason, other apoE-isoform specific mechanisms likely exist to explain the apparent influence of <italic>APOE </italic>alleles on such a broad spectrum of neurologic diseases; indeed, several have been proposed including synaptic stabilization, biologically active proteolytic fragments of apoE, anti-oxidant activity, and nitric oxide (NO) production [<xref ref-type="bibr" rid="B13">13</xref>-<xref ref-type="bibr" rid="B16">16</xref>]. ApoE also has an immune modulatory function, at least in the peripheral adaptive immune response to some bacteria and viruses [<xref ref-type="bibr" rid="B17">17</xref>]. We have recently shown that microglia from mice with targeted replacement (TR) of the mouse apoE gene with the coding sequences of human APOE alleles activated with LPS display an apoE isoform-specific innate immune response and result in apoE isoform-specific paracrine damage to neurons, both of which are dependent on p38 mitogen-activated protein kinase (p38MAPK) -mediated signaling.</p><p>One commonly used approach to investigate selectively innate immune response in neurodegeneration is to use a specific stimulus, lipopolysaccharide (LPS) [<xref ref-type="bibr" rid="B18">18</xref>-<xref ref-type="bibr" rid="B25">25</xref>]. LPS specifically activates CD14/Toll-like receptor (TLR) 4 co-receptors with subsequent increased gene transcription mediated through a bifurcated pathway that is dependent on both nuclear factor-kappa B (NF-κB) and p38MAPK signaling [<xref ref-type="bibr" rid="B26">26</xref>,<xref ref-type="bibr" rid="B27">27</xref>]. Indeed, LPS activation of CD14/TLR4 co-receptors on microglia leads to indirect damage to neurons and oligodendroglia in culture and <italic>in vivo </italic>[<xref ref-type="bibr" rid="B22">22</xref>,<xref ref-type="bibr" rid="B28">28</xref>-<xref ref-type="bibr" rid="B30">30</xref>]. Moreover, a role for CD14/TLR4 co-receptors is now understood to extend well beyond endotoxemia, as they are important in innate immune response to several endogenous ligands [<xref ref-type="bibr" rid="B31">31</xref>]. Indeed, CD14 binds Aβ fibrils and is responsible for most of Aβ–stimulated microglial-mediated neurotoxicity [<xref ref-type="bibr" rid="B32">32</xref>]. In addition, peptides and neoantigens expressed by apoptotic cells also activate this pathway [<xref ref-type="bibr" rid="B33">33</xref>]. Here we tested the hypothesis that innate immune response from CD14/TLR4 activation would show isoform-specific differences in primary cultures of astrocytes from TR APOE mice.</p></sec><sec sec-type="methods"><title>Methods</title><sec sec-type="materials"><title>Materials</title><p>Cell culture solutions and supplies were from GIBCO (Grand Island, NY). Poly-ornithine (0.01%) was from Sigma (St. Louis, MO). 4–15% SDS-polyacrylamide gels were from BioRad (Hercules, CA). LPS and the NO assay kit were from Calbiochem (La Jolla, CA). Primary antibodies used were polyclonal anti-human apoE antibody from Dako Corporation (Carpinteria, CA) and polyclonal anti-glial fibrillary astrocytic protein (GFAP) antibody from Novus Biologicals (Littleton, CO). The NF-κB transcription factor assay kit and purified human HDL were from Chemicon International (Temecula, CA)</p></sec><sec><title>Mice</title><p>Homozygous APOE2, APOE3 and APOE4 targeted replacement (TR) mice 'humanized' at <italic>apoE </italic>were developed by Dr. Maeda and colleagues [<xref ref-type="bibr" rid="B34">34</xref>,<xref ref-type="bibr" rid="B35">35</xref>]. Briefly, human <italic>APOE </italic>genomic fragments were used to replace mouse <italic>apoE </italic>via homologous recombination. All three lines of TR APOE mice contain chimeric genes consisting of mouse 5' regulatory sequences continuous with mouse exon 1 (noncoding) followed by human exons (and introns) 2–4 [<xref ref-type="bibr" rid="B34">34</xref>]. These mice were backcrossed greater than six generations to C67BL/6 genetic background. Mice were housed in an ALAC-approved vivarium and methods approved by a University of Washington International Use and Care of Animals (IACUC) Committee.</p></sec><sec><title>Astrocyte cultures</title><p>Primary cultures of 1-day-old mouse cerebral cortical astrocytes were prepared according to the method of Gebicke-Haerter, <italic>et al</italic>. [<xref ref-type="bibr" rid="B36">36</xref>]. Confluent cultures were used on the 7<sup>th </sup>day <italic>in vitro </italic>(DIV). Our preparations were ≥ 93% pure for astrocytes, as demonstrated by glial fibrillary acidic protein (GFAP) antibody. Astrocytes were exposed to LPS in serum-free medium at a final concentration of 100 ng/ml (20 ng/10<sup>5 </sup>cells). Vehicle control for LPS exposure was PBS.</p></sec><sec><title>Western blot analysis</title><p>Conditioned (serum-free) medium was removed from astrocyte cultures following LPS or vehicle exposure and centrifuged at 13,000 × <italic>g </italic>for 2 min at 4°C to remove cell debris. Equal volumes of conditioned media were diluted with 6X sample buffer (0.35 M Tris, 30% glycerol, 10% SDS, 0.93 g DTT, 1.2 mg bromophenol blue), heated to 95°C for 5 min, subjected to SDS PAGE, transferred to PVDF membranes, and analyzed and quantified as previously described [<xref ref-type="bibr" rid="B37">37</xref>]. Anti-human apoE (Dako) was used at 1:2000 dilution. Secondary antibody was HRP-conjugated anti-rabbit (1:3000).</p></sec><sec><title>NO detection</title><p>NO levels in conditioned media following incubation with LPS or vehicle were measured using a colorimetric NO assay kit (Calbiochem) where nitrate is first converted to nitrite by the NADH-dependent nitrate reductase, followed by nitrite measurement using the Griess Reagent.</p></sec><sec><title>Cytokine measurements</title><p>Conditioned medium following incubation with LPS or vehicle was screened for cytokines with an array method, and selected cytokines further quantified individually by sandwich ELISAs. The bead-based Liquichip™ Mouse 10-Cytokine Kit (Qiagen Inc, Valencia CA) was used to simultaneously screen conditioned media for the following cytokines: GM-CSF, interferon (INF)-γ, interleukin (IL)-1β, -2, -4, -5, -6, -10, -12, and tumor necrosis factor (TNF) –α. This kit uses cytokine antibodies immobilized on LiquiChip™ beads with distinct bead codes, which are added to conditioned media samples. Bead-bound cytokines are detected using a mixture of biotinylated cytokine-specific monoclonal antibodies and Streptavidin-PE. The specific bead code assigned to each of the 10 cytokines enables their unambiguous identification and quantification by a Luminex 100 X-Map reader using Qiagen software. Next, IL-6, IL-1β, and TNF-α from conditioned media were separately quantified by sandwich ELISAs using DuoSet ELISA development kits for each cytokine (R&D Systems, Minneapolis, MN).</p></sec><sec><title>NF-κB activity</title><p>NF-κB activity following incubation with LPS or vehicle was measured using an NF-κB transcription factor assay kit from Chemicon International. Briefly, cells were rinsed with PBS, lysed in Buffer A (10 mM HEPES (pH7.9), 1.5 mM MgCl<sub>2</sub>, 10 mM KCl, 0.5 mM DTT, 0.1% Triton X-100 and protease inhibitor cocktail), and a nuclear extract prepared in Buffer B (20 mM HEPES (pH 7.9), 1.5 mM MgCl<sub>2</sub>, 0.42 M NaCl, 0.2 mM EDTA. 0.5 mM DTT, 1.0% Igepal CA-630, 25% (v/v) glycerol, and protease inhibitor cocktail). Double-stranded biotinylated oligonucleotide containing the flanked consensus sequence for NF-κB was mixed with the nuclear extract and the mixture immobilized on a streptavidin-coated chemiluminescent plate, followed by immunologic detection of the bound NF-κB transcription factor subunits p50 and 065.</p></sec></sec><sec><title>Results</title><p>We have recently reported that LPS activation of TR APOE glial-wt neuron mixed cultures for 24 hours results in apoE isoform-specific paracrine damage to neurons [<xref ref-type="bibr" rid="B30">30</xref>]. For activated microglia, TR APOE4 is more neurotoxic than TR APOE2 or APOE3. For activated astrocytes, which produce much less neurotoxicity than microglia, both TR APOE4 and TR APOE3 are mildly damaging to neurons, while TR APO2 shows no neurotoxic effect. In this previous work, we pursued apoE-isoform specific mechanisms in LPS-activated microglia and showed that these were p38MAPK-dependent. Here, we pursued the basis of apoE isoforms-specific differences in LPS activation of astrocytes from these TR mice.</p><p>We first showed that there was no difference among the three TR APOE astrocytes in the amount of secreted apoE following LPS exposure for up to 24 hours (Figure <xref ref-type="fig" rid="F1">1</xref>), in agreement with our findings for microglia [<xref ref-type="bibr" rid="B30">30</xref>]. We also determined that similar to microglia, there was no difference in medium nitrate plus nitrite levels (a measure of NO secretion) compared to wild type (wt) at 12 or 24 hours after LPS exposure (P > 0.05), although we did observe increased medium nitrate plus nitrite levels in TR APOE4 (205 ± 41 % of wt) but not TR APOE2 (116 + 15% of wt) astrocytes 72 hours after LPS incubation. As with microglia, this temporal mismatch suggests increased NO secretion by TR APOE4 lies distal to the processes underlying the TR-APOE isoform-specific differences in astrocyte-mediated neurotoxicity seen within 24 hours of LPS incubation [<xref ref-type="bibr" rid="B30">30</xref>].</p><fig position="float" id="F1"><label>Figure 1</label><caption><p><bold>ApoE secretion following LPS stimulation. </bold>Mouse cerebral primary astrocyte cultures were incubated in serum-free medium with 100 ng/ml LPS or vehicle (PBS) for 24 hours. 20 μl of conditioned medium was collected at 3, 6, 12, and 24 hrs after exposure, spun briefly, mixed with 4 μl of 6X sample buffer (0.35 M Tris, 30% glycerol, 10% SDS, 0.93 g DTT, and 1.2 mg bromophenol blue), and relative concentrations of apoE determined by Western blotting. Human high-density lipoprotein (hHDL) prepared in the same sample buffer was included as a positive control.</p></caption><graphic xlink:href="1742-2094-3-10-1"/></fig><p>Previously, we observed TR APOE-dependent differences in cytokine secretion by microglia in response to LPS exposure [<xref ref-type="bibr" rid="B30">30</xref>]. Here we measured cytokine secretion in response to LPS in the three TR APOE astrocyte cultures. We screened for changes in medium cytokine concentrations using the LiquiChip™ Mouse 10-Cytokine assay and a Luminex 100 X-Map reader that simultaneously determines 10 mouse cytokines in medium from TR APOE astrocytes. The cytokines quantified were GM-CSF, INF-γ, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12, and TNF-α. Only IL-6 and TNF-α changed significantly following LPS exposure for 12 hr; IL-1β was near the limit of detection for this assay. The magnitude of induction for these cytokines was TR APOE-dependent with IL-6 and TNF-α concentrations following the gradient of TR APOE2 > TR APOE3 > TR APOE4. We confirmed our IL-6 and TNF-α findings with individual ELISAs and extended our analysis to IL-1β, since many others have shown it to be overexpressed and secreted from LPS-stimulated glia; TR APOE-dependence of IL-1β secretion followed the same pattern as the other two cytokines (Figure <xref ref-type="fig" rid="F2">2</xref>).</p><fig position="float" id="F2"><label>Figure 2</label><caption><p><bold>Cytokine secretion following LPS stimulation. </bold>Mouse cerebral primary astrocyte cultures were incubated in serum-free medium with 100 ng/ml LPS or PBS for 12 hr, medium collected, and IL-1β, IL-6 and TNF-α concentrations determined by ELISA. All cytokines were below the limit of detection in PBS-exposed cultures. Data are mean ± SEM (n = 4 to 8 separate cultures per group). One-way ANOVA showed P < 0.05 for all three cytokines. P < 0.05, <sup>^</sup>P < 0.01, or <sup>#</sup>P < 0.001 for Bonferroni-corrected posttests for LPS-incubated TR APOE3 or TR APOE4 <italic>vs</italic>. TR APOE2; <sup>+</sup>P < 0.01 for TR APOE4 <italic>vs</italic>. TR APOE3.</p></caption><graphic xlink:href="1742-2094-3-10-2"/></fig><p>LPS activation of CD14/TLR4 co-receptors leads to subsequent increased gene transcription mediated through a bifurcated pathway that is dependent on NF-κB and p38MAPK signaling. We have previously demonstrated apoE isoform-specific p38MAPK activation following LPS exposure of microglia but not astrocytes [<xref ref-type="bibr" rid="B30">30</xref>]. We therefore determined the activity of two NF-κB subunits, p50 and p65, in astrocytes from TR APOE mice (Figure <xref ref-type="fig" rid="F3">3</xref>). Following LPS exposure, both p50 and p65 activity significantly increased in all 3 genotypes. p50 activity showed an apoE isoform-specific increase, with a larger increase in TR APOE2 than the other two (P < 0.01 for both) and no difference between TR APOE3 and TR APOE4. p65 showed a similar trend in apoE isoform-specific effect; however, this was not significantly different in corrected multiple comparison tests.</p><fig position="float" id="F3"><label>Figure 3</label><caption><p><bold>NF-κB activity following LPS stimulation</bold>. NF-κB p50 and p65 subunit activity was determined in nuclear extracts from mouse cerebral primary astrocyte cultures exposed to vehicle (PBS) or LPS for 12 hr using an NF-κB transcription factor assay kit (Chemicon). Data are expressed as average relative light units (RLU) ± SEM (n = 4 for each group). Two-way ANOVA for p50 data had P < 0.0001 for TR APOE and vehicle <italic>vs</italic>. LPS, but P > 0.05 for interaction between these terms. Two-way ANOVA for p60 data had P < 0.05 for TR APOE and P < 0.0001 for vehicle <italic>vs</italic>. LPS, but P > 0.05 for interaction between these terms. P < 0.01 for Bonferroni-corrected posttests for LPS-exposed TR APOE3 or TR APOE4 <italic>vs</italic>. TR APOE2.</p></caption><graphic xlink:href="1742-2094-3-10-3"/></fig></sec><sec><title>Discussion</title><p>Inheritance of <italic>APOE </italic>alleles is associated with varying clinical outcomes in several neurodegenerative diseases, including AD, PD, ALS, head trauma, multiple sclerosis, and HIV-encephalitis [<xref ref-type="bibr" rid="B2">2</xref>-<xref ref-type="bibr" rid="B10">10</xref>]. Although apoE isoforms likely modulate AD pathogenesis by influencing metabolism of Aβ [<xref ref-type="bibr" rid="B12">12</xref>,<xref ref-type="bibr" rid="B38">38</xref>], the pathophysiologic significance of apoE isoforms appears to go beyond interacting with Aβ since these other diseases of brain are not thought to involve Aβ peptides in their pathogenesis. Indeed, others have suggested more general mechanisms of neurotrophism or neurotoxicity from inheritance of different <italic>APOE </italic>alleles that potentially could contribute to multiple neurologic diseases [<xref ref-type="bibr" rid="B13">13</xref>-<xref ref-type="bibr" rid="B16">16</xref>]. Since activation of innate immunity also is associated with these same diseases, we tested the hypothesis that apoE isoforms may act by modulating glial innate immune response and thereby altering neurotoxicity. Previously, we showed that microglia from TR APOE mice show apoE isoform-specific innate immune activation and paracrine damage to neurons that was greatest with TR APOE4 and dependent on p38MAPK signaling [<xref ref-type="bibr" rid="B30">30</xref>]. Here, we showed that identical activation of astrocytes from these same TR APOE mice had apoE isoforms-specific innate immune response that was greatest with TR APOE2 astrocytes and associated with NF-kB-mediated signaling.</p><p>We used a model of selective activation of CD14/TLR4 co-receptors that is now appreciated to initiate innate immune response to endogenous ligands relevant to neurodegenerative diseases such as Aβ fibrils as well as peptides and neoantigens expressed by apoptotic cells [<xref ref-type="bibr" rid="B32">32</xref>,<xref ref-type="bibr" rid="B33">33</xref>]. LPS activation of CD14/TLR4 co-receptors leads to increased gene transcription through a bifurcated pathway; one arm is NF-κB-dependent and the other is p38MAPK-dependent [<xref ref-type="bibr" rid="B26">26</xref>,<xref ref-type="bibr" rid="B27">27</xref>]. Our data indicated that the intracellular signaling that mediates altered gene transcription in response to LPS is different between astrocytes and microglia expressing TR APOE. Specifically, NF-κB-mediated signaling, which is associated with immune modulation and protection of cells from undergoing apoptosis [<xref ref-type="bibr" rid="B39">39</xref>], was greatest in TR APOE2 astrocytes, the only cell line that did not yield paracrine damage to neurons following activation with LPS [<xref ref-type="bibr" rid="B30">30</xref>]. In contrast, apoE isoforms-specific effects in microglia, including much more extensive paracrine damage to neurons, was associated with p38MAPK signaling [<xref ref-type="bibr" rid="B30">30</xref>]. We speculate that the inverse relationship between low-level neurotoxicity associated with LPS-activated astrocytes that we reported previously [<xref ref-type="bibr" rid="B30">30</xref>] and innate immune activation may be related to diminished NF-κB-dependent trophic factors in TR APOE3 and TR APOE4 astrocytes.</p></sec><sec><title>Conclusion</title><p>Our results suggest that LPS activation of innate immune response in TR APOE glia results in opposing outcomes from microglia and astrocytes as a result of TR APOE-dependent activation of p38MAPK or NF-κB signaling in these two cell types.</p></sec><sec><title>Abbreviations</title><p>AD (Alzheimer's disease); ALS (amyotrophic lateral sclerosis); apo (apolipoprotein); <italic>APOE </italic>(human apoE gene); Aβ (beta amyloid); DIV (days <italic>in vitro</italic>); GFAP (glial fibrillary astrocytic protein); hHDL (human high density lipoprotein); IL (interleukin); INF (interferon); LPS (lipopolysaccharide); NF-κB (nuclear factor kappa B); NO (nitric oxide); p38MAPK (p38 mitogen-activated protein kinase); PD (Parkinson's disease); RLU (relative light units); TLR (Toll-like receptor); TNF (tumor necrosis factor); TR (targeted replacement); wt (wild type).</p></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>IM carried out the experiments described. NM developed the mouse line that was used in all experiments. TJM conceived the study and its design and helped to draft the manuscript. KSM assisted in experimental design, analyzed the data, and drafted the manuscript.</p></sec>
Antibiotic resistance as a global threat: Evidence from China, Kuwait and the United States	<sec><title>Background</title><p>Antimicrobial resistance is an under-appreciated threat to public health in nations around the globe. With globalization booming, it is important to understand international patterns of resistance. If countries already experience similar patterns of resistance, it may be too late to worry about international spread. If large countries or groups of countries that are likely to leap ahead in their integration with the rest of the world – China being the standout case – have high and distinctive patterns of resistance, then a coordinated response could substantially help to control the spread of resistance. The literature to date provides only limited evidence on these issues.</p></sec><sec sec-type="methods"><title>Methods</title><p>We study the recent patterns of antibiotic resistance in three geographically separated, and culturally and economically distinct countries – China, Kuwait and the United States – to gauge the range and depth of this global health threat, and its potential for growth as globalization expands. Our primary measures are the prevalence of resistance of specific bacteria to specific antibiotics. We also propose and illustrate methods for aggregating specific "bug-drug" data. We use these aggregate measures to summarize the resistance pattern for each country and to study the extent of correlation between countries' patterns of drug resistance.</p></sec><sec><title>Results</title><p>We find that China has the highest level of antibiotic resistance, followed by Kuwait and the U.S. In a study of resistance patterns of several most common bacteria in China in 1999 and 2001, the mean prevalence of resistance among hospital-acquired infections was as high as 41% (with a range from 23% to 77%) and that among community- acquired infections was 26% (with a range from 15% to 39%). China also has the most rapid growth rate of resistance (22% average growth in a study spanning 1994 to 2000). Kuwait is second (17% average growth in a period from 1999 to 2003), and the U.S. the lowest (6% from 1999 to 2002). Patterns of resistance across the three countries are not highly correlated; the most correlated were China and Kuwait, followed by Kuwait and the U.S., and the least correlated pair was China and the U.S.</p></sec><sec><title>Conclusion</title><p>Antimicrobial resistance is a serious and growing problem in all three countries. To date, there is not strong international convergence in the countries' resistance patterns. This finding may change with the greater international travel that will accompany globalization. Future research on the determinants of drug resistance patterns, and their international convergence or divergence, should be a priority.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>Zhang</surname><given-names>Ruifang</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" corresp="yes" contrib-type="author"><name><surname>Eggleston</surname><given-names>Karen</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A3" contrib-type="author"><name><surname>Rotimi</surname><given-names>Vincent</given-names></name><xref ref-type="aff" rid="I3">3</xref><email>[email protected]</email></contrib><contrib id="A4" contrib-type="author"><name><surname>Zeckhauser</surname><given-names>Richard J</given-names></name><xref ref-type="aff" rid="I4">4</xref><email>[email protected]</email></contrib>	Globalization and Health	<sec><title/><p>In 1942, the first U.S. patient with streptococcal infection was miraculously cured with a small dose of penicillin. Sixty years later, penicillin-resistant Streptococcus is widespread. Such antimicrobial resistance threatens the health of many throughout the world, since both old and new infectious diseases remain a formidable public health threat.</p><p>Among the issues that merit further scrutiny for understanding the possible spread of antimicrobial resistance, few are as salient as the impact of globalization. Clearly the movement of people and goods around the globe contributes to transmission of disease [<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B2">2</xref>]. To what extent drug resistance and globalization are similarly related remains unclear. The breakout of Severe Acute Respiratory Syndrome (SARS) in the spring of 2003 illustrates how an infectious disease with limited therapeutic options can spread rapidly across national borders. With globalization booming, it is important to understand international patterns of resistance. If countries already experience similar patterns of resistance, it may be too late to worry about international spread. If large countries or groups of countries that are likely to leap ahead in their integration with the rest of the world – China being the standout case – have high and distinctive patterns of resistance, then a coordinated response could help substantially to control the spread of resistance. The literature to date provides only limited evidence on these issues.</p><p>We study the pattern of antibiotic resistance in specific countries to gauge the range and depth of this global health threat. China and the U.S. stand out as good choices for study. Both are world economic powerhouses increasingly responding to the forces of economic globalization. In addition, both are major consumers of antibiotics, with the U.S. also being a leading source of new antibiotics. On the other hand, it would also be interesting to compare patterns of antibiotic resistance in smaller countries that stand relatively distant from these two. Accordingly, we compare the experiences of the U.S. and China with new data on the resistance experience of Kuwait.</p><p>The first section gives brief background on antibiotic resistance and its costs. We then turn to a detailed comparison of surveillance data from China, Kuwait, and the U.S. We conclude with a plea for more research and attention on this critical issue for health and globalization.</p></sec><sec><title>Background: The challenge of antimicrobial resistance</title><p>According to laws of Darwinian evolution, antimicrobial use creates a selection pressure on microorganisms: weak ones are killed, but stronger ones might adapt and survive. When pathogenic microorganisms can multiply beyond some critical mass in the face of invading antimicrobials, treatment outcome is compromised; this phenomenon is referred as antimicrobial resistance (AMR) [<xref ref-type="bibr" rid="B3">3</xref>-<xref ref-type="bibr" rid="B9">9</xref>]. This paper focuses on antibiotic resistance, a major form of AMR.</p><p>Resistance mechanisms may develop over months or years [<xref ref-type="bibr" rid="B6">6</xref>]. Once established, a single resistance mechanism can often allow a bacterium to resist multiple drugs. It remains unclear whether resistance is reversible, and thus whether drug effectiveness is a renewable or non-renewable resource [<xref ref-type="bibr" rid="B10">10</xref>-<xref ref-type="bibr" rid="B15">15</xref>]. Drug resistance raises the cost of treatment for infectious diseases, sometimes manifold, as well as increasing morbidity and mortality from such diseases [<xref ref-type="bibr" rid="B16">16</xref>-<xref ref-type="bibr" rid="B23">23</xref>].</p><p>The greatest long-term threat of AMR is that resistant strains erode drug efficacy over time. The development of drug-resistant <italic>Staphylococci aureus </italic>(SAU) well illustrates the see-saw battle between pathogens and drugs. SAU is a bacterium that harmlessly lives in the human body but can cause infections on wounds or lesions. After the clinical application of penicillin in the 1940s, SAU soon adapted to the treatment mechanism of penicillin, and by the 1950s, almost half of SAU strains had become resistant to penicillin. A new antibiotic, methicillin, was developed in the 1960s. Yet by the late 1970s, methicillin-resistant SAU, i.e. MRSA, again became widespread. Today MRSA has become a major infectious culprit that can only be effectively treated with vancomycin, one of the few last killers of superbugs. Unfortunately, in 1996, a Japanese hospital reported the first case of vancomycin-resistant SAU (VRSA) during surgery on a four-month-old boy. The U.S., France and Hong Kong subsequently all reported VRSA incidents. A few years later in 2000, linezolid was launched as a new antibiotic to combat both MRSA and VRSA. But only one year later, Boston researchers reported the first case of linezolid-resistant MRSA in an 85-year-old man undergoing peritoneal dialysis. After failing to contain his MRSA by linezolid, researchers tried five antibiotics (ampicillin, azithromycin, gentamicin, levofloxacin, and quinupristin-dalfopristin) but the unlucky man eventually died from the uncontrollable infection [<xref ref-type="bibr" rid="B24">24</xref>].</p><p>Resistant pathogens within a hospital or specific community can spread to a nation at large or across national boundaries. Thus, for example, rapidly increasing travel and migration within China probably contributes to the growth of that nation's resistance problem. It may also spur the spread of China's resistance problems overseas as globalization greatly increases travel from and to that nation (see Figure <xref ref-type="fig" rid="F1">1</xref>).</p><fig position="float" id="F1"><label>Figure 1</label><caption><p>Travel to and from China has increased tremendously over the past decade.</p></caption><graphic xlink:href="1744-8603-2-6-1"/></fig></sec><sec sec-type="methods"><title>Methods</title><p>We collected data on drug resistance in China, the U.S. and Kuwait, drawing from published studies, reports from national surveillance systems, and previously unpublished data from a large hospital in Kuwait. Such data must be viewed with caution. Differences between countries arise not only from genuine differences in prevalence, but also from differences in sampling strategies, laboratory processing, and standards for defining a "resistant" strain. Moreover, within-country comparisons across time are biased by measurement error, particularly for small samples. However, analysis of the currently available data does yield some evidence and may help to raise awareness and efforts to improve the data and methods for addressing the problem.</p><p>Our primary measure is the prevalence of resistance by a specific bacterium to a specific drug. The prevalence is calculated as the number of resistant isolates divided by the number of total isolates collected, multiplied by 100. We compute growth rates of resistance to specific bacteria using standard year-on-year growth calculations. Where appropriate, we smooth variance in small-sample data series by using three-year running averages.</p><p>We also develop methods to aggregate specific "bug-drug" data to summarize the resistance pattern for each country. These measures weight resistance rates by (1) the isolation frequency for each bacterium (that is, the proportion of a particular bacterium among all bacteria studied); and, where possible, by (2) the proportion of resistant cases hospital- versus community-acquired; and (3) the frequency with which each drug is used to treat infections caused by each bacterium. (For most calculations, measure (3) is not available.) Finally, we compare and contrast each country's resistance experience and, using the subset of data comparable across the three countries, examine correlations in patterns of resistance.</p><p>These methods represent preliminary steps to gauge whether patterns of antibiotic resistance converge over time amongst countries that currently have little population interchange. Future research would benefit from better surveillance of resistance, more comparable data reporting, data on antibiotic utilization, and further methodological advances in clinically- and policy-relevant aggregation of "bug-drug" data.</p></sec><sec><title>Results</title><sec><title>China</title><p>In 1988, the World Health Organization West Pacific Regional Office set up two antimicrobial resistance surveillance centers in Beijing and Shanghai. Meanwhile, China's Ministry of Health also established the China Nosocomial Infection Surveillance (CNIS) program, which monitors hospital-acquired infections. Unfortunately, most of the surveillance programs in China focus on urban hospitals. We lack data on urban communities and for the rural majority. Nevertheless, the available data allows us to piece together a picture of the extent of antimicrobial resistance in the most populous country in the world.</p><p>To examine AMR development in China, we use annual data from a seven-year (1994–2000) study by China's National Center for Antimicrobial Resistance, which reports resistance levels of ten most prevalent bacteria to a common antibiotic, ciprofloxacin (Table <xref ref-type="table" rid="T1">1</xref>) [<xref ref-type="bibr" rid="B25">25</xref>]. With small sample sizes, the annual measured percentage of isolates found to be resistant varies considerably; to smooth the random variation attributable to small sample size, we use three-year running averages. Some bacteria such as ECO and MRSA have high proportions (60–80%) of resistant strains, whereas the prevalence of resistant strains for others such as PMI is quite low. Almost all but MSSA and PMI have shown considerable growth in resistance over the study period, resulting in an average annual growth rate of about 15%.</p><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>Resistance prevalence of ten common bacteria to Ciprofloxacin in China, 1994–2000</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="right" colspan="12"><bold>unit: %</bold></td></tr></thead><tbody><tr><td align="center">Rank</td><td align="center">Bacter.</td><td></td><td align="center">1994</td><td align="center">1995</td><td align="center">1996</td><td align="center">1997</td><td align="center">1998</td><td align="center">1999</td><td align="center">2000</td><td align="center"><bold><italic>Average Resistance</italic></bold></td><td align="center"><bold><italic>Average Growth Rate</italic></bold></td></tr><tr><td colspan="12"><hr></hr></td></tr><tr><td align="center">1</td><td align="center">Escherichia coli (ECO)</td><td></td><td align="center">53</td><td align="center">49</td><td align="center">60</td><td align="center">61</td><td align="center">60</td><td align="center">63</td><td align="center">62</td><td align="center"><bold><italic>59</italic></bold></td><td align="center"><bold><italic>3</italic></bold></td></tr><tr><td align="center">2</td><td align="center">Pseudomonas aeruginosa (PAE)</td><td></td><td align="center">9</td><td align="center">10</td><td align="center">7</td><td align="center">18</td><td align="center">13</td><td align="center">17</td><td align="center">18</td><td align="center"><bold><italic>13</italic></bold></td><td align="center"><bold><italic>17</italic></bold></td></tr><tr><td align="center">3</td><td align="center">Klebsiella pneumoniae (KPN)</td><td></td><td align="center">2</td><td align="center">4</td><td align="center">7</td><td align="center">8</td><td align="center">14</td><td align="center">17</td><td align="center">18</td><td align="center"><bold><italic>10</italic></bold></td><td align="center"><bold><italic>40</italic></bold></td></tr><tr><td align="center">4</td><td align="center">Staphylococci epidermidis (SEP)</td><td></td><td align="center">22</td><td align="center">33</td><td align="center">34</td><td align="center">35</td><td align="center">41</td><td align="center">40</td><td align="center">46</td><td align="center"><bold><italic>36</italic></bold></td><td align="center"><bold><italic>9</italic></bold></td></tr><tr><td align="center">5</td><td align="center">Staphylococci aureus (SAU)</td><td align="center">MRSA</td><td align="center">47</td><td align="center">65</td><td align="center">74</td><td align="center">88</td><td align="center">83</td><td align="center">78</td><td align="center">76</td><td align="center"><bold><italic>76</italic></bold></td><td align="center"><bold><italic>7</italic></bold></td></tr><tr><td></td><td></td><td align="center">MSSA</td><td align="center">8</td><td align="center">18</td><td align="center">10</td><td align="center">5</td><td align="center">8</td><td align="center">20</td><td align="center">14</td><td align="center"><bold><italic>11</italic></bold></td><td align="center"><bold><italic>8</italic></bold></td></tr><tr><td align="center">6</td><td align="center">Enterococcus faecalis (EFA)</td><td></td><td align="center">25</td><td align="center">34</td><td align="center">28</td><td align="center">34</td><td align="center">32</td><td align="center">45</td><td align="center">45</td><td align="center"><bold><italic>34</italic></bold></td><td align="center"><bold><italic>9</italic></bold></td></tr><tr><td align="center">7</td><td align="center">Enterobacter cloacae (ECL)</td><td></td><td align="center">12</td><td align="center">9</td><td align="center">13</td><td align="center">14</td><td align="center">22</td><td align="center">31</td><td align="center">30</td><td align="center"><bold><italic>18</italic></bold></td><td align="center"><bold><italic>26</italic></bold></td></tr><tr><td align="center">8</td><td align="center">Acinetobacter baumannii (ABA)</td><td></td><td align="center">7</td><td align="center">7</td><td align="center">19</td><td align="center">20</td><td align="center">23</td><td align="center">31</td><td align="center">37</td><td align="center"><bold><italic>20</italic></bold></td><td align="center"><bold><italic>29</italic></bold></td></tr><tr><td align="center">9</td><td align="center">Citrobacter freundii (CFR)</td><td></td><td align="center">10</td><td align="center">21</td><td align="center">20</td><td align="center">17</td><td align="center">22</td><td align="center">26</td><td align="center">26</td><td align="center"><bold><italic>20</italic></bold></td><td align="center"><bold><italic>10</italic></bold></td></tr><tr><td align="center">10</td><td align="center">Proteus mirabilis (PMI)</td><td></td><td align="center">8</td><td align="center">2</td><td align="center">13</td><td align="center">2</td><td align="center">5</td><td align="center">14</td><td align="center">12</td><td align="center"><bold><italic>7</italic></bold></td><td align="center"><bold><italic>10</italic></bold></td></tr><tr><td></td><td></td><td align="center"><bold><italic>Mean</italic></bold></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td align="center"><bold><italic>28</italic></bold></td><td align="center"><bold><italic>15</italic></bold></td></tr><tr><td></td><td></td><td align="center"><bold><italic>Median</italic></bold></td><td></td><td></td><td></td><td></td><td></td><td></td><td></td><td align="center"><bold><italic>20</italic></bold></td><td align="center"><bold><italic>10</italic></bold></td></tr></tbody></table><table-wrap-foot><p>* Based on three-year running averages.</p><p>** Staphylococci aureus (SAU) is further grouped as methicillin susceptible staphylococci aureus (MSSA) and methicillin resistant staphylococci aureus (MRSA).</p></table-wrap-foot></table-wrap><p>Another series of studies by the China Bacterial Resistance Surveillance Study Group focused on resistance prevalence among different patient types, i.e. those with hospital-acquired infections (HAI) versus community-acquired infections (CAI) [<xref ref-type="bibr" rid="B26">26</xref>,<xref ref-type="bibr" rid="B27">27</xref>]. We construct two measures to compare HAI and CAI resistance prevalence. First, by aggregating the seven bacteria, we get a measure γ indexed on the nineteen drugs. γ is calculated by multiplying the resistance rate of each bacterium by its isolation frequency and proportion among HAI (or CAI) infections, and then summing across bacteria. The measure is reported in the last two columns of Table <xref ref-type="table" rid="T2">2</xref> and graphed in Figure <xref ref-type="fig" rid="F2">2</xref>. Second, by aggregating the drugs, we obtain a measure indexed on bacteria. However, because we lack data on how often each drug is used, the best we can do is report the simple average for all drugs (implicitly assuming each drug is used with equal frequency). We name this measure Mean Resistance, shown in the last row in Table <xref ref-type="table" rid="T2">2</xref> and graphed in Figure <xref ref-type="fig" rid="F3">3</xref>.</p><table-wrap position="float" id="T2"><label>Table 2</label><caption><p>Resistance patterns of the seven most common bacteria for Hospital-acquired Infections (HAI) and Community-acquired Infections (CAI), China 2001</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="right" colspan="17">unit: %</td></tr><tr><td></td><td colspan="16"><hr></hr></td></tr><tr><td align="center">Antibiotic(s)</td><td align="center" colspan="2">SAU (n = 176)</td><td align="center" colspan="2">SEP (n = 84)</td><td align="center" colspan="2">ECO (n = 308)</td><td align="center" colspan="2">ECL (n = 78)</td><td align="center" colspan="2">PAE (n = 232)</td><td align="center" colspan="2">KPN (n = 215)</td><td align="center" colspan="2">ABA (n = 191)</td><td align="center" colspan="2">γ</td></tr></thead><tbody><tr><td></td><td align="center">HAI (37)</td><td align="center">CAI (139)</td><td align="center">HAI (14)</td><td align="center">CAI (70)</td><td align="center">HAI (44)</td><td align="center">CAI (264)</td><td align="center">HAI (27)</td><td align="center">CAI (51)</td><td align="center">HAI (95)</td><td align="center">CAI (137)</td><td align="center">HAI (48)</td><td align="center">CAI (167)</td><td align="center">HAI (46)</td><td align="center">CAI (145)</td><td align="center"><bold><italic>HAI</italic></bold>γ<sub><italic>H</italic></sub></td><td align="center"><bold><italic>CAI</italic></bold>γ<sub><italic>C</italic></sub></td></tr><tr><td colspan="17"><hr></hr></td></tr><tr><td align="center">Methicillin</td><td align="center">89</td><td align="center">30</td><td align="center">43</td><td align="center">27</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">n/a</td><td align="center"><bold><italic>11</italic></bold></td><td align="center"><bold><italic>5</italic></bold></td></tr><tr><td colspan="17"><hr></hr></td></tr><tr><td align="center">Ampicillin</td><td align="center">100</td><td align="center">82</td><td align="center">86</td><td align="center">67</td><td align="center">89</td><td align="center">80</td><td align="center">100</td><td align="center">90</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">54</td><td align="center">66</td><td align="center">n/a</td><td align="center">n/a</td><td align="center"><bold><italic>38</italic></bold></td><td align="center"><bold><italic>35</italic></bold></td></tr><tr><td align="center">Amoxicillin</td><td align="center">89</td><td align="center">27</td><td align="center">29</td><td align="center">6</td><td align="center">84</td><td align="center">81</td><td align="center">100</td><td align="center">94</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">90</td><td align="center">95</td><td align="center">48</td><td align="center">50</td><td align="center"><bold><italic>38</italic></bold></td><td align="center"><bold><italic>31</italic></bold></td></tr><tr><td align="center">Ceftizoxime</td><td align="center">87</td><td align="center">28</td><td align="center">14</td><td align="center">7</td><td align="center">32</td><td align="center">25</td><td align="center">96</td><td align="center">86</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">33</td><td align="center">26</td><td align="center">96</td><td align="center">92</td><td align="center"><bold><italic>24</italic></bold></td><td align="center"><bold><italic>16</italic></bold></td></tr><tr><td align="center">Cefaclor</td><td align="center">87</td><td align="center">31</td><td align="center">21</td><td align="center">10</td><td align="center">32</td><td align="center">26</td><td align="center">89</td><td align="center">78</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">33</td><td align="center">25</td><td align="center">65</td><td align="center">57</td><td align="center"><bold><italic>23</italic></bold></td><td align="center"><bold><italic>15</italic></bold></td></tr><tr><td align="center">Cefuroxime</td><td align="center">89</td><td align="center">29</td><td align="center">22</td><td align="center">4</td><td align="center">32</td><td align="center">25</td><td align="center">74</td><td align="center">47</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">29</td><td align="center">23</td><td align="center">57</td><td align="center">41</td><td align="center"><bold><italic>22</italic></bold></td><td align="center"><bold><italic>12</italic></bold></td></tr><tr><td align="center">Cefprozil.</td><td align="center">87</td><td align="center">26</td><td align="center">21</td><td align="center">4</td><td align="center">34</td><td align="center">25</td><td align="center">78</td><td align="center">61</td><td align="center">n/a</td><td align="center">n/a</td><td align="center">33</td><td align="center">23</td><td align="center">94</td><td align="center">86</td><td align="center"><bold><italic>24</italic></bold></td><td align="center"><bold><italic>15</italic></bold></td></tr><tr><td align="center">Ceftazidime</td><td align="center">92</td><td align="center">37</td><td align="center">50</td><td align="center">13</td><td align="center">5</td><td align="center">7</td><td align="center">59</td><td align="center">28</td><td align="center">11</td><td align="center">14</td><td align="center">21</td><td align="center">4</td><td align="center">30</td><td align="center">15</td><td align="center"><bold><italic>19</italic></bold></td><td align="center"><bold><italic>8</italic></bold></td></tr><tr><td align="center">Cefotaxime</td><td align="center">84</td><td align="center">28</td><td align="center">21</td><td align="center">6</td><td align="center">0</td><td align="center">7</td><td align="center">44</td><td align="center">26</td><td align="center">41</td><td align="center">26</td><td align="center">4</td><td align="center">5</td><td align="center">28</td><td align="center">16</td><td align="center"><bold><italic>15</italic></bold></td><td align="center"><bold><italic>8</italic></bold></td></tr><tr><td align="center">Ceftriaxone</td><td align="center">89</td><td align="center">28</td><td align="center">21</td><td align="center">3</td><td align="center">9</td><td align="center">8</td><td align="center">48</td><td align="center">29</td><td align="center">40</td><td align="center">25</td><td align="center">6</td><td align="center">5</td><td align="center">33</td><td align="center">15</td><td align="center"><bold><italic>18</italic></bold></td><td align="center"><bold><italic>8</italic></bold></td></tr><tr><td align="center">Imipenem</td><td align="center">76</td><td align="center">21</td><td align="center">21</td><td align="center">1</td><td align="center">2</td><td align="center">0</td><td align="center">0</td><td align="center">2</td><td align="center">2</td><td align="center">3</td><td align="center">0</td><td align="center">1</td><td align="center">2</td><td align="center">1</td><td align="center"><bold><italic>8</italic></bold></td><td align="center"><bold><italic>2</italic></bold></td></tr><tr><td align="center">Meropenem</td><td align="center">78</td><td align="center">21</td><td align="center">14</td><td align="center">1</td><td align="center">2</td><td align="center">0</td><td align="center">0</td><td align="center">0</td><td align="center">2</td><td align="center">2</td><td align="center">0</td><td align="center">1</td><td align="center">2</td><td align="center">2</td><td align="center"><bold><italic>8</italic></bold></td><td align="center"><bold><italic>2</italic></bold></td></tr><tr><td align="center">Ciprofloxacin</td><td align="center">87</td><td align="center">35</td><td align="center">36</td><td align="center">30</td><td align="center">75</td><td align="center">53</td><td align="center">63</td><td align="center">33</td><td align="center">26</td><td align="center">13</td><td align="center">19</td><td align="center">14</td><td align="center">26</td><td align="center">17</td><td align="center"><bold><italic>29</italic></bold></td><td align="center"><bold><italic>18</italic></bold></td></tr><tr><td align="center">Ofloxacin</td><td align="center">78</td><td align="center">30</td><td align="center">36</td><td align="center">30</td><td align="center">75</td><td align="center">55</td><td align="center">59</td><td align="center">31</td><td align="center">17</td><td align="center">15</td><td align="center">15</td><td align="center">14</td><td align="center">22</td><td align="center">17</td><td align="center"><bold><italic>27</italic></bold></td><td align="center"><bold><italic>18</italic></bold></td></tr><tr><td align="center">Levofloxacin</td><td align="center">46</td><td align="center">7</td><td align="center">29</td><td align="center">10</td><td align="center">68</td><td align="center">52</td><td align="center">33</td><td align="center">20</td><td align="center">22</td><td align="center">15</td><td align="center">10</td><td align="center">11</td><td align="center">13</td><td align="center">12</td><td align="center"><bold><italic>21</italic></bold></td><td align="center"><bold><italic>13</italic></bold></td></tr><tr><td align="center">Sparfloxacin</td><td align="center">89</td><td align="center">39</td><td align="center">50</td><td align="center">40</td><td align="center">75</td><td align="center">56</td><td align="center">63</td><td align="center">33</td><td align="center">43</td><td align="center">31</td><td align="center">25</td><td align="center">16</td><td align="center">15</td><td align="center">14</td><td align="center"><bold><italic>32</italic></bold></td><td align="center"><bold><italic>21</italic></bold></td></tr><tr><td align="center">Moxifloxacin</td><td align="center">5</td><td align="center">2</td><td align="center">14</td><td align="center">3</td><td align="center">64</td><td align="center">43</td><td align="center">22</td><td align="center">18</td><td align="center">43</td><td align="center">27</td><td align="center">4</td><td align="center">8</td><td align="center">13</td><td align="center">15</td><td align="center"><bold><italic>17</italic></bold></td><td align="center"><bold><italic>12</italic></bold></td></tr><tr><td align="center">Gatifloxacin</td><td align="center">30</td><td align="center">1</td><td align="center">14</td><td align="center">4</td><td align="center">36</td><td align="center">25</td><td align="center">7</td><td align="center">6</td><td align="center">23</td><td align="center">17</td><td align="center">6</td><td align="center">6</td><td align="center">15</td><td align="center">14</td><td align="center"><bold><italic>13</italic></bold></td><td align="center"><bold><italic>7</italic></bold></td></tr><tr><td align="center">Gentamicin</td><td align="center">87</td><td align="center">31</td><td align="center">36</td><td align="center">21</td><td align="center">43</td><td align="center">38</td><td align="center">30</td><td align="center">24</td><td align="center">37</td><td align="center">29</td><td align="center">27</td><td align="center">16</td><td align="center">35</td><td align="center">21</td><td align="center"><bold><italic>25</italic></bold></td><td align="center"><bold><italic>16</italic></bold></td></tr><tr><td align="center"><bold><italic>Mean Resistance</italic></bold></td><td align="center"><bold><italic>77</italic></bold></td><td align="center"><bold><italic>28</italic></bold></td><td align="center"><bold><italic>30</italic></bold></td><td align="center"><bold><italic>15</italic></bold></td><td align="center"><bold><italic>42</italic></bold></td><td align="center"><bold><italic>34</italic></bold></td><td align="center"><bold><italic>54</italic></bold></td><td align="center"><bold><italic>39</italic></bold></td><td align="center"><bold><italic>26</italic></bold></td><td align="center"><bold><italic>18</italic></bold></td><td align="center"><bold><italic>23</italic></bold></td><td align="center"><bold><italic>20</italic></bold></td><td align="center"><bold><italic>35</italic></bold></td><td align="center"><bold><italic>28</italic></bold></td><td></td><td></td></tr></tbody></table></table-wrap><fig position="float" id="F2"><label>Figure 2</label><caption><p>Hospital-acquired infections (HAI) are more resistant than community-acquired infections (CAI) to a wide range of antibiotics in China.</p></caption><graphic xlink:href="1744-8603-2-6-2"/></fig><fig position="float" id="F3"><label>Figure 3</label><caption><p>The Seven most common bacteria show higher resistance among hospital-acquired infections (HAI) than community-acquired infections (CAI) in China.</p></caption><graphic xlink:href="1744-8603-2-6-3"/></fig><p>Both measures reinforce the finding that infections acquired in a hospital are often more drug resistant than other (community-acquired) infections. For the seven bacteria, the mean resistance rate of HAI is on average 1.5 times that of CAI in China. For the nineteen drugs, the aggregate measure of resistance for HAI, γ<sub><italic>H</italic></sub>, is on average 1.9 times that for CAI, γ<sub><italic>C</italic></sub>. This pattern is most extreme for infections caused by SAU, where resistance of HAI is two- to three- times that of CAI, depending on which measure is used. (T-tests of the difference between two groups indicate a p-value of less than 0.01 for the γ's and less than 0.09 for the mean resistance). Moreover, the prevalence of drug resistance for both kinds of infections is quite high. Mean resistance of HAI is 41% and that of CAI is 28%.</p></sec><sec><title>United States</title><p>Fairly comprehensive data on resistance trends in the U.S. come from the National Nosocomial Infections Surveillance System (NNIS) for hospital-based resistance, and the U.S. Active Bacterial Core Surveillance (ABC) project, which surveys a population of 16 million to 25 million community residents in 9 states each year [<xref ref-type="bibr" rid="B28">28</xref>-<xref ref-type="bibr" rid="B30">30</xref>]. We use data from an ABC program that surveys <italic>Streptococcus pneumoniae </italic>(SPN) from 1997 to 2002 to examine prevalence and trends (Table <xref ref-type="table" rid="T3">3</xref>). The average growth rate of resistance for this bacterium was 8%, lower than the 15% number for China. Interestingly, unlike the upward resistance trend in China, SPN resistance declined in the last two years of the study period in the US, following an initial rise. Such data should not be interpreted to mean that actual prevalence is permanently declining, since measurement issues engender considerable year-to-year variation in the sample prevalence.</p><table-wrap position="float" id="T3"><label>Table 3</label><caption><p>Non-susceptibilities of <italic>Streptococcus pneumoniae </italic>(SPN) in U.S. communities, 1997–2002</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center" colspan="9">Unit: %</td></tr></thead><tbody><tr><td align="center">Antibiotic</td><td align="center">1997</td><td align="center">1998</td><td align="center">1999</td><td align="center">2000</td><td align="center">2001</td><td align="center">2002</td><td align="center"><bold><italic>Average Resistance</italic></bold></td><td align="center"><bold><italic>Average Growth Rate</italic></bold></td></tr><tr><td colspan="9"><hr></hr></td></tr><tr><td align="center">Penicillin</td><td align="center">25</td><td align="center">24</td><td align="center">27</td><td align="center">28</td><td align="center">26</td><td align="center">21</td><td align="center"><bold><italic>25</italic></bold></td><td align="center"><bold><italic>2</italic></bold></td></tr><tr><td align="center">Cefotaxime</td><td align="center">13</td><td align="center">14</td><td align="center">17</td><td align="center">18</td><td align="center">16</td><td align="center">12</td><td align="center"><bold><italic>15</italic></bold></td><td align="center"><bold><italic>-1</italic></bold></td></tr><tr><td align="center">Erythromycin</td><td align="center">15</td><td align="center">15</td><td align="center">21</td><td align="center">22</td><td align="center">19</td><td align="center">17</td><td align="center"><bold><italic>18</italic></bold></td><td align="center"><bold><italic>4</italic></bold></td></tr><tr><td align="center">TMP/Sulfa</td><td align="center">29</td><td align="center">29</td><td align="center">32</td><td align="center">32</td><td align="center">30</td><td align="center">25</td><td align="center"><bold><italic>30</italic></bold></td><td align="center"><bold><italic>-3</italic></bold></td></tr><tr><td align="center">Levofloxacin</td><td align="center">n/a</td><td align="center">0.2</td><td align="center">0.2</td><td align="center">0.3</td><td align="center">0.7</td><td align="center">0.5</td><td align="center"><bold><italic>0.4</italic></bold></td><td align="center"><bold><italic>39</italic></bold></td></tr><tr><td align="center">Vancomycin</td><td align="center">0</td><td align="center">0</td><td align="center">0</td><td align="center">0</td><td align="center">0</td><td align="center">0</td><td align="center"><bold><italic>18</italic></bold></td><td align="center"><bold><italic>8</italic></bold></td></tr></tbody></table></table-wrap><p>The US NNIS program provides data for inpatients and outpatients. Further, among inpatients, the NNIS differentiates between those in and not in the ICU. For almost every bug-drug pair, resistance prevalence is highest among ICU patients, followed by non-ICU inpatients, with the lowest prevalence among outpatients (Table <xref ref-type="table" rid="T4">4</xref> and Figure <xref ref-type="fig" rid="F4">4</xref>). This pattern seems consistent with clinical reality, since patients in ICUs are more likely to have a weak immune system, either because of prolonged treatment or their own compromised conditions; moreover, many are catheterized, offering a conduit for bacteria.</p><table-wrap position="float" id="T4"><label>Table 4</label><caption><p>Resistance prevalence for selected drug-bug pairs by patient type, U.S. 1999–2002</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center" colspan="5">unit: %</td></tr></thead><tbody><tr><td align="center">Pair</td><td align="center">Bacterium (resistant to) → drug</td><td align="center">ICU patients</td><td align="center">non-ICU inpatients</td><td align="center">Outpatients</td></tr><tr><td colspan="5"><hr></hr></td></tr><tr><td align="center">A</td><td align="center">PAE → Ciprofloxacin/ofloxacin</td><td align="center">32</td><td align="center">25</td><td align="center">23</td></tr><tr><td align="center">B</td><td align="center">PAE → Levofloxacin</td><td align="center">37</td><td align="center">28</td><td align="center">25</td></tr><tr><td align="center">C</td><td align="center">PAE → Imipenem</td><td align="center">18</td><td align="center">12</td><td align="center">9</td></tr><tr><td align="center">D</td><td align="center">PAE → Ceftazidime</td><td align="center">13</td><td align="center">8</td><td align="center">5</td></tr><tr><td align="center">E</td><td align="center">PAE → Piperacillin</td><td align="center">16</td><td align="center">11</td><td align="center">6</td></tr><tr><td align="center">F</td><td align="center">SAU → Methicillin</td><td align="center">47</td><td align="center">38</td><td align="center">23</td></tr><tr><td align="center">G</td><td align="center">Enterococcus spp → Vancomycin</td><td align="center">13</td><td align="center">11</td><td align="center">4</td></tr><tr><td align="center">H</td><td align="center">ECO → Cef3</td><td align="center">1</td><td align="center">1</td><td align="center">0</td></tr><tr><td align="center">I</td><td align="center">ECO → Quinolone</td><td align="center">5</td><td align="center">4</td><td align="center">2</td></tr><tr><td align="center">J</td><td align="center">KPN → Cef3</td><td align="center">6</td><td align="center">5</td><td align="center">2</td></tr><tr><td align="center">K</td><td align="center">Enterobacter spp → Cef3</td><td align="center">26</td><td align="center">21</td><td align="center">10</td></tr><tr><td align="center">L</td><td align="center">Enterobacter spp → Carbapenum</td><td align="center">1</td><td align="center">1</td><td align="center">1</td></tr><tr><td align="center">M</td><td align="center">CNS → Methicillin</td><td align="center">75</td><td align="center">63</td><td align="center">46</td></tr><tr><td align="center">N</td><td align="center">Pneumococcus → Penicillin</td><td align="center">18</td><td align="center">17</td><td align="center">17</td></tr><tr><td align="center">O</td><td align="center">Pneumococcus → Cef3</td><td align="center">7</td><td align="center">8</td><td align="center">6</td></tr><tr><td></td><td align="center"><bold><italic>Mean</italic></bold></td><td align="center"><bold><italic>21</italic></bold></td><td align="center"><bold><italic>17</italic></bold></td><td align="center"><bold><italic>12</italic></bold></td></tr></tbody></table><table-wrap-foot><p>Cef3 (3<sup>rd </sup>generation cephalosporin) = ceftazidime, cefotaxime or ceftriaxone;</p><p>*Quinolone = ciprofloxacin, ofloxacin or levofloxacin.</p></table-wrap-foot></table-wrap><fig position="float" id="F4"><label>Figure 4</label><caption><p>ICU patients have the highest resistance rates in selected drug-bug pairs, followed by non-ICU inpatients and outpatients, U.S. 1999–2002.</p></caption><graphic xlink:href="1744-8603-2-6-4"/></fig><p>Compared with China, the U.S. exhibits more moderate differences in resistance prevalence among different patients. The average prevalence of resistance for ICU, other inpatients, and outpatients in the U.S. are 20%, 17% and 13%, respectively; in China, average resistance for hospital-acquired infections is 41% and that for community-acquired infections is 28%.</p><p>Pooling all patients together (Table <xref ref-type="table" rid="T5">5</xref>), we find the prevalence of resistance and its growth to be 17% and 7% respectively, consistent with our previous observation that the U.S. seems to have both lower resistance prevalence and less dramatic increase in resistance than China does.</p><table-wrap position="float" id="T5"><label>Table 5</label><caption><p>Resistance prevalence of eight common bacteria, U.S. (all patients pooled), 1999–2002</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center" colspan="9">unit: %</td></tr></thead><tbody><tr><td align="center">Bacterium</td><td align="center">Resistant to antibiotic(s)</td><td></td><td align="center">1999</td><td align="center">2000</td><td align="center">2001</td><td align="center">2002</td><td align="center"><bold><italic>Average Resistance</italic></bold></td><td align="center"><bold><italic>Average Growth Rate</italic></bold></td></tr><tr><td colspan="9"><hr></hr></td></tr><tr><td align="center">PAE</td><td align="center">Ciprofloxacin/ofloxacin</td><td></td><td align="center">23</td><td align="center">25</td><td align="center">28</td><td align="center">29</td><td align="center"><bold><italic>26</italic></bold></td><td align="center"><bold><italic>8</italic></bold></td></tr><tr><td></td><td align="center">Levofloxacin</td><td></td><td align="center">29</td><td align="center">30</td><td align="center">31</td><td align="center">30</td><td align="center"><bold><italic>30</italic></bold></td><td align="center"><bold><italic>1</italic></bold></td></tr><tr><td></td><td align="center">Imipenem</td><td></td><td align="center">12</td><td align="center">12</td><td align="center">15</td><td align="center">13</td><td align="center"><bold><italic>13</italic></bold></td><td align="center"><bold><italic>4</italic></bold></td></tr><tr><td></td><td align="center">Ceftazidime</td><td></td><td align="center">8</td><td align="center">8</td><td align="center">9</td><td align="center">9</td><td align="center"><bold><italic>9</italic></bold></td><td align="center"><bold><italic>4</italic></bold></td></tr><tr><td></td><td align="center">Piperacillin</td><td></td><td align="center">10</td><td align="center">10</td><td align="center">11</td><td align="center">12</td><td align="center"><bold><italic>11</italic></bold></td><td align="center"><bold><italic>6</italic></bold></td></tr><tr><td align="center">SAU (MRSA)</td><td align="center">Methicillin</td><td></td><td align="center">32</td><td align="center">35</td><td align="center">38</td><td align="center">39</td><td align="center"><bold><italic>36</italic></bold></td><td align="center"><bold><italic>7</italic></bold></td></tr><tr><td align="center">Enterococcus spp</td><td align="center">Vancomycin</td><td></td><td align="center">11</td><td align="center">8</td><td align="center">10</td><td align="center">10</td><td align="center"><bold><italic>10</italic></bold></td><td align="center"><bold><italic>-1</italic></bold></td></tr><tr><td align="center">ECO</td><td align="center">Cef3</td><td></td><td align="center">1</td><td align="center">1</td><td align="center">1</td><td align="center">1</td><td align="center"><bold><italic>1</italic></bold></td><td align="center"><bold><italic>0</italic></bold></td></tr><tr><td></td><td align="center">Quinolone</td><td></td><td align="center">2</td><td align="center">3</td><td align="center">4</td><td align="center">5</td><td align="center"><bold><italic>4</italic></bold></td><td align="center"><bold><italic>36</italic></bold></td></tr><tr><td align="center">KPN</td><td align="center">Cef3</td><td></td><td align="center">4</td><td align="center">4</td><td align="center">4</td><td align="center">5</td><td align="center"><bold><italic>4</italic></bold></td><td align="center"><bold><italic>8</italic></bold></td></tr><tr><td align="center">Enterobacter spp</td><td align="center">Cef3</td><td></td><td align="center">19</td><td align="center">19</td><td align="center">18</td><td align="center">19</td><td align="center"><bold><italic>19</italic></bold></td><td align="center"><bold><italic>0</italic></bold></td></tr><tr><td></td><td align="center">Carbapenum</td><td></td><td align="center">1</td><td align="center">1</td><td align="center">1</td><td align="center">1</td><td align="center"><bold><italic>1</italic></bold></td><td align="center"><bold><italic>0</italic></bold></td></tr><tr><td align="center">CNS</td><td align="center">Methicillin</td><td></td><td align="center">60</td><td align="center">61</td><td align="center">62</td><td align="center">63</td><td align="center"><bold><italic>62</italic></bold></td><td align="center"><bold><italic>2</italic></bold></td></tr><tr><td align="center">Pneumococcus spp</td><td align="center">Penicillin</td><td></td><td align="center">14</td><td align="center">16</td><td align="center">19</td><td align="center">19</td><td align="center"><bold><italic>17</italic></bold></td><td align="center"><bold><italic>11</italic></bold></td></tr><tr><td></td><td align="center">Cef3</td><td></td><td align="center">5</td><td align="center">8</td><td align="center">7</td><td align="center">7</td><td align="center"><bold><italic>7</italic></bold></td><td align="center"><bold><italic>16</italic></bold></td></tr><tr><td></td><td></td><td align="center"><bold><italic>Mean:</italic></bold></td><td></td><td></td><td></td><td></td><td align="center"><bold><italic>17</italic></bold></td><td align="center"><bold><italic>7</italic></bold></td></tr></tbody></table></table-wrap></sec><sec><title>Kuwait</title><p>There is considerably less detailed data on antibiotic resistance for Kuwait than for China or the U.S. We gathered data on antimicrobial resistance among isolates of eight different bacterial diseases over the most recent five years. The data is based on surveillance from a single large teaching hospital, Mubarak Al-Kabeer Hospital, which serves a catchment area representing about 60% of Kuwait's population. We report that data for the first time here and in a companion paper [<xref ref-type="bibr" rid="B31">31</xref>] (see Tables <xref ref-type="table" rid="T6">6</xref>, <xref ref-type="table" rid="T7">7</xref>, <xref ref-type="table" rid="T8">8</xref>, <xref ref-type="table" rid="T9">9</xref>).The average resistance level for all surveyed bacteria was about 27% from 1999 to 2003 (Table <xref ref-type="table" rid="T10">10</xref>), higher than the 17% for the U.S. and about the same as the 28% China. As for the other two countries, resistance appears to be growing in Kuwait.</p><table-wrap position="float" id="T6"><label>Table 6</label><caption><p>Resistance trend in isolates of <italic>Salmonella </italic>spp. over 5 years in Kuwait</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center">Antibiotic</td><td align="center" colspan="5">Percentage (%) of resistant isolates in:</td></tr><tr><td></td><td colspan="5"><hr></hr></td></tr><tr><td></td><td align="center">1999 (n = 216)</td><td align="center">2000 (n = 215)</td><td align="center">2001 (n = 129)</td><td align="center">2002 (n = 167)</td><td align="center">2003 (n = 165)</td></tr></thead><tbody><tr><td align="center">Amikacin</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr><tr><td align="center">Ampicillin</td><td align="right">6</td><td align="right">12</td><td align="right">7</td><td align="right">25</td><td align="right">26</td></tr><tr><td align="center">Amoxicillin-clavulanate</td><td align="right">5</td><td align="right">10</td><td align="right">7</td><td align="right">2</td><td align="right">0</td></tr><tr><td align="center">Cefotaxime</td><td align="right">0</td><td align="right">1</td><td align="right">0</td><td align="right">1</td><td align="right">0</td></tr><tr><td align="center">Ceftriaxone</td><td align="right">0</td><td align="right">1</td><td align="right">0</td><td align="right">2</td><td align="right">0</td></tr><tr><td align="center">Cefuroxime</td><td align="right">1</td><td align="right">1</td><td align="right">0</td><td align="right">27</td><td align="right">41</td></tr><tr><td align="center">Cephalexin</td><td align="right">2</td><td align="right">10</td><td align="right">37</td><td align="right">57</td><td align="right">50</td></tr><tr><td align="center">Chloramphenicol</td><td align="right">8</td><td align="right">21</td><td align="right">0</td><td align="right">18</td><td align="right">18</td></tr><tr><td align="center">Ciprofloxacin</td><td align="right">0</td><td align="right">0</td><td align="right">14</td><td align="right">10</td><td align="right">16</td></tr><tr><td align="center">TMP/SMX</td><td align="right">8</td><td align="right">8</td><td align="right">10</td><td align="right">20</td><td align="right">20</td></tr><tr><td align="center">Gentamicin</td><td align="right">6</td><td align="right">1</td><td align="right">0</td><td align="right">42</td><td align="right">42</td></tr><tr><td align="center">Imipenem</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr><tr><td align="center">Meropenem</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr><tr><td align="center">Piperacillin</td><td align="right">6</td><td align="right">13</td><td align="right">13</td><td align="right">23</td><td align="right">25</td></tr><tr><td align="center">Piperacillin/tazobactam</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr></tbody></table><table-wrap-foot><p>No ESBL-producing strain has been isolated so far</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T7"><label>Table 7</label><caption><p>Resistance trend in isolates of <italic>Streptococcus pneumoniae </italic>over a 5-year period in Kuwait</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center">Antibiotics</td><td align="center" colspan="5">Percentage (%) of resistant isolates in:</td></tr><tr><td></td><td colspan="5"><hr></hr></td></tr><tr><td></td><td align="center">1999 (n = 78)</td><td align="center">2000 (n = 61)</td><td align="center">2001 (n = 73)</td><td align="center">2002 (n = 66)</td><td align="center">2003 (n = 90)</td></tr></thead><tbody><tr><td align="center">Cefotaxime</td><td align="right">0</td><td align="right">0</td><td align="right">4</td><td align="right">5</td><td align="right">6</td></tr><tr><td align="center">Ceftriaxone</td><td align="right">0</td><td align="right">0</td><td align="right">3</td><td align="right">5</td><td align="right">4</td></tr><tr><td align="center">Cefuroxime</td><td align="right">0</td><td align="right">0</td><td align="right">8</td><td align="right">9</td><td align="right">41</td></tr><tr><td align="center">Cephalexin</td><td align="right">0</td><td align="right">0</td><td align="right">NT</td><td align="right">NT</td><td align="right">NT</td></tr><tr><td align="center">Chloramphenicol</td><td align="right">3</td><td align="right">5</td><td align="right">25</td><td align="right">5</td><td align="right">0</td></tr><tr><td align="center">Erythromycin</td><td align="right">16</td><td align="right">20</td><td align="right">23</td><td align="right">26</td><td align="right">30</td></tr><tr><td align="center">Imipenem</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr><tr><td align="center">Penicillin</td><td align="right">32</td><td align="right">38</td><td align="right">46</td><td align="right">52</td><td align="right">54</td></tr><tr><td align="center">Teicoplanin</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr><tr><td align="center">Vancomycin</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr></tbody></table><table-wrap-foot><p>NT = not tested</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T8"><label>Table 8</label><caption><p>Percentage of <italic>Enterococcus </italic>species resistant to often-tested antibiotics over 5 years in Kuwait</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center">Antibiotic</td><td align="center" colspan="5">Percentage (%) of resistant isolates in:</td></tr><tr><td></td><td colspan="5"><hr></hr></td></tr><tr><td></td><td align="center">1999 (n = 370)</td><td align="center">2000 (n = 335)</td><td align="center">2001 (n = 322)</td><td align="center">2002 (n = 248)</td><td align="center">2003 (n = 212)</td></tr></thead><tbody><tr><td align="center">Ampicillin</td><td align="right">1</td><td align="right">1</td><td align="right">3</td><td align="right">2</td><td align="right">0</td></tr><tr><td align="center">Erythromycin</td><td align="right">59</td><td align="right">78</td><td align="right">77</td><td align="right">75</td><td align="right">92</td></tr><tr><td align="center">Gentamicin</td><td align="right">26</td><td align="right">36</td><td align="right">61</td><td align="right">52</td><td align="right">98</td></tr><tr><td align="center">Nitrofurantoin</td><td align="right">2</td><td align="right">2</td><td align="right">2</td><td align="right">36</td><td align="right">86</td></tr><tr><td align="center">Norfloxacin</td><td align="right">36</td><td align="right">47</td><td align="right">47</td><td align="right">NT</td><td align="right">NT</td></tr><tr><td align="center">Penicillin</td><td align="right">16</td><td align="right">38</td><td align="right">35</td><td align="right">53</td><td align="right">85</td></tr><tr><td align="center">Teicoplanin</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">1</td><td align="right">0</td></tr><tr><td align="center">Vancomycin</td><td align="right">1</td><td align="right">0</td><td align="right">0</td><td align="right">2</td><td align="right">0</td></tr></tbody></table><table-wrap-foot><p>NT = not tested</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T9"><label>Table 9</label><caption><p>Percentage of <italic>Staphylococcus aureus </italic>resistant to often-tested antibiotics over 5 years in Kuwait</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center">Antibiotic</td><td align="center" colspan="5">Percentage (%) of resistant isolates in:</td></tr><tr><td></td><td colspan="5"><hr></hr></td></tr><tr><td></td><td align="center">1999 (n = 648)</td><td align="center">2000 (n = 595)</td><td align="center">2001 (n = 484)</td><td align="center">2002 (n = 420)</td><td align="center">2003 (n = 286)</td></tr></thead><tbody><tr><td align="center">Ampicillin</td><td align="right">96</td><td align="right">100</td><td align="right">98</td><td align="right">96</td><td align="right">98</td></tr><tr><td align="center">Amoxicillin-clavulanic acid</td><td align="right">6</td><td align="right">33</td><td align="right">27</td><td align="right">22</td><td align="right">29</td></tr><tr><td align="center">Cephalexin</td><td align="right">33</td><td align="right">30</td><td align="right">25</td><td align="right">36</td><td align="right">34</td></tr><tr><td align="center">Ciprofloxacin</td><td align="right">10</td><td align="right">35</td><td align="right">30</td><td align="right">45</td><td align="right">50</td></tr><tr><td align="center">Clindamycin</td><td align="right">18</td><td align="right">24</td><td align="right">20</td><td align="right">20</td><td align="right">27</td></tr><tr><td align="center">Cloxacillin</td><td align="right">23</td><td align="right">24</td><td align="right">9</td><td align="right">22</td><td align="right">17</td></tr><tr><td align="center">Erythromycin</td><td align="right">38</td><td align="right">34</td><td align="right">26</td><td align="right">28</td><td align="right">27</td></tr><tr><td align="center">Fusidic acid</td><td align="right">NA</td><td align="right">20</td><td align="right">19</td><td align="right">64</td><td align="right">27</td></tr><tr><td align="center">Gentamicin</td><td align="right">25</td><td align="right">21</td><td align="right">16</td><td align="right">24</td><td align="right">27</td></tr><tr><td align="center">Methicillin</td><td align="right">23</td><td align="right">24</td><td align="right">9</td><td align="right">22</td><td align="right">17</td></tr><tr><td align="center">Penicillin</td><td align="right">95</td><td align="right">95</td><td align="right">99</td><td align="right">96</td><td align="right">99</td></tr><tr><td align="center">Teicoplanin</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr><tr><td align="center">TMP/SMX</td><td align="right">24</td><td align="right">27</td><td align="right">31</td><td align="right">18</td><td align="right">94</td></tr><tr><td align="center">Vancomycin</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td><td align="right">0</td></tr></tbody></table></table-wrap><table-wrap position="float" id="T10"><label>Table 10</label><caption><p>Average Resistance Levels of Major Bacteria in Kuwait, 1999–2003</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center" colspan="11">unit: %</td></tr></thead><tbody><tr><td/><td align="center">ECO</td><td align="center">KPN</td><td align="center">PAE</td><td align="center">SPN</td><td align="center">Shigella spp.</td><td align="center">Salmonella spp.</td><td align="center">Enterococcus spp.</td><td align="center">SAU</td><td align="center"><bold><italic>Average Resistance</italic></bold></td><td align="center"><bold><italic>Average Growth</italic></bold></td></tr><tr><td colspan="11"><hr></hr></td></tr><tr><td align="center">Average Annual Resistance</td><td align="center">13</td><td align="center">8</td><td align="center">5</td><td align="center">31</td><td align="center">45</td><td align="center">65</td><td align="center">37</td><td align="center">8</td><td align="center"><bold><italic>27</italic></bold></td><td align="center"><bold><italic>17</italic></bold></td></tr></tbody></table></table-wrap></sec></sec><sec><title>Discussion: Comparing antibiotic resistance in China, the U.S. and Kuwait</title><p>In China, resistance rates exhibit a clear and rapid upward trend. In the U.S., resistance currently appears to grow at a more leisurely pace. Kuwait seems to be somewhere in between. It is important to note that the pace of growth may depend on the whether resistance to a particular antibiotic has reached a potential equilibrium. As shown in the previous data, the 3% resistance growth rate of ECO against Ciprofloxacin in China (Table <xref ref-type="table" rid="T1">1</xref>), is considerably lower than it is in the other two countries against similar quinolone drugs (Table <xref ref-type="table" rid="T5">5</xref> and Table <xref ref-type="table" rid="T10">10</xref>). This is probably because ECO resistance may have virtually reached equilibrium in China by the beginning of the study period; hence it didn't grow much in subsequent years.</p><p>That resistance does not grow without bound highlights the importance of comparing the current prevalence of resistance in the three countries. After all, the prevalence of resistance reflects the risk of a drug-resistant infection for any given patient. A low rate of growth is small consolation if patients already face a high baseline risk of a acquiring an expensive, debilitating and even potentially untreatable "superbug" infection.</p><p>The prevalence of resistance also substantially differs across countries, although as noted previously, surveillance data is far from ideal in capturing the true scope of the problem. As shown in Table <xref ref-type="table" rid="T11">11</xref>, using the data currently available, China has far higher prevalence of resistance for all the bacteria studied. For example, in China resistance of SPN to one of the oldest antibiotics, erythromycin, reaches 73%, while the figure for Kuwait is only 23%. A challenge for the U.S. is the exceptionally high level of Vancomycin-Resistant Enterococcus spp (VRE). In the U.S., 53% of Shigella spp are resistant to Trimethoprim/Sulfamethoxazole (TMP/SMX), in contrast to 0% in both of the other countries. These examples suggest that severity of resistance may be correlated with volume of usage. Vancomycin is less affordable in both China and Kuwait, presumably resulting in less usage in those countries.</p><table-wrap position="float" id="T11"><label>Table 11</label><caption><p>Resistance rates in China, U.S. and Kuwait, hospital surveillance data for 2001</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center" colspan="6">From Tables 1,2,3,8 and 9; Unit: %</td></tr></thead><tbody><tr><td align="center">Bacterium(a)</td><td align="center">Antibiotic(s)</td><td align="center">Pair</td><td align="center">China</td><td align="center">U.S.</td><td align="center">Kuwait</td></tr><tr><td colspan="6"><hr></hr></td></tr><tr><td align="center">SAU</td><td align="center">Methicillin</td><td align="center">A</td><td align="center">37</td><td align="center">38</td><td align="center">9</td></tr><tr><td align="center">SPN</td><td align="center">Erythromycin</td><td align="center">B</td><td align="center">73</td><td align="center">19</td><td align="center">23</td></tr><tr><td/><td align="center">Cefotaxime</td><td align="center">C</td><td align="center">0</td><td align="center">16</td><td align="center">4</td></tr><tr><td align="center"><italic>Enterococcus spp</italic></td><td align="center">Vancomycin</td><td align="center">D</td><td align="center">4</td><td align="center">10</td><td align="center">0</td></tr><tr><td align="center"><italic>ECO</italic></td><td align="center">Ceftazidime</td><td align="center">E</td><td align="center">9</td><td align="center">1</td><td align="center">5</td></tr><tr><td/><td align="center">Cefotaxime</td><td align="center">F</td><td align="center">18</td><td align="center">1</td><td align="center">1</td></tr><tr><td/><td align="center">Ceftriaxone</td><td align="center">G</td><td align="center">21</td><td align="center">1</td><td align="center">1</td></tr><tr><td/><td align="center">Ciprofloxacin/Ofloxacin</td><td align="center">H</td><td align="center">56</td><td align="center">3</td><td align="center">26</td></tr><tr><td align="center">PAE</td><td align="center">Ceftazidime</td><td align="center">I</td><td align="center">17</td><td align="center">9</td><td align="center">27</td></tr><tr><td/><td align="center">Ciprofloxacin/Ofloxacin</td><td align="center">J</td><td align="center">27</td><td align="center">28</td><td align="center">31</td></tr><tr><td align="center">KPN</td><td align="center">Ceftazidime</td><td align="center">K</td><td align="center">9</td><td align="center">4</td><td align="center">14</td></tr><tr><td/><td align="center">Cefotaxime</td><td align="center">L</td><td align="center">17</td><td align="center">4</td><td align="center">13</td></tr><tr><td/><td align="center">Ceftriaxone</td><td align="center">M</td><td align="center">20</td><td align="center">4</td><td align="center">13</td></tr><tr><td/><td align="center">Ciprofloxacin</td><td align="center">N</td><td align="center">18</td><td align="center">12[27]</td><td align="center">18</td></tr><tr><td align="center"><italic>Salmonella </italic>spp</td><td align="center">Amoxicillin-clavulanate</td><td align="center">O</td><td align="center">10</td><td align="center">4</td><td align="center">7</td></tr><tr><td/><td align="center">Ceftriaxone</td><td align="center">P</td><td align="center">5</td><td align="center">1</td><td align="center">0</td></tr><tr><td/><td align="center">Ciprofloxacin</td><td align="center">Q</td><td align="center">0</td><td align="center">0.4</td><td align="center">10</td></tr><tr><td/><td align="center">TMP/SMX*</td><td align="center">R</td><td align="center">0</td><td align="center">3</td><td align="center">0</td></tr><tr><td/><td align="center">Gentamicin</td><td align="center">S</td><td align="center">10</td><td align="center">2</td><td align="center">0</td></tr><tr><td align="center"><italic>Shigella </italic>spp</td><td align="center">Amoxicillin-clavulanate</td><td align="center">T</td><td align="center">35</td><td align="center">2</td><td align="center">20</td></tr><tr><td/><td align="center">Ceftriaxone</td><td align="center">U</td><td align="center">6</td><td align="center">0</td><td align="center">0</td></tr><tr><td/><td align="center">Ciprofloxacin</td><td align="center">V</td><td align="center">6</td><td align="center">0</td><td align="center">0</td></tr><tr><td/><td align="center">TMP/SMX</td><td align="center">W</td><td align="center">0</td><td align="center">53</td><td align="center">0</td></tr><tr><td/><td align="center">Gentamicin</td><td align="center">X</td><td align="center">18</td><td align="center">0.2</td><td align="center">0</td></tr><tr><td/><td/><td align="center"><bold><italic>Average</italic></bold></td><td align="center"><bold><italic>17</italic></bold></td><td align="center"><bold><italic>7</italic></bold></td><td align="center"><bold><italic>9</italic></bold></td></tr></tbody></table><table-wrap-foot><p> The original U.S. NNIS reported resistance rates to either one of the Cef3 drugs, i.e. ceftazidime, cefotaxime or ceftriaxone. We assume the same rates for each drug.</p><p> Based on surveillance of ICU patients</p><p>* TMP/SMX = Trimethoprim/Sulfamethoxazole</p></table-wrap-foot></table-wrap><p>Table <xref ref-type="table" rid="T12">12</xref> compares the three countries with Japan and Taiwan regarding prevalence of three important drug-resistant bacteria: MRSA, penicillin resistant SPN (PRSP) and vancomycin-resistant <italic>Enterococcus </italic>spp (VRE) [<xref ref-type="bibr" rid="B32">32</xref>-<xref ref-type="bibr" rid="B34">34</xref>]. Interestingly, each country has its own most problematic resistance culprit. For China, MRSA is the biggest threat, where resistance among hospital-acquired infections reaches almost 90%, the highest among the five countries. For the U.S., VRE is high. VRE growth in the U.S. can be traced to the late 1980s and is probably among the highest in the world. For Kuwait, PRSP is considerable. Both Taiwan and Japan are also troubled by at least one of these three resistant bacteria.</p><table-wrap position="float" id="T12"><label>Table 12</label><caption><p>MRSA, PRSP & VRE in Selected Countries</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center" colspan="4">Unit: %</td></tr></thead><tbody><tr><td/><td align="center">MRSA (HAI only)</td><td align="center">PRSP</td><td align="center">VRE</td></tr><tr><td colspan="4"><hr></hr></td></tr><tr><td align="center">China</td><td align="center">89 (2001)</td><td align="center">27 (2001)</td><td align="center">0 (2001)</td></tr><tr><td align="center">U.S.</td><td align="center">16 (2001)</td><td align="center">26 (2001)</td><td align="center">0.3 (1989), 8 (1993), 12.8 (2001) in ICU</td></tr><tr><td align="center">Kuwait</td><td align="center">9 (2001)</td><td align="center">46 (2001)</td><td align="center">0 (2001)</td></tr><tr><td align="center">Japan [33]</td><td align="center">60–80% (1999)</td><td align="center">11–40 (1999)</td><td align="center">n/a</td></tr><tr><td align="center">Taiwan [34]</td><td align="center">n/a</td><td align="center">69 (2000)</td><td align="center">2 (2000)</td></tr></tbody></table></table-wrap><sec><title>Resistance correlations</title><p>How similar or different are resistance patterns in different countries? Does transmission travel across national borders as humans do? If so, do countries' resistance patterns converge? To begin to examine this issue, we construct coefficients of resistance correlation among China, U.S. and Kuwait. We rank resistance rates for 24 bug-drug pairs and define perfect correlation as each bug-drug pair displaying the same resistance rank. Perfect negative correlation exists if the ranks in two countries go in precisely the opposite order. Table <xref ref-type="table" rid="T13">13</xref> reports the correlation coefficient for each pair of countries. The statistic by definition is bounded between -1 and 1, where -1 means perfect disagreement while 1 means perfect agreement. Thus the bigger the statistic, the more correlated two countries' resistance patterns are.</p><table-wrap position="float" id="T13"><label>Table 13</label><caption><p>Ranks of resistance rates in China, U.S. and Kuwait, 2001(Rank correlations at bottom of table)</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center">Bacterium(a)</td><td align="center">Antibiotic(s)</td><td align="center">China</td><td align="center">U.S.</td><td align="center">Kuwait</td></tr></thead><tbody><tr><td align="center">SAU</td><td align="center">Methicillin</td><td align="center"><italic>3</italic></td><td align="center"><italic>2</italic></td><td align="center"><italic>11</italic></td></tr><tr><td align="center">SPN</td><td align="center">Erythromycin</td><td align="center"><italic>1</italic></td><td align="center"><italic>4</italic></td><td align="center"><italic>4</italic></td></tr><tr><td/><td align="center">Cefotaxime</td><td align="center"><italic>21</italic></td><td align="center"><italic>5</italic></td><td align="center"><italic>14</italic></td></tr><tr><td align="center"><italic>Enterococcus spp</italic></td><td align="center">Vancomycin</td><td align="center"><italic>20</italic></td><td align="center"><italic>7</italic></td><td align="center"><italic>17</italic></td></tr><tr><td align="center">ECO</td><td align="center">Ceftazidime</td><td align="center"><italic>15</italic></td><td align="center"><italic>17</italic></td><td align="center"><italic>13</italic></td></tr><tr><td/><td align="center">Cefotaxime</td><td align="center"><italic>8</italic></td><td align="center"><italic>18</italic></td><td align="center"><italic>15</italic></td></tr><tr><td/><td align="center">Ceftriaxone</td><td align="center"><italic>6</italic></td><td align="center"><italic>19</italic></td><td align="center"><italic>16</italic></td></tr><tr><td/><td align="center">Ciprofloxacin/Ofloxacin</td><td align="center"><italic>2</italic></td><td align="center"><italic>13</italic></td><td align="center"><italic>3</italic></td></tr><tr><td align="center">PAE</td><td align="center">Ceftazidime</td><td align="center"><italic>11</italic></td><td align="center"><italic>8</italic></td><td align="center"><italic>2</italic></td></tr><tr><td/><td align="center">Ciprofloxacin/Ofloxacin</td><td align="center"><italic>5</italic></td><td align="center"><italic>3</italic></td><td align="center"><italic>1</italic></td></tr><tr><td align="center">KPN</td><td align="center">Ceftazidime</td><td align="center"><italic>16</italic></td><td align="center"><italic>9</italic></td><td align="center"><italic>7</italic></td></tr><tr><td/><td align="center">Cefotaxime</td><td align="center"><italic>12</italic></td><td align="center"><italic>10</italic></td><td align="center"><italic>8</italic></td></tr><tr><td/><td align="center">Ceftriaxone</td><td align="center"><italic>7</italic></td><td align="center"><italic>11</italic></td><td align="center"><italic>9</italic></td></tr><tr><td/><td align="center">Ciprofloxacin</td><td align="center"><italic>9</italic></td><td align="center"><italic>6</italic></td><td align="center"><italic>6</italic></td></tr><tr><td align="center"><italic>Salmonella </italic>spp</td><td align="center">Amoxicillin-clavulanate</td><td align="center"><italic>13</italic></td><td align="center"><italic>12</italic></td><td align="center"><italic>12</italic></td></tr><tr><td/><td align="center">Ceftriaxone</td><td align="center"><italic>19</italic></td><td align="center"><italic>20</italic></td><td align="center"><italic>18</italic></td></tr><tr><td/><td align="center">Ciprofloxacin</td><td align="center"><italic>22</italic></td><td align="center"><italic>21</italic></td><td align="center"><italic>10</italic></td></tr><tr><td/><td align="center">TMP/SMX</td><td align="center"><italic>23</italic></td><td align="center"><italic>14</italic></td><td align="center"><italic>19</italic></td></tr><tr><td/><td align="center">Gentamicin</td><td align="center"><italic>14</italic></td><td align="center"><italic>15</italic></td><td align="center"><italic>20</italic></td></tr><tr><td align="center"><italic>Shigella </italic>spp</td><td align="center">Amoxicillin-clavulanate</td><td align="center"><italic>4</italic></td><td align="center"><italic>16</italic></td><td align="center"><italic>5</italic></td></tr><tr><td/><td align="center">Ceftriaxone</td><td align="center"><italic>17</italic></td><td align="center"><italic>23</italic></td><td align="center"><italic>21</italic></td></tr><tr><td/><td align="center">Ciprofloxacin</td><td align="center"><italic>18</italic></td><td align="center"><italic>24</italic></td><td align="center"><italic>22</italic></td></tr><tr><td/><td align="center">TMP/SMX</td><td align="center"><italic>24</italic></td><td align="center"><italic>1</italic></td><td align="center"><italic>23</italic></td></tr><tr><td/><td align="center">Gentamicin</td><td align="center"><italic>10</italic></td><td align="center"><italic>22</italic></td><td align="center"><italic>24</italic></td></tr><tr><td align="center" colspan="2"><bold><italic>Correlation Coefficients</italic></bold></td><td align="center"><bold><italic>CHN_US: 0.18</italic></bold></td><td align="center"><bold><italic>US_KW: 0.46</italic></bold></td><td align="center"><bold><italic>CHN_KW: 0.60</italic></bold></td></tr></tbody></table></table-wrap><p>Of course, methods for aggregation and comparing patterns of resistance across countries and over time should be improved, and applied more fruitfully with better data from increased local and global surveillance. But even this preliminary analysis reveals some interesting patterns. For example, resistance rates in China are much more strongly correlated with those in Kuwait than those in the U.S. This correlation pattern suggests that at least in the short run, resistance in a country is more likely to be determined by endogenous factors (such as strictness of practices for prescribing drugs). In the long run, the frequency and magnitude of contacts among nations with different resistance problems is likely to be critical. Because Kuwait and China are relatively isolated countries, it is less surprising that their antibiotic resistance problems show domestic characters. However, as we expect them to be opening more to the world, particularly China, the problem may worsen when these countries can increasingly export and import antibiotic resistance. China, the most populous country in the world and an economy with the highest growth, is particularly likely to exacerbate the problem. As illustrated in Figure <xref ref-type="fig" rid="F1">1</xref>, the number of Chinese departures to overseas destinations has been growing at increasing rates in the past decade and continues to show upward momentum in recent years.</p><p>No doubt, there are also complex interactions with levels of economic well- being. Drugs become more affordable as countries become richer, but they are likely to be given out more carefully, particularly since concerns about resistance also increase. The critical question for policy is whether countries can control their own resistance problems, and also avoid importing the problem from abroad.</p></sec></sec><sec><title>Conclusion</title><p>We have outlined the nature of the antimicrobial resistance problem as an important health and cost issue for three quite disparate nations, and by inference for a broad swath of the world's population. Surprisingly, this issue virtually never receives prominent attention at the national or international level, despite its scope and potentially devastating impact on global public health in the coming decades.</p><p>We examined antimicrobial resistance data for China, Kuwait, and the United States. In each country, we looked at specific infectious agents and their resistance to particular antibiotics or other antimicrobials. Though an upward trend of resistance is found broadly, the patterns of correlation between countries' resistance rates suggest predominantly independent profiles. But we would expect greater convergence as globalization increases contacts between different nations' populations, raising questions about how to coordinate an effective international response [<xref ref-type="bibr" rid="B35">35</xref>].</p><p>Future research should develop better methods of data aggregation, explore the patterns of drug resistance across more countries, analyze the determinants of transmission of drug resistance across national boundaries, and assess how those determinants are progressing. Individuals everywhere would benefit if far greater attention were paid to the problem of antimicrobial resistance.</p></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>RFZ assembled the data, carried out the analysis and drafted the manuscript. KE and RJZ conceived of the study, participated in its design and coordination, and helped to draft the manuscript. VR provided the Kuwait data and helped to draft the manuscript. All authors read and approved the manuscript.</p></sec>
Impaired germ cell development due to compromised cell cycle progression in Skp2-deficient mice	<sec><title>Background</title><p>The gonads are responsible for the production of germ cells through both mitosis and meiosis. Skp2 is the receptor subunit of an SCF-type ubiquitin ligase and is a major regulator of the progression of cells into S phase of the cell cycle, which it promotes by mediating the ubiquitin-dependent degradation of p27, an inhibitor of cell proliferation. However, the role of the Skp2-p27 pathway in germ cell development remains elusive.</p></sec><sec><title>Results</title><p>We now show that disruption of <italic>Skp2 </italic>in mice results in a marked impairment in the fertility of males, with the phenotypes resembling Sertoli cell-only syndrome in men. Testes of <italic>Skp2</italic><sup>-/- </sup>mice manifested pronounced germ cell hypoplasia accompanied by massive apoptosis in spermatogenic cells. Flow cytometry revealed an increased prevalence of polyploidy in spermatozoa, suggesting that the aneuploidy of these cells is responsible for the induction of apoptosis. Disruption of the <italic>p27 </italic>gene of <italic>Skp2</italic><sup>-/- </sup>mice restored germ cell development, indicating that the testicular hypoplasia of <italic>Skp2</italic><sup>-/- </sup>animals is attributable to the antiproliferative effect of p27 accumulation.</p></sec><sec><title>Conclusion</title><p>Our results thus suggest that compromised cell cycle progression caused by the accumulation of p27 results in aneuploidy and the induction of apoptosis in gonadal cells of <italic>Skp2</italic><sup>-/- </sup>mice. The consequent reduction in the number of mature gametes accounts for the decreased fertility of these animals. These findings reinforce the importance of the Skp2-p27 pathway in cell cycle regulation and in germ cell development.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>Fotovati</surname><given-names>Abbas</given-names></name><xref ref-type="aff" rid="I1">1</xref><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Nakayama</surname><given-names>Keiko</given-names></name><xref ref-type="aff" rid="I2">2</xref><xref ref-type="aff" rid="I3">3</xref><email>[email protected]</email></contrib><contrib id="A3" corresp="yes" contrib-type="author"><name><surname>Nakayama</surname><given-names>Keiichi I</given-names></name><xref ref-type="aff" rid="I1">1</xref><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib>	Cell Division	<sec><title>Background</title><p>Infertility affects 10 to 15% of couples with up to half of fertility problems having a genetic etiology [<xref ref-type="bibr" rid="B1">1</xref>]. A major type of infertility is characterized by impaired production of germ cells. Germ cell development begins with the appearance of primordial germ cells at the early stage of embryogenesis. These cells migrate to the genital ridge, where they proliferate extensively by mitosis in both male and female embryos to establish the original pool of germ cells. The germ cells then enter a state of divisional arrest and remain in this state until sexual maturity, when they complete their developmental process by undergoing spermatogenesis in males and folliculogenesis in females. Such gametogenesis is highly sensitive to deficiency of various contributing factors. Targeted disruption of several genes in mice has thus defined key roles for many extracellular and intracellular signaling proteins in germ cell development and reproductive physiology [<xref ref-type="bibr" rid="B2">2</xref>-<xref ref-type="bibr" rid="B5">5</xref>].</p><p>Germ cells undergo both mitosis and meiosis during their development. Progression through the cell cycle requires the activity of two major ubiquitin ligase complexes, the Skp1-cullin-F-box protein (SCF) complex and the anaphase-promoting complex (APC) or cyclosome [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B7">7</xref>]. The F-box protein component of the SCF complex is variable, binds to Skp1 through its F-box motif, and is the subunit responsible for substrate recognition [<xref ref-type="bibr" rid="B8">8</xref>]. The F-box protein Skp2 plays an important role in progression of S phase of the cell cycle by contributing to the ubiquitin-dependent degradation of p27, a major inhibitor of proliferation in mammalian cells [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B9">9</xref>-<xref ref-type="bibr" rid="B11">11</xref>]. Skp2 begins to accumulate in late G<sub>1 </sub>phase of the cell cycle and its abundance is maximal during S and G<sub>2 </sub>phases [<xref ref-type="bibr" rid="B12">12</xref>-<xref ref-type="bibr" rid="B14">14</xref>]. We have previously generated mice that lack Skp2 and shown that the levels of p27 and of various other regulators of the cell cycle are increased in the cells of these animals [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B15">15</xref>-<xref ref-type="bibr" rid="B18">18</xref>]. Although <italic>Skp2</italic><sup>-/- </sup>mice are viable, their somatic cells contain markedly enlarged nuclei and manifest both polyploidy and multiple centrosomes [<xref ref-type="bibr" rid="B6">6</xref>]. Such defects are not apparent in <italic>Skp2</italic><sup>-/-</sup><italic>;p27</italic><sup>-/- </sup>mice, suggesting that they are largely the result of the abnormal accumulation of p27 in the <italic>Skp2</italic><sup>-/- </sup>animals [<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B19">19</xref>].</p><p>We noticed that the fertility of male <italic>Skp2</italic><sup>-/- </sup>mice was reduced. We now show that, unlike other organs of these animals, the testes exhibit massive apoptosis, resulting in the loss of gametes. These characteristics were not apparent in <italic>Skp2</italic><sup>-/-</sup><italic>;p27</italic><sup>-/- </sup>mice, suggesting that accumulation of p27 accounts for the defect in germ cell development in the <italic>Skp2</italic><sup>-/- </sup>animals. These results indicate that regulation of cell cycle progression by the Skp2-p27 pathway is critical for germ cell development in both males and females.</p></sec><sec><title>Results</title><sec><title>Reduced male fertility of Skp2-deficient mice</title><p>The effect of Skp2 deficiency on the male fertility was examined separately by crossing with wild-type C57BL/6 mice. Mating of <italic>Skp2</italic><sup>-/- </sup>male mice with wild type pairs revealed a marked decrease in fertility, as evidenced by reduced litter size, compared with that of <italic>Skp2</italic><sup>+/+ </sup>pairs or <italic>Skp2</italic><sup>+/- </sup>pairs (Fig. <xref ref-type="fig" rid="F1">1A</xref>). The fertility of <italic>Skp2</italic>+/- pairs was also lower than that of wild-type pairs.</p><fig position="float" id="F1"><label>Figure 1</label><caption><p><bold>Fertility and germ cell production in Skp2-deficient male mice. </bold>(A) Fertility rate of male wild-type and mutant mice (2 to 4 months of age) as reflected by litter size when crossed with control C57BL/6 mice. Data are means ± SD for five animals per genotype. <italic>P </italic>< 0.05 versus <italic>Skp2</italic><sup>+/+</sup>, †<italic>P </italic>< 0.05 versus <italic>Skp2</italic><sup>+/-</sup>. (B) Germ cell production in male wild-type and mutant mice. Epididymal sperm in 2-month-old mice were counted. Data are means ± SD for 8–12 mice per genotype. <italic>P </italic>< 0.05 versus <italic>Skp2</italic><sup>+/+</sup>, †<italic>P </italic>< 0.05 versus <italic>Skp2</italic><sup>+/-</sup>.</p></caption><graphic xlink:href="1747-1028-1-4-1"/></fig><p>Epididymal spermatozoa were enumerated for evaluation of germ cell production in adult animals. Male <italic>Skp2</italic><sup>-/- </sup>mice exhibited a markedly reduced number of spermatozoa (azoospermia in some animals) compared with wild-type or heterozygous males (Fig. <xref ref-type="fig" rid="F1">1B</xref>). Furthermore, a substantial number of abnormally large cells and degenerative bodies was apparent in the semen of <italic>Skp2</italic><sup>-/- </sup>males (see Fig. <xref ref-type="fig" rid="F4">4D</xref>). The number of motile spermatzoa achieved >50% of the value for wild-type males in only one of 12<italic>Skp2</italic><sup>-/- </sup>males examined.</p></sec><sec><title>Testicular atrophy in Skp2-deficient mice</title><p>The testes and accessory reproductive organs of <italic>Skp2</italic><sup>+/+</sup>, <italic>Skp2</italic><sup>+/-</sup>, and <italic>Skp2</italic><sup>-/- </sup>mice were further examined for histopathologic abnormalities that might contribute to the reduced fertility of the mutant animals. The testes of <italic>Skp2</italic><sup>-/- </sup>males were markedly hypotrophic and hypoplastic compared with those of <italic>Skp2</italic><sup>+/+ </sup>or <italic>Skp2</italic><sup>+/- </sup>animals (Fig. <xref ref-type="fig" rid="F2">2A</xref>). Given that the body size of <italic>Skp2</italic><sup>-/- </sup>mice is smaller than that of wild-type mice, we normalized testis weight by body weight for each animal; the normalized testis weight was still greatly reduced for <italic>Skp2</italic><sup>-/- </sup>mice compared with that for <italic>Skp2</italic><sup>+/+ </sup>or <italic>Skp2</italic><sup>+/- </sup>animals (Fig. <xref ref-type="fig" rid="F2">2B</xref>). No gross morphological abnormalities were apparent in accessory reproductive glands of <italic>Skp2</italic><sup>-/- </sup>males (data not shown).</p><p>Light and electron microscopic examination of the testes of <italic>Skp2</italic><sup>-/- </sup>mice revealed a progressive loss of spermatogonia with age, resulting in a marked depletion of germ cells in most seminiferous tubules (Fig. <xref ref-type="fig" rid="F2">2E, F</xref>) in comparison with wild-type testes (Fig. <xref ref-type="fig" rid="F2">2C, D</xref>). In advanced stages of germ cell loss, postmeiotic cells were rarely observed in seminiferous tubules of <italic>Skp2</italic><sup>-/- </sup>males, with only Sertoli cells remaining. This phenotype is similar to that of Sertoli cell-only syndrome in humans. Ultrastructural analysis revealed the presence of numerous vacuolated Sertoli cells, resulting from germ cell depletion, in the tubule epithelium of such <italic>Skp2</italic><sup>-/- </sup>males (Fig. <xref ref-type="fig" rid="F2">2F</xref>). Degenerating spermatocytes and clusters of round or elongated spermatids had detached from the seminiferous epithelium and been sloughed off into the tubule lumen (Fig. <xref ref-type="fig" rid="F2">2E</xref>). Multinucleated giant spermatogenic cells, containing from two to five or more nuclei, were also present throughout the seminiferous epithelium (Fig. <xref ref-type="fig" rid="F2">2E</xref>). The number and morphology of Sertoli cells in <italic>Skp2</italic><sup>-/- </sup>mice were otherwise similar to those in wild-type animals, whereas interstitial cellularity was increased, especially for Leydig cells, in the mutant males (Fig. <xref ref-type="fig" rid="F2">2E</xref>). The profile of spermatogenesis in most seminiferous tubules of <italic>Skp2</italic><sup>+/- </sup>mice was similar to that in wild-type animals, although detachment of spermatogenic cells was apparent in a few tubules of the heterozygotes (data not shown).</p></sec><sec><title>Increased apoptosis in the gonads of Skp2-deficient mice</title><p>Cultured embryonic fibroblasts derived from <italic>Skp2</italic><sup>-/- </sup>mice manifest an increased prevalence of apoptosis compared with those derived from wild-type animals [<xref ref-type="bibr" rid="B6">6</xref>]. We therefore performed the TUNEL assay to determine whether apoptosis contributes to the progressive loss of germ cells in <italic>Skp2</italic><sup>-/- </sup>mice. Only a few spermatogenic cells of wild-type males, usually those at the spermatogonial stage, were found to be apoptotic (Fig. <xref ref-type="fig" rid="F3">3A</xref>). In contrast, a large proportion of spermatogenic cells, at almost all stages of spermatogenesis, was apoptotic in <italic>Skp2</italic><sup>-/- </sup>males (Fig. <xref ref-type="fig" rid="F3">3B</xref>). Statistical analysis confirmed that the ratio of apoptotic cells was significantly increased in <italic>Skp2</italic><sup>-/- </sup>mice (Fig. <xref ref-type="fig" rid="F3">3C</xref>). Electron microscopy also revealed the marked increase in the prevalence of apoptosis among spermatogenic cells, including those at the postmeiotic stage, in <italic>Skp2</italic><sup>-/- </sup>males (Fig. <xref ref-type="fig" rid="F3">3D</xref>). The level of apoptosis among supporting cells, including Sertoli, Leydig, and other interstitial cells, did not differ substantially between <italic>Skp2</italic><sup>-/- </sup>and wild-type males.</p></sec><sec><title>Polyploidy and cellular dysmorphism of germ cells of Skp2-deficient mice</title><p>Isolated seminal cells were subjected to analysis of DNA content. Seminal cells from wild-type mice contained a majority of haploid (1<italic>n</italic>) cells, corresponding to normal mature spermatozoa (Fig. <xref ref-type="fig" rid="F4">4A</xref>). However, three out of seven (43%) <italic>Skp2</italic><sup>-/- </sup>males examined manifested a substantial number of polyploid cells in their semen (Fig. <xref ref-type="fig" rid="F4">4C</xref>); the proportion of haploid cells was thus reduced and that of diploid (2<italic>n</italic>), tetraploid (4<italic>n</italic>), and >4<italic>n </italic>cells was increased. Examination of the nuclear morphology of seminal cells by propidium iodide staining revealed a majority of laterally flattened, typically curved, falciform, hock-headed spermatozoa in wild-type males (Fig. <xref ref-type="fig" rid="F4">4B</xref>). In contrast, the seminal cells of <italic>Skp2</italic><sup>-/- </sup>males, in addition to some normally shaped haploid spermatozoa, contained a large number of cells with an abnormal nuclear size and shape (Fig. <xref ref-type="fig" rid="F4">4D</xref>).</p></sec><sec><title>Effects of p27 accumulation in Skp2-defcient mice</title><p>To examine the possible role of p27 in the impaired fertility of <italic>Skp2</italic><sup>-/- </sup>mice, we analyzed <italic>Skp2</italic><sup>-/-</sup><italic>;p27</italic><sup>-/- </sup>double-mutant animals. The gonads of the double-mutant mice showed a general reversal of the marked hypoplasia apparent in parental <italic>Skp2</italic><sup>-/- </sup>mice (Fig. <xref ref-type="fig" rid="F5">5A</xref>). Disruption of the <italic>p27 </italic>gene thus largely restored the pool of germ cells in both testis (Fig. <xref ref-type="fig" rid="F5">5B</xref>). A small number of germ cell-depleted seminiferous tubules was still apparent in double-mutant males, however. The fertility rate of double-mutant males (n = 6) when crossed with C57BL/6 females was also increased to 5.66 ± 1.03 pups per litter (see Fig. <xref ref-type="fig" rid="F1">1A</xref>). The double-mutant females were completely sterile, however, which is a characteristic of p27 deficiency [<xref ref-type="bibr" rid="B20">20</xref>].</p><p>The normalization of the reproductive systems of the double-mutant mice was apparent from early stages of germ cell development, especially in male animals. Although there was a substantial reduction in the number of gonocyte nests in the testes of <italic>Skp2</italic><sup>-/- </sup>embryos (Fig. <xref ref-type="fig" rid="F5">5C</xref>), the gonocyte reserves in double-mutant embryos (Fig. <xref ref-type="fig" rid="F5">5E</xref>) were similar to those in wild-type embryos (Fig. <xref ref-type="fig" rid="F5">5D</xref>). In addition, the amount of p27 in the embryonic testis at the early stage of germ cell development (15.5 days postcoitum) was markedly increased in <italic>Skp2</italic><sup>-/- </sup>mice compared with wild-type animals (Fig. <xref ref-type="fig" rid="F5">5F</xref>).</p></sec></sec><sec><title>Discussion</title><p>We have shown that male <italic>Skp2</italic><sup>-/- </sup>mice manifest a markedly reduced fertility. The gonads of male mutant mice exhibited a pronounced hypoplasia that was independent of the reduced body size of these animals. This gonadal hypoplasia was likely attributable to the lack of Skp2-dependent degradation of p27 by the proteasome [<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B19">19</xref>]. Indeed, we have now shown that p27 accumulates in the testis of <italic>Skp2</italic><sup>-/- </sup>embryos at the early stage of gonocyte development. An important function of p27 is to restrain progression of the cell cycle, and this protein accumulates in response to many antiproliferative signals [<xref ref-type="bibr" rid="B21">21</xref>]. We and others previously showed that disruption of the <italic>p27 </italic>gene removes this brake and results in uncontrolled cell proliferation and hyperplasia of most organs, especially the testis and ovary [<xref ref-type="bibr" rid="B20">20</xref>,<xref ref-type="bibr" rid="B22">22</xref>,<xref ref-type="bibr" rid="B23">23</xref>]. In contrast, the accumulation of p27 in <italic>Skp2</italic><sup>-/- </sup>mice likely exerts an antiproliferative effect on embryonic germ cells, resulting in gonadal hypoplasia. This notion was supported by the observation that germ cell production was restored from an early stage of gonocyte development in <italic>Skp2</italic><sup>-/-</sup><italic>;p27</italic><sup>-/- </sup>double-mutant mice.</p><p>The marked depletion of spermatogenic cells apparent in adult <italic>Skp2</italic><sup>-/- </sup>males is reminiscent of that in infertile men with Sertoli cell-only syndrome [<xref ref-type="bibr" rid="B24">24</xref>]. The mutant male mice manifest hyperplasia of Leydig and other interstitial cells adjacent to the germ cell-depleted seminiferous tubules. A similar phenotype has also been described in human testicular disorders characterized by germ cell depletion and is thought to be a histological marker of testicular failure in men [<xref ref-type="bibr" rid="B25">25</xref>]. The accelerated depletion of germ cells in adult <italic>Skp2</italic><sup>-/- </sup>mice appears to be mediated by apoptosis, which is responsible for the removal of cells with abnormalities, such as an incorrect DNA content [<xref ref-type="bibr" rid="B26">26</xref>,<xref ref-type="bibr" rid="B27">27</xref>]. Such physiological apoptosis is a normal feature of both testis [<xref ref-type="bibr" rid="B28">28</xref>] and ovary, in the latter of which atretic follicles are prevented by apoptosis from achieving the final steps of follicular growth [<xref ref-type="bibr" rid="B28">28</xref>-<xref ref-type="bibr" rid="B30">30</xref>]. Adult <italic>Skp2</italic><sup>-/- </sup>mice manifested a greatly increased frequency of apoptosis during gametogenesis, however. In the testis, cells at various stages of spermatogenesis were found to be apoptotic, resulting in germ cell depletion.</p><p>The increased level of apoptosis apparent in the gonads of <italic>Skp2</italic><sup>-/- </sup>mice might be attributable to polyploidy, which was previously detected in somatic organs, including the liver, kidneys, and lungs, of these animals without any evident accompanying functional defects [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B14">14</xref>]. In general, abnormalities of nuclear DNA content during gametogenesis disturb the development of any resulting zygote, leading to infertility or embryo loss due to aneuploidy [<xref ref-type="bibr" rid="B31">31</xref>,<xref ref-type="bibr" rid="B32">32</xref>]. We now show that both germ cells and their supporting cells of <italic>Skp2</italic><sup>-/- </sup>mice are affected by polyploidy. In male mutant animals, polyploidy of spermatozoa likely contributed to the morphological abnormalities of these cells, given that morphologically abnormal (such as macrocephalic) spermatozoa have been shown to be polyploid both in mice [<xref ref-type="bibr" rid="B33">33</xref>] and in infertile men [<xref ref-type="bibr" rid="B34">34</xref>-<xref ref-type="bibr" rid="B38">38</xref>]. Polyploidy and consequent apoptosis are also likely to be responsible for detachment of spermatogonia and spermatocytes from the seminiferous epithelium and their formation of giant multinucleated structures in the tubules of <italic>Skp2</italic><sup>-/- </sup>males. Similar structures have been shown to represent syncytia of degenerating spermatids [<xref ref-type="bibr" rid="B39">39</xref>-<xref ref-type="bibr" rid="B42">42</xref>]. Aneuploidy has also been proposed to underlie the induction of spermatogenic cell apoptosis and the development of Sertoli cell-only syndrome in men [<xref ref-type="bibr" rid="B43">43</xref>].</p><p>The number of ova recoverable from <italic>Skp2</italic><sup>-/- </sup>female mice was too small to determine the presence of aneuploidy (data not shown). However, the high prevalence of apoptosis among ovarian granulosa cells was accompanied by an increased frequency of polyploidy in the mutant animals. Given the important role of granulosa cells in maintaining the female germ cells [<xref ref-type="bibr" rid="B44">44</xref>], follicles with apoptotic granulosa cells likely fail to progress to later stages of development and eventually undergo follicular atresia.</p><p>Impaired progression of the cell cycle caused by the absence of Skp2, an important controller of S phase, is likely responsible for the development of polyploidy in <italic>Skp2</italic><sup>-/- </sup>mice. Various chemical agents such as colchicines and vinblastine similarly affect both mitotic and meiotic cell division [<xref ref-type="bibr" rid="B45">45</xref>] and induce cell cycle delays in both somatic [<xref ref-type="bibr" rid="B46">46</xref>] and germ [<xref ref-type="bibr" rid="B47">47</xref>] cells, resulting in aneuploidy. Some of these chemicals delay progression of S or G<sub>2 </sub>phases of the cell cycle and thereby prolong cell cycle time [<xref ref-type="bibr" rid="B48">48</xref>,<xref ref-type="bibr" rid="B49">49</xref>]. Given that the abundance of Skp2 is maximal during S and G<sub>2 </sub>phases [<xref ref-type="bibr" rid="B12">12</xref>-<xref ref-type="bibr" rid="B14">14</xref>], it might be expected that its absence in these phases would lead to aneuploidy through a similar mechanism. Aneuploidy may also develop in the germ cells of <italic>Skp2</italic><sup>-/- </sup>mice by a mechanism similar to that operative in somatic cells of these animals; that is, endoreplication caused by the accumulation of p27 [<xref ref-type="bibr" rid="B14">14</xref>]. Indeed, the lack of p27 degradation during G<sub>2 </sub>phase in S<italic>kp2</italic><sup>-/- </sup>cells may result in suppression of Cdc2 activity and consequent inhibition of entry into M phase [<xref ref-type="bibr" rid="B14">14</xref>].</p><p>p27 antagonizes the activity of not only Cdc2, but also Cdk2 and Cdk4. Interestingly, mutant mice that lack the activity of Cdk2 or Cdk4 also exhibit the abnormalities in the gonadal development. <italic>Cdk2</italic><sup>-/- </sup>males and females are sterile with a severe atrophy of the gonads [<xref ref-type="bibr" rid="B50">50</xref>,<xref ref-type="bibr" rid="B51">51</xref>]. Cyclin E2-deficient males displayed reduced fertility, with approximately 50% of males being sterile [<xref ref-type="bibr" rid="B52">52</xref>]. <italic>Cyclin E2</italic><sup>-/- </sup>males displayed reduced testicular size and greatly reduced sperm counts, as compared with wild-type littermates. Cyclin E2-deficient females develop normally and are fully fertile. <italic>Cdk4</italic><sup>-/- </sup>mice showed reproductive dysfunction associated with hypoplastic seminiferous tubules in the testis and perturbed corpus luteum formation in the ovary [<xref ref-type="bibr" rid="B53">53</xref>,<xref ref-type="bibr" rid="B54">54</xref>]. Furthermore, cyclin D2-deficient females are sterile owing to the inability of ovarian granulosa cells to proliferate normally in response to follicle-stimulating hormone (FSH), whereas mutant males display hypoplastic testes [<xref ref-type="bibr" rid="B55">55</xref>]. Overall, these data suggest that Skp2-dependent control of p27 abundance plays a critical role in the regulation of the activity of Cdk2 and Cdk4, which is particularly important for the normal gonadal development.</p></sec><sec><title>Conclusion</title><p>Our results suggest that Skp2, as an important regulator of S phase of the cell cycle, plays a key role in establishment of the original pool of gametic cells by mitosis during early embryogenesis as well as in the proliferation and maturation of these cells at later stages of development. Our findings reinforce the importance both of cell cycle regulators in germ cell development and of impaired function of such factors in fertility problems.</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Animals</title><p>Skp2-deficient mice were generated by homologous recombination in embryonic stem cells as described previously [<xref ref-type="bibr" rid="B6">6</xref>]. Sexually mature mice from 2 to 12 months of age were used for experimental procedures. Pairs of <italic>Skp2</italic><sup>+/-</sup><italic>;p27</italic><sup>+/- </sup>mice were mated to produce <italic>Skp2</italic><sup>-/-</sup><italic>;p27</italic><sup>-/- </sup>animals [<xref ref-type="bibr" rid="B14">14</xref>]. All animal studies conformed with the Kyushu University Animal Experimentation Act. For evaluation of fertility, pairs of <italic>Skp2</italic><sup>+/+</sup>, <italic>Skp2</italic><sup>+/-</sup>, or <italic>Skp2</italic><sup>-/- </sup>mice were housed together for 8 weeks and then separated. Cages were monitored daily for the presence of seminal plugs, and the number and size of litters were recorded.</p></sec><sec><title>Histopathology</title><p>For light microscopic analysis, tissue samples were fixed in either Bouin's fixative or 4% paraformaldehyde, dehydrated, treated with xylene, embedded in paraffin, and sectioned at a thickness of 5 μm. After removal of paraffin, the sections were dehydrated, rehydrated, and stained with hematoxylin-eosin. They were then examined under a Nikon Eclipse E800 microscope with either Nomarski or phase-contrast optics; images were photographed with a Hamamatsu 3CCD digital camera (model 7780). For electron microscopy, samples were fixed in glutaraldehyde immediately after resection and were then embedded in resin. Sections with a thickness of 80 nm were collected on copper grids and counterstained with lead citrate and uranyl acetate. They were then observed with a JEOL (JEM 2000) electron microscope at a voltage of 80 kV.</p></sec><sec><title>TUNEL assay</title><p>For detection of apoptotic cells, tissue fixed in 4% paraformaldehyde or Bouin's solution was processed for the TUNEL (terminal deoxyribonucleotidyl transferase-mediated dUTP-biotin nick-end labeling)<sup>1 </sup>assay essentially as described previously [<xref ref-type="bibr" rid="B56">56</xref>]. Apoptotic cells were visualized with the chromogen 3,3'-diaminobenzidine tetrahydrochloride (Sigma, St. Louis, MO).</p></sec><sec><title>Isolation of spermatozoa</title><p>Mature male mice were killed by neck dislocation and the vas deferens and caudal epididymis were immediately exposed through a lower abdominal incision, excised, and washed briefly in phosphate-buffered saline (PBS). The tissue was then incubated in PBS for 30 min at 37°C and briefly massaged to promote the active exit of sperm from the epididymis. The released spermatozoa were isolated by centrifugation at 200 × <italic>g </italic>for 20 min, washed three times with PBS, counted with a hemocytometer attached to a light microscope, and analyzed for DNA content and morphology.</p></sec><sec><title>Flow cytometric analysis of DNA content</title><p>Isolated spermatozoa or granulosa cells (1 × 10<sup>6</sup>) were washed in PBS, resuspended in 200 μl of PBS, and fixed by the gradual addition of 800 μl of 100% ice-cold ethanol. The fixed cells were washed twice with PBS, resuspended in 1 ml of PBS, treated with RNase (40 μg/ml) for 10 min at room temperature, and then stained with propidium iodide (25 μg/ml). The DNA content of the cells was determined by flow cytometry with a FACSCaliber instrument and CellQuest software (Becton Dickinson, San Jose, CA). The morphology of propidium iodide-stained spermatozoa was also examined with a fluorescence microscope.</p></sec><sec><title>Immunoblot analysis</title><p>Embryonic testes were excised under a dissecting microscope, homogenized, and lysed in the presence of protease inhibitors. The tissue lysate (30 μg of protein) was fractionated by SDS-polyacrylamide gel electrophoresis, and the separated proteins were transferred to a nitrocellulose membrane and subjected to immunoblot analysis with rabbit polyclonal antibodies to mouse p27 (Santa Cruz Biotechnology, Santa Cruz, CA) and horseradish peroxidase-conjugated secondary antibodies. Immune complexes were detected with enhanced chemiluminescence reagents (Amersham Biosciences, Little Chalfont, UK). Immunodetection of β-actin was performed as a control for protein loading.</p></sec><sec><title>Statistical analysis</title><p>Data are presented as means ± SD. The significance of differences between means was determined by ANOVA test. A <italic>P </italic>value of <0.05 was considered statistically significant.</p></sec></sec><sec><title>Abbreviations</title><p>TUNEL: terminal deoxyribonucleotidyl transferase-mediated dUTP-biotin nick-end labeling, PBS: phosphate-buffered saline</p></sec><sec><title>Authors' contributions</title><p>FA performed all analyses of germ cell development. KN and KIN generated <italic>p27</italic><sup>-/-</sup>, <italic>Skp2</italic><sup>-/-</sup>, and <italic>Skp2</italic><sup>-/-</sup><italic>;p27</italic><sup>-/- </sup>mice. KIN is the principal investigator who gave advice in designing the study and edited the manuscript. All authors read and approved the final manuscript.</p><fig position="float" id="F2"><label>Figure 2</label><caption><p><bold>Testicular morphology of Skp2-deficient mice. </bold>(A) Macroscopic comparison of the testes of <italic>Skp2</italic><sup>+/+</sup>, <italic>Skp2</italic><sup>+/-</sup>, and <italic>Skp2</italic><sup>-/- </sup>mice at 2 months of age. Scale bar, 1 mm. <italic>B</italic>, Ratio of the weight of testes to body weight in male mice of the three genotypes at 2 months of age. Data are means ± SD for 6–8 animals of each genotype. <italic>P </italic>< 0.05 versus <italic>Skp2</italic><sup>+/+</sup>, †<italic>P </italic>< 0.05 versus <italic>Skp2</italic><sup>+/-</sup>. (C-F) Representative light (C, E) and electron (D, F) micrographs of testicular sections of wild-type (C, D) and Skp2<sup>-/- </sup>(E, F) males at 2–4 months of age. Note the pronounced loss of spermatogenic cells (asterisks in F), leaving only Sertoli cells, and the hyperplasia of the interstitial cellular population (asterisks in E) apparent adjacent to severely degenerated tubules in <italic>Skp2</italic><sup>-/- </sup>testis. Degenerating spermatocytes and clusters of round or elongated spermatids had detached from the seminiferous epithelium and been sloughed off into the tubule lumen of <italic>Skp2</italic><sup>-/- </sup>males (open arrowheads in E). Multinucleated giant spermatogenic cells were also present throughout the seminiferous epithelium of <italic>Skp2</italic><sup>-/- </sup>testis (closed arrowheads in E). Scale bars, 100 μm (C, E) or 5 μm (D, F).</p></caption><graphic xlink:href="1747-1028-1-4-2"/></fig><fig position="float" id="F3"><label>Figure 3</label><caption><p><bold>Increased level of apoptosis in the gonads of <italic>Skp2</italic><sup>-/- </sup>mice. </bold>(A, B) TUNEL staining of testicular sections of <italic>Skp2</italic><sup>+/+ </sup>(A) or <italic>Skp2</italic><sup>-/- </sup>(B) mice at 4 months of age. Arrowheads indicate apoptotic cells. Scale bars, 100 μm. (C) The number of apoptotic spermatogenic cells per 100 Sertoli cells. <italic>P </italic>< 0.05 versus <italic>Skp2</italic><sup>+/+</sup>. (D) Ultrastructural image of typical apoptotic figures (arrowheads) at late postmeiotic stages of spermatogenesis in a <italic>Skp2</italic><sup>-/- </sup>mouse at 4 months of age. Scale bar, 5 μm.</p></caption><graphic xlink:href="1747-1028-1-4-3"/></fig><fig position="float" id="F4"><label>Figure 4</label><caption><p><bold>Polyploidy of epididymal sperm and ovarian granulosa cells of <italic>Skp2</italic><sup>-/- </sup>mice. </bold>(A, C) Flow cytometric analysis of the DNA content of seminal cells isolated from wild-type (A) and <italic>Skp2</italic><sup>-/- </sup>(C) males at 4 months of age. (B, D) Nuclear morphology of seminal cells isolated from wild-type (B) and <italic>Skp2</italic><sup>-/- </sup>(D) mice and stained with propidium iodide. Scale bars, 50 μm.</p></caption><graphic xlink:href="1747-1028-1-4-4"/></fig><fig position="float" id="F5"><label>Figure 5</label><caption><p><bold>Role of p27 accumulation in the gonadal hypoplasia of Skp2-deficient mice. </bold>(A) Macroscopic comparison of the testes of wild-type, <italic>Skp2</italic><sup>-/-</sup>, <italic>p27</italic><sup>-/-</sup>, and <italic>Skp2</italic><sup>-/-</sup><italic>;p27</italic><sup>-/- </sup>mice. (B) Histology of the testis of <italic>Skp2</italic><sup>-/-</sup><italic>;p27</italic><sup>-/- </sup>mice (4 months of age) showing pronounced recovery of germ cell production compared with that apparent in parental <italic>Skp2</italic><sup>-/- </sup>mice. Scale bars, 100 μm. Arrowheads in B indicate seminiferous tubules with a deficiency of germ cells. (C-E) Histology of embryonic testis (15.5 days postcoitum). Severe deficiency of gonocytes (arrowheads) was evident in <italic>Skp2</italic><sup>-/- </sup>embryos (C) compared with wild-type embryos (D). However, the gonocyte population had recovered substantially in <italic>Skp2</italic><sup>-/-</sup>;<italic>p27</italic><sup>-/- </sup>embryos (E). Scale bars, 100 μm. (F) Immunoblot analysis of lysates of the testes of <italic>Skp2</italic><sup>-/- </sup>or wild-type embryos (15.5 days postcoitum) with antibodies to p27 and to β-actin (loading control).</p></caption><graphic xlink:href="1747-1028-1-4-5"/></fig></sec>
Neuroactive substances specifically modulate rhythmic body contractions in the nerveless metazoon <italic>Tethya wilhelma </italic>(Demospongiae, Porifera)	<sec><title>Background</title><p>Sponges (Porifera) are nerve- and muscleless metazoa, but display coordinated motor reactions. Therefore, they represent a valuable phylum to investigate coordination systems, which evolved in a hypothetical Urmetazoon prior to the central nervous system (CNS) of later metazoa. We have chosen the contractile and locomotive species <italic>Tethya wilhelma </italic>(Demospongiae, Hadromerida) as a model system for our research, using quantitative analysis based on digital time lapse imaging. In order to evaluate candidate coordination pathways, we extracorporeally tested a number of chemical messengers, agonists and antagonists known from chemical signalling pathways in animals with CNS.</p></sec><sec><title>Results</title><p>Sponge body contraction of <italic>T. wilhelma </italic>was induced by caffeine, glycine, serotonine, nitric oxide (NO) and extracellular cyclic adenosine monophosphate (cAMP). The induction by glycine and cAMP followed patterns varying from other substances. Induction by cAMP was delayed, while glycine lead to a bi-phasic contraction response. The frequency of the endogenous contraction rhythm of <italic>T. wilhelma </italic>was significantly decreased by adrenaline and NO, with the same tendency for cAMP and acetylcholine. In contrast, caffeine and glycine increased the contraction frequency. The endogenous rhythm appeared irregular during application of caffeine, adrenaline, NO and cAMP. Caffeine, glycine and NO attenuated the contraction amplitude. All effects on the endogenous rhythm were neutralised by the washout of the substances from the experimental reactor system.</p></sec><sec><title>Conclusion</title><p>Our study demonstrates that a number of chemical messengers, agonists and antagonists induce contraction and/or modulate the endogenous contraction rhythm and amplitude of our nerveless model metazoon <italic>T. wilhelma</italic>. We conclude that a relatively complex system of chemical messengers regulates the contraction behaviour through auto- and paracrine signalling, which is presented in a hypothetical model. We assume that adrenergic, adenosynergic and glycinergic pathways, as well as pathways based on NO and extracellular cAMP are candidates for the regulation and timing of the endogenous contraction rhythm within pacemaker cells, while GABA, glutamate and serotonine are candidates for the direct coordination of the contractile cells.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>Ellwanger</surname><given-names>Kornelia</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" corresp="yes" contrib-type="author"><name><surname>Nickel</surname><given-names>Michael</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	Frontiers in Zoology	<sec><title>Background</title><p>Sponges are nerveless and muscleless multicellular animals, which emerged early during the evolution of the metazoa [<xref ref-type="bibr" rid="B1">1</xref>]. Therefore, sponges represent valuable model systems to conclude upon a hypothetical Urmetazoa, from which the metazoan bodyplan evolved [<xref ref-type="bibr" rid="B2">2</xref>-<xref ref-type="bibr" rid="B4">4</xref>]. Despite the fact, that sponges do not posses muscles and a nervous system, they are able to react upon external stimuli [<xref ref-type="bibr" rid="B5">5</xref>-<xref ref-type="bibr" rid="B7">7</xref>], to move [<xref ref-type="bibr" rid="B7">7</xref>-<xref ref-type="bibr" rid="B10">10</xref>], to contract [summarised in 7], and display diurnal rhythms [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B11">11</xref>]. The earliest description of such sponge behaviour dates back to Aristotle, who mentioned contraction of living sponges, when they were touched and collected by humans for the production of bath sponges [<xref ref-type="bibr" rid="B12">12</xref>].</p><p>The putative coordination mechanisms have been discussed in details over the last century, with a main focus on the topic in the 1960s and 1970s [<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B13">13</xref>-<xref ref-type="bibr" rid="B15">15</xref>], which lead to the general acceptance that sponges do not possess a nervous system. However, questions for nervous-system like coordination mechanisms have been raised again from time to time due to new findings, as action potentials in the Hexactinellida [<xref ref-type="bibr" rid="B16">16</xref>-<xref ref-type="bibr" rid="B18">18</xref>] or the characterisation of a GABA/glutamate-like receptor from <italic>Geodia cydonium </italic>[<xref ref-type="bibr" rid="B19">19</xref>].</p><p>Jones [<xref ref-type="bibr" rid="B8">8</xref>] presented one of the most comprehensive discussions of coordination systems in sponges. As alternative hypothetical signal conduction systems in sponges he named: (a) local fall of pressure inside the canal system, which would be transmitted through the system to induce contraction in distant parts of the sponge; (b) mechanical conduction of contraction forces through the interconnected contractible sponge tissue (presumably the continuous pinacoderm); (c) a chemical messenger, distributed by the aquiferous system; (d) a chemical messenger diffusing through the mesohyle; (e) distribution of action potentials by intercellular junctions. Even though Jones seems to prefer the mechanical mechanism based hypotheses, he states that "clearly the mechanism cannot be a simple one" (p. 41). However, though mechanical conduction may explain the transmission of a physical information, it can not explain how endogenous rhythms are generated and coordinated. This is especially of importance concerning the diurnal patterns found in sponges [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B11">11</xref>]. In this context, chemical messenger based systems, acting via specific ligand-receptor interactions, seem to be more likely. In fact, a number of transmitters, chemical messengers, receptor agonists and antagonist have been tested on sponges. The first experiments date back to Lendenfeld at the End of the 19<sup>th </sup>century [<xref ref-type="bibr" rid="B20">20</xref>]. They were followed by others, partly with contradictory results, e.g. positive reports upon acetylcholine in sponges by some investigators [<xref ref-type="bibr" rid="B21">21</xref>,<xref ref-type="bibr" rid="B22">22</xref>] and negative reports by others [<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B23">23</xref>-<xref ref-type="bibr" rid="B25">25</xref>]. However, this inconsistencies of results can be partly explained by technical reasons and the fact that varying markers, sponge species and sponge reactions (contraction, internal current, state of the oscule) were regarded. In addition, some of these experiments suffered from the lack of precise quantitative measurement methods.</p><p>To overcome these problems, we have established a model system, based on the contractile and locomotive sponge <italic>Tethya wilhelma </italic>(Demospongiae, Hadromerida; Fig. <xref ref-type="fig" rid="F1">1</xref>), which has been described from an aquarium habitat recently and can be maintained and reared in the aquarium [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B26">26</xref>-<xref ref-type="bibr" rid="B30">30</xref>]. The use of digital time lapse imaging of the sponge body, in conjunction with quantitative image analysis, allows us to record and analyse the behaviour and reactions of the sponge quantitatively by measuring changes of the projected body area (for the principle see in Fig. <xref ref-type="fig" rid="F1">1</xref>). Using this methodology, one of us recently demonstrated that <italic>T. wilhelma </italic>displays a regular endogenous contraction rhythm, which is diurnal and can be disturbed by external events. This integrative behaviour characterises <italic>T. wilhelma </italic>as a valuable model system for research on coordination in basal metazoa [<xref ref-type="bibr" rid="B7">7</xref>]. Furthermore, we have shown that the contraction of T. wilhelma is inducible by chemical messengers and we characterised the specificity, dose response and kinetics of contraction for the neuroactive amino acids γ-amino butyric acid (GABA) and glutamate [<xref ref-type="bibr" rid="B31">31</xref>], giving evidence for the presence of a GABAergic coordination mechanism in sponges.</p><fig position="float" id="F1"><label>Figure 1</label><caption><p><bold>Expanding specimen of <italic>Tethya wilhelma </italic></bold>Overlay of a time-lapse series of outlines of the projected body area of <italic>T. wilhelma</italic>, representing an expansion from fully contracted state at t = 0 min to fully expanded state at t = 45 min. In addition, the 8-bit image of the contracted sponge body at t = 0 min is displayed on top of the overlay, to give an impression of the sponge and it's body extensions. Like the body outlines, the measurement of the projected area is computed automatically using the software ImageJ 57 and own macro functions. Outline and overlay were computed using ImageJ, too.</p></caption><graphic xlink:href="1742-9994-3-7-1"/></fig><fig position="float" id="F2"><label>Figure 2</label><caption><p><bold>Contraction patterns of <italic>Tethya wilhelma </italic>during extracorporeal acetylcholine and nicotine application </bold>(a - b): Projected area of two specimens of <italic>T. wilhelma </italic>before and during application (grey background) of acetylcholine (a) and nicotine (b) respectively. Neither of the two substances induced contraction. See additional file 1 (Movie_S1.mov) for the time-lapse movie of Fig. 2b. Filled arrowheads indicate the points in time when stock solutions of the substances were injected into the experimental reactor circulation system to reach final concentrations of 40 μM, 592 μM and 2.8 mM for acetylcholine and 25 μM and 50 μM for nicotine.</p></caption><graphic xlink:href="1742-9994-3-7-2"/></fig><p>In the present study we evaluated the effects of a number of substances upon our model sponge, by means of induction of contraction and interference with endogenous contraction rhythm and amplitude. We have chosen substances acting via ligand-receptor systems, which are known to be involved in the regulation of behaviours of metazoa with central nervous systems (CNS): acetylcholine and nicotine, caffeine, glycine, adrenaline, serotonine, nitric oxide and cyclic AMP. Most of these substances have been previously tested in various sponges under varying conditions and set-ups by various sponge researchers as shown above. In contrast, we used a semi-closed experimental reactor, under reproducible experimental conditions. The aim of the present study was to demonstrate that contraction, endogenous rhythm and strength of contraction are specifically modulated by several neuroactive chemical substances. Consequently, we aimed at adding more details to our hypothetical chemical messenger-based coordination model, presented before [<xref ref-type="bibr" rid="B31">31</xref>]. In addition, the present study should help us to identify ligand-receptor system candidates, which deserve more detailed physiological and molecular characterisation in <italic>T. wilhelma </italic>and eventually other sponge models.</p></sec><sec><title>Results</title><sec><title>Acetylcholine and nicotine</title><p>For acetylcholine, we tested concentrations between 40 μM and 2.8 mM (Fig. <xref ref-type="fig" rid="F1">1a</xref>), for nicotine, we tested concentrations from 25 μM to 50 μM (Fig <xref ref-type="fig" rid="F1">1b</xref>). Neither acetylcholine, nor nicotine induced contractions or altered the endogenous contraction rhythm or amplitude on first sight. In both cases, the endogenous contractions are rhythmical and equally strong, running as waves over the body of the sponge from the basal attachment area to the apical part of <italic>Tethya </italic>(see <xref ref-type="supplementary-material" rid="S1">Additional file 1</xref>: Movie_S1.mov). However, comparing contraction cycle durations (D<sub>cc</sub>) statistically, revealed a slightly but significantly prolonged cycle duration for acetylcholine exposure, whereas no significant difference was found for nicotine (Tab. <xref ref-type="table" rid="T1">1</xref>).</p></sec><sec><title>Caffeine</title><p>In all experiments, caffeine induced contractions at concentrations of 257 μM and 515 μM respectively. In contrast to the previously reported almost immediate induction by glutamate and GABA [<xref ref-type="bibr" rid="B31">31</xref>], the caffeine-induced contraction is delayed by several minutes in most cases (Fig <xref ref-type="fig" rid="F3">3a, b, d</xref>), though immediate onset may occur, too (Fig. <xref ref-type="fig" rid="F3">3b</xref>). In one case, we found a spasm-like reaction of <italic>T. wilhelma </italic>upon caffeine exposure. This spasm resembles very much the spasms, which can be induced by GABA [<xref ref-type="bibr" rid="B31">31</xref>]. After regular contractions and expansions the sponge is stimulated by caffeine, contracts and remains in an almost contracted state, with a fast progression of minor expansions and contractions (Fig. <xref ref-type="fig" rid="F3">3b</xref>). The time-lapse recording shows that the sponge tissue does not contract in a temporally coordinated way (see <xref ref-type="supplementary-material" rid="S2">Additional file 2</xref>: Movie_S2.mov). Local contractions and expansions of parts of the sponge tissue occur concurrently, shaking the sponge in a spasm-like manner.</p><fig position="float" id="F3"><label>Figure 3</label><caption><p><bold>Contraction patterns of <italic>Tethya wilhelma</italic>during extracorporeal caffeine application </bold>(a - d): Projected area of four specimens of <italic>T. wilhelma </italic>before, during (grey background) and after caffeine application. Caffeine induced contractions in all experiments. A spasm-like pattern representing an extremely prolonged phase in contracted state was observed once (b; see additional file 2: Movie_S2.mov for the time-lapse movie of Fig. 3b). Long-term application resulted in an irregular contraction pattern in some cases (c, d; see additional file 3: Movie_S3.mov for the time-lapse movie of Fig. 3c). Filled arrowheads indicate the points in time when stock solutions of the substances were injected into the experimental reactor circulation system to reach final concentrations of 515 μM (a - c) and 257 μM (d) respectively. Open arrowheads indicate the wash out of the substance from the experimental reactor system.</p></caption><graphic xlink:href="1742-9994-3-7-3"/></fig><p>In addition, caffeine triggered long-term effects by altering both, D<sub>cc </sub>and the strength of the overall contraction. The D<sub>cc </sub>is shortened significantly during long-term caffeine exposure (Tab. <xref ref-type="table" rid="T1">1</xref>), but the effect is suspended by washout of caffeine from the experimental reactor. The same applies for the alterations of the overall contraction strength during caffeine-exposure. The amplitude of the recorded changes in projected body area is attenuated (Fig. <xref ref-type="fig" rid="F3">3c, d</xref>). In the most prominent case <italic>T. wilhelma </italic>reached a semi-contracted state after several hours and displayed only weak contractions (Fig. <xref ref-type="fig" rid="F3">3c</xref>). The time-lapse recording reveals a behaviour of the sponge tissue similar to the spasm-like state, but the effect is weaker and the endogenous rhythm continues (see <xref ref-type="supplementary-material" rid="S3">Additional file 3</xref>: Movie_S3.mov).</p></sec><sec><title>Glycine</title><p>Like caffeine, glycine induced contraction and altered D<sub>cc </sub>and the overall strength of contraction. The induced contraction is characterised by a bi-phasic appearance (Fig. <xref ref-type="fig" rid="F4">4</xref>). Time-lapse recordings reveal that in the first phase, the sponge contracts only in the basal part of the body. Consequently, the sponge drops towards the substrate and the tissue of the whole body starts to contract quite simultaneously in the following second phase (see <xref ref-type="supplementary-material" rid="S4">Additional file 4</xref>: Movie_S4.mov, time index 69 – 90 min).</p><fig position="float" id="F4"><label>Figure 4</label><caption><p><bold>Contraction patterns of <italic>Tethya wilhelma </italic>during extracorporeal glycine application </bold>(a - b): Projected area of two specimens of <italic>T. wilhelma </italic>before and during glycine application (grey background). Glycine induced bi-phasic contractions in all experiments and attenuated the contraction amplitude (b; see additional file 4: Movie_S4.mov for the time-lapse movie of Fig. 4b). Filled arrowheads indicate the points in time when stock solutions of the substances were injected into the experimental reactor circulation system to reach final concentrations of 1.3 mM.</p></caption><graphic xlink:href="1742-9994-3-7-4"/></fig><p>The endogenous D<sub>cc </sub>is significantly shortened during glycine exposure (Tab. <xref ref-type="table" rid="T1">1</xref>), but restores after wash out of the substance. The amplitude of the recorded changes in projected body area may be attenuated (Fig. <xref ref-type="fig" rid="F4">4b</xref>). The attenuation differs from the one observed for caffeine, since the overall appearance of the contraction is very similar to the non-induced endogenous contractions, except that the volume changes during contraction and expansion are lower (see <xref ref-type="supplementary-material" rid="S4">Additional file 4</xref>: Movie_S4.mov, time index 90 – 378 min).</p></sec><sec><title>Adrenaline and serotonine</title><p>Adrenaline did not induce contraction in <italic>T. wilhelma</italic>. Nevertheless, the endogenous rhythm is strongly disturbed by adrenaline, leading to longer irregular D<sub>cc </sub>during exposure (Fig. <xref ref-type="fig" rid="F5">5a</xref>, Tab. 1). The time-lapse recording shows that <italic>T. wilhelma </italic>remains in the expanded phase over a longer period. No local contractions on the surface are observable. Elongation and retraction of body extensions as well as bud-formation are not affected by adrenaline (see <xref ref-type="supplementary-material" rid="S5">Additional file 5</xref>: Movie_S5.mov). We found that the effect on D<sub>cc </sub>was stronger at 44 μM than at 22 μM adrenaline (Fig. <xref ref-type="fig" rid="F5">5b</xref>).</p><fig position="float" id="F5"><label>Figure 5</label><caption><p><bold>Contraction patterns of <italic>Tethya wilhelma</italic>during extracorporeal adrenaline and serotonine application </bold>(a - d): Projected area of four specimens of <italic>T. wilhelma </italic>before, during (grey background) and after application of adrenaline (a - b) and serotonine (c - d) respectively. Adrenaline did not induce contractions, but resulted in prolonged phases of expansion (a, b; see additional file 5: Movie_S5.mov for the time-lapse movie of Fig. 5c). Serotonine induced contractions at higher concentrations (c), but not at lower concentrations (d). Filled arrowheads indicate the points in time when stock solutions of the substances were injected into the experimental reactor circulation system to reach final concentrations of 44 μM (a) and 22 μM (b) for adrenaline and 65 μM (d), 260 μM and 520 μM (c) for serotonine. Open arrowheads indicate the wash out of the substance from the experimental reactor system.</p></caption><graphic xlink:href="1742-9994-3-7-5"/></fig><p>In contrast, serotonine immediately induced contractions at concentrations of 260 μM and 520 μM, but not at 65 μM (Fig. <xref ref-type="fig" rid="F5">5c, d</xref>). However, we found no attenuation of the strength of contraction and no significant alteration of the D<sub>cc </sub>(Tab. <xref ref-type="table" rid="T1">1</xref>).</p></sec><sec><title>Nitric oxide</title><p>The NO-releasing substance NOC-12 was found to induce contractions instantly after application (Fig. <xref ref-type="fig" rid="F6">6</xref>). In addition, strong effects, both on D<sub>cc </sub>and the strength of contraction were recorded. The D<sub>cc </sub>is significantly increased and appears to be more irregular (Tab. <xref ref-type="table" rid="T1">1</xref>, Fig. <xref ref-type="fig" rid="F6">6b</xref>). The contraction strength is lowered, with lower volume at expanded phase and higher volume at contracted phase. During this semi-expanded periods, <italic>T. wilhelma </italic>displays almost no local contraction activity (see <xref ref-type="supplementary-material" rid="S6">Additional file 6</xref>: Movie_S6.mov). The contractions, which occur within the time of exposure to NO, are similar to non-induced endogenous contractions, except of the attenuated amplitude: they seem to be triggered locally and spread over the sponge body in a wave like manner.</p><fig position="float" id="F6"><label>Figure 6</label><caption><p><bold>Contraction patterns of <italic>Tethya wilhelma</italic>during extracorporeal nitric oxide application (via release by NOC-12) </bold>(a - b): Projected area of two specimens of <italic>T. wilhelma </italic>before and during NOC-12 application (grey background). The nitric oxide released by NOC-12 induced immediate contractions in all experiments, attenuated the contraction amplitude and disturbed the rhythm of endogenous contractions (see additional file 6: Movie_S6.mov for the time-lapse movie of Fig. 6b). Filled arrowheads indicate the points in time when stock solutions of the substances were injected into the experimental reactor circulation system to reach final concentrations of 5.7 μM (a) and 22.7 μM (b) for NOC-12.</p></caption><graphic xlink:href="1742-9994-3-7-6"/></fig></sec><sec><title>Cyclic AMP</title><p>The application of cAMP resulted in contractions, either with immediate onset (Fig. <xref ref-type="fig" rid="F7">7a</xref>) or delayed onset (Fig. <xref ref-type="fig" rid="F7">7b</xref>). In either case there is a tendency for an altered D<sub>cc </sub>(Tab. <xref ref-type="table" rid="T1">1</xref>) and the typical pattern was slightly disturbed, as displayed in one case by an immediate second contraction after the initial one, without a typical rest in the expanded state (Fig. <xref ref-type="fig" rid="F7">7b</xref>). Even though the rhythm is slightly disturbed, the overall appearance of the contractions is not altered during application of cAMP. Like in endogenous contractions, it seems to be triggered locally in the basal part of the sponge and to spread from there towards the apical part (see <xref ref-type="supplementary-material" rid="S6">Additional file 6</xref>: Movie_S6.mov).</p></sec><sec><title>General observations and control experiments</title><p>In general, most of the sponges responded to the washout of the substances, too. In some cases relatively strong contractions occurred (Fig. <xref ref-type="fig" rid="F5">5a</xref>; <xref ref-type="fig" rid="F6">6b</xref>; <xref ref-type="fig" rid="F7">7c</xref>), in others, <italic>T. wilhelma </italic>responded only slightly (Fig. <xref ref-type="fig" rid="F3">3d</xref>; <xref ref-type="fig" rid="F6">6a</xref>). In all cases, water temperature and other physical characters of the replacing artificial sea water were as close as possible to the conditions in the chamber. In almost all cases, the sponge expanded to a higher volume after a wash in comparison to the expanded volume during the experiment (Fig. <xref ref-type="fig" rid="F3">3c, d</xref>; <xref ref-type="fig" rid="F7">7a, b</xref>) and in some cases even higher than prior to the experiment (Fig. <xref ref-type="fig" rid="F5">5a</xref>; <xref ref-type="fig" rid="F6">6a, b</xref>). The control experiments show, that injection of small volumes of seawater alone into the system does not result in contractions. This is also reflected by those substances, which do not induce contraction, like acetylcholine and nicotine (Fig. <xref ref-type="fig" rid="F3">3</xref>) or adrenaline (Fig. <xref ref-type="fig" rid="F5">5a, b</xref>). The pH value of the seawater inside the experimental reactor is slightly altered by some of the substances. Differences of the pH (Δ pH) higher than ± 0.15 were observed only for cAMP (Δ pH = -1.16 ± 0.02; N = 3). An analogous pH change, imitated by applying a corresponding amount of HCl to <italic>T. wilhelma </italic>did not result in contractions (data not shown). Temperature changes within the range of ± 0.5°C, which occurred during the experiments did not affect the sponges in control experiments. The oxygen concentration was kept on a saturated level throughout all experiments, to exclude any influence.</p></sec></sec><sec><title>Discussion</title><sec><title>Effects of neuroactive substances on sponges</title><p>We applied substances, which are well-known to be neuroactive in metazoans with a CNS, to our nerveless model metazoon, the sponge <italic>Tethya wilhelma</italic>. The sponge responded by contractions as well as altered rhythm and/or attenuated contraction amplitude to a number of these substances. The results are summarized in Table <xref ref-type="table" rid="T2">2</xref> and compared to previous reports concerning the Porifera.</p><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>Statistical analysis of contraction cycle durations in <italic>Tethya wilhelma </italic>Results of a Mann-Whitney U-test on differences between endogenous and influenced contraction cycle durations (D<sub>cc</sub>) of <italic>T. wilhelma </italic>(each comparison within one experimental series).</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left"><bold>Substance</bold></td><td align="center"><bold>Concentration</bold></td><td align="center"><bold>Figure</bold></td><td align="center"><bold>D</bold><sub>cc </sub><bold>endogenous [min]</bold></td><td align="center"><bold>D</bold><sub>cc </sub><bold>substance-exposed [min]</bold></td><td align="center"><bold>U-test</bold></td><td align="center"><bold>significant difference in Dcc?</bold></td></tr></thead><tbody><tr><td align="left">Acetycholine</td><td align="center">2.8 mM</td><td align="center">2a</td><td align="center">78.3 ± 5.3 (N = 9)</td><td align="center">86.9 ± 2.3 (N = 9)</td><td align="center">p < 0.001</td><td align="center">yes</td></tr><tr><td align="left">Nicotine</td><td align="center">25 – 50 μM</td><td align="center">2b</td><td align="center">101.3 ± 8.6 (N = 4)</td><td align="center">111.3 ± 10.6 (N = 4)</td><td align="center">p > 0.01</td><td align="center">no</td></tr><tr><td align="left">Caffeine</td><td align="center">515 μM</td><td align="center">3a</td><td align="center">93.1 ± 7.6 (N = 6)</td><td align="center">85.3 ± 7.3 (N = 6)</td><td align="center">p > 0.01</td><td align="center">no</td></tr><tr><td></td><td align="center">515 μM</td><td align="center">3b</td><td align="center">105.0 ± 7.7 (N = 5)</td><td align="center">44.12 ± 21.0 (N = 5)</td><td align="center">p < 0.01</td><td align="center">yes</td></tr><tr><td></td><td align="center">515 μM</td><td align="center">3c</td><td align="center">85.5 ± 19.3 (N = 14)</td><td align="center">52.5 ± 17.2 (N = 18)</td><td align="center">p < 0.001</td><td align="center">yes</td></tr><tr><td></td><td align="center">257 μM</td><td align="center">3d</td><td align="center">113.3 ± 34.6 (N = 9)</td><td align="center">78.8 ± 23.5 (N =)</td><td align="center">p < 0.01</td><td align="center">yes</td></tr><tr><td align="left">Glycin</td><td align="center">1.3 mM</td><td align="center">4a</td><td align="center">105.5 ± 23.5 (N = 6)</td><td align="center">54.2 ± 9.1 (N = 14)</td><td align="center">p < 0.001</td><td align="center">yes</td></tr><tr><td></td><td align="center">1.3 mM</td><td align="center">4b</td><td align="center">93.0 ± 10.7 (N = 9)</td><td align="center">42.7 ± 5.5 (N = 10)</td><td align="center">p < 0.001</td><td align="center">yes</td></tr><tr><td align="left">Adrenaline</td><td align="center">44 μM</td><td align="center">5a</td><td align="center">88.7 ± 3.4 (N = 7)</td><td align="center">154.8 ± 58.7 (N = 4)</td><td align="center">p < 0.01</td><td align="center">yes</td></tr><tr><td></td><td align="center">22 μM</td><td align="center">5b</td><td align="center">88.5 ± 7.9 (N = 8)</td><td align="center">109.8 ± 6.0 (N = 2)</td><td align="center">n.d.<sup>a</sup></td><td align="center">n.d.<sup>a</sup></td></tr><tr><td align="left">Serotonine</td><td align="center">260 – 520 μM</td><td align="center">5d</td><td align="center">115.3 ± 14.0 (N = 5)</td><td align="center">96.3 ± 10.0 (N = 5)</td><td align="center">p > 0.01</td><td align="center">no</td></tr><tr><td></td><td align="center">65 μM</td><td align="center">5c</td><td align="center">105.0 ± 8.8 (N = 5)</td><td align="center">102.9 ± 2.8 (N = 4)</td><td align="center">p > 0.01</td><td align="center">no</td></tr><tr><td align="left">(by NOC-12)</td><td align="center">5.7 μM</td><td align="center">6a</td><td align="center">75.8 ± 13.0 (N = 8)</td><td align="center">116.0 ± 23.8 (N = 8)</td><td align="center">p < 0.01</td><td align="center">yes</td></tr><tr><td></td><td align="center">22.7 μM</td><td align="center">6b</td><td align="center">95.0 ± 9.4 (N = 6)</td><td align="center">238.0 ± 58.2 (N = 5)</td><td align="center">p < 0.01</td><td align="center">yes</td></tr><tr><td align="left">cAMP</td><td align="center">288 mM</td><td align="center">7a</td><td align="center">108.0 ± 11.8 (N = 4)</td><td align="center">83.3 ± 5.3 (N = 2)</td><td align="center">n.d.<sup>a</sup></td><td align="center">n.d.<sup>a</sup></td></tr><tr><td></td><td align="center">288 mM</td><td align="center">7b</td><td align="center">105.5 ± 23.5 (N = 6)</td><td align="center">73.5 ± 3.5 (N = 2)</td><td align="center">n.d.<sup>a</sup></td><td align="center">n.d.<sup>a</sup></td></tr></tbody></table><table-wrap-foot><p><sup>a </sup>not determined, not enough data available</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T2"><label>Table 2</label><caption><p>Summary of induced effects on <italic>Tethya wilhelma</italic>Overview regarding induced contraction, endogenous contraction rhythm alteration and/or amplitude attenuation by substances used in the present and previous study, including effects on sponges, reported by other authors before.</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left"><bold>Substance</bold></td><td align="center"><bold>Contraction induced</bold></td><td align="center"><bold>Rhythm altered</bold></td><td align="center"><bold>Amplitude altered</bold></td><td align="left"><bold>Signalling pathway known to be influenced</bold></td><td align="left"><bold>Other reports on sponges</bold></td></tr></thead><tbody><tr><td align="left">Acetycholine</td><td align="center">-</td><td align="center">slightly slower rhythm</td><td align="center">-</td><td align="left">ionotrophic or metabotrophic acetylcholine receptor</td><td align="left">Increased rhythm and intensity of contraction in <italic>Euspongia officinalis </italic>[22], no effect in <italic>Cliona celata </italic>[5]. Acetylcholine esterase activity in <italic>Sycon </italic>sp. 21, <italic>Hippospongia </italic>[35] and <italic>T. wilhelma </italic>[26]</td></tr><tr><td align="left">Nicotine</td><td align="center">-</td><td align="center">-</td><td align="center">-</td><td align="left">agonist at ionotrophic acetylcholine receptor</td><td align="left">Loss of current at high (cytotoxic) concentrations in <italic>Cliona celata </italic>[5]</td></tr><tr><td align="left">Caffeine</td><td align="center">X<sup>a</sup></td><td align="center">faster rhythm</td><td align="center">attenuated</td><td align="left">antagonist at metabotrphic adenosine receptors</td><td align="left">-</td></tr><tr><td align="left">GABA<sup>b</sup></td><td align="center">X</td><td align="center">n.d.</td><td align="center">-</td><td align="left">ionotrophic or metabotrophic GABA receptors</td><td align="left">No effect in <italic>Cliona celata </italic>[5].<break/>GABA/mgluR-like receptor gene in <italic>Geodia cydonium </italic>[19]. High specificity in inducing contractions in <italic>T. wilhelma </italic>(Ellwanger et al., submitted).</td></tr><tr><td align="left">Glutamate<sup>b</sup></td><td align="center">X</td><td align="center">n.d.</td><td align="center">-</td><td align="left">metabotrophic glutamate receptors</td><td align="left">Induces intracellular Ca2+ increase. GABA/mGluR-like receptor gene in <italic>Geodia cydonium </italic>[19]. Lower specificity than GABA in inducing contractions in <italic>T. wilhelma </italic>[31]</td></tr><tr><td align="left">Glycin</td><td align="center">X</td><td align="center">faster rhythm</td><td align="center">attenuated</td><td align="left">metabotrophic glycin receptors</td><td align="left">-</td></tr><tr><td align="left">Adrenaline</td><td align="center">-</td><td align="center">very slow, irregular rhythm</td><td align="center">-</td><td align="left">metabotrophic adrenergic receptor family</td><td align="left">Slight current reduction in <italic>Cliona celata </italic>[5].<break/>Histochemical detection in <italic>Sycon </italic>sp. [21]; increased rhythm and intensity of contraction in <italic>Euspongia officinalis </italic>[22]</td></tr><tr><td align="left">Serotonine</td><td align="center">X</td><td align="center">-</td><td align="center">-</td><td align="left">ionotrophic or metabotrophic serotonine receptors</td><td align="left">No effect in <italic>Cliona celata </italic>[5]. Histochemical detection in <italic>Sycon </italic>sp. [21], and <italic>Tedania ignis </italic>[42]</td></tr><tr><td align="left">(by NOC-12)<sup>c</sup></td><td align="center">X</td><td align="center">very slow, irregular rhythm</td><td align="center">attenuated</td><td align="left">intracellular soluble guanylate cyclase</td><td align="left">NOS activity in heat stress response of <italic>Axinella polypoides </italic>and <italic>Petrosia ficiformis </italic>[48]</td></tr><tr><td align="left">cAMP<sup>d</sup></td><td align="center">X</td><td align="center">tendency for slower, irregular rhythm</td><td align="center">-</td><td align="left">metabotrophic cAMP receptor</td><td align="left">Reduced locomotion an altered shape in dissociated <italic>Clathrina cerebrum </italic>cells [51];<break/>Release of <italic>Spongilla lacustris </italic>gemmules from dormancy [50]</td></tr></tbody></table><table-wrap-foot><p><sup>a </sup>delayed contraction</p><p><sup>b </sup>detailed data and results presented elswhere [31]</p><p><sup>c </sup>extracellular application, membrane permeable</p><p><sup>d </sup>extracellular application, membrane impermeable</p></table-wrap-foot></table-wrap><p>Acetylcholine (ACh) is one of the best investigated transmitters and the related muscarinic and nicotinic receptors were the first for which the pharmacology was elucidated [<xref ref-type="bibr" rid="B32">32</xref>]. The nicotinic receptors are ligand-gated ion channels, the muscarinic types are metabotrophic systems. Both play important roles in the CNS of all phyla, in which they have been investigated [<xref ref-type="bibr" rid="B33">33</xref>,<xref ref-type="bibr" rid="B34">34</xref>]. Specific acetylcholine esterase (AChE) activity has been reported for the sponges <italic>Sycon </italic>sp. [<xref ref-type="bibr" rid="B21">21</xref>], <italic>Hippospongia communis </italic>[<xref ref-type="bibr" rid="B35">35</xref>]. and also for <italic>T. wilhelma </italic>[<xref ref-type="bibr" rid="B26">26</xref>]. On the other hand, no AChE activity was found by investigators in the first half of the 20th century, though several sponge specimens have been investigated: <italic>Spongilla lacustris </italic>[<xref ref-type="bibr" rid="B23">23</xref>], <italic>Scypha </italic>sp. [<xref ref-type="bibr" rid="B24">24</xref>], <italic>Leuconia asperta </italic>and <italic>Syphonocalina crassa </italic>[<xref ref-type="bibr" rid="B25">25</xref>]. On the other hand, according to Pavans de Ceccatty [<xref ref-type="bibr" rid="B22">22</xref>], the application of ACh increases the rhythm and intensity of contraction in <italic>H. communis</italic>. In contrast to this results and our own previous finding of AChE activity in <italic>T. wilhelma</italic>, our present investigations do not reveal a significant correlation between ACh and contraction. ACh does not induce contraction. Nevertheless, the slight, but statistically significant effect of ACh on contraction cycle duration provides evidence that ACh plays a role in <italic>T. wilhelma</italic>, but is not directly involved in the coordination of contraction. However, it may be involved in the regulation of body extension formation and retraction [<xref ref-type="bibr" rid="B26">26</xref>], which has to be proven. Our present results show that nicotine does not affect contraction and rhythm in any way. Therefore, we can conclude that at least ionotrophic ACh-receptors are most likely not involved in regulation of contraction in <italic>T. wilhelma</italic>.</p><p>Caffeine is a non-specific antagonist of adenosine-receptors in the mammalian brain [<xref ref-type="bibr" rid="B36">36</xref>]. In addition, as stated by Fredholm and co-authors, it is the „most widely consumed behaviourally active substance in the world" [<xref ref-type="bibr" rid="B36">36</xref>]. In this context, it is strange, that to our knowledge, no coffee-consuming sponge scientists has ever reported any caffeine tests upon sponges. Our own tests on <italic>T. wilhelma </italic>display a strong effect: it induces contractions, increases the endogenous rhythm and attenuates the amplitude of endogenous contractions. Interestingly, in most experiments, the onset of caffeine-induced contraction is decelerated. Except for cAMP (see below), we have not observed such a pattern in any other contraction-inducing substance. Consequently, we conclude from the decelerated onset, that caffeine does not act directly upon the contractile cells. From the disturbed endogenous rhythm and contraction amplitude, it seems likely that caffeine interferes with the regulation of endogenous contractions. The spasm-like reaction (Fig. <xref ref-type="fig" rid="F3">3b</xref>) and the chronotropic effect, viz. the significantly faster, though slightly irregular contraction rhythm, underline this hypothesis. Consequently, a putative adenosine receptor may be part of the triggering mechanism of endogenous contractions. In mammals, adenosine receptors are ubiquitous and important throughout all kind of tissues and cell types and they have been shown to cause chronotropic effects in the heart [review in 37]. In addition, adenosine receptors, which are all metabotrophic, link the ATP-driven energy metabolism, to the signalling pathways of the cells. The adenosine signalling pathway strongly interacts with other messenger substances [<xref ref-type="bibr" rid="B38">38</xref>]. It is very likely that such a central signalling mechanism evolved early and plays an important role in invertebrates, too [<xref ref-type="bibr" rid="B33">33</xref>].</p><p>Glycine induces effects similar to caffeine in <italic>T. wilhelma</italic>: it stimulates contraction, fastens the rhythm of contraction, while attenuating the amplitude. Beside GABA and glutamate, for which we have characterised the effects on <italic>T. wilhelma </italic>in details elsewhere [<xref ref-type="bibr" rid="B31">31</xref>], glycine is one of the important amino acid transmitter in the vertebrate brain [<xref ref-type="bibr" rid="B39">39</xref>]. Its transmitter function has also been demonstrated in invertebrates, e.g. <italic>Hydra vulgaris</italic>, where it acts as an inhibitory transmitter, too [<xref ref-type="bibr" rid="B40">40</xref>]. In our experiments <italic>T. wilhelma </italic>showed a biphasic response, with a characteristic initial dropping of the sponge towards the substrate, followed by the decelerated body contraction. This indicates that glycine most likely acts upon cells which are involved in the regulation of endogenous contractions, not upon the contractile cells directly. This is underlined by the altered rhythm and amplitude. It has been demonstrated before that amino acids play a role in the intercellular signalling of sponges:. Recently, we characterised the kinetics of GABA- and glutamate-induced contractions in <italic>T. wilhelma </italic>[<xref ref-type="bibr" rid="B31">31</xref>]. In relation with the finding of a GABA/glutamate-like receptor gene in <italic>Geodia cydonium </italic>[<xref ref-type="bibr" rid="B19">19</xref>] and the observation of a slower current under GABA exposure for <italic>Cliona celata </italic>[<xref ref-type="bibr" rid="B5">5</xref>], we conclude that the GABAergic signalling pathway exists in sponges. Our first indications for a glycinergic system based on the results of the present study will have to proven by further molecular and biochemical investigations.</p><p>The catecholamine adrenaline is another substance, which has been tested on sponges previously: by Emson on <italic>Cliona celata</italic>, observing a slight reduction in current [<xref ref-type="bibr" rid="B5">5</xref>], and by Lentz who specifically stained adrenaline-containing bipolar and multipolar cell types in <italic>Sycon </italic>sp., using histochemical methods [<xref ref-type="bibr" rid="B21">21</xref>]. The result of our present study underlines that adrenaline plays a role in sponges. The strict limitation of effects to the elongation of endogenous contraction cycle duration in <italic>T. wilhelma </italic>suggests that adrenalin is directly involved or alt least interfering with the regulation of the endogenous contraction rhythm. In mammals, adrenaline is involved in the upregulation of genes of the molecular clock and is therefore involved in the feedback loop regulation of circadian rhythm [<xref ref-type="bibr" rid="B41">41</xref>]. In this context, it is interesting that Reiswig and Nickel reported in independent publications on circadian rhythms in two different <italic>Tethya </italic>species [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B11">11</xref>]. In conclusion, adrenaline might directly be involved in the circadian regulation in <italic>Tethya</italic>.</p><p>The effect of the biogenic amine serotonine upon sponges has also been tested previously. It has been detected twice using histochemical and imunohistochemical methods in <italic>Sycon </italic>sp. and <italic>Tedania ignis </italic>[<xref ref-type="bibr" rid="B21">21</xref>,<xref ref-type="bibr" rid="B42">42</xref>]. However, Emson found no effect upon <italic>Cliona celata </italic>[<xref ref-type="bibr" rid="B5">5</xref>]. In contrast, serotonine did immediately induce contraction in <italic>T. wilhelma </italic>in the present study at concentrations of 260 μM and higher. The effect is limited, since the rhythm of endogenous contractions is not affected. However, the widespread finding of serotonine in all phyla of the metazoa, acting via ionotrophic and metabotrophic receptor types, points to an early evolution of this central signalling system [<xref ref-type="bibr" rid="B33">33</xref>,<xref ref-type="bibr" rid="B34">34</xref>]. This is also supported by molecular data [<xref ref-type="bibr" rid="B43">43</xref>].</p><p>The gas nitric oxide (NO) is a central molecule in the regulation of a high number of physiological processes in vertebrates [<xref ref-type="bibr" rid="B44">44</xref>], but NO also plays a role in many invertebrate phyla [<xref ref-type="bibr" rid="B45">45</xref>,<xref ref-type="bibr" rid="B46">46</xref>]. It acts upon soluble guanylate cyclase and is released by nitric oxide synthase (NOS) during conversion of l-arginine to l-citruline. This reaction is mainly regulated by intracellular Ca<sup>2+</sup>. NO diffuses directly from cell to cell and its range is limited by its short half life [<xref ref-type="bibr" rid="B47">47</xref>]. NO is supposed to be an evolutionary old messenger, which acts auto- and paracrine [<xref ref-type="bibr" rid="B33">33</xref>]. This is supported by the finding of NOS in the sponge <italic>Axinella polypoides</italic>, where it is involved in the temperature signalling cascade [<xref ref-type="bibr" rid="B48">48</xref>]. Here we showed that NO induces contraction in <italic>T. wilhelma</italic>, but also modulates the endogenous contraction rhythm and amplitude. Due to its fast diffusion, the short half life and the limited range, NO is an optimal candidate for a contraction-inducing messenger, which acts auto- and paracrine and my also be involved in a feedback loop to putative triggering cells, involved in the timing of the endogenous rhythm.</p><p>Cyclic AMP also induced contractions in <italic>T. wilhelma </italic>in the present study. It is of interest that we did not use a membrane permeable form of cAMP. Consequently, the observed effect has to be regarded as a specific extracellular signalling event, which is most likely modulated by a specific cAMP receptor, similar to <italic>Dictyostelium discoideum </italic>[<xref ref-type="bibr" rid="B49">49</xref>]. It has been shown before that extracellular cAMP plays a regulative roll in Porifera: it is involved in the regulation of production and development of gemmules in freshwater sponges [<xref ref-type="bibr" rid="B50">50</xref>], and it directly affects the locomotion of dissociated cells of <italic>Clathrina cerebrum </italic>[<xref ref-type="bibr" rid="B51">51</xref>]. It may therefore link cellular mobility and body locomotion with contraction in sponges. All three phenomena are important aspects in the life of <italic>Tethya </italic>[<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B10">10</xref>,<xref ref-type="bibr" rid="B26">26</xref>,<xref ref-type="bibr" rid="B52">52</xref>].</p></sec><sec><title>Ligand-based coordination pathways in sponges</title><p>The question whether sponges possess a nervous (-like) system or not, has been discussed in detail and is generally answered in the negative [<xref ref-type="bibr" rid="B4">4</xref>], for an overview see [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B15">15</xref>]. Even the reports of action potentials in the Hexactinellida did not alter this conclusion, since a syncytial tissue is involved in this there [<xref ref-type="bibr" rid="B16">16</xref>-<xref ref-type="bibr" rid="B18">18</xref>]. The most detailed discussion is given by Jones [<xref ref-type="bibr" rid="B8">8</xref>], who also discussed alternative coordination mechanisms, based on mechanical signal propagation, which he preferred to chemical messenger based mechanisms. However, our present reports of specific induction of body contraction and the modulation of endogenous contraction rhythm and amplitude by various chemical messengers or receptor agonists and antagonists in our model system <italic>T. wilhelma</italic>, strongly suggests the chemical messenger hypothesis. This hypothesis is supported by or recent detailed kinetic characterisation of the contractile response of <italic>T. wilhelma </italic>upon GABA and glutamate [<xref ref-type="bibr" rid="B31">31</xref>], as well as the previous reports on effects of messengers, agonists and antagonists upon sponges (see Tab. 1 and text above). In addition the molecular characterisation of a putative metabotrophic GABA/glutamate-like receptor, clearly shows the presence of messenger based signalling systems in sponges [<xref ref-type="bibr" rid="B19">19</xref>]. Consequently, we support the idea of an extensive chemical messenger based signalling system in sponges [<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B53">53</xref>]. From our results, the question raises, why so many messenger substances are effective upon a biological system, which obviously has only few behavioural choices? The answer is simple. For a filtering, sessile organism like a sponge the adequate answer to many stimuli or environmental states is to stop filtration and eventually to renew the internal water content of the aquiferous system quickly. Our model system <italic>Tethya wilhelma </italic>most likely alters the body volume by contraction of the pinacoderm of the complex aquiferous system, in order to exchange the water [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B54">54</xref>]. It seems to be favourable for the sponge to exchange the water regularly, which is reflected by the endogenous contraction rhythm. Hence, we conclude upon the presence at least one signalling pathway in conjunction with the regulation an endogenous rhythm. Taking into account that the pacemakers of the endogenous rhythm are most likely not the contractile cells themselves, <italic>Tethya </italic>requires at least two signalling pathways to coordinate the pacemakers via a auto- and paracrine signal and to trigger the contraction of the pinacocytes by a paracrine messenger (Fig. <xref ref-type="fig" rid="F8">8</xref>).</p><fig position="float" id="F7"><label>Figure 7</label><caption><p><bold>Contraction patterns of <italic>Tethya wilhelma</italic>during extracorporeal cAMP application </bold>(a - b): Projected area of two specimens of <italic>T. wilhelma </italic>before and during cAMP application (grey background). Cyclic AMP induced immediate (a) and delayed (b) contractions and slightly disturbed the rhythm of endogenous contractions (see additional file 7: Movie_S7.mov for the time-lapse movie of Fig. 7b). Filled arrowheads indicate the points in time when stock solutions of the substances were injected into the experimental reactor circulation system to reach final concentrations of 288 mM for NOC-12.</p></caption><graphic xlink:href="1742-9994-3-7-7"/></fig><fig position="float" id="F8"><label>Figure 8</label><caption><p><bold>Hypothetical signalling pathways in <italic>Tethya wilhelma </italic>involved in coordination of contractions upon external stimuli and endogenous signals </bold>An external stimulus (1) at a putative receptor cell (grey) in the pinacoderm triggers the release of a signal substance (2) which diffuses through the mesohyle (light blue) of the sponge and triggers the contraction (C!) of contractile pinacocytes (blue) via a specific receptor and an intracellular signalling pathway. Eventually, stimulated pinacocytes release a second signal substance (3), which may further diffuse through the mesohyle or be distributed by currents in the canal system. Such a secondary signal would amplify the reaction speed upon external and internal triggering of contraction. The endogenous contraction rhythm may be controlled by numerous trigger cells (red) distributed in the mesohyle. These cells are supposed to release an auto-/paracrine signal substance (4), which diffuses through the mesohyle to coordinate the release of a signal substance (5), which diffuses to the pinacocytes and triggers contraction and eventually results in a signal amplification like shown in step (3). Signal substances (2) and (5) are likely not identical to allow independent specific coordination of contraction upon endo- and exogenous stimuli. See Additional file 8: Movie_S8.mov for a stepwise presentation of the hypothetical model.</p></caption><graphic xlink:href="1742-9994-3-7-8"/></fig><p>On the other hand it has been shown that sponges react upon mechanical stimuli [<xref ref-type="bibr" rid="B5">5</xref>-<xref ref-type="bibr" rid="B7">7</xref>] or changes of external water currents, e.g. during washout events in the present study (see above). Our previous studies on GABA also showed that attenuation and desensitisation mechanisms are part of the regulation systems. Consequently, if contraction would be coordinated by only one signalling system, the sponge would not be able to respond upon further stimuli during periods of attenuation. Therefore we conclude that several paracrine ligand-receptor system based signalling pathways are involved in the coordination of external signals down to the contractile effector cells (Fig. <xref ref-type="fig" rid="F8">8</xref>). Whether these paracrine signals spread through the mesohyle or if one or more are additionally or even exclusively distributed through the aquiferous system will have to be determined in future investigations.</p></sec></sec><sec><title>Conclusion</title><p>Our study demonstrates that a number of chemical messengers, agonists and antagonists induce contraction and/or modulate the endogenous contraction rhythm and amplitude of our nerveless model metazoon, the sponge <italic>Tethya wilhelma</italic>. From our results in combination with previous investigations on signalling and coordination in sponges, we conclude that a relatively complex system of chemical messengers regulates the contraction behaviour through auto- and paracrine signalling, acting upon specific receptor systems. As a consequence of the living conditions of sessile filtering organisms, it seems to be most likely that several different signalling pathways are involved in the coordination of contraction, since this single behaviour is the adequate answer to a number of external stimuli. Furthermore, the present screening on potential messenger systems displays again the value of <italic>T. wilhelma </italic>as a model system to understand the early evolution of signalling systems during the evolution of multicellularity in the basal metazoa. Basically we expect to find the same messenger-receptor-systems, which are part of the nervous systems and paracrine regulation systems of other metazoan phyla. From the present and a previous study, we conclude that an adrenergic, adenosynergic and glycinergic pathways, as well as pathways based on NO and extracellular cAMP are candidates for the regulation and timing of the endogenous contraction rhythm within pacemaker cells, while GABA, glutamate and serotonine are candidates for the direct coordination of the contractile cells. However, most likely these systems act on a different scale and in varied contexts, in comparison to the well-investigated systems in animals with a CNS. This will have to be unravelled by further molecular and physiological characterisations of the signalling systems in <italic>T. wilhelma </italic>and other sponge model systems, e.g. the freshwater sponges.</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Sponges</title><p>Specimens of the sponge <italic>Tethya wilhelma </italic>Sarà et al. 2001 (Tethyidae, Hadromerida, Demospongiae) were obtained from the type location in the aquarium of the zoological-botanical garden 'Wilhelma' in Stuttgart [<xref ref-type="bibr" rid="B27">27</xref>]. For experiments the sponges were maintained in a 180 l aquarium, running at 26°C, using artificial seawater [<xref ref-type="bibr" rid="B55">55</xref>], at a light-dark cycle of 12/12 hours. Sponges were fed four to five times a week using suspended commercial invertebrate food (Artificial Plancton, Aquakultur Genzel, Fellbach, Germany, www.aquakultur-genzel.de), by pipetting several ml of suspension to each sponge. Seawater was exchanged at a rate of around 10 % of the total aquarium volume each three to four weeks.</p></sec><sec><title>Experimental reactor</title><p>All experimental manipulations were carried out in a 250 ml closed experimental reactor. It consisted of an aerated experimental chamber, designed on the principles of airlift reactors, connected to a temperature regulation unit (F25, Julabo, Seelbach, Germany). Oxygen level and temperature were monitored using a multi-sensor system (P4, WTW, Weilheim, Germany), controlled by a computer-software (MultiLab Pilot 3.0, WTW). A built-in optical glass filter (Ø 49 mm, D.K. Enterprises, India) allowed proper imaging.</p></sec><sec><title>Digital time-lapse imaging and image analysis</title><p>Digital images of the sponge specimens were taken at a resolution of 2048 × 1536 pixels at regular intervals of 180 s (pre- and post-experimental monitoring) and 30 s (monitoring for induced contraction), resulting in an image-data accumulation rate of 60 megabytes per hour and 360 megabytes per hour, respectively (uncompressed 8-bit image data). A Nikon Coolpix 990E digital camera in manual macro focus and exposure mode was used to acquire greyscale images. The camera was connected to a Nikon SB 24 flash unit, set to manual mode (24 mm, output 1/16). The camera was controlled by a PC, using USB connection cable and the software DC_RemoteShutter V 2.3.0 in conjunction with DC_TimeTrigger V. 1.0 [<xref ref-type="bibr" rid="B56">56</xref>]. Images were downloaded, saved on the PC and erased on the CF-card of the camera instantly after being taken. A reference image including a scale bar placed next to the sponge was taken for each experimental series, to allow scaling. In all cases a black background was used to maximise contrast.</p><p>Image analysis was performed using ImageJ 1.30 to 1.34 (NIH, Washington, USA), based on built in functions [<xref ref-type="bibr" rid="B57">57</xref>]. The measurement of projected areas bases on the contrast difference between sponge (whitish) and background (black). All images were scaled using the reference image. A threshold value between 50 and 90 was applied to the 8-bit images and the absolute projected area of the sponge was measured using ImageJ's built-in measurement tool. Each time-lapse series was loaded as an image stack into ImageJ. Time-lapse movies were prepared based on built in functions of ImageJ. A macro was programmed to measure semi-automatically. Measurement results were written to a text file and further computed using Excel 2000. We performed Mann-Whitney U-tests on contraction cycle duration (D<sub>cc</sub>) of endogenous contraction periods and substance-exposed periods for single sponge individuals using the Excel add-on WinSTAT [<xref ref-type="bibr" rid="B58">58</xref>]. D<sub>cc </sub>was measured as time between two subsequent contractions, defined by body volume minima [<xref ref-type="bibr" rid="B7">7</xref>].</p></sec><sec><title>Test substance application</title><p>Diluted stock solutions of the test substances were injected into the experimental reactor to reach final concentrations between 5.7 μM and 1.3 mM depending on the substance. For our tests, the following substances were applied: acetylcholine (acytlcholine chloride, Sigma-Aldrich A6625); adrenalin ([± ]-epinephrine, Sigma-Aldrich E1635); cAMP (cyclic adensosine-3'-5'-monophosphoric acid monosodium salt, Boehringer 102296); caffeine (Sigma Aldrich C5-3); glycine (Merck 104201); nicotine ([S]- [-]-nicotine, Sigma-Aldrich 18,637-6); NOC-12 (3-ethyl-3- [ethylaminoethyl]-1-hydroxy-2-oxo-1-triazene, Sigma-Aldrich E3145); serotonin (serotonin creatinine sulfate complex, Sigma-Aldrich H7752).</p><p>Care was taken not to inject the solutions directly to the sponge specimen, but allow mixing in the circulating current of the system. For each experiment, substances were only applied during a phase of expansion in-between two subsequent endogenous rhythmic contractions, but the relative time within the contraction cycle varied among the experiment, for image data-processing reasons. The sponges long-term reaction was recorded. After the experiment the reactor was perfused by fresh, aerated artificial seawater at 26°C.The same specimens were eventually used for another experiment after retaining a normal contraction rhythm for at least several hours.</p></sec><sec><title>Control experiments and control measurements</title><p>Prior to the experiment, each specimen used was allowed to acclimatise to the experimental reactor for several hours. Experiments were only started after the sponge specimens displayed typical regular contraction patterns as described previously [<xref ref-type="bibr" rid="B7">7</xref>]. During all of the experiments, temperature and oxygen-level were monitored and recorded. Changes in pH due to application of substances were measured. Control experiments were carried out to test the sponge's reaction to pH-changes.</p></sec></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interest.</p></sec><sec><title>Authors' contributions</title><p>KE designed and performed the experiments, analysed the data, prepared figure drafts and revised the manuscript. MN designed the study and principal experiments, performed data analysis, prepared final figures, drafted, wrote and revised the manuscript.</p></sec><sec sec-type="supplementary-material"><title>Supplementary Material</title><supplementary-material content-type="local-data" id="S1"><caption><title>Additional file 1</title><p><bold>Time-lapse movie of </bold><bold><italic>Tethya wilhelma </italic>reacting upon extracorporal nicotine application </bold>The quicktime-movie Movie_S1.mov represents the time-lapse image series of the time period displayed as a contraction pattern graph in Fig. <xref ref-type="fig" rid="F2">2b</xref>. The time span of extracorporeal application of the substance is indicted in the movie as well as the elapsed time. Time-lapse 2300-fold. For more details refer to the results section of the manuscript.</p></caption><media xlink:href="1742-9994-3-7-S1.mov" mimetype="video" mime-subtype="quicktime"><caption><p>Click here for file</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="S2"><caption><title>Additional file 2</title><p><bold>Time-lapse movie of </bold><bold><italic>Tethya wilhelma</italic></bold><bold>reacting upon extracorporal caffeine application (spasm-like contraction behaviour) </bold>The quicktime-movie Movie_S2.mov represents the time-lapse image series of the time period displayed as a contraction pattern graph in Fig. <xref ref-type="fig" rid="F3">3b</xref>. The time span of extracorporeal application of the substance is indicted in the movie as well as the elapsed time. Time-lapse 2300-fold. For more details refer to the results section of the manuscript.</p></caption><media xlink:href="1742-9994-3-7-S2.mov" mimetype="video" mime-subtype="quicktime"><caption><p>Click here for file</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="S3"><caption><title>Additional file 3</title><p><bold>Time-lapse movie of </bold><bold><italic>Tethya wilhelma</italic></bold><bold>reacting upon extracorporal caffeine application (attenuated amplitude, local contractions) </bold>The quicktime-movie Movie_S3.mov represents the time-lapse image series of the time period displayed as a contraction pattern graph in Fig. <xref ref-type="fig" rid="F3">3c</xref>. The time span of extracorporeal application of the substance is indicted in the movie as well as the elapsed time. Shifts in perspective are a result of unintentional slight shifts of the camera during the experiment. Time-lapse 2300-fold. For more details refer to the results section of the manuscript.</p></caption><media xlink:href="1742-9994-3-7-S3.mov" mimetype="video" mime-subtype="quicktime"><caption><p>Click here for file</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="S4"><caption><title>Additional file 4</title><p><bold>Time-lapse movie of </bold><bold><italic>Tethya wilhelma</italic></bold><bold>reacting upon extracorporal glycin application </bold>The quicktime-movie Movie_S4.mov represents the time-lapse image series of the time period displayed as a contraction pattern graph in Fig. <xref ref-type="fig" rid="F4">4b</xref>. The time span of extracorporeal application of the substance is indicted in the movie as well as the elapsed time. Time-lapse 2300-fold. For more details refer to the results section of the manuscript.</p></caption><media xlink:href="1742-9994-3-7-S4.mov" mimetype="video" mime-subtype="quicktime"><caption><p>Click here for file</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="S5"><caption><title>Additional file 5</title><p><bold>Time-lapse movie of </bold><bold><italic>Tethya wilhelma</italic></bold><bold>reacting upon extracorporal adrenaline application </bold>The quicktime-movie Movie_S5.mov represents the time-lapse image series of the time period displayed as a contraction pattern graph in Fig. <xref ref-type="fig" rid="F5">5a</xref>. The time span of extracorporeal application of the substance is indicted in the movie as well as the elapsed time. Shifts in perspective are a result of unintentional slight shifts of the camera during the experiment. Time-lapse 2300-fold. For more details refer to the results section of the manuscript.</p></caption><media xlink:href="1742-9994-3-7-S5.mov" mimetype="video" mime-subtype="quicktime"><caption><p>Click here for file</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="S6"><caption><title>Additional file 6</title><p><bold>Time-lapse movie of </bold><bold><italic>Tethya wilhelma</italic></bold><bold>reacting upon extracorporal nitric oxide application (via NOC-12) </bold>The quicktime-movie Movie_S6.mov represents the time-lapse image series of the time period displayed as a contraction pattern graph in Fig. <xref ref-type="fig" rid="F6">6b</xref>. The time span of extracorporeal application of the substance is indicted in the movie as well as the elapsed time. Shifts in perspective are a result of unintentional slight shifts of the camera during the experiment. Time-lapse 2300-fold. For more details refer to the results section of the manuscript.</p></caption><media xlink:href="1742-9994-3-7-S6.mov" mimetype="video" mime-subtype="quicktime"><caption><p>Click here for file</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="S7"><caption><title>Additional file 7</title><p><bold>Time-lapse movie of </bold><bold><italic>Tethya wilhelma</italic></bold><bold>reacting upon extracorporal cyclic AMP application </bold>The quicktime-movie Movie_S7.mov represents the time-lapse image series of the time period displayed as a contraction pattern graph in Fig. <xref ref-type="fig" rid="F7">7b</xref>. The time span of extracorporeal application of the substance is indicted in the movie as well as the elapsed time. Shifts in perspective are a result of unintentional slight shifts of the camera during the experiment. Time-lapse 2300-fold. For more details refer to the results section of the manuscript.</p></caption><media xlink:href="1742-9994-3-7-S7.mov" mimetype="video" mime-subtype="quicktime"><caption><p>Click here for file</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="S8"><caption><title>Additional file 8</title><p><bold>Hypothetical signalling pathways in <italic>Tethya wilhelma</italic></bold><bold>involved in coordination of contractions upon external stimuli and endogenous signals </bold>The quicktime-movie Movie_S7.mov represents a step by step presentation of the hypothetical coordination model given in Figure <xref ref-type="fig" rid="F8">8</xref>. An external stimulus (1) at putative receptor cell (grey) in the pinacoderm triggers the release of a signal substance (2) which diffuses through the mesohyle (light blue) of the sponge and triggers the contraction (C!) of contractile pinacocytes (blue) via a specific receptor and an intracellular signalling pathway. Eventually, stimulated pinacocytes release a second signal substance (3), which may further diffuse through the mesohyle or be distributed by currents in the canal system. Such a secondary signal would amplify the reaction speed upon external and internal triggering of contraction. The endogenous contraction rhythm may be controlled by numerous trigger cells (red) distributed in the mesohyle. These cells are supposed to release an auto-/paracrine signal substance (4), which diffuses through the mesohyle to coordinate the release of a signal substance (5), which diffuses to the pinacocytes and triggers contraction and eventually results in a signal amplification like shown in step (3). Signal substances (2) and (5) are likely not identical to allow independent specific coordination of contraction upon endo- and exogenous stimuli.</p></caption><media xlink:href="1742-9994-3-7-S8.mov" mimetype="video" mime-subtype="quicktime"><caption><p>Click here for file</p></caption></media></supplementary-material></sec>
Active Stat3 is required for survival of human squamous cell carcinoma cells in serum-free conditions	<sec><title>Background</title><p>Squamous cell carcinoma (SCC) of the skin is the most aggressive form of non-melanoma skin cancer (NMSC), and is the single most commonly diagnosed cancer in the U.S., with over one million new cases reported each year. Recent studies have revealed an oncogenic role of activated signal transducer and activator of transcription 3 (Stat3) in many human tumors, especially in those of epithelial origin, including skin SCC. Stat3 is a mediator of numerous growth factor and cytokine signaling pathways, all of which activate it through phosphorylation of tyrosine 705.</p></sec><sec><title>Results</title><p>To further address the role of Stat3 in skin SCC tumorigenesis, we have analyzed a panel of human skin-derived cell lines ranging from normal human epidermal keratinocytes (NHEK), to non-tumorigenic transformed skin cells (HaCaT), to highly tumorigenic cells (SRB1-m7 and SRB12-p9) and observed a positive correlation between Stat3 phosphorylation and SCC malignancy. We next determined the role of Stat3 activity in cell proliferation and viability under serum-free culture conditions. This was accomplished by suppressing Stat3 activity in the SRB12-p9 cells through stable expression of a dominant negative acting form of Stat3β, which contains a tyrosine 705 to phenylalanine mutation (S3DN). The S3DN cells behaved similar to parental SRB12-p9 cells when cultured in optimal growth conditions, in the presence of 10% fetal calf serum. However, unlike the SRB12-p9 cells, S3DN cells underwent apoptotic cell death when cultured in serum-free medium (SFM). This was evidenced by multiple criteria, including accumulation of sub-G1 particles, induced PARP cleavage, and acquisition of the characteristic morphological changes associated with apoptosis.</p></sec><sec><title>Conclusion</title><p>This study provides direct evidence for a role for Stat3 in maintaining cell survival in the conditions of exogenous growth factor deprivation produced by culture in SFM. We also propose that delivery of the S3DN gene or protein to tumor cells could induce apoptosis and/or sensitize those cells to the apoptotic effects of cancer therapeutic agents, raising the possibility of using S3DN as an adjunct for treatment of skin SCC.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>Yin</surname><given-names>Weihong</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Cheepala</surname><given-names>Satish</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" contrib-type="author"><name><surname>Roberts</surname><given-names>Jennifer N</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A4" contrib-type="author"><name><surname>Syson-Chan</surname><given-names>Keith</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A5" contrib-type="author"><name><surname>DiGiovanni</surname><given-names>John</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A6" corresp="yes" contrib-type="author"><name><surname>Clifford</surname><given-names>John L</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	Molecular Cancer	<sec><title>Introduction</title><p>Non-melanoma skin cancer (NMSC) is the most common cancer in the U.S., with over a million new cases of the two most common forms, squamous and basal cell carcinoma, anticipated in 2004 [<xref ref-type="bibr" rid="B1">1</xref>]. The more clinically aggressive form, squamous cell carcinoma (SCC) [<xref ref-type="bibr" rid="B2">2</xref>], has been increasing in incidence since the 1960s at annual rates from 4% to as much as 10% in recent years [<xref ref-type="bibr" rid="B3">3</xref>]. About 95% of skin SCC cases are diagnosed at an early stage and are easily controlled. Unlike early stage SCC, advanced SCC is aggressive, often resistant to local therapy, requires repeated surgical resections and courses of radiotherapy, and accounts for approximately 2000 U.S. deaths each year [<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B4">4</xref>]. Advanced disease- and treatment-related morbidity have a profound impact on patients' quality of life, frequently producing cosmetic deformity, loss of function, and psychosocial problems. Improved control of advanced skin SCC is clearly necessary and will rely on a thorough understanding of the molecular basis for skin SCC progression.</p><p>Signal transducers and activators of transcription (Stat) proteins, a family of latent cytoplasmic transcription factors, are expressed in many cell types and, in response to a wide variety of extracellular polypeptides, regulate the transcription of a broad spectrum of genes that are critically involved in cytokine signaling [<xref ref-type="bibr" rid="B5">5</xref>], cell proliferation and development [<xref ref-type="bibr" rid="B6">6</xref>], and tumorigenesis [<xref ref-type="bibr" rid="B7">7</xref>-<xref ref-type="bibr" rid="B9">9</xref>]. Upon binding of extracellular ligands, cell surface receptors oligomerize and activate associated Janus kinases (JAKs), which in turn phosphorylate Stats on a single critical tyrosine residue located adjacent to an -SH2 (src homology domain 2) domain. The Stats then dimerize via reciprocal -SH2 domain phosphorylation site interactions and translocate to the nucleus where they regulate gene expression by direct DNA binding or by associating with other transcription factors [<xref ref-type="bibr" rid="B10">10</xref>,<xref ref-type="bibr" rid="B11">11</xref>]. The activity of Stats can be abolished by mutation of this critical tyrosine [<xref ref-type="bibr" rid="B12">12</xref>,<xref ref-type="bibr" rid="B13">13</xref>].</p><p>Among the seven known members of mammalian Stat family, Stat3 has been most strongly implicated in tumorigenesis [<xref ref-type="bibr" rid="B7">7</xref>-<xref ref-type="bibr" rid="B9">9</xref>]. Elevated levels of Stat3 activity have been observed in a number of human cancers and cancer cell lines [<xref ref-type="bibr" rid="B9">9</xref>]. In cancers of epithelial origin, Stat3 is constitutively activated in head and neck squamous cell carcinoma (HNSCC) [<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B15">15</xref>], breast cancer cell lines [<xref ref-type="bibr" rid="B16">16</xref>,<xref ref-type="bibr" rid="B17">17</xref>], ovarian cancer cell lines [<xref ref-type="bibr" rid="B18">18</xref>], and lung cancer cell lines [<xref ref-type="bibr" rid="B19">19</xref>]. In particular, Stat3 plays a critical role in the development of skin cancer [<xref ref-type="bibr" rid="B20">20</xref>]. In an experimental two-stage mouse skin chemical carcinogenesis model it has been shown that Stat3 is constitutively activated in skin tumors [<xref ref-type="bibr" rid="B21">21</xref>], and that activated Stat3 is indispensable for both the initiation and the promotion stages of epithelial carcinogenesis [<xref ref-type="bibr" rid="B22">22</xref>]. The critical role of Stat3 in skin tumor development was further supported by data obtained from a transgenic mouse model in which a constitutively active mutant of Stat3 called Stat3C (7), was expressed in skin under the control of the keratin-5 promoter [<xref ref-type="bibr" rid="B23">23</xref>]. These mice have a skin phenotype closely resembling psoriasis in humans and, when subjected to the two-stage skin chemical carcinogenesis protocol, rapidly developed carcinomas, bypassing the papilloma stage that normally takes place in this model [[<xref ref-type="bibr" rid="B23">23</xref>], Chan et al, submitted].</p><p>Apoptosis or programmed cell death, is mediated through two major pathways, the extrinsic and intrinsic [<xref ref-type="bibr" rid="B24">24</xref>,<xref ref-type="bibr" rid="B25">25</xref>]. The extrinsic pathway is primarily triggered by the binding of extra-cellular death ligands (e.g. TNFα, TRAIL and FasL) to their cognate membrane death receptors. The intrinsic pathway is often initiated by cellular stresses such as withdrawal of survival factors, direct DNA damage (e.g. UV exposure, cytotoxic drugs), and is characterized by the disruption of mitochondrial membrane integrity, an event regulated by Bcl-2 protein family members [<xref ref-type="bibr" rid="B26">26</xref>,<xref ref-type="bibr" rid="B27">27</xref>]. There are more than 20 known members of the Bcl-2 family which, based on their functions in regulating apoptosis, can be divided into an anti-apoptotic 'Bcl-2-like' group, (Bcl-2, Bcl-XL, Bcl-w, Bfl-1/A1 and Mcl-1 etc) and a pro-apoptotic group (Bax, Bak, Bok, Bcl-Xs, Bad, Bid, Bik/Nbk, Bim, Hrk, Bmf, Noxa and Puma etc). It has been reported that Stat3 can regulate transcription of several Bcl-2 family proteins, such as Bcl-2 and Bcl-xL [<xref ref-type="bibr" rid="B28">28</xref>-<xref ref-type="bibr" rid="B31">31</xref>], Bax [<xref ref-type="bibr" rid="B32">32</xref>], and Mcl-1 [<xref ref-type="bibr" rid="B33">33</xref>-<xref ref-type="bibr" rid="B35">35</xref>].</p><p>At early stages of tumor development, tumor cells often have to face and survive harsh physiological micro-environments, such as lack of nutrition and/or blood supply, survival factor insufficiency, and hypoxia, which generally lead to apoptosis in normal cells [<xref ref-type="bibr" rid="B36">36</xref>]. In fact, it has been well accepted that one of the six hallmarks of tumor cells is the reduced or complete loss of dependence on exogenous growth factor stimulation for survival and proliferation [<xref ref-type="bibr" rid="B37">37</xref>]. Stats are the first family of transcription factors found to be directly activated upon growth factor receptor stimulation. Stat3 is thought to confer protection against apoptosis in many transformed or tumor cells. Several studies in which Stat3 activity is either blocked by anti-sense oligonucleotides, small interfering RNA or expression of dominant negative Stat3 isoforms [<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B32">32</xref>,<xref ref-type="bibr" rid="B38">38</xref>-<xref ref-type="bibr" rid="B40">40</xref>], or elevated by expression of Stat3C [<xref ref-type="bibr" rid="B41">41</xref>], have shown an inverse correlation between Stat3 activity and induced apoptosis.</p><p>In this study we have examined the activity of Stat3 in several human skin-derived cell lines, ranging from non-transformed to highly malignant, and observed a positive correlation between malignancy and constitutive Stat3 phosphorylation. In addition, we have generated human skin SCC cell lines with reduced Stat3 activity by stably expressing a dominant negative acting form of Stat3β, hereafter referred to as S3DN [<xref ref-type="bibr" rid="B42">42</xref>-<xref ref-type="bibr" rid="B44">44</xref>]. The S3DN cells, unlike the parental SRB12-p9 cells, undergo apoptosis in the conditions of exogenous growth factor deprivation produced by culture in serum free medium (SFM).</p></sec><sec><title>Results</title><sec><title>Constitutive phosphorylation of Stat3 positively correlates with malignancy of human skin SCC cell lines</title><p>Stat3 transcriptional activity can be induced by the phosphorylation of a single tyrosine residue (Tyr705) by receptor associated Janus kinases. Increased Stat3 activation is often associated with either elevated constitutive levels of Stat3 protein or increased Stat3 tyrosine phosphorylation [<xref ref-type="bibr" rid="B14">14</xref>]. To further elucidate the role of Stat3 in skin cancer development, we first examined the Stat3 status in NHEKs, non-tumorigenic, spontaneously transformed keratinocyte cells (HaCaT) and two aggressive skin SCC cell lines (SRB1-m7 and SRB12-p9). In order to avoid detecting a high level of background Stat3 phosphorylation, the HaCaT and SRB cells were switched from their normal growth in medium which contains 5 and 10% FCS, respectively, to medium containing 0.5% serum for two days, followed by an additional two hours culture in SFM. We have previously reported that low concentrations of IFN-α efficiently induce Stat3 phosphorylation in SRB12-p9 cells [<xref ref-type="bibr" rid="B45">45</xref>]. In order to compare the inducibility of Stat3 phosphorylation between normal, premalignant and malignant skin cells, they were treated with 100 international units (IU)/ml IFN-α for 30 min, and then whole cell protein was extracted and subjected to western blot probing with a Stat3α- or a Stat3 phospho-tyrosine 705-specific antibody. All cell lines expressed comparable steady state levels of Stat3α protein with or without IFN-α treatment (Fig. <xref ref-type="fig" rid="F1">1</xref>, upper panel). However, in the absence of IFN-α treatment, phosphorylated Stat3α (phospho-Stat3α) was only observed in the tumorigenic skin SCC cell lines (Fig. <xref ref-type="fig" rid="F1">1</xref>, lanes labeled SRB1-m7 and SRB12-p9, middle panel). Treatment with IFN-α, which has been previously shown to induce Stat3 phosphorylation and DNA binding [<xref ref-type="bibr" rid="B44">44</xref>], induced phosphorylation in the HaCaT cells, but not in the NHEK cells (Fig. <xref ref-type="fig" rid="F1">1</xref> middle panels, NHEK and HaCaT lanes). <bold>The upper band in lanes 3–5 is non-specific and is sometimes observed with this antibody</bold>.</p><fig position="float" id="F1"><label>Figure 1</label><caption><p>Stat3 phosphorylation correlates with human skin SCC malignancy. NHEK cells were grown for 48 hours in KGM-2 growth media and HaCaT, SRB12-p9 and SRB1-m7 cells were grown for 48 hours in normal growth media containing 0.5% serum, followed by culture for 2 hours in SFM and then treated for 30 min with 100 IU/ml IFN-α. Whole cell extracts were purified, subjected to Western blotting and sequentially probed with antibodies specific to Stat3α, tyrosine 705 phosphorylated Stat3α (Stat3-p) (middle panel), and β-actin as a control for well loading.</p></caption><graphic xlink:href="1476-4598-5-15-1"/></fig></sec><sec><title>Establishment of cell lines stably expressing a dominant negative form of Stat3 (S3DN)</title><p>To explore the role of Stat3 in skin cell malignancy, we over-expressed both the wild type Stat3α and a dominant negative form of Stat3β in one of the tumorigenic SCC cell lines, SRB12-p9. It has been reported that Stat3β, a naturally occurring Stat3 splice variant that has a truncated C-terminus, can function as a dominant negative form of Stat3 and inhibit its transcriptional activity [<xref ref-type="bibr" rid="B42">42</xref>,<xref ref-type="bibr" rid="B43">43</xref>]. It was subsequently shown that substituting the critical Jak kinase tyrosine phosphorylation site with phenylananine generated a form of Stat3 that could block DNA binding by all endogenous forms of Stat3 (Stat3β-Y705F) [<xref ref-type="bibr" rid="B44">44</xref>]. We established SRB12-p9 cell clones expressing either the FLAG-tagged wild type Stat3α protein (S3WT) or FLAG-tagged Stat3β-Y705F (S3DN) (Fig. <xref ref-type="fig" rid="F2">2A</xref>). Over-expressing the S3WT resulted in higher Stat3 DNA binding activity than in parental cells as determined by EMSA (Fig. <xref ref-type="fig" rid="F2">2B</xref>, compare lane 4 with lane 1, white triangle), while the DNA binding activity in cells expressing the S3DN was reduced (Fig. <xref ref-type="fig" rid="F2">2B</xref>, compare lane 3 to lanes 1 and 2). The specificity of DNA binding was confirmed by the elimination of the shifted band upon addition of excess unlabelled Stat3 probe (Fig. <xref ref-type="fig" rid="F2">2B</xref>, compare lane 5 with lane 4). We note that the EMSA band intensities in Fig. <xref ref-type="fig" rid="F2">2B</xref> are lower than those typically detected for lysates of cells transiently transfected with expression constructs for Stats (see [<xref ref-type="bibr" rid="B44">44</xref>]). In the present study we assayed for endogenous and stably expressed Stat3 isoforms, which are less abundant than in transiently transfected cells.</p><fig position="float" id="F2"><label>Figure 2</label><caption><p>Establishment of cell clones stably expressing S3DN and S3WT proteins. (A) SRB12-p9 cells were transfected with expression vectors as described in materials and methods. Whole cell extracts from cell clones (number shown above each lane) were subjected to Western blotting and probed sequentially with antibodies to the FLAG octapeptide, Stat3 (recognizing both α and β isoforms), and β-actin. (B) SRB12-p9, Neo, S3DN2 and S3WT6 cells (lane 1, 2, 3 and 4 respectively) were serum-starved for 2 days. Nuclear extracts were prepared and the EMSA assay was performed as described in Materials and Methods. In lane 5 nuclear extract from S3WT6 was pre-incubated with excess cold (unlabeled) Stat3 probe prior to the addition of labeled probe. Open arrow, expected location for Stat3α homodimer; closed arrow, unbound probe; dot, non-specific band. Results are representative of three separate experiments.</p></caption><graphic xlink:href="1476-4598-5-15-2"/></fig></sec><sec><title>Stat3 activity is required for cell survival in SFM</title><p>Stressful physiological conditions, such as a shortage of nutrients or growth factors, often result in apoptotic cell death in non-transformed cells but not in transformed cells. The S3WT and S3DN cell lines allowed determination of the role of Stat3 on cell viability during culture in SFM, a well-established experimental stress condition. Cells were grown in the presence or absence of FCS for 4 days and cell viability measured by the MTT assay at 1 day intervals. Doubling times for cells growing at a subconfluent density (between days 1 and 3) were calculated based on the viability data shown in Fig. <xref ref-type="fig" rid="F3">3A</xref>. In the standard 10% FCS-containing medium there was no major difference in the cell viability and calculated doubling times for parental SRB12-p9 cells and control SRB12-p9 cells stably transfected with the empty pSG5 expression vector and the pKJ1 neomycin resistance vector (Neo) (Fig. <xref ref-type="fig" rid="F3">3A</xref>, black bars and Table <xref ref-type="table" rid="T1">1</xref>). The S3WT cells (represented by clones 6 and 10) had an increase in doubling time of approximately 2 hours and one of the S3DN cell lines (S3DN5) was increased by 3 hours, indicating no correlation between Stat3 expression pattern and doubling time in standard media (Fig. <xref ref-type="fig" rid="F3">3A</xref> and Table <xref ref-type="table" rid="T1">1</xref>). However, in SFM (Fig. <xref ref-type="fig" rid="F3">3A</xref>, white bars) the two S3DN cell lines showed no increase in cell number between days 2 and 3 and by the 4<sup>th </sup>day had a reduction in cell viability, indicating cell death. Visual inspection revealed that S3DN cells cultured in SFM became rounded and detached from the plate, two events that typically accompany apoptotic cell death (Fig. <xref ref-type="fig" rid="F3">3B</xref>, compare upper and lower S3DN2 panels). In contrast, the SRB12-P9, Neo and S3WT cells showed only a slight reduction in cell viability compared to cells grown in FCS-containing medium (Fig. <xref ref-type="fig" rid="F3">3A</xref> compare white bars to black bars and table <xref ref-type="table" rid="T1">1</xref>). From this data we conclude that Stat3 activity is required for SCC cell survival in SFM. This response was consistently observed for SRB12-p9 cells and multiple independent Neo (n = 2), S3DN (n = 3) and S3WT (n = 3) cell clones, for 4 experiments.</p><fig position="float" id="F3"><label>Figure 3</label><caption><p>Stat3 activity is required for survival in SFM. (A) SRB12-p9, Neo and two representative independent S3WT (S3WT6, S3WT10) and S3DN clones (S3DN2, S3DN5) were cultured for the indicated times in 10% FCS-containing media or SFM and cell viability was determined by the MTT assay (see materials and methods). White and black bars indicate the mean absorbance at 540 nm for cells grown in SFM or 10% FCS-containing media, respectively. Error bars indicate SEM for triplicate cultures. Results shown are representative of >4 assays that included 2 independent S3WT and S3DN cell clones. (B) SRB12-p9, Neo, S3DN2 and S3WT6 cells were cultured for 5 days in 10% FCS-containing media or SFM (upper and lower rows, respectively). Results shown are representative of 3 independent S3DN and S3WT cell clones from 4 separate experiments. Cells were photographed on a phase contrast microscope (100×).</p></caption><graphic xlink:href="1476-4598-5-15-3"/></fig><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>Expression of wild type Stat3α (S3WT column) and dominant negative Stat3β (S3DN column) protein does not correlate with cell doubling time. Cell doubling time was calculated as described in Materials and Methods using the proliferation data shown in Fig. 3A. + symbol refers to the approximate level of S3WT or S3DN protein, as determined by Western blotting. N/A: not applicable.</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td></td><td></td><td align="center" colspan="2">Doubling Time</td></tr><tr><td align="left">Cell line</td><td align="center" colspan="2">Relative protein level</td><td align="center" colspan="2">Day 1 through Day 3 ime</td></tr><tr><td></td><td align="center">S3WT</td><td align="center">S3DN</td><td align="center">10% FCS</td><td align="center">-FCS</td></tr></thead><tbody><tr><td align="left">SRB12-p9</td><td align="center">+</td><td align="center">-</td><td align="center">15.6</td><td align="center">18.6</td></tr><tr><td align="left">P9 neo</td><td align="center">+</td><td align="center">-</td><td align="center">15.4</td><td align="center">17.9</td></tr><tr><td align="left">S3DN2</td><td align="center">+</td><td align="center">++</td><td align="center">15.3</td><td align="center">N/A</td></tr><tr><td align="left">S3DN5</td><td align="center">+</td><td align="center">++</td><td align="center">18.4</td><td align="center">N/A</td></tr><tr><td align="left">S3WT6</td><td align="center">+++</td><td align="center">-</td><td align="center">17.4</td><td align="center">20</td></tr><tr><td align="left">S3WT10</td><td align="center">+</td><td align="center">-</td><td align="center">17.4</td><td align="center">18.6</td></tr></tbody></table></table-wrap></sec><sec><title>Expression of S3DN protein causes apoptosis in cells grown in SFM</title><p>The reduced viability of the S3DN cells grown in SFM could be due to either cell death alone or to a combination of cell death and reduced proliferation. There is abundant evidence for a role for Stat3 both in regulating proliferation and in suppressing apoptotic cell death [<xref ref-type="bibr" rid="B7">7</xref>-<xref ref-type="bibr" rid="B9">9</xref>]. In order to determine whether expression of the S3DN or S3WT proteins could also influence proliferation, we compared the cell cycle profiles of the SRB12-p9, Neo, S3DN and S3WT cells. Cells were cultured under conditions identical to that shown in Fig. <xref ref-type="fig" rid="F3">3B</xref> and their cell cycle profiles determined by flow cytometry. In FCS-containing media all cell populations had similar percentages in the G1, S and G2 phases of the cell cycle (Fig. <xref ref-type="fig" rid="F4">4A</xref>, upper panels). When grown in SFM for 4 days several small differences were observed. SRB12-P9, S3DN2 and S3WT6 cells showed reductions in the percentage of cells in S phase, <bold>the S3DN2 cells showed an increase in the percentage of cells in the G2 phase</bold>, and SRB12-P9 and S3WT6 cells showed an increase in the percentage of cells in the G1 phase (Fig. <xref ref-type="fig" rid="F4">4A</xref>, -FCS panels). However the most prominent difference was the greater number of DNA containing particles in the sub-G1 size range observed for the S3DN2 cells in both FCS-containing media (12% of total particle count compared to approximately 5% for the other cell lines) and in SFM (28.5% compared to 2–3% for the other cells, Fig <xref ref-type="fig" rid="F4">4A</xref>). The presence of sub-G1 particles is a hallmark of apoptosis for cultured cells. This finding was reproducible for multiple Neo, S3DN and S3WT cell clones and is consistent with the results shown in Fig. <xref ref-type="fig" rid="F3">3A</xref> and <xref ref-type="fig" rid="F3">3B</xref>, strongly suggesting that S3DN expression primarily affects the control of apoptosis and not proliferation in this model.</p><fig position="float" id="F4"><label>Figure 4</label><caption><p>Expression of S3DN protein leads to apoptosis in SFM. SRB12-p9, Neo, S3DN2 and S3WT6 cells were cultured in 10% FCS-containing media or SFM for 4 days. (A) Cells and floating material were collected, fixed in 70% ethanol, stained with propidium iodide and analyzed by flow cytometry. The percentage of cells in the G1/0, S and G2 phases of the cell cycle are indicated for each cell line and treatment condition. Numbers to the left of the G1/0 peak indicate the percentage of sub-G1/0 particles from the total population of gate particles (excluding clumped cells). (B) Whole cell extracts were resolved on an SDS polyacrylamide gel, transferred to nitrocellulose and sequentially probed with antibodies specific for c-PARP, Bax and β-actin. Results shown in A and B are representative of 2 Neo and 3 S3DN and S3WT clones, in 2 independent experiments.</p></caption><graphic xlink:href="1476-4598-5-15-4"/></fig><p>In order to further confirm that the loss of cell viability of the S3DN cells grown in SFM is due to increased apoptosis, we assayed for the appearance of c-PARP protein, the cleaved from of PARP, in cells treated as in Fig. <xref ref-type="fig" rid="F4">4A</xref>. PARP cleavage facilitates irreversible cellular disassembly and provides a highly sensitive indicator of apoptosis [<xref ref-type="bibr" rid="B46">46</xref>]. The appearance of c-PARP was most pronounced in the S3DN2 cells grown in SFM, as determined by western blotting with a c-PARP-specific antibody (Fig. <xref ref-type="fig" rid="F4">4B</xref>, upper panel). A low level of c-PARP was detectable in S3DN2 cells grown in FCS-containing media, as well as in Neo and S3WT6 cells grown in SFM (Fig. <xref ref-type="fig" rid="F4">4B</xref>, upper panel).</p><p>It has been reported in other cell culture systems and animal models that Stat3 regulates apoptosis via modulation of transcription of several of the Bcl-2 family proteins, including Bcl-2, Bcl-xL, Bax and Mcl-1 [<xref ref-type="bibr" rid="B28">28</xref>-<xref ref-type="bibr" rid="B35">35</xref>]. Here we observe enhanced Bax protein expression for S3DN2 grown in SFM, but not for the other cells, suggesting a possible role for Bax in this apoptotic effect (Fig. <xref ref-type="fig" rid="F4">4B</xref>, middle panel). As in 4A, these results are representative of several Neo, S3DN and S3WT cell clones in at least 2 independent experiments per clone.</p></sec></sec><sec><title>Discussion</title><p>Elevated Stat3 activity has been observed in numerous spontaneous and experimentally established mammalian cancers, demonstrating a critical role in tumorigenesis [<xref ref-type="bibr" rid="B7">7</xref>-<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B17">17</xref>]. In this study we provide direct evidence that Stat3 activity, as indicated by phosphorylation at tyrosine 705, positively correlates with malignancy in human skin-derived cell lines. Suppression of Stat3 activity, through forced expression of the S3DN protein, in human skin SCC cells blocks their growth factor- and/or other serum factor-independence, a major characteristic of malignancy.</p><p>Recent studies have provided convincing evidence for a critical role for Stat3 in every stage of mouse skin cancer development, from promoting the survival of initiated cells to conferring late-stage malignant characteristics such as enhanced motility and invasiveness [<xref ref-type="bibr" rid="B21">21</xref>,<xref ref-type="bibr" rid="B22">22</xref>]. In parallel with these studies we sought to develop a human skin SCC model in which Stat3 activity is stably suppressed, in order to assess the contribution of activated Stat3 to the malignant phenotype in human disease. The SRB12-p9 cell line was originally derived from an aggressive skin SCC tumor. These cells were chosen for stable transfection of the S3DN protein because, in addition to having constitutive and IFN-α inducible Stat3 phosphorylation, they are highly tumorigenic upon subcutaneous injection into nude mice and can be readily transfected and selected for stable gene expression [<xref ref-type="bibr" rid="B45">45</xref>].</p><p>The S3DN protein is a unique Stat3 blocking reagent. It consists of Stat3β, τhe short alternative splice product of the Stat3 gene, bearing a point mutation at the critical tyrosine 705 phosphorylation site [<xref ref-type="bibr" rid="B44">44</xref>], that has been FLAG-tagged to allow distinguishing it from the endogenous Stat3β. Stat3β was previously shown to act in a dominant negative manner to suppress the transcriptional activity of Stat3α [<xref ref-type="bibr" rid="B42">42</xref>]. However Stat3β can be transcriptionally active under conditions where Stat3α is not, through interaction with the N-terminal segment of c-jun, [<xref ref-type="bibr" rid="B43">43</xref>]. Unlike Stat3β, the S3DN protein lacks detectable DNA binding or transcriptional activity in EMSA and transient transfection reporter assays, respectively [<xref ref-type="bibr" rid="B44">44</xref>], and would therefore not be expected to induce Stat3β-specific effects. The S3DN protein should be able to form non-functional heterodimers with endogenous forms of Stat3α or β however, blocking their ability to enter the nucleus and/or bind DNA. Alternatively, S3DN may interfere with endogenous Stat3 activity at another level, such as the phosphorylation by JAK kinases, where S3DN may occupy the Stat3 docking sites on the cytoplasmic domains of growth factor and cytokine receptors, thereby blocking phosphorylation of endogenous Stat3α. This later possibility is less likely since we do not observe a reduced level of phospho-Stat3α in the S3DN cells compared to SRB12-p9 or Neo cells (data not shown). Also, recent evidence has emerged indicating that unphosphorylated Stat3α can drive expression of several genes, including some well known oncoproteins, through a novel mechanism that is distinct from that of phosphorylated Stat3α [<xref ref-type="bibr" rid="B47">47</xref>]. It therefore cannot be formally ruled out that S3DN, even though it cannot be phosphorylated, could itself have effects not involving interaction with endogenous Stats.</p><p>Although the precise mechanism of suppression of Stat3 activity by S3DN is unclear, its expression in the SRB12-p9 cells reduced binding of Stat3α to DNA (Fig. <xref ref-type="fig" rid="F2">2B</xref>) and was predicted to inhibit the constitutive Stat3 activity, thereby suppressing proliferation and possibly de-repressing apoptotic signals. To our surprise, the initial characterization of the S3DN stable transfectants indicated no obvious effects on proliferation rate or viability compared to the parental SRB12-p9 cells (Table <xref ref-type="table" rid="T1">1</xref>). Similarly, forced overexpression of the S3WT protein did not produce an increase in proliferation rate, but rather the opposite occurred, with an approximately 2 hour increase in cell doubling time observed for 2 of the S3WT clones (Table <xref ref-type="table" rid="T1">1</xref>). While this latter result is difficult to explain, the very short doubling time for SRB12-p9 cells (15–16 hours) suggests that further increases in proliferation rate may be limited by other intrinsic factors such as nutrient and biomolecule availability. The lack of consistent reduction in cell proliferation rate for the S3DN cells could be explained by an insufficient amount of S3DN protein expression necessary to block endogenous Stat3α activity under these culture conditions. The highest expressing S3DN clone, S3DN5, shows approximately equal signal for the Stat3 α and β bands, indicating roughly equal amounts of both proteins (Fig. <xref ref-type="fig" rid="F2">2A</xref>, middle panel, lanes labeled S3DN2 and 5). While this is a substantial increase over the wild type amount of Stat3β protein (compare levels of Stat3β between S3DN and S3WT cells in Fig. <xref ref-type="fig" rid="F2">2A</xref>, middle panel), it appears insufficient to illicit the effects on cell proliferation and/or cell viability that would be consistent with a suppression of endogenous Stat3α activity.</p><p>In an attempt to determine whether expression of S3DN could affect any Stat3α regulated cellular processes, we grew the cells in SFM, an experimental stress condition known to suppress cell growth and induce apoptosis in many cell lines. Indeed, it was only after depriving the S3DN cells of FCS that a dramatic effect was observed. Several independently selected S3DN clones underwent cell death induction when grown in SFM, which was not observed with SRB12-p9, Neo or S3WT clones. This effect was characterized by cell rounding and detachment from the plate followed by disintegration into subcellular particles, all characteristics of apoptotic cell death. This effect was quantified using the MTT cell viability assay. A reduction in cell viability for the S3DN cell lines after 2 days in SFM indicated that cell death was induced, in addition to a possible reduction in proliferation rate. In contrast, the SRB12-p9 and Neo cells remained viable and continued to proliferate, with only an approximately 2–3 hour increase in doubling time.</p><p>The induction of cell death could entirely account for the reduced viability of S3DN cells grown in SFM. However, because Stat3 is also a key regulator of cell proliferation [<xref ref-type="bibr" rid="B7">7</xref>-<xref ref-type="bibr" rid="B9">9</xref>], reduced proliferation may also contribute to this effect. Comparison of the cell cycle profiles of SRB12-p9, Neo, S3DN and S3WT cells indicated that, in FCS-containing media, the distribution of cells in the G1/0, S and G2 phases of the cell cycle was similar in all cases. One exception was the higher amount of sub-G1 DNA-containing particles for the S3DN2 cell line than for the other cell lines (12% compared to approximately 5% for the other cell lines, Fig. <xref ref-type="fig" rid="F4">4A</xref> upper row), suggesting a higher rate of apoptosis even in the presence of 10% FCS. Four days of culture in SFM resulted in an increase to 28.5% sub-G1 particles for the S3DN2 cells, but not the other cell lines. It should be noted that the data in the S3DN2 -FCS panel represents the relatively small population of S3DN2 cells remaining alive after 4 days in SFM (see corresponding panels in Fig. <xref ref-type="fig" rid="F3">3B</xref>). Surprisingly, these cells still exhibited a similar percentage of cells in the G1/0 phase to those growing in FCS-containing media, indicating that S3DN expression induces cell death, but does not cause an accumulation of cells in G1/0. However there was a reduction in the percent of cells in S phase with a proportional increase in the G2 fraction for the S3DN2 cells in SFM, indicating potential blocks at both the G1 to S phase transition and at mitosis. Thus the contribution of reduced proliferation to the overall effect of the S3DN expression on these cells is minor compared to the apoptotic effect. This conclusion is further supported by our finding that the pro-apoptotic Bcl-2 family member, Bax, is upregulated in the S3DN cells grown in SFM and this effect is accompanied by accumulation of c-PARP.</p><p>The results of the present study are consistent with an anti-apoptotic role for Stat3 in human skin SCC and are also in agreement with much of the predicted role for Stat3 derived from recent mouse skin tumorigenesis studies [<xref ref-type="bibr" rid="B21">21</xref>-<xref ref-type="bibr" rid="B23">23</xref>]. In addition, it has been demonstrated that gene therapy with Stat3β was effective in suppressing tumor growth in an <italic>in-vivo </italic>mouse melanoma model [<xref ref-type="bibr" rid="B40">40</xref>]. This effect was associated with induction of the secreted death ligand TRAIL, which could induce apoptosis and cell cycle arrest of adjacent non-transfected cells [<xref ref-type="bibr" rid="B48">48</xref>]. Other investigators, using a complimentary approach to assessing Stat3 function, have demonstrated that expression of the constitutively active Stat3C protein in fibroblasts can protect them from UV-induced apoptosis [<xref ref-type="bibr" rid="B41">41</xref>].</p></sec><sec><title>Conclusion</title><p>We have demonstrated that suppression of Stat3 signaling through forced expression of the S3DN protein in human skin SCC cells blocks their growth factor- and/or other serum factor-independence. Unlike the parental SRB12-p9 cells, the S3DN cells remain viable only when cultured under optimal growth conditions, in nutrient medium supplemented with 10% FCS. The SFM culture condition is likely to provide a closer approximation to the inhospitable conditions found in tumor microenvironments <italic>in-vivo</italic>. This raises the exciting possibility of using S3DN as an adjunct therapy for treatment of skin SCC. We propose that delivery of the S3DN gene or protein to tumor cells could induce apoptosis directly and/or sensitize tumor cells to the apoptosis-inducing effects of cancer therapeutic agents. Future studies are planned to explore this possibility. These include assessing the effect of S3DN expression in <italic>in-vivo </italic>tumor models. We will begin by comparing the malignant properties of the S3DN cells with Neo control cells in mouse subcutaneous injection tumorigenicity assays. This study, together with our <italic>in-vivo </italic>studies in mouse skin, supports the hypothesis that Stat3 is a central regulator of apoptosis and proliferation in malignant skin cells.</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Cell culture and generation of stable S3DN expressing SRB12-p9 cell lines</title><p>The origin and culture of the human skin SCC cell line SRB12-p9 was described previously [<xref ref-type="bibr" rid="B45">45</xref>]. HaCaT cells were cultured according to [<xref ref-type="bibr" rid="B49">49</xref>] in Dulbecco's Modified Eagles Media (DMEM)-low glucose media, supplemented with 5% fetal calf serum (FCS). Normal human epidermal keratinocytes (NHEK) cells were purchased from Clonetics (now Cambrex BioScience, Walkersville, MD) and grown in KGM-2 media (Cambrex) supplemented with hEGF (0.1 ng/ml), insulin (5.0 μg/ml), hydrocortisone (0.5 μg/ml, calcium (0.15 mM), epinephrine (5 ng/ml), transferrin (10 μg/ml), Gentamycin (50 μg/ml), Amphotericin-B (50 ng/ml) and 2 ml bovine pituitary extract. All cells were grown in a humidified atmosphere with a 5% CO2 concentration.</p><p>The pSG5-Stat3β-Y705F-FLAG expression vector was constructed by ligating the mouse Stat3β-Y705F cDNA cassette into the EcoRI site of the pSG5 expression vector [<xref ref-type="bibr" rid="B50">50</xref>], followed by sequential PCR-based site-specific mutagenesis using the following oligonucleotides: 5'-TCATTGATGCAGTTTGGAAACTCGAGTAACACTCTGTGAGCTGATA-3' and 5'-TATCAGCTCACAGAGTGTTACTCGAGTTTCCAAACTGCATCAATGA-3', to introduce an Xho I site immediately upstream of the stop codon at position 723. The FLAG epitope extension was created by inserting a DNA fragment, consisting of the following complimentary oligonucleotide pair: 5'-TCGAAGACTACAAAGACGATGACGATAAATAGTAGTGATGACTTGTCATCATCGTCCTTATAATCAGATCTT-3' and 5'-TCGAAAGATCTGATTATAAGGACGATGATGACAAGTCATCACTACTATTTATCGTCATCGTCTTTGTAGTCT-3', into the newly generated Xho I site. The pSG5-Stat3α-FLAG plasmid was constructed by ligating the mouse Stat3α cassette into the EcoRI site of the pSG5 expression vector, followed by sequential PCR-based site-specific mutagenesis using the following oligonucleotides: 5'-CTTCTGGTTTCAGCTCCTCGAGCATGGGGGAGGTAGCACACT-3' and 5'-AGTGTGCTACCTCCCCCATGCTCGAGGAGCTGAAACCAGAAG-3', to introduce an Xho I site immediately upstream of the stop codon. The FLAG epitope extension was created by inserting the same complimentary oligonucleotide pair used for the Stat3β construct. The Stat3β-Y705F and Stat3α cDNA cassettes were a gift from T. Schaefer (Meso Scale Discovery). The resulting plasmids were linearized with Aat II restriction enzyme and electroporated along with the bacterial neomycin phosphotransferase gene expression vector pKJ1 [<xref ref-type="bibr" rid="B51">51</xref>], and stably expressing cell clones were isolated essentially as previously described [<xref ref-type="bibr" rid="B45">45</xref>]. 5 × 10<sup>6 </sup>cells suspended in 800 μl PBS were electroporated with 5 μg of linearized, purified pSG5-Stat3β-Y705F-FLAG or pSG5-Stat3α-FLAG and 0.5 μg pKJ1 with a Bio-Rad Gene Pulser set at 200 V and 960 μF. Cells were then plated at a density of approximately 1 × 10<sup>6 </sup>cells/10 cm culture plate, and after 24 h subjected to neomycin selection (300 μg/ml G418 sulfate, GIBCO/BRL, Rockville, MD) for up to 14 days. Individual colonies were isolated, propagated, and divided into two aliquots, one for freezing and the other for expansion and western blotting.</p></sec><sec><title>Reagents</title><p>Recombinant human interferon alpha (IFN-α) was purchased from Serotec Inc. (Raleigh, NC). Antibodies against human Stat3, Stat3 phospho-tyrosine 705, and human β-actin were supplied by Santa Cruz Biotechnology (Santa Cruz, CA). Antibodies specific for Bax, cleaved poly [ADP-ribose] polymerase (PARP) were purchased from Cell Signaling Technology (Beverly, MA). Antibody against the FLAG epitope was from Sigma-Aldrich (St. Louis, MO).</p></sec><sec><title>Western blotting and electrophoretic mobility shift assay (EMSA)</title><p>Whole cell extract purification from stably transfected cells and western blotting was performed as described previously [<xref ref-type="bibr" rid="B45">45</xref>], with minor modifications. Total cellular protein was prepared using RIPA lysis buffer (150 mM NaCl, 10 mM Tris-HCl pH 7.5, 1 mM EDTA, 1% NP-40, 1 mM dithiothreitol) supplemented with Complete protease inhibitor cocktail (Roche, Indianapolis, IN) according to manufacturer-provided instructions. Extracted protein was quantified using the Bio-Rad Protein Assay kit (Hercules, CA). Proteins were separated by SDS acrylamide gel electrophoresis and transferred to nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany). Blots were blocked with 5% milk powder for 1 h at room temperature, followed by incubation for 1 h with antibodies for Stat3, Stat3 phospho-tyrosine 705, the cleaved form of PARP, Bax and human β-actin. Blots were then washed with PBS/0.05% Tween and incubated with horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature, followed by an additional 3 washes with PBS/0.05% Tween. Chemiluminescence detection was performed according to the manufacturer's instructions (Amersham Life Sciences Inc., Piscataway, NJ) followed by autoradiography.</p><p>Nuclear extracts were prepared from approximately 1 × 10<sup>7 </sup>cells. Briefly, cells were trypsinized and washed with ice-cold PBS. Pelleted cells were resuspended in ice-cold buffer A containing (10 mM HEPES pH 7.9, 10 mM KCl, 1.5 mM MgCl<sub>2</sub>, 0.5 mM DTT, 0.2 mM PMSF, and 0.05% NP-40). After centrifugation at 1,500 g for 3 min at 4°C the pellet was resuspended in buffer C (20 mM HEPES, 420 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM DTT, 0.2 mM PMSF, and 25% (v/v) glycerol) for 30 minutes at 4°C, followed by centrifugation at 15,000 g for 20 min at 4°C. EMSA was performed with the Gel-Shift kit (Cat No. AY1046, Panomics, Redwood, CA) according to provided protocol. Briefly, nuclear extracts (3 μg) were incubated with 10 ng biotin-labeled Stat3 response element probe (5'-GATCCTTCTGGGAATTCCTAGATC-3') derived from the c-fos gene promoter [<xref ref-type="bibr" rid="B52">52</xref>], in a binding buffer (both provided in the kit, final volume of 10 μl) containing 1 μg poly (dI-dC) for 40 min at 18°C. In the competitive EMSA, nuclear extracts were incubated for 5 min at room temperature with 20 ng unlabeled probe (provided by kit) prior to the addition of the biotin-labeled probe. The reactions were loaded on a 6% polyacrylamide non-denaturing PAGE gel in 0.5 × TBE buffer and electrophoresed for 1.5 h at 150 V before being transferred to BiodyneB membranes (PallGelman Lab, Ann Arbor, MI) and detected by the reagents provided with the kit.</p></sec><sec><title>Cell viability and cell cycle assays</title><p>Cell viability was determined by the MTT (3- [4, 5-dimethyltiazol 2-yl]-2,5-diphenyltetraolium) bromide assay [<xref ref-type="bibr" rid="B53">53</xref>], essentially as previously described [<xref ref-type="bibr" rid="B54">54</xref>]. Briefly, cells were plated in triplicate wells (1000 cells per well) in 100 μl growth media in 96-well plates and subjected to serum starvation the following day, for the indicated durations. At the appropriate times a solution of MTT (20 μl of a 12 mM solution in PBS) was added and incubated for 1 hour at 37°C. The cells were washed gently with PBS, and 100 μl of dimethylsulfoxide was added to the wells followed by mild shaking to dissolve the MTT precipitate. Absorbance was measured for each well using a Wallac Victor3 1420 Multilabel multiwell plate reader (Perkin-Elmer) at a wavelength of 540 nm. Mean absorbance values and standard errors based on the mean (SEM) were calculated for triplicate cultures. Cell doubling time was calculated according to the following formula: doubling time (hrs) = hrs in culture ÷ (log [A540<sup>F</sup>/A540<sup>I</sup>]/log2), where A540<sup>F </sup>and A540<sup>I </sup>are the mean absorbance values of triplicate cultures from the MTT assay at the end and beginning of the time span measured, respectively. The cell cycle profile of control and serum starved cells was determined by cell cycle flow cytometry based on cellular DNA content, using an FACSCalibur Cell Sorter (Becton Dickenson) essentially as described previously [<xref ref-type="bibr" rid="B55">55</xref>]. Cells were serum-starved for the indicated durations, trypsinized, collected and pelleted together with the material floating in the medium. Cells were fixed in cold 70% ethanol, resuspended in PBS at a density of > 10<sup>6 </sup>cells/ml followed by RNase A (1 mg/ml) treatment, addition of propidium iodide (20 μg/ml final concentration) and analysis by flow cytometry. The percentage of cells at different phases of the cell cycle was determined from the raw data using the ModFit LT v 3.0 software package (Verity Software House Inc.).</p></sec></sec><sec><title>Abbreviations used</title><p>NMSC, non-melanoma skin cancer; SCC, squamous cell carcinoma; Stat, signal transducer and activator of transcription; SH2, Src homology domain 2; JAKs, Janus-activated kinases; HNSCC, head and neck SCC; S3DN, Stat3β Y705F dominant negative protein; SFM, serum-free media; DMEM, Dulbecco's Modified Eagles Media; FCS, fetal calf serum; NHEK, normal human epidermal keratinocytes; hEGF, human epidermal growth factor; IFN-α, recombinant human interferon alpha; c-PARP, cleaved form of poly [ADP-ribose] polymerase; EMSA, electrophoretic mobility shift assay; TBE, Tris borate electrophoresis buffer; MTT, 3- [4, 5-dimethyltiazol 2-yl]-2,5-diphenyltetraolium; IU, international units; S3WT, Stat3α wild type protein; Neo, SRB12-p9 cells stably transfected with the empty pSG5 expression vector and the pKJ1 neomycin resistance vector</p></sec><sec><title>Authors' contributions</title><p>WY carried out Western blotting, EMSA, cell viability and cell cycle analyses and contributed to the draft of the manuscript. SC and JR participated in the generation and characterization of the stably transfected cell lines. KS-C contributed to the initial study design and participated in the characterization of the stably transfected cells. JD contributed to the initial study design and to the draft of the manuscript. JC conceived of the study, coordinated the study, participated in the generation of the stably transfected cell lines and contributed to the draft of the manuscript.</p></sec>
Efficacy of topical cobalt chelate CTC-96 against adenovirus in a cell culture model and against adenovirus keratoconjunctivitis in a rabbit model	<sec><title>Background</title><p>Adenovirus (Ad), associated with significant morbidity, has no topical treatment. A leading CTC compound (CTC-96), a Co<sup>III </sup>chelate, was found to have potent <italic>in vitro </italic>and <italic>in vivo </italic>antiviral efficacy against herpes viruses. In this study CTC-96 is being tested for possible anti-Adenovirus activity.</p></sec><sec sec-type="methods"><title>Methods</title><p>The biological anti-adenovirus activity of CTC-96 in concentrations from 5 to 250 ug/ml, was evaluated initially by viral inactivation (viral exposure to CTC-96 followed by dilution and inoculation of cells), virucidal (viral exposure to CTC-96 and inoculation of cells without dilution) and antiviral (effect of CTC-96 on previously adsorbed virus) plaque assays on HeLa (human cervical carcinoma), A549 (human lung carcinoma) and SIRC (rabbit corneal) cells. After verifying the antiviral activity, New Zealand White rabbits were infected with Ad-5 into: 1) the anterior cul-de-sac scarifying the conjunctiva (Group "C+"); 2) the anterior cul-de-sac scarifying the conjunctiva and cornea (Group "CC+"); 3) the stroma (Group "CI+"). Controls were sham-infected ("C-", "CC-", "CI-"). Other rabbits, after "CC", were treated for 21 days with: 1) placebo, 9x/day ("-"); 2) CTC-96, 50 ug/ml, 9x/day ("50/9"); CTC-96, 50 ug/ml, 6x/day ("50/6"); CTC-96, 25 ug/ml, 6x/day ("25/6"). All animals were monitored via examination and plaque assays.</p></sec><sec><title>Results</title><p><italic>In </italic>vitro viral inactivation, virucidal and antiviral assays all demonstrated CTC-96 to be effective against Adenvirus type 5 (ad-5). The <italic>in vivo </italic>model of Ad keratoconjunctivitis most similar to human disease and producing highest viral yield was "CC". All eyes (6/6) developed acute conjunctivitis. "CI" yielded more stromal involvement (1/6) and iritis (5/6), but lower clinical scores (area × severity). Infection via "C" was inconsistent (4/6). Fifty (50) ug/ml was effective against Ad-5 at 6x, 9x dosings while 25 ug/ml (6x) was only marginally effective.</p></sec><sec><title>Conclusion</title><p>CTC-96 demonstrated virucidal activity against Ad5 in tissue culture with HeLa, A549 and SIRC cell lines.</p><p>Animal Model Development: 1) "CC" produced conjunctival infection with occasional keratitis similar to human disease; "CI" yielded primarily stromal involvement; 2) "C" consistently produced neither conjunctivitis nor keratitis.</p><p>CTC Testing: 1) Conjunctivitis in all eyes; 2) Resolution fastest in "50/9" ("50/9". "50/6" > "25/6" > "-"); 3) Efficacy in "50/6" was not statistically different than "50/9"; 4) Conjunctival severity was lower in treatment groups then controls; 5) Little corneal or intra-ocular changes were noted.</p></sec>	<contrib id="A1" equal-contrib="yes" corresp="yes" contrib-type="author"><name><surname>Epstein</surname><given-names>Seth P</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" equal-contrib="yes" contrib-type="author"><name><surname>Pashinsky</surname><given-names>Yevgenia Y</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" equal-contrib="yes" contrib-type="author"><name><surname>Gershon</surname><given-names>David</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A4" equal-contrib="yes" contrib-type="author"><name><surname>Winicov</surname><given-names>Irene</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A5" equal-contrib="yes" contrib-type="author"><name><surname>Srivilasa</surname><given-names>Charlie</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A6" equal-contrib="yes" contrib-type="author"><name><surname>Kristic</surname><given-names>Katarina J</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A7" equal-contrib="yes" contrib-type="author"><name><surname>Asbell</surname><given-names>Penny A</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	BMC Ophthalmology	<sec><title>Background</title><p>Adenovirus is the most common external ocular viral infection worldwide[<xref ref-type="bibr" rid="B1">1</xref>]. Although not permanently blinding, ocular adenoviral infections are associated with significant patient morbidity, including symptomatic distress, and corneal changes causing visual disturbances that can last months to years. About one half of the over 50 serotypes of human adenovirus are known to cause ocular disease in patients[<xref ref-type="bibr" rid="B1">1</xref>]. Currently there are no specific efficacious antiviral agents for topical or systemic treatment of Adenoviral infections[<xref ref-type="bibr" rid="B2">2</xref>].</p><p>The studies of the pathogenesis and treatment of ocular adenovirus infections have been limited due to the narrow host range exhibited by human adenoviruses. It has been previously determined that one serotype of human adenovirus, adenovirus type 5 (Ad-5), has the ability to extend its host range to permit replication in the eyes of New Zealand rabbits[<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B4">4</xref>]. These and other studies have shown human adenovirus type 5 (Ad-5) to present clinically within 24 to 48 hours of innoculum in rabbits[<xref ref-type="bibr" rid="B5">5</xref>] and last for approximately 16 days post-innoculum (mean duration of shedding)[<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B6">6</xref>].</p><p>A number of cobalt complexes (CTC compounds) have been identified that exhibit potent <italic>in vitro </italic>and <italic>in vivo </italic>activity against herpes group viruses[<xref ref-type="bibr" rid="B7">7</xref>]. Most significantly, the mode of action of these novel compounds differ from currently available antiviral nucleoside analogs and protease inhibitors[<xref ref-type="bibr" rid="B8">8</xref>] (and unpublished data: Redox Pharmaceutical Corporation). One CTC compound in particular, CTC-96, has already been shown to exhibit pronounced efficacy in the topical therapy of HSV-1-induced epithelial and stromal disease in the rabbit eye[<xref ref-type="bibr" rid="B7">7</xref>].</p><p>In this study we evaluate the efficacy of topical CTC-96 against adenovirus infection in tissue culture on both human and rabbit cell lines as well as against ocular adenovirus infection in New Zealand White rabbits.</p></sec><sec sec-type="methods"><title>Methods</title><sec sec-type="materials"><title>Materials</title><sec><title>Adenovirus</title><p>Human Adenovirus type 5 [<xref ref-type="bibr" rid="B9">9</xref>-<xref ref-type="bibr" rid="B11">11</xref>] (AD-5) was obtained from the American Type Culture Collection (VR-5; ATCC, Manassas, VA) and propagated on human cervical carcinoma cell monolayers ["HeLa cells" (CCL-2); ATCC, Manassas, VA]. Stabilized viral stocks grown in EMEM [Vitacells Eagle Minimum Essential Medium (30–2003; ATCC, Manassas, VA)] supplemented with 10% fetal calf serum (FCS), 100 units/ml penicillin, 0.1 mg/ml streptomycin and 2 mM L-glutamine (all from Sigma chemical Co., St. Louis, MO) and containing approximately 9 × 10<sup>8 </sup>plaque forming units/ml (pfu/ml) were produced and stored at -80°C for up to 1 year. Cells were infected at a Multiplicity of Infection (MOI) of 10. Eyes were infected with approximately 2 × 10<sup>7 </sup>pfu of virus.</p></sec><sec><title>Cells</title><p>1) Human cervical carcinoma cells [HeLa cells (CCL-2); ATCC, Manassas, VA] were grown and maintained in EMEM supplemented with FCS, penicillin, streptomycin and L-glutamine as described above.</p><p>2) Human lung carcinoma cells [A549 cells (CCL-185); ATCC, Manassas, VA] were grown and maintained in Hams F-12 K Medium (30–2004; ATCC, Manassas, VA) supplemented with 10% fetal calf serum, 100 units/ml penicillin, 0.1 mg/ml streptomycin and 2 mM L-glutamine (all from Sigma chemical Co., St. Louis, MO).</p><p>3) Rabbit corneal cells [SIRC (CCL-60); ATCC, Manassas, VA] were grown and maintained in EMEM supplemented with 10% fetal bovine serum, 100 units/ml penicillin, 0.1 mg/ml streptomycin and 2 mM L-glutamine (all from Sigma chemical Co., St. Louis, MO).</p></sec><sec><title>Pharmaceuticals</title><p>Doxovir™, CTC-96 (Figure <xref ref-type="fig" rid="F1">1</xref>) was synthesized according to a standard Redox Pharmaceutical Corporation procedure[<xref ref-type="bibr" rid="B12">12</xref>]. CTC-96 solution (stable in aqueous solutions at 4°C for >6 months and for several weeks at 37°C) was prepared from crystalline compound which is completely stable for over 2.5 years when stored desiccated at 0°C. The structure was confirmed by the manufacturer utilizing <sup>1</sup>H and <sup>13</sup>C nuclear magnetic resonance spectroscopy, mass spectrometry, and elemental analysis[<xref ref-type="bibr" rid="B12">12</xref>]. Using these techniques, purity was determined to be ≥ 99%, with no detectable impurities[<xref ref-type="bibr" rid="B12">12</xref>]. Prior to use, drug concentrations in all CTC-96 solutions were confirmed by HPLC against an independently prepared CTC-96 standard. The activity of the drug solutions was confirmed in an anti-Herpes Simplex-1 assay[<xref ref-type="bibr" rid="B9">9</xref>]. Formulations of this stabilized active compound at the appropriate concentrations were prepared, filter-sterilized and supplied by the manufacturer. Samples were stored refrigerated and protected from light prior to use.</p><fig position="float" id="F1"><label>Figure 1</label><caption><p>The general structure of Doxovir™ (CTC-96).</p></caption><graphic xlink:href="1471-2415-6-22-1"/></fig></sec><sec><title>Animals</title><p>New Zealand White Rabbits weighing 4–5 pounds were purchased from H.A.R.E. Industries (Hewitt, NJ). Animals were designated both SPF (specific pathogen free) and VAF (viral antigen free) and were handled in accordance with NIH guidelines. At the conclusion of each experiment, the rabbits were euthanized by pentobarbital injection (Sleepaway, Fort Dodge Labs, Fort Dodge, IA).</p><p>Rabbits were housed in the pathogen containment area (BSL-2) of the Mount Sinai Medical Center Animal Facility. Newly received animals were allowed to acclimate to the facility and to daily handling for one week prior to infection. All rabbits were examined to verify no clinical or culture evidence of viral disease (slit lamp biomicroscopy and viral cultures of corneal swabs followed by standardized plaque assays).</p></sec></sec><sec sec-type="methods"><title>Procedures</title><sec><title>I. <italic>In vitro </italic>studies</title><sec><title>A. <italic>In vitro </italic>viral inactivation assays (direct drug-virus {high titer} neutralization with post-neutralization CTC dilution)</title><p>Varying concentrations of CTC-96 or placebo were mixed with concentrated human Adenovirus type 5 (Ad5; 1.75 × 10<sup>7 </sup>pfu) and incubated at 37°C for 60 minutes (min) under atmospheric conditions. Aliquots of the virus/drug suspension were then diluted 500 fold and plated in triplicate on aspirated HeLa, A549 and SIRC cell monolayers to initiate infection (6–8 × 10<sup>5 </sup>cells/cm2 in T-25 culture flasks containing 9.9 mls fresh culture medium giving a total volume of 10 mls with an MOI of 10). The low final concentration of the CTC-96 (1:500) has no inhibitory effect on Adenovirus type 5 growth. All cell lines were grown and maintained in their appropriate medium.</p><p>The cell monolayers, having been infected with the virus, were incubated for 24 hours (hr) at 37°C and 5% CO2 and then washed, scraped, sonicated, centrifuged and the supernatant serially diluted. These serial dilutions were plated onto indicator HeLa, A549 and SIRC cell monolayers and adsorbed for 60 min. The supernatant was then aspirated and a methylcellulose overlay placed over the cells which were then incubated at 37°C. Due to differences in replication rates of the Adenovirus in the various cell lines, HeLa cells were incubated for 3 days, A549 cells for 5 days and the SIRC cells for 7 days. All were counterstained with 1% methylene blue, allowed to dry and the plaques counted under a phase contrast microscope in a masked fashion.</p></sec><sec><title>B. <italic>In vitro </italic>virucidal assays (direct drug-virus neutralization without post-neutralization CTC dilution)</title><p>Varying concentrations of CTC-96 or placebo were mixed with concentrated human Adenovirus type 5 (Ad5; 1.75 × 10<sup>7 </sup>pfu) and incubated at 37°C for 60 minutes (min). Media of the HeLa, A549 and SIRC cell monolayers (at 80% confluency) was decanted and replaced with the above suspension. These cell monolayers, having now been infected with the virus (6–8 × 10<sup>5 </sup>cells/cm2 in T-25 culture flasks containing 9.9 mls fresh culture medium giving a total volume of 10 mls with an MOI of 10), were incubated for 24 hours (hr) at 37°C and 5% CO<sub>2 </sub>in medium containing CTC-96 at the indicated concentrations and then washed, scraped, sonicated, centrifuged and the supernatant serially diluted.</p><p>The serial dilutions were then plated onto indicator HeLa, A549 and SIRC cell monolayers and adsorbed for 60 min as described above. The supernatant was then aspirated and a methylcellulose overlay placed over the cells which were then incubated at 37°C. All were counterstained with 1% methylene blue, allowed to dry and the plaques counted under a phase contrast microscope in a masked fashion.</p></sec><sec><title>C. <italic>In vitro </italic>antiviral activity assays (effect of CTC-96 on virus previously adsorbed onto cells)</title><p>Adenovirus type 5 (1.75 × 10<sup>7 </sup>pfu) was adsorbed onto preconfluent HeLa, A549 and SIRC cell monolayers for 60 min at 37°C. Various concentrations of CTC-96 or placebo in medium were then added to the culture and the monolayers were subsequently incubated for 24 hr at 37°C and 5% CO<sub>2</sub>. Monolayers were then washed, scraped, sonicated, centrifuged, the supernatants serially diluted and ultimately stained and the plaques counted as described above under the section for the viral inactivation assay.</p></sec><sec><title>D. Procedure</title><p>The biological anti-adenovirus activity of freshly solubilized CTC-96 was evaluated by the standard antiviral, virucidal and antiviral plaque reduction assays as described above. Seven concentrations of CTC-96 were used with each cell line and in each assay: 0, 5, 10, 25, 50, 100 and 250 ug/ml. Both negative (0 ug/ml CTC-96, no Ad-5) and positive [0 ug/ml CTC-96, addition of Ad-5 (MOI = 10)] controls were included. All were diluted in the appropriate medium for each of the cell lines.</p></sec><sec><title>E. Data analysis</title><p>All clinical and virus recovery data were analyzed for statistical differences, utilizing the mean ± standard deviation for each of the various groups in computer-generated two-tailed bivariant Student's t tests (GB-STAT, New England Software, Inc., College Station, TX, U.S.A.; SAS, SAS Institute Inc., Cary, NC, U.S.A.; and SPSS, SPSS Inc., Chicago, IL, U.S.A.)[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>]. Individual Fisher exact tests were also performed (GB-STAT, SAS and SPSS)[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>], as well as an overall chi-squared analysis (GB-STAT, SAS and SPSS)[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>] and a correlation coefficient[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>]. Two-tailed significance was established at a confidence level of 0.05 ≤ P ≤ 0.95.</p></sec></sec><sec><title>II. Animal studies</title><sec><title>A. Methods of viral infection (6 groups)</title><p>1. Experimental</p><p>a. Installation of 20 ul containing 4 × 10<sup>7 </sup>pfu Ad-5 into the anterior cul-de-sac and scarifying the conjunctiva (4 scratches) with a 25-gauge needle (Group "C+": 8 eyes).</p><p>b. Installation of 20 ul containing 4 × 10<sup>7 </sup>pfu Ad-5 into the anterior cul-de-sac and scarifying the conjunctiva and cornea (8 scratches: 4 to conjunctiva, 4 to cornea) with a 25-gauge needle (Group "CC+": 8 eyes).</p><p>c. Injection of 20 ul containing 4 × 10<sup>7 </sup>pfu Ad-5 into the corneal stroma with a 25-gauge needle (Group "CI+": 8 eyes).</p><p>Sham-infected animals of all groups were attempted as controls:</p><p>2. Controls</p><p>a. Installation of 20 ul containing sterile HeLa media [Eagles Minimum Essential Medium (EMEM; ATCC, Manassas, VA) + 10% fetal calf serum (Hyclone, Logan, UT) + 1% penicillin/streptomycin (Sigma Chemical Company, St. Louis, MO) + 1% l-glutamine (Sigma Chemical Company, St. Louis, MO)] into the anterior cul-de-sac and scarifying the conjunctiva (4 scratches) with a 25-gauge needle (Group "C-": 8 eyes).</p><p>b. Installation of 20 ul containing sterile HeLa media into the anterior cul-de-sac and scarifying the conjunctiva and cornea (8 scratches: 4 to conjunctiva, 4 to cornea) with a 25-gauge needle (Group "CC-": 8 eyes).</p><p>c. Injection of 20 ul containing HeLa media into the corneal stroma with a 25-gauge needle (Group "CI-": 8 eyes).</p><p>3. Procedure</p><p>The eyes were randomly sorted into groups, examined by a masked observer (ophthalmologist) by slit-lamp biomicroscopy [Days 1 (preinfection), 3, 5, 7, 10, 12, 15, 17, 19, 23, 26, 30 post-infection (pi)] and scored using the grading system for rabbit conjunctival disease (Table <xref ref-type="table" rid="T1">1</xref>) derived from that originally described by Wander <italic>et al</italic>[<xref ref-type="bibr" rid="B15">15</xref>] for herpes simplex virus (HSV).</p><table-wrap position="float" id="T1"><label>Table 1</label><caption><p><italic>In Vitro</italic>Plaque Assays. Human Adenovirus type 5 (Ad-5) viral titers [plaque forming units (pfu)/preliminary flask] in Virucidal, Antiviral and Viral Inactivation assays in HeLa (human cervical carcinoma), A549 (human cervical carcinoma) and SIRC (rabbit corneal) cells.</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="center"><bold>Com-pound</bold></td><td align="center"><bold>Cell Line</bold></td><td align="center"><bold>Conc.</bold></td><td align="center" colspan="3"><bold>Virucidal</bold></td><td align="center" colspan="3"><bold>Antiviral</bold></td><td align="center" colspan="3"><bold>Viral Inactivation</bold></td></tr></thead><tbody><tr><td></td><td></td><td></td><td align="center"><bold>Average</bold></td><td></td><td align="center"><bold>Standard Deviation</bold></td><td align="center"><bold>Average</bold></td><td></td><td align="center"><bold>Standard Deviation</bold></td><td align="center"><bold>Average</bold></td><td></td><td align="center"><bold>Standard Deviation</bold></td></tr><tr><td colspan="12"><hr></hr></td></tr><tr><td></td><td></td><td align="center"><bold>(μg/ml)</bold></td><td align="center"><bold>(PFU)</bold></td><td></td><td align="center"><bold>(PFU)</bold></td><td align="center"><bold>(PFU)</bold></td><td></td><td align="center"><bold>(PFU)</bold></td><td align="center"><bold>(PFU)</bold></td><td></td><td align="center"><bold>(PFU)</bold></td></tr><tr><td colspan="12"><hr></hr></td></tr><tr><td align="center">Negative Control (None)</td><td align="center">HeLa</td><td align="center">0</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td align="center">A549</td><td align="center">0</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td align="center">SIRC</td><td align="center">0</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td align="center">CTC-96</td><td align="center">HeLa (human cervix)</td><td align="center">0</td><td align="center">5.00 × 10<sup>9</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">4.00 × 10<sup>10</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">5.33 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td></tr><tr><td></td><td></td><td align="center">5</td><td align="center">4.67 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">1.77 × 10<sup>10</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">4.67 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td></tr><tr><td></td><td></td><td align="center">10</td><td align="center">3.67 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">1.33 × 10<sup>10</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">2.47 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td></tr><tr><td></td><td></td><td align="center">25</td><td align="center">3.67 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">1.47 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">1.17 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td></tr><tr><td></td><td></td><td align="center">50</td><td align="center">6.67 × 10<sup>1</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>1</sup></td><td align="center">1.00 × 10<sup>2</sup></td><td align="center">±</td><td align="center">1.00 × 10<sup>2</sup></td><td align="center">3.33 × 10<sup>1</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>1</sup></td></tr><tr><td></td><td></td><td align="center">100</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td></td><td align="center">250</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td align="center">A549 (human lung)</td><td align="center">0</td><td align="center">3.07 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">2.47 × 10<sup>10</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">3.33E+09</td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td></tr><tr><td></td><td></td><td align="center">5</td><td align="center">2.77 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">1.17 × 10<sup>10</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">2.77E+09</td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td></tr><tr><td></td><td></td><td align="center">10</td><td align="center">2.27 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">8.67 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">1.47E+09</td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td></tr><tr><td></td><td></td><td align="center">25</td><td align="center">2.10 × 10<sup>9</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">9.33 × 10<sup>7</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>6</sup></td><td align="center">6.67E+08</td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td></tr><tr><td></td><td></td><td align="center">50</td><td align="center">6.67 × 10<sup>1</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>1</sup></td><td align="center">6.67 × 10<sup>1</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>1</sup></td><td align="center">3.33E+01</td><td align="center">±</td><td align="center">5.77 × 10<sup>1</sup></td></tr><tr><td></td><td></td><td align="center">100</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td></td><td align="center">250</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td align="center">SIRC (rabbit cornea)</td><td align="center">0</td><td align="center">5.67 × 10<sup>8</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">3.67 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">5.33 × 10<sup>8</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td></tr><tr><td></td><td></td><td align="center">5</td><td align="center">4.67 × 10<sup>8</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">3.67 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>8</sup></td><td align="center">3.00 × 10<sup>8</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td></tr><tr><td></td><td></td><td align="center">10</td><td align="center">4.33 × 10<sup>8</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">3.10 × 10<sup>9</sup></td><td align="center">±</td><td align="center">1.00 × 10<sup>8</sup></td><td align="center">3.33 × 10<sup>8</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td></tr><tr><td></td><td></td><td align="center">25</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">2.03 × 10<sup>9</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>7</sup></td><td align="center">1.00 × 10<sup>7</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td></td><td align="center">50</td><td align="center">3.33 × 10<sup>1</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>1</sup></td><td align="center">3.33 × 10<sup>1</sup></td><td align="center">±</td><td align="center">5.77 × 10<sup>1</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td></td><td align="center">100</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr><tr><td></td><td></td><td align="center">250</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">0.00 × 10<sup>0</sup></td><td align="center">±</td><td align="center">0.00 × 10<sup>0</sup></td></tr></tbody></table></table-wrap><p>Corneal swabs (Viral Culturettes, Baxter, Deerfield, IL) were taken for viral plaque assay twice per week [Days 1 (preinfection), 3, 5, 7, 10, 12, 15, 17, 19, 23, 26 and 30 pi]. Briefly, sterile culturettes were gently swabbed across the corneas (one per cornea) and the swabbed culturettes immediately eluted with 0.5 ml of Hanks balanced salt solution and serial 1:10 dilutions prepared for virus assays. Neither sterile culturettes nor cultures received any other manipulation. Culturettes were not frozen prior to titration. Plaque assays were performed in triplicate to ensure reproducibility.</p></sec><sec><title>B. CTC testing</title><p>1. Infection</p><p>The conjunctivae and corneas were infected with 20 ul containing 4 × 10<sup>7 </sup>pfu of virus into the anterior cul-de-sac and then scarified with a 25-gauge needle (8 scratches: 4 to conjunctiva, 4 to cornea: exactly as in Group "CC+" above)[<xref ref-type="bibr" rid="B16">16</xref>].</p><p>2. Treatment Groups (4 groups)</p><p>The animals were sorted into treatment groups of matched disease severities on day 7 post-infection, when all eyes showed disease signs, averaging 2–3, suggestive of acute conjunctivitis.</p><p>Grouped animals were treated, both eyes identically, with instillation of a 20 ul drop per eye of:</p><p>a. Placebo treatment (formulation without active ingredient), nine times per day [9x/day ("-"): 2 eyes],</p><p>b. CTC-96, 50 ug/ml, 9x/day ("50/9"), 2 eyes,</p><p>c. CTC-96, 50 ug/ml, six times per day [6x/day ("50/6")], 2 eyes,</p><p>or</p><p>d. CTC-96, 25 ug/ml, 6x/day ("25/6"), 2 eyes.</p><p>The formulation consisted of: 45 mg/ml mannitol, 60 ug/ml benzalkonium chloride, 2 mg/ml 2-methylimidazole in aqueous solution (pH 7.4). Drug treatments were spread out over each day (7:30 a.m. to 9:30 p.m.), for 21 days of treatment (both eyes of each animal treated identically).</p><p>3. Procedure</p><p>The eyes were examined in a masked fashion by an ophthalmologist. They were graded clinically (Table <xref ref-type="table" rid="T1">1</xref>) and corneal swabs (Viral Culturettes, Baxter, Deerfield, IL) were taken for viral plaque assay twice per week [Days 7 (pretreatment on day of treatment initiation), 10, 14, 17, 20, 25, 28, 31, 35 and 38 post-infection (pi; Days 1, 3, 7, 10, 13, 18, 21, 24, 28 and 31 post-treatment initiation (pti)]. Briefly, sterile culturettes were gently swabbed across the corneas (one per cornea) and the swabbed culturettes immediately eluted with 0.5 ml of Hanks balanced salt solution and serial 1:10 dilutions prepared for virus assays. Neither sterile culturettes nor cultures received any other manipulation. Culturettes were not frozen prior to titration. Plaque assays were performed in triplicate to ensure reproducibility.</p><p>a. Standardized Viral Plaque Assays</p><p>Titers of active virions of adenovirus, scored on preconfluent HeLa cell culture monolayers and expressed as plaque forming units (pfus) were determined according to established procedures[<xref ref-type="bibr" rid="B17">17</xref>]. Monolayers were grown and maintained in "HeLa media" [Vitacell= s Eagle Minimum Essential Medium (30–2003; ATCC, Manassas, VA) + 10% fetal calf serum + 1% penicillin/streptomycin + 1% l-glutamine (all from Sigma chemical Co., St. Louis, MO). Three days after inoculation (at 37°C and 5% CO<sub>2</sub>) with serial ten-fold dilutions of analyte, plaques had formed and the plates were stained with 1% methylene blue, allowed to dry and the plaques counted under a phase contrast microscope in a masked fashion in that the examiner had no knowledge of the treatment protocol for the individual animals. The animals were examined in a unique, random order for each examination.</p><p>b. Ocular Grading System</p><p>Clinical disease was divided up into: conjunctival, corneal epithelial and stromal disease and iritis. Corneal epithelial disease was subdivided into SPK, epithelial adenovirus symptoms, pannus, and epithelial defect. Stromal disease was subdivided into: edema, melting, neovascularization and infiltrate. Both the area and severity of each subdivision was graded individually from 0 to +4. Area of involvement was represented in increasing amounts of 25% (0 = "normal cornea", +1 = ≤ 25%, +2 = >25%, ≤ 50%, +3 = > 50%, ≤ 75%, +4 = > 75%, ≤ 100%). While severity of each subdivision was graded while the severity of each subdivision was individually graded from 0 (normal cornea) to +4 (severe).</p><p>c. Data Analysis</p><p>aa. Clinical Data</p><p>While conjunctival, corneal epithelial, stromal and iritis data was collected clinically, for simplicity only conjunctival data was graphed as other findings were insignificant (≈0). Clinical scores (area × severity) were calculated and utilized for all statistical calculations.</p><p>Non parametric methods of statistical analysis were used because disease severity is graded by assigning numerical scores[<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B16">16</xref>] to observed lesions. The arithmetic means of such arbitrarily scaled scores should not be analyzed by parametic methods such as the student's t test. Instead, rank medians[<xref ref-type="bibr" rid="B18">18</xref>,<xref ref-type="bibr" rid="B19">19</xref>], a Wilcoxson-Mann-Whitney rank-sum test[<xref ref-type="bibr" rid="B19">19</xref>,<xref ref-type="bibr" rid="B20">20</xref>], as well as both Spearman[<xref ref-type="bibr" rid="B19">19</xref>,<xref ref-type="bibr" rid="B21">21</xref>] and Kendall[<xref ref-type="bibr" rid="B19">19</xref>,<xref ref-type="bibr" rid="B22">22</xref>] rank correlations, as per standard non-parametric statistics, were used for analysis of the rank nonparametric assigned scores because non-parametric statistical techniques permit evaluation of significance of chemotherapeutic efficacies when measured by changes in severity of disease signs. The results presented were considered significant at a rank median of ≤ 0.05 units.</p><p>bb. Virus Recovery Data</p><p>Viral titers (per ml) were analyzed for statistical differences, utilizing the mean ± standard deviation for each of the various groups in computer-generated two-tailed bivariant Student's t tests[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>] (GB-STAT, New England Software, Inc., College Station, TX, U.S.A.; SAS, SAS Institute Inc., Cary, NC, U.S.A.; and SPSS, SPSS Inc., Chicago, IL, U.S.A.)[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>] at all time points. Individual Fisher exact tests[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>] were also performed, as well as an overall chi-squared analysis[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>] and a correlation coefficient[<xref ref-type="bibr" rid="B13">13</xref>,<xref ref-type="bibr" rid="B14">14</xref>]. Two-tailed significance was established at a confidence level of 0.05 ≥ P ≥ 0.95.</p></sec></sec></sec></sec><sec><title>Results</title><sec><title>I. <italic>In vitro </italic>studies</title><p>Viral inactivation, virucidal and antiviral efficacy studies in tissue culture demonstrated CTC-96 to be effective against human Adenovirus type 5 in a dose-dependent fashion (Tables <xref ref-type="table" rid="T1">1</xref> and <xref ref-type="table" rid="T2">2</xref>). CTC-96 at doses ≥ 50 ug/ml shows high levels of antiviral efficacy in the viral inactivation, virucidal and antiviral (post-infection) models, against human Adenovirus type 5 in all three cell lines used. The standardized Viral Inactivation, Virucidal and Antiviral assays showed viral yield to be almost zero at a concentration of 50 ug/ml and virtually zero at 100 ug/ml for the two human cell types [HeLa & A549; virucidal and antiviral: 6.67 × 10<sup>1 </sup>pfu/flask (8-fold inactivation of virions: Table <xref ref-type="table" rid="T1">1</xref>)]. In rabbit corneal cells (SIRC), viral yield reached undetectable levels even at 50 ug/ml (Tables <xref ref-type="table" rid="T1">1</xref> and <xref ref-type="table" rid="T2">2</xref>). There was a precipitous, dose dependent inhibition between 25 and 50 ug/ml of CTC-96 and a modest effect at still lower concentrations (i.e. 2.47 × 10<sup>9 </sup>pfu at 10 ug/ml and 1.17 × 10<sup>9 </sup>pfu at 25 ug/ml; Table <xref ref-type="table" rid="T1">1</xref>) demonstrating, as previously mentioned, dose dependence.</p><table-wrap position="float" id="T2"><label>Table 2</label><caption><p>ED 50s & 90s of Doxovir™ to Adenovirus Type 5. Effective dose (ED) 50s and 90s of Doxovir™ (CTC-96) against Adenovirus type 5 in HeLa (human cervical carcinoma), A549 (human cervical carcinoma) and SIRC (rabbit corneal) cells during Virucidal, Antiviral and Viral Inactivation assays.</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td></td><td align="center" colspan="3">(MIC 50 (μg/ml)</td><td align="center" colspan="3">(MIC 90 (μg/ml)</td></tr></thead><tbody><tr><td align="center"><bold>HeLa (human cervical carcinoma)</bold></td><td align="center"><bold>Virucidal</bold></td><td align="center">33.0</td><td align="center">±</td><td align="center">6.9</td><td align="center">46.6</td><td align="center">±</td><td align="center">9.7</td></tr><tr><td></td><td align="center"><bold>Antiviral</bold></td><td align="center">4.5</td><td align="center">±</td><td align="center">0.1</td><td align="center">21.8</td><td align="center">±</td><td align="center">1.2</td></tr><tr><td></td><td align="center"><bold>Viral Inactivation</bold></td><td align="center">16.5</td><td align="center">±</td><td align="center">2.5</td><td align="center">43.0</td><td align="center">±</td><td align="center">1.3</td></tr><tr><td align="center"><bold>A549 (human lung carcinoma)</bold></td><td align="center"><bold>Virucidal</bold></td><td align="center">31.7</td><td align="center">±</td><td align="center">1.9</td><td align="center">46.3</td><td align="center">±</td><td align="center">2.6</td></tr><tr><td></td><td align="center"><bold>Antiviral</bold></td><td align="center">4.7</td><td align="center">±</td><td align="center">0.2</td><td align="center">20.8</td><td align="center">±</td><td align="center">1.9</td></tr><tr><td></td><td align="center"><bold>Viral Inactivation</bold></td><td align="center">4.2</td><td align="center">±</td><td align="center">1.1</td><td align="center">37.5</td><td align="center">±</td><td align="center">6.5</td></tr><tr><td align="center"><bold>SIRC (rabbit cornea)</bold></td><td align="center"><bold>Virucidal</bold></td><td align="center">13.6</td><td align="center">±</td><td align="center">2.3</td><td align="center">22.7</td><td align="center">±</td><td align="center">3.5</td></tr><tr><td></td><td align="center"><bold>Antiviral</bold></td><td align="center">27.3</td><td align="center">±</td><td align="center">8.3</td><td align="center">45.5</td><td align="center">±</td><td align="center">1.6</td></tr><tr><td></td><td align="center"><bold>Viral Inactivation</bold></td><td align="center">13.1</td><td align="center">±</td><td align="center">2.9</td><td align="center">23.9</td><td align="center">±</td><td align="center">4.9</td></tr></tbody></table></table-wrap><p>In all assays, the compound exhibited consistently effective antiviral effects against adenoviral infections in HeLa, A549 and SIRC cell types (Tables <xref ref-type="table" rid="T1">1</xref> and <xref ref-type="table" rid="T2">2</xref>) and statistically significant differences were observed in all assays (Viral Inactivation, Virucidal and Antiviral) between each of the three cell types. These differences were fairly constant for the various cell lines. For example, A549 viral titers were consistently approximately half (i.e. approx 50%) those of the identical assays run on HeLa cells (see Table <xref ref-type="table" rid="T1">1</xref>), while the titers of the rabbit corneal cells (SIRC) were consistently one sixth (i.e. approx 17%) those of the identical assays run on human lung carcinoma (A549; see Table <xref ref-type="table" rid="T1">1</xref>).</p></sec><sec><title>II. Animal studies</title><sec><title>A. Methods of viral infection</title><p>Clinical biomicroscopic examination and standardized antiviral plaque assay showed rabbit ocular infection with human Adenvirus type 5 via the corneal and conjunctival scarification method (Group "CC") produced an acute conjunctivitis (6/6 eyes) with occasional keratitis that appeared similar to human disease (Figure <xref ref-type="fig" rid="F2">2</xref>). The CI group produced lower conjunctival clinical scores than the CC group, but did yield rare stromal involvement (1/6 eyes) and greater iritis (5/6 eyes; Figure <xref ref-type="fig" rid="F2">2</xref>). Infection via the conjunctival scarification method (Group "C+") yielded conjunctivitis in 4/6 eyes (Figure <xref ref-type="fig" rid="F2">2</xref>). None of the control animals ("C-", "CC-", "CI-") developed any signs of conjunctivitis or keratitis.</p><fig position="float" id="F2"><label>Figure 2</label><caption><p>Pearson rank sum medians ± nonparametric equivalents of standard deviation of conjunctival clinical scores (clinical score = area × severity) of conjunctivae of model development animals by group. Group "C": corneal infection, conjunctival scarification; Group "CC": corneal infection, conjunctival and corneal scarification; Group "CI": stromal infection. "+" animals ("C+", "CC+", "CI+") were infected with active virus, "-" were sham-infected ("C-", "CC-", "CI-").</p></caption><graphic xlink:href="1471-2415-6-22-2"/></fig><p>The plaque assays matched the clinical disease (Figure <xref ref-type="fig" rid="F3">3</xref>). The corneal and conjunctival scarification group (CC) developed viral titers by day 3 post-infection comparable to those of the corneal injection group (CI; Figure <xref ref-type="fig" rid="F3">3</xref>). Bearing this out, the conjunctival scarification group (C) developed viral titers significantly less (other than on day 5 p.i.) than either of the other two groups (Figure <xref ref-type="fig" rid="F3">3</xref>). While viral clearance was achieved in the conjunctival scarification group (C) at day 12 p.i., it was not achieved until day 15 p.i. for CC or 17 for CI (Figure <xref ref-type="fig" rid="F3">3</xref>) in the other two.</p><fig position="float" id="F3"><label>Figure 3</label><caption><p>Averages ± standard deviations of infectious virus titers [plaque forming units (pfu)/ml] of model development animals by group. Group "C": corneal infection, conjunctival scarification; Group "CC": corneal infection, conjunctival and corneal scarification; Group "CI": stromal infection. "+" animals ("C+", "CC+", "CI+") were infected with active virus, "-" were sham-infected ("C-", "CC-", "CI-").</p></caption><graphic xlink:href="1471-2415-6-22-3"/></fig><p>Since the corneal and conjunctival scarification group (CC) consistently developed acute conjunctivitis similar to human disease this model was chosen to evaluate the CTC treatment for acute adenovirus conjunctivitis.</p></sec><sec><title>B. CTC testing</title><p>Both clinical biomicroscopic examination and standardized antiviral plaque assay showed CTC-96 at all concentrations and frequencies tested was effective against ocular disease caused by adenovirus type 5. The majority of clinical signs of infection were in the conjunctiva, with only minimal and occasional signs observable in the corneal epithelium and/or stroma.</p><p>Clinically, placebo-treated eyes initially showed a gradual worsening in clinical scores (up to a rank sum of 8) until Day 14 pi (Day 7 pti), after which they slowly resolved for another 3 weeks (Figure <xref ref-type="fig" rid="F4">4</xref>).</p><fig position="float" id="F4"><label>Figure 4</label><caption><p>Pearson rank sum medians ± nonparametric equivalents of standard deviations of conjunctival clinical scores (clinical score = area × severity) of CTC-treated animals by group. Placebo (squares, solid line); 50 ug/ml, 9x/day ["50/9" (triangles, dotted line)]; 50 ug/ml, 6x/day ["50/6" (diamonds, mixed solid/dotted line)]; 258 ug/ml, 6x/day ["25/6" (asterisks/circles, dashed line)].</p></caption><graphic xlink:href="1471-2415-6-22-4"/></fig><p>Animals treated with CTC-96 at 50 ug/ml, showed a steady improvement in clinical scores (Figure <xref ref-type="fig" rid="F4">4</xref>) in both of the dosing frequencies evaluated (9x/day and 6x/day). No adverse events were noted in any of the eyes and the clinical scores did not exceed 4–5 with these treatments. Statistical significance as compared to the placebo group was achieved for [Days 10, 14, 17, 20, 25 and 28 pi (Days 1, 3, 7, 10, 13, 18 and 21 pti; 50/9: Wilcoxin-Mann: p ≤ 0.03; 50/6: Wilcoxin-Mann: p ≤ 0.04). The marginal effect that the increased frequency of 9x/day appeared to have had over that of 6x/day was not statistically significant. Clinically, conjunctivae of both frequencies (9x/day and 6x/day: Figure <xref ref-type="fig" rid="F4">4</xref>) of CTC-96 at 50 ug/ml resolved on Days 28 – 31 pi (Days 21–24 pti) and were statistically indistinguishable in clinical severity until resolution (Wilcoxin-Mann: p ≥ 0.80).</p><p>In all cases, CTC-96 at 25 ug/ml, 6x/day, was effective against human adenovirus type 5 as compared to sham-treated eyes. Although only statistically significant at Days 10, 14 and 20 pi (Days 3, 7 and 13 pti), while the placebo-treated eyes were still visibly resolving clinically at Day 38 pi (the last exam day: Day31 pti), CTC-96 at 25 ug/ml, 6x/day, had visibly resolved by Day 35 (Day 28 pti; Figure <xref ref-type="fig" rid="F4">4</xref>).</p><p>The clinical observations in which placebo-treated eyes initially showed a gradual worsening, after which they slowly resolved, was confirmed by the plaque assays (Figure <xref ref-type="fig" rid="F5">5</xref>)]. Viral clearance was not achieved until Day 28 pi (Day 21 pti; Figure <xref ref-type="fig" rid="F5">5</xref>) in the placebo-treated eyes.</p><fig position="float" id="F5"><label>Figure 5</label><caption><p>Averages ± standard deviations of plaque assay viral titers [plaque forming units (pfu)/ml] of CTC-treated animals by group. Placebo (squares, solid line); 50 ug/ml, 9x/day ["50/9" (triangles, dotted line)]; 50 ug/ml, 6x/day ["50/6" (diamonds, mixed solid/dotted line)]; 258 ug/ml, 6x/day ["25/6" (asterisks/circles, dashed line)].</p></caption><graphic xlink:href="1471-2415-6-22-5"/></fig><p>The viral plaque assays also demonstrated that CTC-96 at 50 ug/ml in both of the dosing frequencies evaluated (9x/day and 6x/day), eliminated virus by Day 20 pi (Day 13 pti; Figure <xref ref-type="fig" rid="F5">5</xref>), whereas 25 ug/ml, 6x/day was significantly less effective [t test: p ≥ 0.01 (50 ug/ml, 9x/d > 50 ug/ml, 6x/d: p ≥ 0.01; 25 ug/ml: p = 0.02)]. While viral clearance occurred at Day 20 pi (Day 13 pti) at both of the dosage frequencies of 50 ug/ml (Figure <xref ref-type="fig" rid="F5">5</xref>), viral titers were already significantly lowered (ie there was a significant difference between the two) at Day 10 pi (Day 3 pti) as compared to that of the control [t test: p = 0.02 (50 ug/ml, 9x/d < 50 ug/ml, 6x/d)].</p><p>At 25 ug/ml, 6x/day, CTC-96 was effective against human adenovirus type 5 as compared to placebo-treated eyes and viral clearance in this group was not achieved until Day 28 pi (Day 21 pti), as compared to Day 31 pi (Day 24 pti) in the placebo-treated eyes (Figure <xref ref-type="fig" rid="F5">5</xref>). Thus, in conclusion, although some therapeutic effects of CTC-96 at 25 ug/ml are observed at 6x/day, it was considerably inferior to 50 ug/ml (50 ug/ml, 9x/day. 50 ug/ml, 6x/day >> 25 ug/ml, 6x/day).</p></sec></sec></sec><sec><title>Discussion</title><p>This study demonstrates that the cobalt chelate CTC-96 (designated Doxovir™) is an effective antiviral agent against human Adenovirus type 5 both in cell culture and in rabbits. Adenovirus type 5 infection of human and rabbits cells was effectively treated by CTC-96 and primary ocular human adenovirus type 5-induced conjunctivitis in a New Zealand White rabbit model was effectively treated by topical administration with CTC-96. Treatment considerably reduced both disease severity and corneal surface viral titers, significantly shortening the course of the disease and the symptoms in parallel. The data presented here indicate that against Ad-5, six (6) daily applications of CTC-96 at 50 ug/ml are approximately as effective as nine (9) daily applications. Six (6) daily applications of 25 ug/ml CTC-96 are inferior to either of the other two, but significantly better than placebo. It is most likely that this demonstrates a dose responsiveness, although the possibility of a low sample population size producing the observed effect cannot be entirely discounted.</p><p>Currently there are no specific antiviral agents for topical or systemic treatment of adenoviral infections[<xref ref-type="bibr" rid="B2">2</xref>]. Antiviral agents active against HSV and interferons have been used in the treatment of Adenovirus infections with only limited effect[<xref ref-type="bibr" rid="B23">23</xref>,<xref ref-type="bibr" rid="B24">24</xref>]. Thus far, previous recent <italic>in vitro</italic>experiments with the cobalt chelate CTC-96, in addition to showing strong anti-herpes simplex efficacy[<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B8">8</xref>], have demonstrated anti-human adenovirus type 5 efficacy of CTC-96[<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B6">6</xref>]. In addition, newer nucleoside analogs have recently been found to exhibit some inhibitory activity against Adenovirus <italic>in vitro</italic>and <italic>in vivo </italic>[<xref ref-type="bibr" rid="B24">24</xref>-<xref ref-type="bibr" rid="B27">27</xref>].</p><p>Topical cidofovir, a nucleoside analog, was shown to be an effective topical agent against against adenovirus in non-human systems and may have some effect in man[<xref ref-type="bibr" rid="B28">28</xref>,<xref ref-type="bibr" rid="B29">29</xref>], however it has proven to be too toxic in human eyes. Interestingly, despite the relatively short experience with the drug, cidofovir-resistant adenovirus variants have already been observed[<xref ref-type="bibr" rid="B30">30</xref>]. Since CTC-96 has been found to have a very different mechanism of action than nucleoside analogs, it has been shown to be effective against drug-resistant mutants of herpes viruses (unpublished results: Redox Pharmaceutical Corporation).</p><p>In the past it has been difficult to define a reliable animal model of adenovirus ocular disease. For the most part previous studies of the pathogenesis and treatment of ocular adenovirus infections have been limited by the narrow host range exhibited by human adenoviruses. About half of the greater than 50 serotypes of human adenovirus are known to cause ocular disease in humans [<xref ref-type="bibr" rid="B3">3</xref>-<xref ref-type="bibr" rid="B6">6</xref>] yet only three have been previously found to have the ability to extend their host range to permit replication in the eyes of New Zealand rabbits [<xref ref-type="bibr" rid="B3">3</xref>-<xref ref-type="bibr" rid="B6">6</xref>].</p><p>We developed a simple, reliable method of inducing adenovirus infection that mimics human disease using a rabbit model. Previous animal models of adenovirus employed the injection of active virus into the corneal stroma first described by Gordon, <italic>et al</italic>[<xref ref-type="bibr" rid="B3">3</xref>]. This "corneal injection method" is technically difficult and neither mimics nor even approximates the usual routes of human infection. For this reason, we evaluated three (3) methods of infecting the ocular surface with adenovirus to determine which most mimicked human adenovirus ocular disease, and was repeatable and easy to create: 1) the "conjunctival scarification method" involving the installation of 20 ul containing 2 × 10<sup>7 </sup>pfu Ad-5 into the anterior cul-de-sac with a scarified conjunctiva (Group "C+"); 2) the "corneal and conjunctival scarification method" involving the installation of 20 ul containing 2 × 10<sup>7 </sup>pfu Ad-5 into the anterior cul-de-sac after both the conjunctiva and cornea were scarified (Group "CC+"); and 3) the "corneal injection method" involving the intrastromal injection of 20 ul containing 2 × 10<sup>7 </sup>pfu of Ad-5 into the cornea (Group "CI+"). Concurrent controls with conjunctival scarification (Group "C-") or corneal and conjunctival scarification (Group "CC-") did not develop acute conjunctivitis or keratitis as evaluated by a masked observer.</p><p>When compared both clinically and via the quantification of viral titers (standardized plaque assay), infection via the corneal and conjunctival scarification method (Group "CC") produced a conjunctival infection with occasional keratitis most similar to human disease. It was easier and more reliable than the corneal injection (Group "CI") or the conjunctival scarification methods. The corneal injection method utilized by other authors[<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B4">4</xref>] yields more stromal involvement than the "CC" group ["CI": 2/6 eyes (33%) versus "CC": 0/6 eyes (0%)] but is more difficult to create.</p><p>The results presented here demonstrate that adenovirus infection via the corneal and conjunctival scarification method (Group "CC") produced the best rabbit model to study conjunctival infection, while the corneal injection method (Group "CI") produced more stromal involvement and iritis making it the model of choice to study stromal disease and iritis.</p><p>As previously mentioned, a number of cobalt complexes (CTC compounds) were found to exhibit potent <italic>in vitro </italic>and <italic>in vivo </italic>activity against herpes group viruses [<xref ref-type="bibr" rid="B7">7</xref>]. The CTC compound in this study, CTC-96, has already been shown to exhibit pronounced efficacy in the treatment of HSV-1 in cell culture studies on vero cells[<xref ref-type="bibr" rid="B7">7</xref>] as well as in topical therapy of HSV-1-induced epithelial and stromal disease in the rabbit eye[<xref ref-type="bibr" rid="B7">7</xref>]. Safety data and toxicity studies have shown the agent to exhibit minimal signs of toxicity and to not be readily absorbed during topical application[<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B8">8</xref>] (and unpublished data: Redox Pharmaceutical Corporation). This previous research discounted the possibility that the effect of CTC-96 is due to toxicity by evaluating the effects of CTC-96 both in tissue culture [5–500 ug/ml (no toxicity ≤50 ug/ml, trace and transient at 100 ug/ml): Vero (African Green monkey kidney fibroblast cell line), HeLa (human cervical cancer cell line), A549 (human alveolar carcinoma cell line), SIRC (immortalized rabbit corneal epithelial cell line) and HeP2 (human laryngeal squamous cancer cell line)[<xref ref-type="bibr" rid="B5">5</xref>-<xref ref-type="bibr" rid="B8">8</xref>] and unpublished observations] and systemic applications (IP and IV) in mice (unpublished observations), as well as via topical administration of 20 ul drops 3x, 4x, 5x, 6x and 9x/day on normal uninfected rabbit (New Zealand White rabbits) eyes [total of 240 eyes,10–500 ug/ml, checked clinically and histologically[<xref ref-type="bibr" rid="B7">7</xref>] (and unpublished observations)] although it is still possible that diseased eyes may be more sensitive than normal ones. Phase II studies for the treatment of HSV-1 are presently underway.</p><p>There is a noticeable difference in the efficacious dose against Adenovirus than that against Herpesvirus, where concentrations as low as 5 ug/ml were found to possess strong antiviral activity <italic>in vitro</italic>[<xref ref-type="bibr" rid="B7">7</xref>] and in rabbit eyes with an IC<sub>50 </sub>for HSV-1 <italic>in vitro</italic>of 0.7 ug/ml (Winicov, Redox unpublished observations). Current findings on human Adenovirus type 5 (Ad-5) demonstrated that a concentration of 50 ug/ml of CTC-96 is required to obtain full antiviral efficacy. We attribute at least part of this difference to the effect of the drug on membrane proteins of enveloped viruses (e.g. HSV-1 and HIV) rather than capsid or intra-capsid proteins. CTC-96 interacts with a membrane glycoprotein of HIV that results in virus inactivation (in preparation). Since Adenovirus does not possess a membrane, it is possible that the inhibitory action of the drug involves binding to one or more capsid or intra-capsid proteins that are less accessible to the drug than membrane surface proteins in enveloped viruses.</p><p>Thus, CTC-96 has a completely different mode of action than nucleotide analogs as well as other DNA replication inhibitors. The virucidal effect of CTC-96 can be due to various modes of action of the drug, but the major activity of the drug entails disruption of protein conformation through binding to methyl imidazole nitrogens of specific histidines in several protein types[<xref ref-type="bibr" rid="B31">31</xref>]. But the compound is also known to be a superoxide scavenger[<xref ref-type="bibr" rid="B32">32</xref>] (and unpublished data: Redox Pharmaceutical Corporation).</p><p>In conclusion, the corneal and conjunctival scarification method of adenovirus infection has proven to be a usable model of ocular Ad-5 infection in the rabbit as it approximates the signs and symptoms of adenvirus-derived conjunctivitis in human eyes. In addition, topical CTC-96 has demonstrated significant anti-viral activity against human adenovirus type 5 (Ad-5) in these <italic>in vivo </italic>models both by clinical examination and viral plaque assays.</p><p>Clearly more studies will be needed to evaluate completely the efficacy and safety of CTC-96 and clinical trials in humans will be necessary to determine the safety and efficacy of CTC-96 for treating human adenoviral keratoconjunctivitis.</p></sec><sec><title>Conclusion</title><sec><title>I. <italic>In vitro </italic>studies</title><p>A. No toxic effects observed in HeLa, A549 or SIRC cells for CTC-96 at the therapeutic concentration of 10 ug/ml.</p><p>B. CTC-96 demonstrated virucidal activity against Adenovirus type 5 in tissue culture with HeLa, A549 and SIRC cell lines.</p><p>1. In all cases, peak efficacy was observed in concentrations ≥ 50 μg/ml, while lower concentrations showed decreasing dose-responsiveness.</p><p>2. Greater viral replication/yield and antiviral activity was observed in HeLa >A549 > SIRC.</p></sec><sec><title>II. Animal studies</title><sec><title>A. In the Model Development experimentation (testing of the method of infection)</title><p>1. Infection via corneal and conjunctival scarification method (Group "CC") produced a conjunctival infection with occasional keratitis similar to human disease.</p><p>2. The corneal injection method (Group "CI") yielded primarily stromal involvement.</p><p>3. The conjunctival scarification method (Group "C") did not consistently produce acute conjunctivitis.</p><p>4. The conjunctival scarification method (Group "C") did not produce any keratitis.</p></sec><sec><title>B. In the CTC Efficacy experimentation (testing of the efficacy of human adenovirus type 5 in a NZW rabbit model)</title><p>1. Conjunctivitis was seen in all eyes by day 8 post-infection.</p><p>2. Resolution was fastest in treatment group "50/9" where most animals resolved by Day 21 post-treatment initiation (pti):</p><p>"50/9" ≈ "50/6" > "25/6" > "-".</p><p>2. The efficacy of CTC-96 in treatment group A50/6" was not statistically less effective than that observed in treatment group A50/9": most animals resolving by Day 21 pti.</p><p>3. The degree of severity of the conjunctiva was lower in all treatment groups as compared to the control eyes.</p><p>4. Little corneal or intra-ocular changes were noted in any of the treatment groups.</p></sec></sec></sec><sec><title>Competing interests</title><p>Supported by Redox Pharmaceutical Corp., grants from Research to Prevent Blindness, Inc., New York, NY; a research grant from Turobiner-Finley Grant Fund, New York, NY; and in part by EY01867 from the National Eye Institute, National Institutes of Health, Bethesda, MD.</p></sec><sec><title>Authors' contributions</title><p>SPE participated in the design, experimentation and drafted the manuscript. YYP participated in the experimentation. DG conceived of the study with PAA. IW participated in the coordination of the study. CS participated in the experimentation. KJK participated in the experimentation. PAA conceived of the study with DG. All authors read and approved the final manuscript.</p></sec><sec><title>Pre-publication history</title><p>The pre-publication history for this paper can be accessed here:</p><p><ext-link ext-link-type="uri" xlink:href="http://www.biomedcentral.com/1471-2415/6/22/prepub"/></p></sec>
Combined IL-21 and Low-Dose IL-2 therapy induces anti-tumor immunity and long-term curative effects in a murine melanoma tumor model	<sec><title>Background</title><p>In vivo studies have recently demonstrated that interleukin 21 (IL-21) enhances the anti-tumor function of T-cells and NK cells in murine tumor models, and the combined use of IL-21 and IL-15 has resulted in prolonged tumor regression and survival in mice with previously established tumors. However, the combined anti-tumor effects of IL-21 and low dose IL-2 have not been studied even though IL-2 has been approved for human use, and, at low dose administration, stimulates the proliferation of memory T cells, and does not significantly increase antigen-induced apoptosis or regulatory T cell (Treg) expansion. This study examined whether recombinant IL-21 alone or in combination with low-dose IL-2 could improve the in vivo anti-tumor function of naïve, tumor-antigen specific CD8<sup>+ </sup>T cells in a gp100<sub>25–33 </sub>T cell receptor transgenic pmel murine melanoma model.</p></sec><sec sec-type="methods"><title>Methods</title><p>Congenic C57BL/6 (Ly5.2) mice bearing subcutaneous B16F10 melanoma tumors were sublethally irradiated to induce lymphopenia. After irradiation naive pmel splenocytes were adoptively transferred, and mice were immunized with bone marrow-derived dendritic cells pulsed with human gp100<sub>25–33 </sub>(hgp100<sub>25–33</sub>). Seven days after vaccination groups of mice received 5 consecutive days of intraperitoneal administration of IL-2 alone (20 × 10<sup>3 </sup>IU), IL-21 alone (20 μg) or IL-21 and IL-2. Control animals received no cytokine therapy.</p></sec><sec><title>Results</title><p>IL-21 alone and IL-2 alone both delayed tumor progression, but only IL-21 significantly augmented long-term survival (20%) compared to the control group. However, combination therapy with IL-21 and IL-2 resulted in the highest long-term (>150 days) tumor-free survival frequency of 46%. Animals that were tumor-free for > 150 days demonstrated tumor-specific protection after rechallenge with B16F10 melanoma cells. At peak expansion (21 days post vaccination), the combination of IL-21 plus IL-2 resulted in a 2- to 3-fold higher absolute number of circulating tumor antigen-specific pmel CD8<sup>+ </sup>T cells than was stimulated by IL-2 or IL-21 alone. Pmel CD8<sup>+ </sup>T cells were predominantly partitioned into central memory (CD62L<sup>+</sup>/CD127<sup>+</sup>) or effector-memory (CD62L<sup>-</sup>/CD127<sup>+</sup>) phenotypes by day 28-post vaccination in IL-21 + IL-2 treated mice.</p></sec><sec><title>Conclusion</title><p>These observations support the potential use of IL-21 and low-dose IL-2 therapy in combination with a tumor-antigen vaccine and lymphopenic conditioning in future cancer clinical trials to maintain high numbers of anti-tumor memory CD8<sup>+ </sup>T cells with the potential to sustain long term tumor regression and survival.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>He</surname><given-names>Hong</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Wisner</surname><given-names>Preya</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" contrib-type="author"><name><surname>Yang</surname><given-names>Guojun</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A4" contrib-type="author"><name><surname>Hu</surname><given-names>Hong-Ming</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A5" contrib-type="author"><name><surname>Haley</surname><given-names>Dan</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A6" contrib-type="author"><name><surname>Miller</surname><given-names>William</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A7" contrib-type="author"><name><surname>O'Hara</surname><given-names>Aisling</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A8" contrib-type="author"><name><surname>Alvord</surname><given-names>W Gregory</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A9" contrib-type="author"><name><surname>Clegg</surname><given-names>Christopher H</given-names></name><xref ref-type="aff" rid="I3">3</xref><email>[email protected]</email></contrib><contrib id="A10" contrib-type="author"><name><surname>Fox</surname><given-names>Bernard A</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A11" contrib-type="author"><name><surname>Urba</surname><given-names>Walter J</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A12" corresp="yes" contrib-type="author"><name><surname>Walker</surname><given-names>Edwin B</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	Journal of Translational Medicine	<sec><title>Background</title><p>IL-21 has potent immunomodulatory effects on T cells and NK cells [<xref ref-type="bibr" rid="B1">1</xref>]. However, the current understanding of the effects of IL-21 on the regulation of both antigen-independent homeostatic and antigen-stimulated proliferation and activation of naive and memory CD8<sup>+ </sup>T cells is confounded by conflicting data. IL-21 has been reported both to synergize with IL-15 to increase homeostatic antigen-independent proliferation of naive and memory CD8<sup>+ </sup>T cells [<xref ref-type="bibr" rid="B2">2</xref>,<xref ref-type="bibr" rid="B3">3</xref>], and to inhibit IL-15 induced homeostatic proliferation of memory T cells [<xref ref-type="bibr" rid="B4">4</xref>]. While evidence suggests IL-21 enhances primary antigen-stimulated T-cell proliferation and functional activation [<xref ref-type="bibr" rid="B3">3</xref>-<xref ref-type="bibr" rid="B6">6</xref>], there is disagreement over whether this effect occurs during priming [<xref ref-type="bibr" rid="B5">5</xref>], or during the late expansion phase of antigen-specific proliferation [<xref ref-type="bibr" rid="B6">6</xref>]. Recent data suggests IL-21 primarily increases the frequency and absolute number of in vitro stimulated antigen-specific CD8<sup>+ </sup>T cells by enhancing the rate of proliferation rather than by reducing apoptosis [<xref ref-type="bibr" rid="B5">5</xref>]. Alternatively, other studies indicate IL-21 maintains increased numbers of antigen-stimulated effector and long-term memory CD8<sup>+ </sup>T cells by reducing apoptosis while maintaining a low rate of cell division [<xref ref-type="bibr" rid="B6">6</xref>]. These conflicting data may be attributed to differences in the in vitro and in vivo models employed in the individual studies. However, they suggest that much work remains in the effort to elucidate the mechanism of IL-21 regulation of T-cell immune responses, including any potential role it may play in the regulation of T-cell mediated anti-tumor immunity through the induction of functionally active tumor-specific effector and memory T cells (Reviewed in [<xref ref-type="bibr" rid="B1">1</xref>]).</p><p>In vivo data demonstrate that IL-21 enhances the anti-tumor function of T-cells and NK cells. Several studies have shown that genetically engineered IL-21-secreting murine tumors activate potent NK and CD8<sup>+ </sup>T cell mediated anti-tumor responses [<xref ref-type="bibr" rid="B7">7</xref>-<xref ref-type="bibr" rid="B9">9</xref>]. IL-21 was also compared to IL-2 and IL-15 injected intraperitoneally (IP) subsequent to challenge with the ova-expressing E.G7 thymoma in either naive C57BL/6 mice, or in mice infused with OT-1 transgenic CD8<sup>+ </sup>T cells [<xref ref-type="bibr" rid="B6">6</xref>]. In both models, IL-2 and IL-15 delayed tumor growth, but only IL-21 resulted in significant prevention of tumor progression and improved survival beyond 50 days from tumor challenge. The combination of IL-21 and IL-15 was also administered IP subsequent to human gp100<sub>25–33 </sub>(hgp100<sub>25–33</sub>) peptide vaccination of lymphopenic, tumor-bearing (B16F10 melanoma) C57BL/6 mice [<xref ref-type="bibr" rid="B3">3</xref>]. Before vaccination mice received in vitro stimulated (IVS) pmel transgenic (Tg) CD8<sup>+ </sup>T cells, which have a T-cell receptor specific for the murine H-2D<sup>b </sup>restricted gp100<sub>25–33 </sub>peptide of the B16F10 melanoma associated gp100 protein, and strongly cross react with hgp100<sub>25–33 </sub>[<xref ref-type="bibr" rid="B10">10</xref>]. The combination of IL-21 + IL-15 resulted in prolonged tumor regression and survival out to 32 days. Mice treated with the vaccine plus IL-15 or IL-21 alone all died of tumor within 32 days of treatment. No data was presented demonstrating long-term survival (>60 days) of mice treated with IL-21 + IL-15. Notably, IL-2 was not tested in vivo in combination with IL-21 in this study primarily because the combination of these two cytokines failed to drive antigen-independent homeostatic proliferation of murine CD8<sup>+ </sup>splenocytes in vitro [<xref ref-type="bibr" rid="B3">3</xref>], and the concern that high-dose IL-2 has been shown to decrease memory CD8<sup>+ </sup>T-cell function by inducing regulatory T-cells and increasing antigen-driven apoptosis [<xref ref-type="bibr" rid="B11">11</xref>-<xref ref-type="bibr" rid="B14">14</xref>]. However, other data have demonstrated that the in vitro maintenance of antigen-stimulated murine T cells in low-dose IL-2 resulted in the proliferative expansion of cells with a memory phenotype similar to that induced by IL-15; these cells were capable of long-term survival in vivo, and potent proliferative and functional responses after secondary in vivo challenge [<xref ref-type="bibr" rid="B15">15</xref>]. Thus, treatment with IL-21 + low-dose IL-2 after the adoptive transfer of pmel Tg CD8<sup>+ </sup>T cells into mice with established B16 melanoma tumors (pmel Tg/B16 model) might yield anti-tumor effects comparable to or greater than those observed with IL-21+IL-15.</p><p>The pmel Tg/B16 melanoma model has been employed repeatedly in recent studies to model immunotherapeutic strategies for treatment of pre-existing disease. Most of these experiments have focused on the adoptive transfer of antigen-educated pmel CD8<sup>+ </sup>splenocytes stimulated in vitro with hgp100<sub>25–33 </sub>peptide and IL-2, and the subsequent in vivo expansion of these highly activated cells in tumor-bearing mice with hgp100<sub>25–33 </sub>vaccination and follow-on cytokine therapy [<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B16">16</xref>-<xref ref-type="bibr" rid="B18">18</xref>]. While this strategy has resulted in significant tumor regression, little data has been presented demonstrating long-term curative effects (≥ 100 days) with IL-2 alone [<xref ref-type="bibr" rid="B16">16</xref>,<xref ref-type="bibr" rid="B17">17</xref>], IL-15 alone [<xref ref-type="bibr" rid="B18">18</xref>], or the combination of IL-15 and IL-21 [<xref ref-type="bibr" rid="B3">3</xref>] – except in mice with small tumors (≤ 10 mm<sup>2</sup>) at the time of treatment, or after the adoptive transfer of very large numbers (10<sup>7</sup>) of IVS pmel CD8<sup>+ </sup>T cells [<xref ref-type="bibr" rid="B16">16</xref>]. The adoptive transfer of highly activated IVS T-cells may not produce long-term anti-tumor curative effects due to functional and proliferative exhaustion [<xref ref-type="bibr" rid="B19">19</xref>-<xref ref-type="bibr" rid="B21">21</xref>]. This conclusion is supported by a series of provocative experiments using the pmel Tg/B16 model in which the adoptive transfer of naive or "early" IVS-effector pmel CD8<sup>+ </sup>T cells into lymphopenic tumor-bearing recipient mice resulted in much more durable tumor regression following hgp100<sub>25–33 </sub>vaccination and IL-2 therapy compared to the poor therapeutic effects observed using IVS expanded, highly stimulated intermediate or late-stage effector T cells [<xref ref-type="bibr" rid="B22">22</xref>]. Thus, as described herein, the adoptive transfer of naïve, tumor antigen-specific CD8<sup>+ </sup>T cells in this model prior to vaccine and IL-21 + low-dose IL-2 therapy may provide a better opportunity to study methods of generating both effector and long-term anti-tumor memory T-cell function than can be achieved with highly activated IVS T cells.</p><p>There have been few published studies describing the anti-tumor effects of recombinant IL-21 in conjunction with hgp100<sub>25–33 </sub>vaccination in B16F10 tumor-bearing lymphopenic mice receiving naive pmel Tg CD8<sup>+ </sup>T cells. There have, to our knowledge, been no reports describing the potential synergy of IL-21 + low-dose IL-2 in this or other tumor models. Herein we describe the long-term curative effect of combined IL-21 + low-dose IL-2 cytokine therapy in lymphopenic, tumor-bearing C57BL/6 mice infused with naive pmel splenocytes and vaccinated with the hgp100<sub>25–33 </sub>melanoma peptide.</p></sec><sec sec-type="materials\|methods"><title>Materials and methods</title><sec><title>Mice</title><p>Congenic C57BL/6 (Ly5.2) mice (Charles River Laboratories, Inc. NCI-Frederick) bred at the Earle A. Chiles Research Institute (EACRI) served as recipient mice for tumor inoculation and pmel Tg CD8<sup>+ </sup>T-cell adoptive transfer in all experiments. Pmel Tg mice express a TCR specific for an H-2D<sup>b </sup>epitope (gp100<sub>25–33</sub>) of the melanoma-associated gp100 protein on the C57BL/6 (Ly5.1) background [<xref ref-type="bibr" rid="B10">10</xref>,<xref ref-type="bibr" rid="B16">16</xref>]. Virtually all (>90%) CD8<sup>+</sup>T cells in pmel Tg mice were Vβ13<sup>+</sup>, and were also distinguishable by a monoclonal antibody specific for a point mutation of the CD45 epitope (CD45.2). Pmel Tg mice were the generous gift of Dr. Nicholas Restifo (Surgery Branch, NCI-NIH) and were bred at EACRI. Male or female mice (10–16 weeks) were used in separate experiments. All mice were treated according to the regulations and guidelines of the Institutional Animal Care and Use Committee.</p></sec><sec><title>Tumor cell lines, tumor inoculation and in vivo measurement</title><p>The B16 F10 melanoma and 3LL (Lewis Lung) cell lines were obtained from ATCC. Both cell lines were maintained in complete medium consisting of RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum (Biofluids, Rockville, MD), 0.03% L-glutamine, 100 μg/ml streptomycin, 100 μg/ml penicillin, 50 μg/ml gentamicin sulfate and 50 mmol 2-mercaptoethanol. Tumor was established by injecting mice subcutaneously in the flank with 2 × 10<sup>5 </sup>B16 F10 melanoma cells in 0.1 ml of phosphate buffered balanced salt solution (PBS). In tumor rechallenge experiments mice were injected subcutaneously in the opposite flank with 5.0 × 10<sup>5 </sup>B16 F10 melanoma cells or 5.0 × 10<sup>4 </sup>3LL tumor cells in 0.1 ml of PBS. Tumor growth was monitored three times a week by measurement of two perpendicular diameters using a digital caliper. The products of the perpendicular diameters plotted for multiple animals are presented as the mean mm<sup>2 </sup>± SEM. Mice were sacrificed when tumors exceeded 200 mm<sup>2</sup>. Survival was analyzed by using Kaplan-Meier statistics.</p></sec><sec><title>Adoptive cell transfer</title><p>Splenocytes from pmel transgenic mice were depleted of erythrocytes with ACK lysing buffer, and 4 × 10<sup>6 </sup>were adoptively transferred via I.V. injection into C57BL/6 (Ly5.2) mice. Pmel spleen cells contained on average 20%–25% (~1 × 10<sup>6</sup>) naive CD45.2<sup>+ </sup>CD8<sup>+ </sup>T cells.</p></sec><sec><title>Peptides</title><p>The H-2D<sup>b</sup>-restricted hgp100<sub>25–33 </sub>peptide, (KVPRNQDWL) was used as the CD8<sup>+ </sup>immunogen [<xref ref-type="bibr" rid="B16">16</xref>], and the I-A<sup>b</sup>-restricted epitope of Plasmodium falciparum (NANPNVDPNANP), hereafter referred to as "NV", was used as a CD4<sup>+</sup>directed helper peptide [<xref ref-type="bibr" rid="B23">23</xref>,<xref ref-type="bibr" rid="B24">24</xref>]in all experiments. The peptides were made by Invitrogen Inc. and purified by reverse-phase high-performance liquid chromatography. Purity of >99% was confirmed by mass spectrometry.</p></sec><sec><title>Preparation of dendritic cells</title><p>Bone marrow-derived murine dendritic cells (DCs) were generated as described previously [<xref ref-type="bibr" rid="B25">25</xref>]. Briefly, cells from the femur of C57BL/6 mice were grown at 1 × 10<sup>6 </sup>cells/ml in RPMI 1640 complete medium supplemented with 25 ng/ml of murine GM-CSF (PeproTech, Rocky Hill, NJ). Fresh medium supplemented with GM-CSF was added on day 3, and all loosely adherent cells were transferred to petri dishes on day 6. Three days later, nonadherent and loosely adherent cells were harvested, washed, and frozen in 10% DMSO and 90% FCS in liquid nitrogen at 10<sup>7 </sup>cells per vial. Frozen DCs were thawed and pulsed for 2 hours at 37°C with hgp100<sub>25–33 </sub>(1 μg/ml), and 1 μg/ml of the NV peptide in complete medium. DCs were washed three times with PBS before injection. DC purity and maturation were confirmed by flow cytometry analysis after staining with antibodies against MHC class I (H-2D<sup>b</sup>, H-2K<sup>b</sup>), class II (I-A<sup>b</sup>), CD11c, CD40, CD80, CD86. (BD Biosciences PharMingen, San Diego, CA.)</p></sec><sec><title>Vaccine and cytokine therapy regimen</title><p>After B16F10 tumor cells were injected subcutaneously on day 0, groups of 5–10 mice per test group were sublethally irradiated (500 cGy) 7 days later to induce lymphopenia. All animals were then infused IV with 4 × 10<sup>6 </sup>pmel splenocytes, which equated to the transfer of 8 × 10<sup>5 </sup>- 1 × 10<sup>6 </sup>naive CD45.2<sup>+</sup>/CD8<sup>+ </sup>pmel T cells. Immediately after cell transfer, mice were immunized subcutaneously with 2 × 10<sup>6 </sup>C57BL/6 DCs pulsed with hgp100<sub>25–33 </sub>(1 μg/ml) and NV peptide (1 μg/ml). The hgp100<sub>25–33 </sub>and NV peptide pulsed DC immunization was the standard vaccine in all experiments. Seven days after pmel T cell transfer and vaccination, mice received 5 consecutive days (days 14–18 post tumor inoculation) of IP injection of recombinant human IL-2 at 20 × 10<sup>3 </sup>IU/dose (Chiron Corporation, Emoryville, CA), murine IL-21 at 20 μg/dose (ZymoGenetics Corporation, Seattle, WA), or the combination of IL-2 and IL-21 at the same dosage used for single cytokine injections.</p></sec><sec><title>Flow cytometry</title><p>Except where noted, cell surface staining of peripheral blood lymphocytes and spleen cells was performed using BD Biosciences PharMingen (San Diego, CA) reagents. Single-cell suspensions were incubated with anti-mouse CD16/32 (eBioscience, San Diego, CA) to block Fc receptors. Cells were then stained with allophycocyanin (APC) conjugated anti-CD8 and fluorescein isothiocyanate (FITC)-labeled anti-CD45.2, phycoerythrin (PE) anti-CD62L and PE-Cy7 anti-CD127 (eBiosciences). Data were acquired on a FACS Calibur (BD Biosciences, San Jose, CA.) and analyzed using Cellquest Pro software (version 4.0.1). Peripheral blood and tissue derived-lymphocytes were analyzed by direct ex vivo interrogation of fresh samples. The frequency of pmel T cells was determined by measuring the frequency of total gated CD8<sup>+ </sup>T cells, which were also CD45.2<sup>+</sup>. CD8<sup>+ </sup>T cells from recipient Ly5.2 mice could be distinguished from donor pmel CD8<sup>+ </sup>T-cells since recipient T cells did not express the CD45.2 point mutation, and thus were not stained with the anti-CD45.2 antibody. "Flow-Count" fluorospheres (Beckman-Coulter; Miami, Fl.) were added to each sample before event acquisition to determine the absolute cell count according to the manufacturer's instructions. A minimum of 8000 "Flow Count" beads were counted for each sample collected. The absolute number of pmel CD8<sup>+ </sup>T cells per μl of blood was determined by first calculating the ratio of beads counted by flow divided by the total number of beads added to the sample, and multiplying this ratio times the known volume of the test sample of blood; this determined the fraction of the total test sample volume (in μls) analyzed. The total sample volume analyzed (in μls) was then divided into the total analysis pmel cell count to give pmel<sup>+</sup>CD8<sup>+ </sup>T cells/μl.</p><p>Cytokine flow cytometry (CFC) analysis was performed by stimulating 10<sup>6 </sup>spleen cells suspended in 250 ul of complete medium in a microtiter well (96 well plate) with hgp100<sub>25–33 </sub>peptide (1 ug/ml). Splenocytes were cultured with antigen and brefeldin A (10 ug/ml) for 5 hours at 37°C. Cells were washed twice in PBS and incubated with anti-mouse CD16/32 (FcR-blocking) monoclonal antibody (eBiosciences). Cells were then stained with APC-conjugated CD8 and FITC-labeled CD45.2 (eBiosciences) antibodies for 30 minutes at 4°C, washed 2× in PBS, and fixed and permeablized with 100 ul Cytofix/Cytoperm™ buffer/well (BD Biosciences Pharmingen) for15 minutes at room temperature. Cells were washed 2× in PermWash™ buffer (BD Biosciences Pharmingen) and resuspended in 100 ul of PermWash™ buffer. PE-conjugated anti-IL-2, IFNγ or TNFγ specific antibodies (BD Biosciences Pharmingen) were added at optimal dilution and incubated for one hour at 4°C; PE-conjugated isotype controls were used to determine non-specific cytoplasmic background staining. Cells were washed 2× with PBS and analyzed by flow cytometry. Cytokine positive cell frequencies were determined for pre-gated CD8<sup>+</sup>/CD45.2<sup>+</sup>pmel T cells.</p></sec><sec><title>Statistical analysis</title><p>Phenotype data in this study were analyzed using repeated measures analysis of variance (ANOVA), analysis of covariance (ANCOVA), linear hierarchical mixed-effects regression models, nonparametric (distribution free) tumor growth analyses, and simple and advanced graphical techniques [<xref ref-type="bibr" rid="B26">26</xref>-<xref ref-type="bibr" rid="B28">28</xref>]. Animal survival data were analyzed with standard log-rank statistics and Kaplan-Meier plots. Pair wise <italic>a posteriori </italic>comparisons among treatment conditions were performed with standard post hoc tests (i.e. Tukey's test). Interpretations of tumor growth results from follow-up tests were consistent with those obtained from global analyses. Hence, for simplicity we report probability (p) values obtained from follow-up nonparametric Wilcoxon tests at specific time points. All tests were two-sided; probability values less than 0.05 were considered significant.</p></sec></sec><sec><title>Results</title><sec><title>Combined IL-21 + IL-2 therapy significantly enhanced anti-tumor immunity</title><p>Seven days after subcutaneous injection of 2 × 10<sup>5 </sup>B16F10 melanoma cells C57BL/6 mice were sublethally irradiated, infused with naive pmel splenocytes and vaccinated with hgp100<sub>25–33 </sub>pulsed DCs. Seven days later therapy was initiated with 5 daily IP injections (days 14–18 post tumor inoculation) of IL-2 alone, IL-21 alone or both cytokines. The optimal dose of IL-21 (20 μg/dose) was established previously [<xref ref-type="bibr" rid="B6">6</xref>]; and the optimal IL-2 dose (20 × 10<sup>3 </sup>IU/injection) was established by titration (data not shown). Figure <xref ref-type="fig" rid="F1">1</xref> shows the cumulative mean tumor size from three separate experiments during the first four weeks following tumor inoculation (≥ 15 mice/group). On day 14, just prior to cytokine administration, the mean tumor size for all vaccinated mice (N = 73) was 26 mm<sup>2 </sup>(± 2.4 mm). Thus, all vaccinated mice had established disease prior to cytokine therapy. Global analysis showed that each therapy, including the vaccine alone, reduced the rate of tumor growth compared to the irradiation only control. By day 21, vaccination combined with IL-21 + IL-2 treatment inhibited tumor growth significantly better than vaccination alone (p = 0.0013) and vaccination combined with IL-2 (p = 0.0044) or with IL-21 (p = 0.044). By day 28, vaccination combined with IL-21+ IL-2 again inhibited tumor growth significantly better than vaccination alone (p < 0.0001) and vaccination combined with IL-2 (p = 0.0006) or with IL-21 (p = 0.0033). By day 30, vaccination combined with IL-21 + IL-2 similarly inhibited tumor growth significantly better than vaccination alone (P < 0.0001), but produced higher statistically significant inhibition than vaccination combined with IL-2 (p = 0.0002) or with IL-21 (p = 0.0024) when compared to day 28. Mice in the irradiation only (lymphopenic) control group exhibited rapid tumor growth; the mean tumor size for all animals reached > 200 mm<sup>2 </sup>30 days after tumor inoculation. Similarly, unvaccinated tumor-bearing control mice which were irradiated and treated with IL-21 or IL-21 + IL-2 only had comparable rapid tumor growth (data not shown). Mice treated with IL-21 + IL-2 after vaccination were the only animals that experienced no significant increase in tumor size (Figure <xref ref-type="fig" rid="F1">1</xref>).</p><fig position="float" id="F1"><label>Figure 1</label><caption><p>B16F10 melanoma tumor growth in hgp100<sub>25–33 </sub>vaccinated and cytokine treated lymphopenic mice. Tumor size is presented as the mean (mm<sup>2</sup>) ± SEM of the cumulative analysis of replicate mice (≥ 15) for each test group from 3 separate experiments. Cytokines were administered for 5 consecutive days beginning on day 14 when the average tumor size for vaccinated mice from all test groups in 3 experiments was 26 mm<sup>2 </sup>± 2.4 mm (N = 73 mice). By day 30 IL-21 + low-dose IL-2 treated mice showed significant inhibition of tumor growth compared to the vaccine control (p < 0.0001), or compared to vaccinated mice treated with IL-2 (p = 0.0002) or IL-21 only (p = 0.0024).</p></caption><graphic xlink:href="1479-5876-4-24-1"/></fig></sec><sec><title>IL-21 + IL-2 improved long-term tumor-free survival</title><p>Having shown that treatment with IL-21+ low-dose IL-2 significantly delayed early tumor growth, we compared the effects of the three different cytokine regimens on long-term tumor regression and overall survival beyond 30 days. Figure <xref ref-type="fig" rid="F2">2</xref> shows individual tumor growth curves and long-term survival for all mice in each test group collected from three separate experiments. The large majority of tumor-bearing mice that received sublethal irradiation only (Figure <xref ref-type="fig" rid="F2">2A</xref>) had progressing tumors ≥ 200 mm<sup>2 </sup>by day 28. Control animals that received sublethal irradiation and the hgp100<sub>25–33 </sub>vaccine exhibited a short delay in tumor progression, but >80% of the mice had tumors ≥ 200 mm<sup>2 </sup>by day 35. There were no long-term survivors among mice whose treatment did not include a cytokine (Figure <xref ref-type="fig" rid="F2">2A, B</xref>). Approximately half of the mice treated with low dose IL-2 demonstrated a modest delay of tumor growth to ≥ 200 mm<sup>2 </sup>until approximately day 42, and there was one long-term survivor (Figure <xref ref-type="fig" rid="F2">2C</xref>). IL-21 treatment resulted in an even longer delay (Figure <xref ref-type="fig" rid="F2">2D</xref>) in tumor progression, with 50% of animals reaching day 49 with tumors < 200 mm<sup>2</sup>, and 3/15 mice from the three experiments survived long-term. Mice treated with both IL-21 and IL-2 exhibited the best anti-tumor response. Among 24 mice from three separate experiments approximately 63% survived to day 49 with tumors < 200 mm<sup>2</sup>, and 46% (11/24) remained tumor free 63 days after tumor inoculation (Figure <xref ref-type="fig" rid="F2">2E</xref>). The cumulative long-term survival of mice from all groups from three experiments followed out to 150 days post tumor inoculation showed that IL-21 + IL-2 treated mice had significantly better survival (46%) compared to all other groups (Figure <xref ref-type="fig" rid="F2">2F</xref>). Combined therapy with IL-21 and IL-2 resulted in improved survival compared to vaccination only (p << 0.0001), vaccination plus IL-2 (p << 0.0001), or vaccination plus IL-21 (p = 0.035). The IL-21 only effect was significantly better than the vaccine only control (p = 0.030), but there was no significant difference between the long-term survival response to IL-21 alone and IL-2 alone (p = 0.80). All surviving mice in the IL-21 + IL-2 test group were tumor-free at day 150-post tumor inoculation.</p><fig position="float" id="F2"><label>Figure 2</label><caption><p>Rate of tumor growth and long-term survival in tumor-bearing mice after hgp100<sub>25–33 </sub>vaccination and cytokine therapy. Cumulative analysis of tumor growth rates are depicted for all replicate mice from 3 separate experiments in which mice were irradiated only (A), irradiated and vaccinated (B), or irradiated, vaccinated and treated with IL-2 (C), IL-21 (D), or IL-21 + IL-2 (E). Both IL-21 and IL-21 + IL-2 therapy delayed tumor progression in >50% of treated mice until day 42 while tumors in control and IL-2 treated mice all progressed more rapidly. IL-21 + IL-2 treated mice had a significant survival advantage (46%) over IL-21 (p = 0.035), or IL-2 (p << 0.0001) treated mice; and IL-21 therapy alone (20% survival) was significantly better than the vaccine only (p = 0.03) (F).</p></caption><graphic xlink:href="1479-5876-4-24-2"/></fig></sec><sec><title>Tumor-Specific Protective Immunity in Long-Term Survivor Mice</title><p>To determine whether long-term tumor-specific protective immunity developed after treatment with IL-21 + IL-2, mice surviving tumor free for 150 days were subsequently rechallenged subcutaneously with 5 × 10<sup>5 </sup>B16F10 tumor cells or 5 × 10<sup>4 </sup>3LL tumor cells. Tumor size was monitored three times per week out to 39 days (Figure <xref ref-type="fig" rid="F3">3</xref>). All 4 mice injected with control 3LL cells developed tumor (mean size >150 mm<sup>2</sup>) by day 32, while all 5 mice challenged with B16F10 were tumor-free to day 39, and remained so for another 150 days (data not shown). Additional control groups included naive congenic C57BL/6 mice inoculated with B16F10 (5 mice) or 3LL (5 mice) tumor cells. The mean tumor size for both tumors in naive control mice reached 200 mm<sup>2 </sup>before day 25. Thus, the regimen of adoptive cell transfer with lymphopenic conditioning and tumor antigen vaccination followed by IL-21 + IL-2 therapy was curative in mice with established tumors, and produced a long-term tumor-specific memory response capable of protecting against a secondary challenge with the original tumor.</p><fig position="float" id="F3"><label>Figure 3</label><caption><p>Tumor-specific protective immunity in long-term survivor mice. IL-21 + IL-2 treated mice that survived disease free for 150 days were rechallenged with B16F10 (5 × 10<sup>5</sup>) or 3LL(5 × 10<sup>4</sup>) tumor cells. All 3LL challenged mice (4/4) exhibited mean tumor growth to ≥ 200 mm<sup>2 </sup>by day 32; all B16F10 challenged mice (5/5) were protected. Naïve mice challenged with the same B16F10 (5/5) or 3LL (5/5) tumor cells had rapid mean tumor growth to ≥ 200 mm<sup>2 </sup>by day 25.</p></caption><graphic xlink:href="1479-5876-4-24-3"/></fig></sec><sec><title>IL-21 + IL-2 therapy induced high numbers of circulating pmel CD8<sup>+ </sup>T cells</title><p>Representative data from one of three experiments demonstrated that IL-21 + IL-2 therapy in the pmel Tg/B16F10 model induces higher absolute numbers of circulating pmel CD8<sup>+ </sup>T cells than either IL-2 or IL-21 therapy alone (Figure <xref ref-type="fig" rid="F4">4</xref>). Pooled blood from 5–10 mice in each cytokine group and the vaccine only control were stained for CD8/CD127/CD62L/CD45.2 expression, and the absolute number of circulating pmel CD8<sup>+ </sup>T cells was determined using counting beads as described. Analysis of variance (ANOVA) was performed on cell counts made at weekly intervals from 7–28 days post vaccine administration from pooled samples for each group; there was a significant difference between test groups (p = 0.029). A follow-up (post hoc) Tukey test showed that the absolute number of circulating pmel CD8<sup>+ </sup>T cells in peripheral blood was significantly higher after combined cytokine therapy than following IL-2 (p < 0.05), or IL-21 (p < 0.05) therapy alone at 14, 21, and 28 days after vaccination. The expansion of circulating pmel cells appeared to peak at day 21 and go through a contraction phase from day 21 to day 28 in mice treated with IL-21+ IL-2, and for mice in the vaccine only control group. Cell numbers remained relatively constant from day 14–28 for mice treated with IL-21 or IL-2 alone.</p><fig position="float" id="F4"><label>Figure 4</label><caption><p>Quantitation of absolute numbers of pmel CD8<sup>+ </sup>T cells. Pooled blood from mice in the vaccine control and all three cytokine experimental groups (≥ 5 mice/group) was analyzed by flow cytometry to determine the absolute number of circulating pmel-1 CD8<sup>+ </sup>T cells/μL of blood. Vaccinated mice treated with IL-21+ IL-2 had the highest total pmel CD8<sup>+ </sup>cell numbers at all time points. Data are shown from a representative experiment and show that IL-21 + IL-2 resulted in significantly higher circulating pmel CD8<sup>+ </sup>T cell numbers than was produced by IL-2 (p = 0.05), or IL-21 (p = 0.05) alone.</p></caption><graphic xlink:href="1479-5876-4-24-4"/></fig></sec><sec><title>Cell surface phenotype of pmel CD8<sup>+ </sup>T cells from cytokine-treated mice</title><p>Detailed pmel CD8<sup>+ </sup>T cell subset analysis (CD8/CD127/CD62L/CD45.2) was performed on pooled peripheral blood samples from mice used in the experiment described in Figure <xref ref-type="fig" rid="F4">4</xref>. Recent reports have identified CD127 (IL-7α receptor) expression as an important descriptor for long-lived memory T cells which may arise early after in vivo priming [<xref ref-type="bibr" rid="B29">29</xref>-<xref ref-type="bibr" rid="B34">34</xref>]. These studies also indicated that CD62L and CD127 staining partitioned antigen-specific cells into two distinct memory T cell subsets which, by functional criteria, were similar to central memory T cells (Tcm) with a CD62L<sup>+</sup>/CD127<sup>+ </sup>phenotype, and effector memory (Tem) cells with a CD62L<sup>-</sup>/CD127<sup>+ </sup>phenotype [<xref ref-type="bibr" rid="B32">32</xref>,<xref ref-type="bibr" rid="B33">33</xref>]. Effector T cells (Te) were CD62L<sup>-</sup>/CD127<sup>- </sup>and were characterized by poor proliferation response, immediate effector function, and high levels of antigen-induced apoptosis [<xref ref-type="bibr" rid="B33">33</xref>]. Figure <xref ref-type="fig" rid="F5">5</xref> shows dot plots of CD127 versus CD62L staining of pre-gated CD8<sup>+</sup>/CD45.2<sup>+ </sup>peripheral blood pmel T cells obtained on days 14, 21 and 28 from each cytokine therapy group and vaccine only control mice. The frequency of circulating pmel CD8<sup>+ </sup>T cells with a Tcm phenotype on day 14 was no different in IL-2, and IL-21 + IL-2 treated mice or in the vaccine controls. Only IL-21 treated mice showed a small increase in Tcm frequency, which was 15% higher than that produced in control mice. Tem frequencies following treatment with IL-2 or with IL-21 + IL-2 were 30% and 63% higher respectively than in control mice. The frequencies of both Tcm and Tem pmel CD8<sup>+ </sup>T cells decreased on day 21 for all experimental groups compared to day 14 values. The day 21 frequency of Tem pmel CD8<sup>+ </sup>T cells (CD62L<sup>-</sup>/CD127<sup>+</sup>) was similar for all test groups except IL-21 + IL-2, which maintained approximately a 70% higher Tem percentage. The frequencies of Tcm (CD62L<sup>+</sup>/CD127<sup>+</sup>) on day 21 indicated IL-2 sustained a slightly higher percentage than the control (22%); IL-21 stimulation resulted in a 54% higher frequency of Tcm CD8<sup>+ </sup>T cells compared to the control, and IL-21 + IL-2 treated mice maintained a Tcm frequency 35% higher than the control. By day 28 the control and IL-2 induced Tcm frequencies were equal to each other but higher than on day 21, while the IL-21 effect (38% increase) and the IL-21 + IL-2 effect (25% increase) on Tcm expression were both similarly elevated compared to the control and IL-2 induced percentages. Combined IL-21 and low-dose IL-2 therapy sustained the circulating Tem frequency at a level that was 2-fold higher than the control group or the effect induced by IL-2 alone or IL-21 alone. The cell density plots for IL-21+IL-2 on all three days also suggested an overall higher absolute number of circulating Tcm and Tem CD8<sup>+</sup>/CD45.2<sup>+ </sup>T cells than was induced by any other treatment. Generally, lower percentages of circulating effector cells (CD62L<sup>-</sup>/CD127<sup>-</sup>) were present at day 21 and day 28 for IL-21 and IL-21+IL-2 treated mice compared to both IL-2 treated animals and control mice. The upper left quadrant of each 2-parameter histogram contained CD62L<sup>+</sup>/CD127<sup>- </sup>T cells, and their frequency was generally lower in the IL-21 and IL-21+IL-2 treated animals. Presently there is no correlated functional data to determine the lineage relationship of this population to Tcm or Tem T cells or their functional properties. Overall the data in Figure <xref ref-type="fig" rid="F5">5</xref> suggest IL-21+IL-2 induced higher frequencies of Tem cells compared to IL-2 or IL-21 alone on all three days; and by the early contraction phase of the pmel CD8<sup>+ </sup>T cell response (day 28) IL-21 and IL-21 + low-dose IL-2 induced equally higher percentages of tumor-specific Tcm cells than were stimulated in the control or IL-2 treated mice.</p><fig position="float" id="F5"><label>Figure 5</label><caption><p>CD27 vs CD62L 2 parameter phenotype analysis of pmel CD8<sup>+ </sup>T cells. Lymphocytes from pooled blood (5–10 mice/group) in a representative experiment were analyzed for CD127 and CD62L staining. Flow cytometry phenotype analysis of pmel CD8<sup>+ </sup>T cells on days 14, 21 and 28 post vaccination indicated higher frequencies of antigen specific Tcm (CD62L<sup>+</sup>/CD127<sup>+</sup>) were induced at all time points by IL-21 alone compared to the control and IL-2 stimulated PBMCs; IL-21 + IL-2 stimulated comparably high Tcm frequencies on days 21 and 28. IL-21+IL-2 stimulated the highest frequency of pmel CD8<sup>+ </sup>Tem (CD62L<sup>-</sup>/CD127<sup>+</sup>) at all time points.</p></caption><graphic xlink:href="1479-5876-4-24-5"/></fig></sec><sec><title>IL-21+ IL-2 induced proliferation of Tem and Tcm pmel CD8<sup>+ </sup>T cells</title><p>The frequencies of Tcm and Tem T cells on days 14, 21 and 28 (Figure <xref ref-type="fig" rid="F5">5</xref>), and the total number of circulating pmel cells determined for each group (Figure <xref ref-type="fig" rid="F4">4</xref>) were subsequently used to calculate the absolute number of Tem (Figure <xref ref-type="fig" rid="F6">6A</xref>) and Tcm (Figure <xref ref-type="fig" rid="F6">6B</xref>) pmel CD8<sup>+ </sup>T cells for each experimental group. Tem numbers on each day were significantly higher for IL-21+ IL-2 treated mice than for all other groups (p = 0.011) as determined by ANCOVA, and declined from day 14 to day 28 (Figure <xref ref-type="fig" rid="F6">6A</xref>). By contrast, the absolute number of Tcm cells for all groups was equally low at day 14 and increased incrementally for each group over time (Figure <xref ref-type="fig" rid="F6">6B</xref>). IL-21 + IL-2 induced Tcm cell numbers increased at a significantly higher rate compared to all other test groups (p = 0.029) as determined by ANCOVA. Thus, IL-21+ IL-2 treatment maintained Tem absolute numbers at all time points that were 2- to 4-fold greater than those produced by the vaccine control or by IL-2 or IL-21 therapy alone. Similarly, Tcm absolute numbers induced by IL-21+ IL-2 were 2- to 4-fold higher than those stimulated in any other test group by day 28. Recent data have demonstrated that the adoptive transfer of pmel CD8<sup>+ </sup>Tcm (characterized as CD62L<sup>+</sup>/CCR7<sup>+</sup>), or "early effector" T cells (CD62L<sup>dim</sup>/CD127<sup>+</sup>/CCR7<sup>+</sup>) produced a potent anti-tumor response subsequent to tumor-antigen hgp100<sub>25–33 </sub>vaccination and high-dose IL-2 therapy in the pmel Tg/B16 therapeutic model [<xref ref-type="bibr" rid="B22">22</xref>,<xref ref-type="bibr" rid="B35">35</xref>]. The data in Figure <xref ref-type="fig" rid="F6">6B</xref> suggest IL-21+ low-dose IL-2 in vivo therapy promotes the long-term maintenance of CD8<sup>+ </sup>T cells with a similar Tcm (CD62L<sup>+</sup>/CD127<sup>+</sup>) or "early effector" – like phenotype (CD62L<sup>dim/-</sup>/CD127<sup>+</sup>).</p><fig position="float" id="F6"><label>Figure 6</label><caption><p>Quantitation of absolute numbers of Tem and Tcm CD8<sup>+ </sup>T cells. PBMCs were examined by flow cytometry to determine the absolute number of Tem (CD62L<sup>-</sup>/CD127<sup>+</sup>) pmel CD8<sup>+ </sup>T cells (A), and Tcm (CD62L<sup>+</sup>/CD127<sup>+</sup>) pmel CD8<sup>+ </sup>T cells (B) on days 14, 21, and 28 after vaccination and cytokine treatment. IL-21 + IL-2 supported the highest level of circulating Tem cells on all three days compared to all other experimental groups (p = 0.011), and stimulated the greatest increase of Tcm cells on days 21 and 28 (p = 0.029). Data was collected from pooled blood from ≥ 5 mice in each test group.</p></caption><graphic xlink:href="1479-5876-4-24-6"/></fig></sec><sec><title>Functional phenotype of CD8<sup>+ </sup>T cells from IL-21 + IL-2 treated tumor regressor mice</title><p>By day 42 previously regressing tumors in a subset of IL-21 + IL-2 treated mice began to grow again, reaching a mean tumor size of 25–50 mm<sup>2</sup>. All tumors of this size on day 42–49 continued to progress to ≥ 200 mm<sup>2 </sup>by day 63, whereupon the mice were sacrificed. However, approximately 50% of all IL-21+ IL-2 treated mice in three experiments did not exhibit tumor growth, and by day 42–49 manifested complete regression of the tumor. These animals remained tumor free out to >150 days. Day 42 splenocytes from two "progressor", and from two "regressor" mice treated with IL-21 + IL-2 were pooled separately, and were stimulated in vitro with hgp100<sub>25–33</sub>; cells were then analyzed for IFNγ, IL-2 and TNF-α production using a standard five-hour CFC assay. After IVS cells were fixed, permeablized, and stained with CD8, CD45.2 and cytokine-specific antibodies. Spleens from regressor mice had more IFNγ, IL-2 and TNF-α positive CD8<sup>+</sup>/CD45.2<sup>+ </sup>pmel T cells than spleens from progressor animals (Figure <xref ref-type="fig" rid="F7">7A</xref>). Regressor mice had 4.7-fold, 4.8-fold, and 2.5-fold more IL-2<sup>+</sup>, IFNγ<sup>+ </sup>and TNF-α<sup>+ </sup>pmel CD8<sup>+ </sup>T cells, respectively, than progressor mice with rapidly growing tumors. The data in Figure <xref ref-type="fig" rid="F7">7A</xref> indicates that tumor regression in mice receiving IL-21+ IL-2 therapy was associated with an increase in the absolute number of pmel CD8<sup>+ </sup>T cells producing Tc1 cytokines compared to the numbers found in mice with growing tumors. The presence of higher numbers of Tc1 cytokine<sup>+ </sup>CD8<sup>+ </sup>T cells in regressor compared to progressor mice was attributable to both a 2-fold greater absolute number of pmel CD8<sup>+ </sup>T cells in regressor mice than were measured in progressor animals (data not shown) and to higher frequencies of pmel CD8<sup>+ </sup>T cells which were positive for Tc1 cytokines in regressor mice at 42 days (Figure <xref ref-type="fig" rid="F7">7B</xref>).</p><fig position="float" id="F7"><label>Figure 7</label><caption><p>Cytokine flow cytometry analysis of the functional phenotype of pmel CD8<sup>+ </sup>splenocytes from IL-21 + IL-2 treated mice. Spleen cells were pooled from 2 mice with progressing tumors (≥ 50 mm<sup>2</sup>) and from 2 mice with regressing tumors (<10 mm<sup>2</sup>) 42 days after vaccination. Cytokine flow cytometry analysis shows the absolute number of pmel CD8<sup>+ </sup>T cells per 10<sup>6 </sup>splenocytes (A), and the frequency of pmel CD8<sup>+ </sup>cells (B) that are positive for IL-2, IFNγ, and TNFγ production. Data show mice with regressing tumors had 4.7-fold, 4.8-fold and 2.5-fold more IL-2<sup>+</sup>, IFN-γ<sup>+</sup>, and TNF-α<sup>+ </sup>pmel CD8<sup>+ </sup>T cells respectively, and higher frequencies of Tc1 cytokine<sup>+ </sup>cells than progressor mice.</p></caption><graphic xlink:href="1479-5876-4-24-7"/></fig></sec></sec><sec><title>Discussion and conclusions</title><p>We report for the first time the anti-tumor effects of a treatment strategy that combines lymphopenic conditioning, adoptive transfer of naive antigen-specific CD8<sup>+ </sup>T cells, tumor antigen-specific vaccination, and cytokine treatment with IL-21 + low-dose IL-2. Using a well established murine model of pre-existing disease, the pmel Tg/B16 model, our results show that combined IL-21 + low-dose IL-2 therapy delayed B16F10 melanoma growth in the majority of mice, and resulted in a significant increase in tumor-free survival (46%) out to ≥ 150 days compared to mice treated with IL-21 or IL-2 alone. Surviving mice exhibited tumor-specific protective immunity since they resisted B16F10 rechallenge but succumbed to the unrelated 3LL tumor. These data are in contrast to other studies in the pmel Tg/B16 model, which employed the adoptive transfer of highly activated IVS pmel CD8<sup>+ </sup>T cells (stimulated with hgp100<sub>25–33 </sub>peptide and IL-2 or IL-15) rather than naive pmel T cells [<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B17">17</xref>,<xref ref-type="bibr" rid="B18">18</xref>]. After lymphopenic conditioning and T-cell transfer to tumor-bearing mice, therapy was administered in these studies using different combinations of hgp100<sub>25–33 </sub>vaccination and cytokine treatment. Thus, Lou et al [<xref ref-type="bibr" rid="B17">17</xref>] vaccinated with hgp100<sub>25–33 </sub>peptide-pulsed DCs, and administered high-dose IL-2 (1.2 × 10<sup>6 </sup>IU/day) for three days beginning the day of vaccination. Klebanoff, et al [<xref ref-type="bibr" rid="B18">18</xref>] vaccinated with a recombinant fowl pox virus encoding hgp100<sub>25–33 </sub>(rFPhgp100) after adoptive transfer of IVS pmel CD8<sup>+ </sup>T cells grown in IL-2 or in IL-15 cultures. Follow-on cytokine therapy consisted of high-dose IL-2 (1.2 × 10<sup>6 </sup>IU/day) for three days. In the only published study using IL-21 in the pmel Tg/B16 model Zeng and coworkers [<xref ref-type="bibr" rid="B3">3</xref>] also vaccinated with rFPhgp100<sub>25–33 </sub>following adoptive transfer of IL-2 cultured pmel T cells, and animals were treated over three days with IL-15 (20 μg/day), IL-21 (20 μg/day), or both cytokines at the same doses. None of these treatment strategies resulted in long-term inhibition of tumor growth beyond 30–35 days after treatment unless very large numbers (4–6 × 10<sup>6</sup>) of IVS pmel CD8<sup>+ </sup>T cells were adoptively transferred [<xref ref-type="bibr" rid="B17">17</xref>]. Furthermore, in contrast to our results, none of these studies presented data describing long-term tumor-free survival ≥ 150 days with any combination of transferred T cells, vaccine administration and cytokine therapy. Another group compared the ability of IL-2, IL-15 or IL-21 to augment anti-tumor immunity in C57BL/6 mice challenged with the OVA-expressing E.G7 thymoma [<xref ref-type="bibr" rid="B6">6</xref>]. Cytokine therapy was initiated 48 hours after tumor inoculation – well before the establishment of vascularized tumors, and was administered every other day for two weeks thereafter. In this model IL-21 (20 μg/day) did result in prolonged tumor-free survival out to 100 days in approximately 25–30% of treated mice. Neither low-dose IL-2 (2 × 10<sup>3 </sup>IU and 20 × 10<sup>3 </sup>IU/day) nor IL-15 (5 μg and 50 μg/day) produced a survival frequency that was as high. All animals that survived beyond 100 days were protected from rechallenge with E.G7 tumor cells. Importantly, this report also showed that the 2-week cytokine administration could begin as late as day 12 following tumor inoculation and still produce a therapeutic response [<xref ref-type="bibr" rid="B6">6</xref>]. This study also demonstrated the ability of IL-21 to augment immunity to the OVA-expressing E.G7 tumor in the absence of previously antigen-educated CD8<sup>+ </sup>T cells. However, the experimental design did not test the anti-tumor effects of IL-21 in a model of established bulky disease such as the pmel Tg/B16 system, in which the cognate tumor antigen is a relatively weak self-tumor antigen rather than a strongly immunogenic foreign protein. To date only the combined therapeutic effects of IL-21+ low-dose IL-2 reported here have resulted in long-term survival of tumor-free mice in the pmel Tg/B16 model. As noted, in most of the previous studies using the pmel Tg/B16 melanoma model very high doses (1.2 × 10<sup>6 </sup>IU/day) of IL-2 were used in combination with the adoptive transfer of large numbers (up to 10<sup>7</sup>) of highly activated IVS pmel CD8<sup>+ </sup>T cells [<xref ref-type="bibr" rid="B16">16</xref>-<xref ref-type="bibr" rid="B18">18</xref>] However, treatment was initiated after very large, bulky (50–100 mm<sup>2</sup>) tumors were established. By contrast, cytokine therapy was initiated in our experiments when vascularized tumors were smaller (26 mm<sup>2</sup>). Long-term survival using our current IL-21 + IL-2 therapeutic model may decrease with larger established tumors; experiments designed to test and optimize IL-21 +low-dose IL-2 therapy in mice with larger vascularized tumors are ongoing.</p><p>The H-2D<sup>b</sup>-restricted NV peptide was used in all experiments as a putative CD4 "helper" antigen source [<xref ref-type="bibr" rid="B23">23</xref>,<xref ref-type="bibr" rid="B24">24</xref>] However, our recent experiments suggest NV peptide vaccination produces only a modest increase in circulating pmel CD8<sup>+ </sup>T cell numbers over that produced by hgp100<sub>25–33 </sub>alone (unpublished data – H.M. Hu, EACRI). CD4 helper T cell function in our DC-based vaccine system may have been provided by fetal calf serum (FCS) proteins/peptides associated with the DCs in the vaccine – both DCs and the B16F10 melanoma cells used in the tumor inoculum were cultured in medium containing FCS. The potential for FCS antigen-induced CD4 T cell activation is also suggested by other hgp100<sub>25–33</sub>-pulsed DC vaccine studies in the pmel Tg/B16 model in which anti-tumor therapeutic effect was achieved in the absence of any known source of CD4-specific antigen [<xref ref-type="bibr" rid="B17">17</xref>].</p><p>There is limited information on IL-21-mediated changes in the cell surface and functional phenotype of CD8<sup>+ </sup>T-cells associated with tumor immunity. In this report, we show that IL-21 alone and IL-21+ low-dose IL-2 increased the in vivo frequency of Tcm T cells as defined by concomitant CD62L<sup>+</sup>/CD127<sup>+ </sup>staining[<xref ref-type="bibr" rid="B29">29</xref>], compared to the vaccine only control or to cells from IL-2-treated mice (Figure <xref ref-type="fig" rid="F5">5</xref>). IL-21 + IL-2 treatment also sustained the highest percentage of Tem (CD62L<sup>-</sup>/CD127<sup>+</sup>) pmel CD8<sup>+ </sup>T cells [<xref ref-type="bibr" rid="B29">29</xref>] compared to any other test group. This observation correlated with the result that IL-21+ IL-2 therapy also produced the highest absolute number of circulating pmel CD8<sup>+ </sup>T cells during the beginning (day 14), peak (day 21) and end (day 28) of the hgp100<sub>25–33 </sub>induced expansion of pmel CD8<sup>+ </sup>T cells (Figure <xref ref-type="fig" rid="F4">4</xref>). As a consequence, IL-21 + IL-2 therapy produced the highest absolute number of circulating pmel CD8<sup>+ </sup>Tem cells at each time point of the expansion and early contraction phases of the anti-hgp100<sub>25–33 </sub>immune response, and also resulted in the highest absolute number of Tcm T cells in the peripheral blood (Figure <xref ref-type="fig" rid="F6">6</xref>). Although Tem absolute numbers decreased from day 14 to day 28, and Tcm numbers increased over this same period the data do not indicate if this is attributable in anyway to innate differences in IL-21 + IL-2 induced Tem vs Tcm proliferation. Future in vitro and in vivo studies will examine the relative proliferative potential of cognate antigen-driven purified populations of pmel CD8<sup>+ </sup>Tcm and Tem cells. Notably, the tumor-specific protective memory response observed in long-term survivor mice was also associated with high frequencies of Tcm in the spleen and lymph nodes. Polychromatic (8 color) flow cytometry analysis of pmel CD8<sup>+ </sup>splenocytes from IL-21 + IL-2 treated long-term survivor mice (> 150 days) indicated approximately 10%–12% of all CD8<sup>+ </sup>T cells in the spleen and lymph nodes were pmel CD8<sup>+ </sup>T cells, and 43% and 70% of these T cells in the lymph nodes and spleen respectively expressed a Tcm phenotype (CD62L<sup>+</sup>/CD127<sup>+</sup>/CD27<sup>+</sup>/CD28<sup>+</sup>)(unpublished data). Recent data have demonstrated that enriched pmel CD8<sup>+ </sup>Tcm T cells conferred a potent in vivo anti-tumor recall response upon adoptive transfer – leading to the eradication of large established tumors in the pmel Tg/B16 model [<xref ref-type="bibr" rid="B35">35</xref>]. Thus, as described herein, IL-21 + IL-2 therapy favored the in vivo expansion and maintenance of such tumor antigen-specific Tcm T cells from naive pmel CD8<sup>+ </sup>T cell precursors which were capable of providing long-term protective immunity. Preliminary phenotype analysis of IL-21 + low-dose IL-2 and hgp100<sub>25–33 </sub>IVS pmel splenocytes similarly indicates the combination of both cytokines results in the proliferation of pmel Tg CD8<sup>+ </sup>T cells with a Tcm (CD62L<sup>+</sup>/CD127<sup>+</sup>/CD27<sup>+</sup>/CD28<sup>+</sup>) phenotype (unpublished data – EACRI). This further suggests that IL-21 + low-dose IL-2 might be effective in supporting the cognate antigen-driven in vitro expansion of rare self-tumor antigen Tcm CD8<sup>+ </sup>T cells from autologous PBMCs – perhaps in concert with procedures to remove tumor antigen-specific regulatory T cell effects. In addition to increasing tumor-antigen directed expansion and maintenance of Tcm CD8<sup>+ </sup>T cells, data presented here demonstrates that IL-21 + IL-2 therapy also stimulated the expression of anti-tumor memory CD8<sup>+ </sup>T cells with a Tc1 cytokine functional phenotype in mice with regressing tumors. By day 42 post-tumor inoculation IL-21 + IL-2 treated tumor-regressor mice had much higher absolute numbers of IL-2<sup>+</sup>, IFNγ<sup>+ </sup>and TNF-α<sup>+ </sup>pmel CD8<sup>+ </sup>splenocytes than IL-21 + IL-2 treated tumor-progressor mice with rapidly growing tumors (Figure <xref ref-type="fig" rid="F7">7</xref>). The relative absence of CD8<sup>+ </sup>splenocytes with a Tc1 cytokine functional phenotype in tumor progressor mice may be attributable in part to a simple lack of sustained B16F10 tumor antigen stimulation. Previous published data from our institute using the pmel Tg/B16 model has demonstrated loss of MHC class I and gp100 expression in a high percentage of tumors in experimental animals with progressive disease due to immunoediting[<xref ref-type="bibr" rid="B36">36</xref>]. Overall our data suggest that IL-21 + low-dose IL-2-mediated tumor regression and long-term survival was associated with elevated absolute numbers of tumor antigen-specific Tem and Tcm CD8<sup>+ </sup>T cells with a Tc1 functional phenotype.</p><p>IL-21 + low-dose IL-2 induction of increased in vivo expression of Tc1 cytokine<sup>+ </sup>Tcm and Tem CD8<sup>+ </sup>T cells is a central observation for any putative mechanistic explanation of the anti-tumor therapeutic synergy produced using both cytokines. This observation examined in the context of recent studies describing the broad inhibitory effects of IL-21 on B lymphocytes, NK cells and DCs (Reviewed in [<xref ref-type="bibr" rid="B37">37</xref>]) suggests a possible mechanism for the IL-21+ low-dose IL-2 induced expansion of Tcm CD8<sup>+ </sup>T cells similar to what may occur during the late contraction phase of an ongoing immune response. Thus, CD4 T cell (Th2)-derived IL-21[<xref ref-type="bibr" rid="B38">38</xref>] inhibits anti-IgM and IL-4-mediated B cell proliferation [<xref ref-type="bibr" rid="B2">2</xref>] and enhances apoptosis of activated B cells [<xref ref-type="bibr" rid="B39">39</xref>,<xref ref-type="bibr" rid="B40">40</xref>], inhibits DC maturation and the ability of DCs to prime T cells [<xref ref-type="bibr" rid="B41">41</xref>,<xref ref-type="bibr" rid="B42">42</xref>]; and increases NK maturation and cytolytic function while inhibiting NK proliferation [<xref ref-type="bibr" rid="B43">43</xref>]. Similarly, while IL-21 has generally been described as an inducer of antigen stimulated IFNγ and CTL function in CD8<sup>+ </sup>T cells [<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B44">44</xref>,<xref ref-type="bibr" rid="B45">45</xref>], our unpublished data in both murine and human in vitro experiments indicate IL-21 alone does not drive high levels of cognate antigen stimulated CTL proliferation. In vivo data presented here suggest that IL-21 works best in concert with low-dose IL-2 to support both optimal anti-tumor effector CTL function and expansion of CD8<sup>+ </sup>memory T cells. Paradoxically, IL-21 + low-dose IL-2 treatment produced unexpectedly low frequencies of circulating effector pmel CD8<sup>+ </sup>T cells (Figure <xref ref-type="fig" rid="F4">4</xref>) – perhaps due to cell trafficking to the primary tumor and sites of metastatic disease. Ongoing experiments are directed at the in situ analysis of the phenotype, cell number, and functional properties of tumor invasive pmel CD8<sup>+ </sup>T cells in IL-21 + low dose IL-2 treated mice. Such studies should shed light on whether combined cytokine therapy induces trafficking of increased numbers of antigen-specific cytolytic CD8<sup>+ </sup>T cells to the tumor site. Taken together, these observations suggest a dual role for IL-21 in dampening ongoing innate and adaptive immune responses (as might occur during the contraction phase of antigen-specific immunity), while concomitantly augmenting the expansion of antigen-driven long-term memory T cells. Recently published data demonstrate IL-21 treated cultures of antigen-stimulated cells produce CD8<sup>+ </sup>T cells with increased TCR binding affinity [<xref ref-type="bibr" rid="B5">5</xref>]. Other data show the IL-21 receptor (IL-21R) is upregulated upon TCR engagement [<xref ref-type="bibr" rid="B46">46</xref>,<xref ref-type="bibr" rid="B47">47</xref>]. Thus, Th2-derived IL-21 may "rescue" memory T cells by blocking further antigen-driven effector T cell differentiation through presently unknown mechanisms, while simultaneously increasing antigen binding affinity. The resulting increase in binding of low-levels of residual antigen (present during the contraction phase) could increase IL-21R expression and up-regulate increased IL-21 binding to memory T cells – thus further skewing the shift to memory T cells. Memory CD8<sup>+ </sup>T cells in turn express amplified levels of CD122 [<xref ref-type="bibr" rid="B48">48</xref>], which may facilitate increased high affinity IL-2R formation if other IL-2R chains (CD25 and CD132) are available, and potentially increase binding of the low levels of IL-2 which may be present during immunological contraction. The combination of the "arrested" Tcm memory phenotype (maintained by IL-21), and low dose IL-2-induced proliferation of these cells could result in increased Tcm cell numbers. Thus, IL-21 + low-dose IL-2 therapy may favor the enhanced expansion of long-term memory (Tcm) CD8<sup>+ </sup>T cells through augmented memory T cell "rescue" and expansion mechanisms similar to those which may normally be present during the late contraction phase of an ongoing immune response. This concept is supported by the results of our study and other reports [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B8">8</xref>], which suggest IL-21 induced tumor immunity is most effective when IL-21 is administered several days (4–12 days) after initial tumor-antigen activation of T cells. The late acting anti-tumor effects of IL-21, and, in our study, IL-21 + IL-2 therapy suggests IL-21 may be acting directly on tumor antigen educated CD8<sup>+ </sup>T cells rather than modulating early APC and/or T cell function associated with primary tumor antigen stimulation. Preliminary in vitro studies with purified naïve and hgp100<sub>25–33 </sub>stimulated pmel CD8<sup>+ </sup>T cells support this conclusion (data not shown).</p><p>In summary, the results of this study suggest that IL-21 + low-dose IL-2 cytokine therapy may "rescue" and augment proliferation of tumor antigen-specific memory CD8<sup>+ </sup>T cells, and thus provide an important new strategic component for effective cancer immunotherapy. The combined use of both cytokines may be most effective in the context of lymphopenic conditioning, tumor antigen-specific vaccination, and the adoptive transfer of autologous tumor-specific Tcm or "early" effector T cells. Ongoing work at our institute continues to focus on developing procedures for the IL-21 + low-dose IL-2-directed in vivo and in vitro expansion and maintenance of such Tcm and "early" effector subpopulations from small numbers of normally tolerized autologous self-tumor antigen-specific precursors.</p></sec><sec><title>Competing interests</title><p>Christopher Clegg is a scientist employed by Zymogenetics – all other authors declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>HH designed the in vivo experiments and worked with PW and DH to perform the flow cytometry analysis of PBMC and splenocyte samples. GY assisted in all aspects of tumor inoculation, lymphopenic conditioning, tumor measurement and tissue harvest. HMH, CHC, and AOH provided key reagents and assisted in the design of controls which utilized these reagents in the experiments. WM was directly involved in drafting and revising the manuscript, and WGA performed all the statistical analysis. EBW designed and supervised all aspects of the experimental strategy and acquisition of data, and, with BAF and WJU, was involved in data interpretation and the critical review and revision of the manuscript.</p></sec>
Integrated biclustering of heterogeneous genome-wide datasets for the inference of global regulatory networks	<sec><title>Background</title><p>The learning of global genetic regulatory networks from expression data is a severely under-constrained problem that is aided by reducing the dimensionality of the search space by means of clustering genes into putatively <italic>co-regulated </italic>groups, as opposed to those that are simply <italic>co-expressed</italic>. Be cause genes may be co-regulated only across a subset of all observed experimental conditions, <italic>biclustering </italic>(clustering of genes <italic>and </italic>conditions) is more appropriate than standard clustering. Co-regulated genes are also often functionally (physically, spatially, genetically, and/or evolutionarily) associated, and such <italic>a priori </italic>known or pre-computed associations can provide support for appropriately grouping genes. One important association is the presence of one or more common cis-regulatory motifs. In organisms where these motifs are not known, their <italic>de novo </italic>detection, integrated into the clustering algorithm, can help to guide the process towards more biologically parsimonious solutions.</p></sec><sec><title>Results</title><p>We have developed an algorithm, cMonkey, that detects putative co-regulated gene groupings by integrating the biclustering of gene expression data and various functional associations with the <italic>de novo </italic>detection of sequence motifs.</p></sec><sec><title>Conclusion</title><p>We have applied this procedure to the archaeon <italic>Halobacterium </italic>NRC-1, as part of our efforts to decipher its regulatory network. In addition, we used cMonkey on public data for three organisms in the other two domains of life: <italic>Helicobacter pylori, Saccharomyces cerevisiae</italic>, and <italic>Escherichia coli</italic>. The biclusters detected by cMonkey both recapitulated known biology and enabled novel predictions (some for <italic>Halobacterium </italic>were subsequently confirmed in the laboratory). For example, it identified the <italic>bacteriorhodopsin </italic>regulon, assigned additional genes to this regulon with apparently unrelated function, and detected its known promoter motif. We have performed a thorough comparison of cMonkey results against other clustering methods, and find that cMonkey biclusters are more parsimonious with all available evidence for co-regulation.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>Reiss</surname><given-names>David J</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" corresp="yes" contrib-type="author"><name><surname>Baliga</surname><given-names>Nitin S</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" corresp="yes" contrib-type="author"><name><surname>Bonneau</surname><given-names>Richard</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib>	BMC Bioinformatics	<sec><title>Background</title><p>The statistical elucidation of genetic regulatory networks from experimental data (commonly mRNA expression levels) is an important problem that has been the center of a large body of work [<xref ref-type="bibr" rid="B29">29</xref>,<xref ref-type="bibr" rid="B43">43</xref>]. Because this problem is <italic>underconstrained </italic>(the number of free parameters is far greater than the dimensionality of the data), many efforts include some means for dimensionality reduction. A common practice for reducing the dimensionality of this problem space has been to <italic>cluster </italic>genes into <italic>co-expressed </italic>groups based on their expression profiles, prior to network inference. Such a practice has the additional advantage that, if done properly, the signal-to-noise in the data can thereby be reduced through signal averaging. The genes in such clusters are often assumed to be <italic>co-regulated, i.e</italic>. to share the same regulatory controls, thereby implying biological relevance for such a pre-clustering step. However, gene transcript levels can be correlated either by chance (due to experimental noise or systematic error) or because of indirect effects, and therefore they might not actually be directly co-regulated. The integration of additional biologically-relevant evidence into a clustering procedure may be used to provide constraints on the identification of groups of co-regulated genes.</p><p>Co-regulated genes are often functionally (physically, spatially, genetically, and/or evolutionarily) linked [<xref ref-type="bibr" rid="B33">33</xref>,<xref ref-type="bibr" rid="B34">34</xref>,<xref ref-type="bibr" rid="B58">58</xref>,<xref ref-type="bibr" rid="B63">63</xref>,<xref ref-type="bibr" rid="B66">66</xref>,<xref ref-type="bibr" rid="B67">67</xref>]. For example, genes whose products form a protein complex are likely to be co-regulated. Other types of associations among genes, or their protein products, that (can) imply functional couplings include (a) presence of common cis-regulatory motifs; (b) co-occurrence in the same metabolic pathway (s); (c) cis-binding to common regulator(s); (d) physical interaction; (e) common ontology; (f) paired evolutionary conservation among many organisms; (g) common synthetic phenotypes upon joint deletion with a third gene; (h) sub-cellular co-location; and (i) proximity in the genome, or in bacteria and archaea, operon co-occurrence. These associations can be either derived experimentally or computationally (either pre-computed ahead-of-time, <italic>e.g</italic>. [<xref ref-type="bibr" rid="B23">23</xref>,<xref ref-type="bibr" rid="B60">60</xref>,<xref ref-type="bibr" rid="B62">62</xref>], or on-the-fly during the clustering process); indeed it is common practice to use one or more of these associations as a <italic>post-facto </italic>measure of the biological quality of a gene cluster. However, it is important to note that some of these data types, to varying degrees, can contain a high rate of false positives, or may imply relationships that have no implication for co-regulation. Therefore in their consideration as evidence for co-regulation, these different sources of evidence should be treated as priors, with appropriately different weights, based upon prior knowledge (or assumptions) of their quality and/or relevance.</p><p>Because a biological system's interaction with its environment is complex and gene regulation is multi-factorial, genes might not be co-regulated across all experimental conditions observed in any comprehensive set of transcript or protein levels. Also, genes can be involved in multiple different processes, depending upon the state of the organism during a given experiment. Therefore, a biologically-motivated clustering method should be able to detect patterns of co-expression across subsets of the observed experiments, and to place genes into multiple clusters. So-called <italic>biclustering </italic>(clustering both genes and experimental conditions), is a widely studied problem and many different approaches to it have been published [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B25">25</xref>,<xref ref-type="bibr" rid="B52">52</xref>,<xref ref-type="bibr" rid="B76">76</xref>,<xref ref-type="bibr" rid="B80">80</xref>,<xref ref-type="bibr" rid="B86">86</xref>,<xref ref-type="bibr" rid="B98">98</xref>]. Unlike standard clustering methods, most biclustering algorithms place genes into more than one cluster. Because biclustering is an NP-hard problem [<xref ref-type="bibr" rid="B25">25</xref>], no solution is guaranteed to find the optimal set of biclusters. However, many of these procedures have successfully demonstrated the value of biclustering when applied to real-world biological data (<italic>e.g</italic>. [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B56">56</xref>,<xref ref-type="bibr" rid="B88">88</xref>]).</p><p>We have previously described a procedure, the INFERELATOR [<xref ref-type="bibr" rid="B22">22</xref>], for learning global regulatory influences from expression data using continuous models of transcript levels. For this analysis (and most regulatory network inference algorithms), a pre-clustering step is desired to reduce the dimensionality of the data and enable noise reduction through signal averaging of clustered gene profiles. Low-level (but still significantly coherent) changes in expression of the clusters play an important role in constraining the model parameters, and the inclusion of these conditions in the biclusters can be important. Thus, a trade-off needs to be found between including as many experiments as possible in each cluster (to increase the constraints on the model parameters), while enforcing that these experiments be co-expressed. Different biclustering methods have different models of a "perfect" bicluster; for example constant rows/columns, coherent values, coherent "evolution" [<xref ref-type="bibr" rid="B56">56</xref>]. For our modeling purposes, only methods which derive biclusters with coherent, or correlated, gene profiles, such as those of Cheng and Church [<xref ref-type="bibr" rid="B25">25</xref>], Yang <italic>et al</italic>. [<xref ref-type="bibr" rid="B98">98</xref>], and Lazzeroni and Owen [<xref ref-type="bibr" rid="B53">53</xref>] are suitable. For example, algorithms which identify biclusters with constant levels of activation and/or repression [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B86">86</xref>] and/or which discretize the data [<xref ref-type="bibr" rid="B80">80</xref>] do not contain low or intermediate-levels of expression changes to constrain the regulatory network inference; indeed they often do not generate biclusters with many experimental conditions at all. Our analysis and previous reviews [<xref ref-type="bibr" rid="B6">6</xref>] of the Cheng and Church (CC) algorithm [<xref ref-type="bibr" rid="B25">25</xref>] show that it is not suitable for large-scale expression analysis. It, and the Plaid models of Lazzeroni and Owen [<xref ref-type="bibr" rid="B53">53</xref>] both produce biclusters that focus on low-variance sub-matrices of the expression data. The FLOC algorithm of Yang <italic>et al</italic>. [<xref ref-type="bibr" rid="B98">98</xref>] (an update to the CC algorithm which can handle missing values) provided the early inspiration for this work, which is essentially a re-formulation of their <italic>δ</italic>-cluster model with a basic probabilistic model for the expression data. This enables a more rigorous and intuitive integration of the model of expression data with models for the additional data types, as well as with prior distributions for constraining bicluster sizes and redundancy.</p><p>Guided by these motivations and requirements, we herein describe an algorithm that detects genes putatively co-regulated over subsets of experimental conditions by integrating the biclustering of gene expression data and multiple gene association networks with the <italic>de novo </italic>detection of cis-regulatory motifs. We applied this method to a global expression data set collected for the archaeon <italic>Halobacterium </italic>NRC-1, to find co-regulated gene sets as part of our ongoing efforts to model its regulatory network, and we present detailed evidence for the biological utility of this procedure as part of our modeling procedure. In addition, we used cMonkey to compute co-regulated gene clusters for three additional organisms in the two remaining domains of life: <italic>Helicobacter pylori, Saccharomyces cerevisiae</italic>, and <italic>Escherichia coli</italic>. The biclusters are presented to the biologist using the interactive visualization tools, the <italic>Gaggle </italic>[<xref ref-type="bibr" rid="B79">79</xref>] and <italic>Cytoscape </italic>[<xref ref-type="bibr" rid="B78">78</xref>], at our web site [<xref ref-type="bibr" rid="B4">4</xref>].</p></sec><sec><title>Results</title><p>In this section we summarize the results of the application of our algorithm to four organisms, and describe its usefulness as a first step in our modeling of the <italic>Halobacterium </italic>regulatory network in conjunction with the Inferelator [<xref ref-type="bibr" rid="B22">22</xref>]. We perform a detailed analysis of its capabilities and assess its global performance, both internally and in comparison to other biclustering methods. The complete set of biclusters for all organisms are available for exploration using <italic>Cytoscape </italic>and the <italic>Gaggle </italic>[<xref ref-type="bibr" rid="B78">78</xref>,<xref ref-type="bibr" rid="B79">79</xref>] at our web site [<xref ref-type="bibr" rid="B4">4</xref>].</p><sec><title>The bacteriorhodopsin regulon in Halobacterium</title><p>The induction of phototrophic growth of <italic>Halobacterium NRC-1 </italic>under anaerobic conditions triggers the synthesis of bacteriorhodopsin (bR; a complex of the protein Bop and retinal), a light-driven proton pump that is further assembled into a purple membrane. Br is the major component of <italic>Halobacterium </italic>phototrophy, one of two anaerobic ATP generation pathways utilized by the organism [<xref ref-type="bibr" rid="B14">14</xref>]. Four genes responsible for bR synthesis (Bop and isoprenoid synthesis genes), <italic>bop, brp, bat</italic>, and <italic>crtB1</italic>, are co-regulated by Bat [<xref ref-type="bibr" rid="B13">13</xref>] through a common transcription factor motif that was characterized by saturation mutagenesis (the Bat UAS) [<xref ref-type="bibr" rid="B12">12</xref>]. This is the most well-studied regulon in <italic>Halobacterium</italic>, and the only one whose cis-regulatory motif has been experimentally verified. Bicluster #11 (Fig. <xref ref-type="fig" rid="F1">1</xref>) recapitulates much of what is known about this regulon, including all four bR genes, and a very close match to the Bat UAS (Figure <xref ref-type="fig" rid="F2">2</xref>). The additional genes in this bicluster are consistent with the co-regulation of bR with anaerobic respiration, including phytoene synthases, members of a DMSO-related operon [<xref ref-type="bibr" rid="B64">64</xref>], alcohol dehydrogenases, and an iron transporter. While the Bat UAS is not found upstream of many of these latter genes, a second significant motif (which was found upstream of the bR operon as well) was identified by cMonkey upstream of these genes.</p><p>Table <xref ref-type="table" rid="T1">1</xref> shows correlations between and among genes containing the putative Bat UAS (denoted "bR genes") and between and among genes containing the 2nd detected motif (denoted "DMSO genes"), over experiments within and outside the bicluster. While the bR genes are as tightly correlated with each other <italic>outside </italic>the bicluster as they are with the DMSO genes <italic>inside </italic>the bicluster, they are significantly less-correlated with the DMSO genes outside the bicluster (<italic>p </italic>¡ 0.026; paired <italic>t</italic>-test), and vice versa (<italic>p </italic>¡ 0.00095). This suggests that cMonkey partitioned the experiments between those in which the regulator which binds to the 2nd motif is controlling most of the genes in the bicluster (thereby causing them to appear tightly co-expressed) while over the conditions outside the bicluster, Bat is active, binds to the UAS, and bifurcates the regulation of the two sets of genes. Thus, cMonkey identified a novel relationship between phototrophy and DMSO (two of the four ATP-generating pathways available to <italic>Halobacterium</italic>), implying that the organism produces energy simultaneously via these two pathways under some environmental conditions.</p><p>The bicluster also includes <italic>cdc48a</italic>, which encodes a cell-division cycle – associated protein, with a strong match to the Bat UAS. We note that initial studies of the Bat UAS suggested that the regulatory sequences of as many as 108 genes contain instances of the motif [<xref ref-type="bibr" rid="B12">12</xref>]; clearly not all of these instances are active over the experiments used here. No similar bicluster, in terms of completeness of gene membership or similarity of motifs detected (via MEME [<xref ref-type="bibr" rid="B10">10</xref>]) to the Bat UAS, was found using other bi/clustering methods (see below for a list of methods attempted). When the cMonkey motif-detection component was turned off (see below), the UAS was not detected.</p></sec><sec><title>SirR as a regulator of transport processes in Halobacterium</title><p>cMonkey detected a bicluster (#76, Figure <xref ref-type="fig" rid="F3">3</xref>) primarily composed of transporter genes, including two phosphate transport systems, Co(II) transporters, a Mn(II) uptake system, glycerol phosphate transporters, and two peptide transport systems. While the phosphate, peptide, and Mn(II) transport systems might have been included in the bicluster by virtue of their functional associations, the glycerol phosphate and Co(II) transport system genes appear to have been included due to a strong match in the biclusters' putative motif #1. We can hypothesize that motif #1, which is present upstream to 24 out of the bicluster's 30 genes, is responsible for the high degree of expression correlation over ~150 conditions in this bicluster. None of the other bi/clustering methods tested identified a cluster containing the complete set of these transporters that enabled the generation of this type of model of the joint regulation of transporter activity in <italic>Halobacterium</italic>.</p><p>A potential advantage of the inclusion of <italic>de novo </italic>motif detection as part of the cMonkey biclustering procedure is that, for transcription factors that are not autoregulated, motif detection can break the causal symmetry between regulator targets and regulators controlling those targets. For example, an activator and several of its targets might seem co-expressed (and would therefore be placed in the same bicluster) when considering expression data alone. The absence of the regulator's binding site from its upstream sequence could, however, cause cMonkey to exclude the regulator from the bicluster, and thus assist any subsequent regulatory network inference on that bicluster. Although the above case is somewhat idealized, we find specific examples where motif detection correctly separates co-regulated groups from the co-expressed super sets that merge regulators and their targets together. SirR was predicted to regulate bicluster #76 [<xref ref-type="bibr" rid="B22">22</xref>] and this relationship was confirmed via a <italic>sirR </italic>knockout experiment [<xref ref-type="bibr" rid="B49">49</xref>]. SirR is annotated as an iron-dependent regulator in <italic>Staphylococcus epidermis </italic>and <italic>Staphylococcus aureus </italic>and is associated with Mn and Fe stress response in other microbial systems [<xref ref-type="bibr" rid="B44">44</xref>]. While <italic>sirR </italic>is correlated with the bicluster (Pearson correlation of 0.77, versus 0.69–0.92 for the genes in the bicluster), it was omitted from the bicluster by cMonkey, in part due to the poor match of <italic>sirR's </italic>upstream sequence to the bicluster's significant motif #1. While PhoU and Prpl (the transcriptional regulators that were included in bicluster #76) are also putative regulators of genes in bicluster 76, the inclusion of motif detection (along with the high stringency for co-expression used by cMonkey) suggests that SirR may have a more general role in the regulation of these transporter genes than PhoU and Prpl.</p></sec><sec><title>Regulation of flagellar biosynthesis in E. coli and H. pylori</title><p>In <italic>E. coli</italic>, the repertoire of more than 50 genes that encode proteins involved in motility (flagellar and chemotaxis system) are regulated in a cascade that can be separated into three classes. These regulatory classes correspond to the ordering of the genes' temporal requirement during flagellar assembly [<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B26">26</xref>,<xref ref-type="bibr" rid="B47">47</xref>]. Class-2 genes are regulated by an RpoD/<italic>σ</italic><sup>70 </sup>and FlhDC activation complex, and encode flagellar structural and assembly proteins and two regulators (<italic>fliA </italic>and <italic>flgM</italic>)<italic>. fliA </italic>and <italic>flgM </italic>subsequently activate the Class-3 operons (which include chemotaxis signaling and flagellar activation/motion-associated genes) [<xref ref-type="bibr" rid="B26">26</xref>]. cMonkey detected a bicluster in <italic>E. coli </italic>(Fig. <xref ref-type="fig" rid="F4">4</xref>) that is enriched in flagellar biosynthesis genes (including the regulator <italic>flgM</italic>)<italic>; </italic>most of these genes' upstream sequences contain motifs (#1 and #2 in Fig. <xref ref-type="fig" rid="F4">4</xref>) that correspond to the known promoter binding site for this activator complex [<xref ref-type="bibr" rid="B26">26</xref>]. While several other bi/clustering methods (see below for details), such as <italic>k</italic>-means and SAMBA, detected clusters that were enriched in both flagellar- and chemotaxis-associated genes, we were unable to detect the <italic>σ</italic><sup>70</sup>/FlhDC binding motif in any of these clusters due to the presence of many additional unrelated sequences that added noise to the search. The cMonkey bicluster included only two (of 11) annotated "chemotaxis"-related genes (which are all in Class-3, and do not contain the detected motif), whereas the larger SAMBA bicluster, for example, did not discriminate between these two related functions (containing 9 of the 11 genes). If MEME [<xref ref-type="bibr" rid="B10">10</xref>] is run independently on upstream sequences of the flagellar function-annotated genes (43 in all), it detects the <italic>σ</italic><sup>70</sup>/FlhDC binding motif in ~20 of them, while it does not detect a motif for the 11 chemotaxis-annotated genes (nor in the combined set of 54 sequences). This analysis suggests that while many genes in both Class-2 and Class-3 are co-expressed in the <italic>E. coli </italic>data, cMonkey can correctly separate the two classes on the basis of motif detection and association networks.</p><p>The <italic>H. pylori </italic>cluster in Fig. <xref ref-type="fig" rid="F5">5</xref> is also highly enriched in Class-2 flagellar-associated genes, many of which are associated with the RpoN/<italic>σ</italic><sup>54</sup>-regulated flagellar regulon [<xref ref-type="bibr" rid="B65">65</xref>]. The most significant motif detected in this cluster corresponds to the RpoN binding site: 5'-GGaa-N5-tttGCtT-3' [<xref ref-type="bibr" rid="B65">65</xref>] that is similar to the <italic>σ</italic><sup>70 </sup>binding site in <italic>E. coli </italic>[<xref ref-type="bibr" rid="B26">26</xref>]. Other biclustering algorithms identified biclusters in the <italic>H. pylori </italic>data containing some of the same genes as this cMonkey bicluster, however most of those clusters contain > 50 additional genes (several with > 200), and thus the RpoN-binding motif was undetectable for clusters generated by any of these methods. Individual clusters found using hierarchical clustering (<italic>k = </italic>300) and fc-means (<italic>k = </italic>50) on the <italic>H. pylori </italic>data had matches to this motif, suggesting that because the data set is small (~60 experiments), biclustering is not always necessary here. However, neither of these respective clusters were as complete in their list of genes with the RpoN-binding motif as was the cMonkey version (6 of 6 for the hierarchical clustering cluster, and 12 of 19 for the <italic>k</italic>-means cluster, versus 14 of 15 for cMonkey). The similarity in function and putative regulatory motifs for these two orthologous biclusters points to the potential future use of algorithms such as cMonkey for cross-species analyses of gene regulation [<xref ref-type="bibr" rid="B46">46</xref>,<xref ref-type="bibr" rid="B85">85</xref>].</p></sec><sec><title>A novel putative ricin-like toxin in H. pylori</title><p>The integrated analysis of the full set of biclusters in the context of additional biological knowledge (such as detailed annotations for individual genes) can result in biological insights into the combined roles of multiple biological modules. Such an analysis requires the presentation and integration of cMonkey biclusters with the visualization and exploration tools <italic>Cytoscape </italic>[<xref ref-type="bibr" rid="B78">78</xref>] and the <italic>Gaggle </italic>[<xref ref-type="bibr" rid="B79">79</xref>] (see below for details). An illustrative example in <italic>H. pylori </italic>involves a group of biclusters containing CAG pathogenicity genes. It has been hypothesized that a drop in pH may act as a signal to induce genes encoding several virulence factors including CagA (Cag26), which upon injection into target cells plays a role in the early events of gastric colonization. A known promoter motif TTTTAA [<xref ref-type="bibr" rid="B61">61</xref>,<xref ref-type="bibr" rid="B94">94</xref>] appears conserved upstream to several of these pH-induced genes. Several biclusters were detected which contain this motif and numerous pathogenicity island genes, including <italic>cag8, cag12</italic>, and <italic>virB11</italic>, which encode type IV secretion system proteins and <italic>flaA </italic>and <italic>flaB</italic>, which encode key flagellin subunits [<xref ref-type="bibr" rid="B32">32</xref>]. Other processes represented in these biclusters include outer membrane biogenesis (<italic>omp</italic>5, <italic>omp9</italic>, <italic>omp29</italic>) and peptidoglycan biosynthesis (<italic>murC, murF and murG</italic>)- which have all been implicated as important for pathogenesis [<xref ref-type="bibr" rid="B81">81</xref>,<xref ref-type="bibr" rid="B95">95</xref>]. Through the analysis of these related biclusters and their common motif, we identified a novel putative ricin-like toxin among the un-annotated <italic>H. pylori </italic>genes (HP1028) [<xref ref-type="bibr" rid="B79">79</xref>].</p></sec><sec><title>Biclusters in S. cerevisiae</title><p>The algorithm detected many strongly significant biclusters in <italic>S. cerevisiae</italic>, many of which with known or previously-observed cis-regulatory motifs, and combinations thereof. Some examples of these are included in [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>]; all cMonkey-generated yeast biclusters may be viewed and explored using <italic>Cytoscape </italic>and the <italic>Gaggle </italic>[<xref ref-type="bibr" rid="B78">78</xref>,<xref ref-type="bibr" rid="B79">79</xref>] at our web site [<xref ref-type="bibr" rid="B4">4</xref>]. Histograms of the positions of the detected motifs in the yeast upstream sequences show a marked peak near -150 bp, which hints that many of the motifs identified by cMonkey for <italic>S. cerevisiae </italic>are functional, since the motifs are actually searched for in the first 500 bp upstream of each gene [see <xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figure Twelve].</p></sec><sec><title>Validation and comparisons with available methods</title><sec><title>Tracking the cMonkey optimization</title><p>By tracking the mean progression of all biclusters during their optimization, we can quantify the degree to which the biclusters improved with regard to each model component (data type). Examples of such measures for <italic>Halobacterium </italic>are shown in Fig. <xref ref-type="fig" rid="F6">6</xref>. The scores shown are mean bicluster residual [<xref ref-type="bibr" rid="B98">98</xref>], the mean motif log-p-value [<xref ref-type="bibr" rid="B10">10</xref>], and mean log <italic>p</italic>-values of mutual clustering coefficient in certain association networks [<xref ref-type="bibr" rid="B37">37</xref>]. It is clear that most of these measures greatly improve (<italic>i.e</italic>. decrease) throughout the optimization, even though the procedure is <italic>not </italic>optimizing any of the "scores" that are plotted in Fig. <xref ref-type="fig" rid="F6">6</xref>; rather it is optimizing a joint discriminative model that includes terms which are related to these measures. We obtain similar trends in cMonkey runs on all organisms [see <xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figure Ten].</p></sec><sec><title>Testing the cMonkey model</title><sec><title>Tests of data integration</title><p>We tested whether cMonkey is correctly optimizing the joint model with respect to the different data types by varying the weights which parameterize the influence of each of those data types on the joint model (the default for these mixing parameters is set such that the three major data types have roughly equivalent influence). When we down-weight the mixture parameter for a given data type and thus eliminate its influence on the bicluster optimization, as expected, we find that this down-weighted component is poorly-optimized. At the same time, the remaining components are almost always optimized better. Thus each model component serves to regularize the bicluster model, preventing the biclusters from being over-fit to one or more individual subsets of the data. Not surprisingly, we also find that when certain components are up-weighted, they are better optimized, at the expense of a somewhat diminished ability to optimize the remaining components. [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figure Fifteen] displays mean measures of bicluster quality (here, residual against motif log-<italic>p</italic>-value) for these different cMonkey runs with weights adjusted in this manner (here, on the <italic>S. cerevisiae </italic>data). These tests show that our inclusion of the three data types results in biclusters that simultaneously satisfy our joint model better than biclusters supported by subsets of the data types (model components). A similar conclusion may be drawn from comparisons of these different cMonkey runs to "external" sources of evidence (see below and [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figures Sixteen and Eighteen]).</p></sec><sec><title>Additional tests of the relationship between multiple data types and model components</title><p>By successively removing individual components of the model, we can also characterize relationships that exist between an individual data type and the others, that have not been removed, by observing the degree to which the optimization of the removed data type still improves. For example, by turning off an individual network <italic>N </italic>(setting <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M1" name="1471-2105-7-280-i1" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msubsup><mml:mi>q</mml:mi><mml:mn>0</mml:mn><mml:mi>N</mml:mi></mml:msubsup></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGXbqCdaqhaaWcbaGaeGimaadabaGaemOta4eaaaaa@3057@</mml:annotation></mml:semantics></mml:math></inline-formula> to zero), we can rank that network with respect to the degree to which it improves (using the scores described above) when the other components (co-expression, motifs, and other networks) are optimized. For example, we find that the operon associations and protein-DNA interaction networks are well-optimized via the indirect optimization of co-expression, while metabolic pathways and phylogenetic profile associations show weaker, but still significant, correlation to co-expression. Protein interaction networks and Rosetta Stone associations appear to be the least-significantly correlated with co-expression, possibly due to their higher false-positive rate. Carrying out this type of analysis on-the-fly could allow us to iteratively update the weighting parameters as cMonkey optimizes the biclusters (so-called "Pareto-front" optimization [<xref ref-type="bibr" rid="B93">93</xref>]).</p></sec><sec><title>Randomization and shuffling tests</title><p>As an alternative to the difficult task of generating biologically realistic "synthetic" data, we chose to randomize the data instead, in order to further assess the significance of patterns discovered by cMonkey. If we completely shuffle an individual data type, then we effectively eliminate any signal that exists in that component but preserve any influence that the noise component of that data type adds to the procedure (possibly interfering with optimization of other model components). The resulting effect is very similar to strongly down-weighting that component of the model, as described above. A more stringent test can be performed by randomizing only the associations between each gene's expression data, its sequence, and its location in the association networks. This preserves the higher-order structure of each data type, but scrambles the mutual support each data type might present to the overall model. On data randomized in this manner, cMonkey is unable to find biclusters that, on average, are as well-optimized (in terms of the "scores" described above) as in the original data. The significance of this result varies depending upon the organism and the quality and amount of data available; on the <italic>Halobacterium </italic>data, this type of data shuffling results in average bicluster residuals ~20% higher, and average motif <italic>p</italic>-values ~1 log<sub>10</sub>-unit higher than in the un-shuffled data. The algorithm does not find significant association subnetworks in any of the shuffled trial runs.</p></sec></sec><sec><title>Comparison of cMonkey with other methods</title><p>In our assessment of cMonkey's performance, we compared cMonkey-generated biclusters against those generated using the following algorithms: Cheng-Church (CC [<xref ref-type="bibr" rid="B25">25</xref>]), Order Preserving Sub-matrix (OPSM [<xref ref-type="bibr" rid="B18">18</xref>]), Iterative Signature (ISA [<xref ref-type="bibr" rid="B19">19</xref>]), xMOTIF [<xref ref-type="bibr" rid="B55">55</xref>], BIMAX [<xref ref-type="bibr" rid="B6">6</xref>], and SAMBA [<xref ref-type="bibr" rid="B86">86</xref>]. We also compared our method to hierarchical clustering and <italic>k</italic>-means clustering [<xref ref-type="bibr" rid="B30">30</xref>] with <italic>k </italic>varying between 10 and 300 (see Methods for details). In addition, we performed these analyses on cMonkey runs with various model parameters up- and down-weighted, as described above, to demonstrate the effect of including various subsets of the cMonkey model components in the comparisons. Additional details on the analysis are provided in the Methods section; supporting figures are shown in [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>]. All bi/clusters generated by the various algorithms are available for interactive exploration via <italic>Cytoscape </italic>and the <italic>Gaggle </italic>[<xref ref-type="bibr" rid="B78">78</xref>,<xref ref-type="bibr" rid="B79">79</xref>] at our web site [<xref ref-type="bibr" rid="B4">4</xref>].</p><sec><title>Comparison in the context of regulatory network inference</title><p>A major motivation of this work is to provide a method for deriving co-regulated groups of genes for use in subsequent regulatory network inference procedures. To do this, we wish to find coherent groups of genes over those conditions with a large amount of variation. In other words, we are hoping to detect submatrices in the expression data matrix which are coherent and simultaneously have high information content or overall variance. In addition, we need to find biclusters with many conditions/observations included, as this increases the significance of each bicluster and also of the subsequently inferred regulatory influences for that bicluster. Some relevant summary statistics of the runs of various algorithms on all four organisms are listed in [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Table Two]. In general we see that cMonkey generates biclusters with a significantly greater number of experiments than the other methods. Even with this additional constraint (<italic>i.e</italic>. including a greater number of experiments in the clusters) and further constraints that cMonkey imposes with the association- and motif- priors, the algorithm in general generates biclusters with a "tighter" profile, as measured by mean bicluster residual [<xref ref-type="bibr" rid="B25">25</xref>]. Thus, we find that biclusters generated by cMonkey are generally better-suited for inference algorithms such as the Inferelator [<xref ref-type="bibr" rid="B22">22</xref>], and potentially other linear or continuous models as well. We tested this by running the Inferelator on biclusters generated by SAMBA [<xref ref-type="bibr" rid="B86">86</xref>] for <italic>Halobacterium </italic>and then comparing the predictive performance of the resultant regulatory network models on newly-collected data, relative to those generated for cMonkey-generated biclusters [<xref ref-type="bibr" rid="B22">22</xref>]. We found that, largely due to the smaller number of experiments included in SAMBA biclusters, the inferred network was significantly less able to predict new experiments (an increase in the predictive error from 0.368 to 0.470; <italic>p</italic>-value of difference by <italic>t</italic>-test = 1.0 × 10<sup>-22</sup>).</p></sec><sec><title>Comparison against external measures</title><p>Defining an unbiased external measure of "success" of a bi/clustering algorithm is a very difficult problem [<xref ref-type="bibr" rid="B30">30</xref>]. In fact, even if a good, unbiased measure were to be found, a comparison of different bi/clustering results in the context of that measure is also not straightforward. We have attempted to estimate various measures of success of different algorithms in various contexts, with regard to sensitivity, selectivity, and two measures of coverage, in order to provide the reader with a fair comparison of cMonkey with other previously published methods. We define the <italic>sensitivity </italic>of a bi/cluster set as the commonly-used fraction of bi/clusters that are significantly enriched with genes that (a) have the same functional annotation in GO [<xref ref-type="bibr" rid="B40">40</xref>] or KEGG [<xref ref-type="bibr" rid="B48">48</xref>], or (b) contain a known cis-regulatory motif [<xref ref-type="bibr" rid="B60">60</xref>], or (c) mimic groups of co-regulated genes, from experiments such as ChlP-chip assays [<xref ref-type="bibr" rid="B39">39</xref>]. These measures are shown for <italic>S. cerevisiae </italic>[<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figure Sixteen (A-D)] and for <italic>Halobacterium </italic>[<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figure Eighteen (A-D)] for the different algorithms. Bi/cluster <italic>specificity </italic>measures how well the bi/clusters segregate genes along the same lines as the different Classes; here, we use a measure of the fraction of genes in each significantly-annotated bi/cluster that have the same significantly-enriched annotation(s) found for that bi/cluster. We use <italic>coverage </italic>to describe two distinct measures: (a) the fraction of all observed genes and experimental conditions in the data which are included in at least one bi/cluster [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Table Two], and (b) the fraction of all groups in a given Class that are significantly enriched in at least one bi/cluster for <italic>S. cerevisiae </italic>[<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figure Sixteen E] and for <italic>Halobacterium </italic>[<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figure Eighteen E]. We should note that it is debatable which of these metrics of bicluster quality represent the best measures of "correctness" for a bi/clustering method. For example, genes that modulate the protein and transcript levels of other proteins might have similar GO functional categories (protein degradation, transcription factor, regulation, etc.) but may be correctly partitioned separately with the processes they individually regulate. It is also important to note that all of these statistical measures of bi/cluster validity contain inherent flaws or biases that correlate strongly with bi/cluster size, overlap degree, and gene coverage. For example, OPSM generated 8 biclusters which excluded less than ~1/2 of all measured genes from its clusters, yet it outperforms all other methods in the sensitivity measure. We have used the false discovery rate (which is larger for bigger clusters) to correct these <italic>p</italic>-values for multiple testing (see Methods), however, we still find a size bias in the corrected scores (which is also seen in previously-written comparisons of biclustering methods, <italic>e.g</italic>. [<xref ref-type="bibr" rid="B6">6</xref>]). In addition to GO and KEGG, we assess bi/clusters against known cis-regulatory motifs [<xref ref-type="bibr" rid="B39">39</xref>,<xref ref-type="bibr" rid="B60">60</xref>], and high-throughput protein-DNA interaction sets [<xref ref-type="bibr" rid="B39">39</xref>]. We included the runs from various test parameterizations of cMonkey in the analysis (see above), so the effect of the different input data sets could be seen. We also divided each tested bicluster set into "BIG" and "SMALL" halves, so that the size-related biases in this measurement may be seen and accounted for in the comparisons (for example, the BIG half of cMonkey's bicluster set have about the same mean number of genes per bicluster as the SMALL half of SAMBA's bicluster set, which therefore makes them more readily comparable [see <xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figures Seventeen and Nineteen]).</p><p>In general, we find that cMonkey performs well in comparison to all other methods when the trade off between sensitivity, specificity, and coverage is considered, particularly in context of the other bulk characteristics (cluster size, residual, etc.). We find that SAMBA also performs well when these measures are considered; however because its biclusters contain on average 3 × more genes than cMonkey's, and far fewer experiments (and therefore SAMBA, like most other methods, cover less of the data space), the direct comparison is difficult. cMonkey, as it was designed to do, covers more of the data space (and therefore more of the different Classes defined above) for each organism, and it is therefore more suitable for our regulatory network learning motivations. In particular, while it includes far more experiments per cluster and restricts its clusters to have significantly tighter co-expression, it still does comparably well when assessed against the external measures due to its data integration. [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Figure Sixteen] shows, for example, that the cMonkey runs carried out with the association networks up-weighted, in particular, do partition the functional classes better (and vice versa when they are turned down). The final judgement is that because cMonkey biclusters do a better job at regenerating the expression data than other methods, and at least a comparable job at recapitulating the external (as well as internal) measures of bicluster quality, they are, overall, more parsimonious with, and more generative of the patterns found in the available data. Thus, cMonkey biclusters are arguably well-suited for the inference of gene co-regulation and regulatory networks, in comparison to available bi/clustering methods.</p></sec></sec></sec><sec><title>Bicluster visualization</title><p>Because a population of biclusters will contain some overlapping elements which can confuse their interpretation, it is important to present them to the biologist in a format that promotes their interpretation and exploration in the context of supporting information, cMonkey automatically generates, for each bicluster, a "bicluster diagram" (example in Figure <xref ref-type="fig" rid="F5">5</xref>), presents to the biologist the bicluster's co-expression pattern, motif logos [<xref ref-type="bibr" rid="B74">74</xref>] and upstream sequence locations (in this study, for as many as three detected motifs), as well as the various functional associations among the bicluster's gene members. We have found that a useful and intuitive visualization scheme for a population of overlapping and often redundant biclusters is via an association network (Figure <xref ref-type="fig" rid="F7">7</xref>) of rectangular bicluster nodes (whose sizes are proportional to their gene/condition membership); analogous to "module networks" published in previous works. We visualize this bicluster network using <italic>Cytoscape </italic>[<xref ref-type="bibr" rid="B78">78</xref>]. Each bicluster is annotated with its gene and condition members, a measure of its co-expression, significant functional annotations (GO [<xref ref-type="bibr" rid="B40">40</xref>], KEGG [<xref ref-type="bibr" rid="B48">48</xref>] and COG [<xref ref-type="bibr" rid="B89">89</xref>]), and significant motifs. Edges are drawn between two biclusters if they contain non-redundant genes which are connected individually in any association networks. Connections are also added between pairs of biclusters that have a large amount of overlap in gene membership, motif similarity, expression correlation, and/or functional annotation. A spring-embedded layout algorithm [<xref ref-type="bibr" rid="B83">83</xref>] is used to spatially organize the network, placing highly-connected (and therefore related) biclusters spatially closer to each other. As a result, groups of biclusters with common function(s), or which lie in adjacent biochemical pathways, may be easily identified in the network, as shown in Figure <xref ref-type="fig" rid="F7">7</xref>. The integration of <italic>Cytoscape </italic>with the <italic>Gaggle </italic>[<xref ref-type="bibr" rid="B79">79</xref>] automatically cross-references biclusters with their respective "bicluster diagrams", and enables searching and browsing of additional biological information (such as expression data submatrices, gene browsers, annotation databases) or further analysis (<italic>e.g</italic>. via direct connection to <italic>R</italic>) of a bicluster's gene members, greatly facilitating their analysis.</p></sec></sec><sec><title>Discussion and conclusion</title><p>The integration of clustering or biclustering of expression data with additional information is a problem of growing interest. The method presented here may be compared favorably with several recently published clustering and biclustering algorithms that have integrated different types of data, including <italic>de novo </italic>detection of sequence motifs [<xref ref-type="bibr" rid="B75">75</xref>], known sequence motifs [<xref ref-type="bibr" rid="B28">28</xref>,<xref ref-type="bibr" rid="B54">54</xref>], and various types of association information [<xref ref-type="bibr" rid="B28">28</xref>,<xref ref-type="bibr" rid="B86">86</xref>,<xref ref-type="bibr" rid="B87">87</xref>]. We have (to date) seen each of these other methods applied primarily to yeast, which is unique in the quantity of data available relative to the complexity of its genome. Many aspects of our method are inspired by these works. cMonkey does not require discretization of expression data, and is therefore capable of capturing patterns in low-level responses, while still being robust to noise due to its integration of different types of biological information. For example, although the <italic>H. pylori </italic>and <italic>E. coli </italic>data was limited in size and quality (with many expression experiments containing only one replicate, and many missing values), we were able to detect several interesting biclusters with significant putative (or known) motifs. In addition, cMonkey includes a greater number of experiments in each bicluster than other methods, while still obtaining a higher amount of correlation among its gene members. Finally, cMonkey is model-based and variables (such as the distribution of bicluster sizes, and the distribution of overlap between biclusters) are parameterized using simple statistical distributions. Therefore, their adjustment is intuitive and understandable, as well as robust to varying data size and quality. In our experience, this is in contrast to other biclustering algorithms, which often require tweaking of <italic>p</italic>-value cutoffs, dimensionless variables, or thresholds, which often result in unpredictable effects on the biclusters' properties.</p><p>We believe that the ability for the cMonkey user to explicitly control the contribution of different data types through their weights opens up many potential uses for the algorithm beyond the basic identification of co-expressed clusters of genes. This flexibility enables the detection of biclusters which stress certain type(s) of biological information over others. Indeed, in many cases it is still not known whether a certain type of pair-wise association between genes is actually correlated with co-expression. Such "guilt-by-association" is often assumed, <italic>e.g</italic>. between co-expression and functional categories [<xref ref-type="bibr" rid="B97">97</xref>], but such conclusions can be controversial [<xref ref-type="bibr" rid="B11">11</xref>], as bioinformatics has "only codified a small proportion of the biological knowledge required to understand microarray data" [<xref ref-type="bibr" rid="B27">27</xref>] (obviously other types of associations, such as operon [<xref ref-type="bibr" rid="B69">69</xref>] or cis-motif co-occurrence are more strongly tied to co-expression). cMonkey users can easily choose to generate tightly-co-expressed biclusters that are strongly supported by evidence provided by one or more other sources of information for their system of interest, and they can do so by including them as highly-weighted components of the bicluster model. For example, they could (a) identify active or co-expressed signaling or regulatory pathways or complexes, as in [<xref ref-type="bibr" rid="B45">45</xref>], by up-weighting protein interaction networks or metabolic networks; (b) reconstruct metabolic pathways, by up-weighting the metabolic network and expression data, as in [<xref ref-type="bibr" rid="B50">50</xref>]; (c) attempt putative <italic>de novo </italic>cis-regulatory motif detection in newly-sequenced genomes (without expression data), by setting the expression weight to zero; (d) assess the quality of complete networks or individual edges in operon associations or protein-DNA interactions, as in [<xref ref-type="bibr" rid="B69">69</xref>], by up-weighting these associations and the expression data. Future improvements to the method could be made to learn the appropriate weights for each data type, from the data (rather than as input parameters), for example by using an unconstrained multi-parametric logistic regression as briefly described in the Methods section, or by adaptively constraining the weights such that no component of the model over-regularizes with respect to the other components (<italic>e.g</italic>. "Pareto-front" optimization [<xref ref-type="bibr" rid="B93">93</xref>]).</p><p>For sake of simplicity, flexibility and statistical transparency, we have used simple models for each of the individual data types and logistic regression to integrate them into a joint model. However, this simplicity comes at the expense of several trade-offs, which could be improved upon. Whereas it may be more appropriate to treat some associations as a property of sets rather than networks, we have treated all the same. Certain types of associations (such as protein-DNA networks and functional annotation classes) could be treated differently. In addition, any confidence values associated with individual edges in some of the networks are currently ignored. While edge weights could currently be included, for example, by dividing the high and low confidence edges into separate networks with different weights, it would be preferable to more cleanly model such association evidence. Third, we have reason to believe that our use of MEME for motif detection may be increasing our sensitivity to noise. The method could benefit from an assessment of different algorithms for detecting motifs in conjunction with biclustering, or the consensus of more than one method can be integrated, as in [<xref ref-type="bibr" rid="B39">39</xref>]. Also, as we move to more complex organisms, we find that multiple motifs cooperate in their regulation function, with conserved patterns, orientations, and upstream locations, such additional motif correlation and positional information may be exploited, with little modification to the current framework, to increase the sensitivity and specificity of identified motif patterns, such as via meta-MEME, [<xref ref-type="bibr" rid="B38">38</xref>] or others [<xref ref-type="bibr" rid="B20">20</xref>,<xref ref-type="bibr" rid="B68">68</xref>]. Also possible is the move toward the integrated multi-species biclustering of expression data, merging the multi-species clustering motivations of [<xref ref-type="bibr" rid="B83">83</xref>] with additional phylogenetic associations and motif detection (as in [<xref ref-type="bibr" rid="B96">96</xref>]).</p><p>Because the goals of the development of cMonkey are unique relative to previous biclustering methods (<italic>i.e</italic>. coupled to a continuous regulatory network inference procedure, such as the Inferelator [<xref ref-type="bibr" rid="B22">22</xref>]), the resulting biclusters have unique characteristics when compared to many previously-published methods. We have shown that the procedure "works harder" to insure that a greater percentage of genes that are observed in the data set are included in at least one cluster, while reducing redundancy between overlapping biclusters and maximizing the number of experiments that are included in each bicluster. Because of these characteristics, standard methods of assessment of biological relevance of cMonkey-generated clusters (<italic>e.g</italic>. by functional annotation over-representation) are far from ideal, as they do not account for varying bicluster sizes, redundancy, and coverage of the data. Choosing the appropriate biclustering procedure for one's needs therefore involves finding a balance of these different bicluster-set properties that returns the desired outcome. As was written by Patrick D'haeseleer, [<xref ref-type="bibr" rid="B30">30</xref>] "There is no one-size-fits-all solution to clustering, or even a consensus of what a 'good' clustering should look like."</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Materials and data</title><sec><title>Expression data</title><p>Expression data for <italic>Halobacterium </italic>were collected by members of the Baliga lab, containing genome-wide measurements of mRNA expression in 292 conditions, as described in full in [<xref ref-type="bibr" rid="B22">22</xref>] and references therein. Expression data for <italic>H. pylori </italic>and <italic>S. cerevisiae </italic>were collected from the Stanford Microarray Database [<xref ref-type="bibr" rid="B3">3</xref>]. Certain experiments such as strain comparisons, genomic DNA, and RNA decay experiments which are unlikely to relate to gene regulation were removed from the sets prior to analysis. This filtering resulted in 58 of an original 250 conditions for <italic>H. pylori </italic>and 667 of 1051 conditions for <italic>S. cerevisiae</italic>. Data for <italic>E. coli </italic>was compiled from publicly-available data provided to us by E. A1m, including 86 conditions. As a pre-processing step, genes were removed from the expression data for which there was not significant (1.5-fold) expression in any of the experiments. The data were then row-normalized (each gene's expression levels normalized to mean = 0, SD = 1). No further pre-processing or filtering of the microarray data was performed.</p></sec><sec><title>Association and metabolic networks</title><p>Genetic associations derived from comparative genomics, such as phylogenetic profile, Rosetta Stone, gene neighbor and gene cluster, were compiled from <italic>Prolinks </italic>[<xref ref-type="bibr" rid="B23">23</xref>] and <italic>Predictome </italic>[<xref ref-type="bibr" rid="B62">62</xref>] for all organisms. These networks include predicted operon "associations," which were also used to identify "unique" regulatory sequences that are to be used in the motif detection. Metabolic network reconstructions from the Kyoto Encyclopedia of Genes and Genomes [<xref ref-type="bibr" rid="B48">48</xref>] were represented as associations between two genes if they participate in a reaction sharing one or more ligands, after removing the most highly-connected ligands, such as water and ATP [<xref ref-type="bibr" rid="B16">16</xref>].</p></sec><sec><title>Interaction networks</title><p><italic>H. pylori </italic>protein-protein interactions were collected from the global experiments of [<xref ref-type="bibr" rid="B70">70</xref>]. <italic>S. Cerevisiae </italic>protein-protein, protein-DNA, and genetic interactions were collected from DIP [<xref ref-type="bibr" rid="B73">73</xref>] and BIND [<xref ref-type="bibr" rid="B9">9</xref>]. The protein-DNA interactions were converted into a network of associations between all pairs of genes whose upstream sequences were found to bind to the same regulator(s).</p></sec><sec><title>Upstream sequences</title><p>Upstream sequences for all organisms were obtained from GenBank using the Regulatory Sequence Analysis Tools (RSAT [<xref ref-type="bibr" rid="B92">92</xref>]). Using these tools, we extracted 1000-bp cis-regulatory sequences. For bacteria and archaea, these sequences were shortened to 500 bp, and then "operon-shifted" using the <italic>gene cluster </italic>(operon association) networks from <italic>Prolinks </italic>[<xref ref-type="bibr" rid="B23">23</xref>] and <italic>Predictome </italic>[<xref ref-type="bibr" rid="B62">62</xref>]. Upstream sequences for genes in the same operon were converted to "operon-shifted" sequences by using the (same) upstream sequence of the first gene in the operon Similar "operon-shifted" upstream sequences were identified using BLASTN [<xref ref-type="bibr" rid="B8">8</xref>] using a 50 bp non-gapped alignment window, to avoid using multiple copies of the same sequence in the motif detection.</p></sec><sec><title>Functional annotations for comparison tests</title><p>Gene ontology (GO) [<xref ref-type="bibr" rid="B40">40</xref>] annotations for each organism were obtained from the European Bioinformatics Institute [<xref ref-type="bibr" rid="B1">1</xref>] and matched to annotation names obtained from the GO web site. KEGG annotations were downloaded from their web site [<xref ref-type="bibr" rid="B2">2</xref>]. Predicted and experimentally-derived DNA binding motifs were obtained for <italic>S. cerevisiae </italic>from [<xref ref-type="bibr" rid="B39">39</xref>,<xref ref-type="bibr" rid="B60">60</xref>], and for <italic>E. coli </italic>from [<xref ref-type="bibr" rid="B71">71</xref>,<xref ref-type="bibr" rid="B72">72</xref>]. When these binding motifs were provided as position weight matrices (PWMs), they were converted into regular expressions, in order to enable rapid scanning of upstream sequences.</p></sec></sec><sec><title>The bicluster model</title><sec><title>Model overview</title><p>Each bicluster is modeled via a Markov chain process, in which the bicluster is iteratively optimized, and its state is updated based upon conditional probability distributions computed using the cluster's previous state. This enables us to define probabilities that each gene or condition belongs in the bicluster, <italic>conditioned upon </italic>the current state of the bicluster, as opposed to requiring us to build a complete (joint) model for the bicluster, <italic>a priori</italic>. The components of this conditional probability are modeled independently (one for each of the different types of information which we are integrating) as <italic>p</italic>-values based upon individual data likelihoods, which are then combined into a regression model to derive the full conditional probability. In this work, three major distinct data types are used (gene expression, upstream sequences, and association networks), and accordingly <italic>p</italic>-values for three such model components are computed: the <italic>expression </italic>component, the <italic>sequence </italic>component, and the <italic>network </italic>component.</p><p>Each bicluster begins as a <italic>seed</italic>, or starting cluster, that is iteratively optimized by adding/removing genes and conditions to/from the cluster by sampling from the conditional probability distribution using a Monte Carlo procedure, to prevent premature convergence. Such an iterative machine learning technique is akin to a Markov chain Monte Carlo (MCMC) process. Additional clusters are seeded and optimized until a given number (<italic>k</italic><sub>max</sub>) of clusters have been generated, or significant optimization is no longer possible. The complete process is shown schematically in Fig. <xref ref-type="fig" rid="F8">8</xref>, and described in detail below.</p><p>In the following discussion, let <italic>i </italic>be an arbitrary gene and <italic>j </italic>an arbitrary experimental condition. A bicluster <italic>k </italic>∈ <bold>K </bold>is fully defined by its set of genes <bold>I</bold><sub><italic>k </italic></sub>and experimental conditions <bold>J</bold><sub><italic>k</italic></sub>. The membership <italic>y</italic><sub><italic>lk </italic></sub>∈ {0, 1} of an arbitrary gene <italic>or </italic>condition <italic>l </italic>in bicluster <italic>k </italic>is an independent Bernoulli indicator variable with conditional probability <italic>p</italic>(<italic>y</italic><sub><italic>lk </italic></sub>= 1).</p></sec><sec><title>The expression component</title><p>The expression data is a set of measurements of genes <italic>i </italic>∈ <bold>I </bold>over experiments <italic>j </italic>∈ <bold>J</bold>, comprising a \|<bold>I</bold>\| × \|<bold>J</bold>\| matrix <italic>x</italic><sub><italic>ij </italic></sub>∈ <bold>X</bold>. Each bicluster <italic>k </italic>defines a \|<bold>I</bold><sub><italic>k</italic></sub>\| × \|<bold>J</bold><sub><italic>k</italic></sub>\| submatrix <italic>x</italic><sub><italic>i'j' </italic></sub>∈ <bold>X</bold><sub><italic>k</italic></sub>: <italic>i' </italic>∈ <bold>I</bold><sub><italic>k </italic></sub>⊂ <bold>I</bold>; <italic>j' </italic>∈ <bold>J</bold><sub><italic>k </italic></sub>⊂ <bold>J</bold>. The variance in the measured levels of condition <italic>j </italic>is <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M2" name="1471-2105-7-280-i2" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msubsup><mml:mi>σ</mml:mi><mml:mi>j</mml:mi><mml:mn>2</mml:mn></mml:msubsup><mml:mo>=</mml:mo><mml:mo>\|</mml:mo><mml:mi>I</mml:mi><mml:msup><mml:mo>\|</mml:mo><mml:mrow><mml:mo>−</mml:mo><mml:mn>1</mml:mn></mml:mrow></mml:msup><mml:mstyle displaystyle="true"><mml:msub><mml:mo>∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>∈</mml:mo><mml:mi>I</mml:mi></mml:mrow></mml:msub><mml:mrow><mml:msup><mml:mrow><mml:mo stretchy="false">(</mml:mo><mml:msub><mml:mi>x</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>−</mml:mo><mml:msub><mml:mover accent="true"><mml:mi>x</mml:mi><mml:mo>¯</mml:mo></mml:mover><mml:mi>j</mml:mi></mml:msub><mml:mo stretchy="false">)</mml:mo></mml:mrow><mml:mn>2</mml:mn></mml:msup></mml:mrow></mml:mstyle></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaiiGacqWFdpWCdaqhaaWcbaGaemOAaOgabaGaeGOmaidaaOGaeyypa0JaeiiFaWhcbeGae4xsaKKaeiiFaW3aaWbaaSqabeaacqGHsislcqaIXaqmaaGcdaaeqaqaaiabcIcaOiabdIha4naaBaaaleaacqWGPbqAcqWGQbGAaeqaaOGaeyOeI0IafmiEaGNbaebadaWgaaWcbaGaemOAaOgabeaakiabcMcaPmaaCaaaleqabaGaeGOmaidaaaqaaiabdMgaPjabgIGiolab+Leajbqab0GaeyyeIuoaaaa@494C@</mml:annotation></mml:semantics></mml:math></inline-formula>, where <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M3" name="1471-2105-7-280-i3" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi>x</mml:mi><mml:mo>¯</mml:mo></mml:mover><mml:mi>j</mml:mi></mml:msub><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:msub><mml:mo>∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>∈</mml:mo><mml:mi>I</mml:mi></mml:mrow></mml:msub><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>/</mml:mo><mml:mo>\|</mml:mo><mml:mi>I</mml:mi><mml:mo>\|</mml:mo></mml:mrow></mml:mstyle></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqWGQbGAaeqaaOGaeyypa0ZaaabeaeaacqWG4baEdaWgaaWcbaGaemyAaKMaemOAaOgabeaakiabc+caViabcYha8Hqabiab=LeajjabcYha8bWcbaGaemyAaKMaeyicI4Sae8xsaKeabeqdcqGHris5aaaa@401D@</mml:annotation></mml:semantics></mml:math></inline-formula>. We compute the mean expression level of condition <italic>j </italic>over the cluster's genes <bold>I</bold><sub><italic>k</italic></sub>, <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M4" name="1471-2105-7-280-i4" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi>x</mml:mi><mml:mo>¯</mml:mo></mml:mover><mml:mrow><mml:mi>j</mml:mi><mml:mi>k</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:msub><mml:mo>∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>∈</mml:mo><mml:msub><mml:mi>I</mml:mi><mml:mi>k</mml:mi></mml:msub></mml:mrow></mml:msub><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo>/</mml:mo><mml:mo>\|</mml:mo><mml:msub><mml:mi>I</mml:mi><mml:mi>k</mml:mi></mml:msub><mml:mo>\|</mml:mo></mml:mrow></mml:mstyle></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqWGQbGAcqWGRbWAaeqaaOGaeyypa0ZaaabeaeaacqWG4baEdaWgaaWcbaGaemyAaKMaemOAaOgabeaakiabc+caViabcYha8Hqabiab=LeajnaaBaaaleaacqWGRbWAaeqaaOGaeiiFaWhaleaacqWGPbqAcqGHiiIZcqWFjbqsdaWgaaadbaGaem4AaSgabeaaaSqab0GaeyyeIuoaaaa@44A8@</mml:annotation></mml:semantics></mml:math></inline-formula>. Then, the likelihood of an arbitrary measurement <italic>x</italic><sub><italic>ij </italic></sub>relative to this mean expression level is</p><p><inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M5" name="1471-2105-7-280-i5" overflow="scroll"> <mml:semantics definitionURL="" encoding=""> <mml:mrow> <mml:mi>p</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>x</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>j</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>=</mml:mo> <mml:mfrac> <mml:mn>1</mml:mn> <mml:mrow> <mml:msqrt> <mml:mrow> <mml:mn>2</mml:mn> <mml:mi>π</mml:mi> <mml:mrow> <mml:mo>(</mml:mo> <mml:mrow> <mml:msubsup> <mml:mi>σ</mml:mi> <mml:mi>j</mml:mi> <mml:mn>2</mml:mn> </mml:msubsup> <mml:mo>+</mml:mo> <mml:msup> <mml:mi>ε</mml:mi> <mml:mn>2</mml:mn> </mml:msup> </mml:mrow> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> </mml:msqrt> </mml:mrow> </mml:mfrac> <mml:mi>exp</mml:mi> <mml:mo>⁡</mml:mo> <mml:mrow> <mml:mo>[</mml:mo> <mml:mrow> <mml:mo>−</mml:mo> <mml:mfrac> <mml:mn>1</mml:mn> <mml:mn>2</mml:mn> </mml:mfrac> <mml:mfrac> <mml:mrow> <mml:msup> <mml:mrow> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>x</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>j</mml:mi> </mml:mrow> </mml:msub> <mml:mo>−</mml:mo> <mml:msub> <mml:mover accent="true"> <mml:mi>x</mml:mi> <mml:mo>¯</mml:mo> </mml:mover> <mml:mrow> <mml:mi>j</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> <mml:mn>2</mml:mn> </mml:msup> <mml:mo>+</mml:mo> <mml:msup> <mml:mi>ε</mml:mi> <mml:mn>2</mml:mn> </mml:msup> </mml:mrow> <mml:mrow> <mml:msubsup> <mml:mi>σ</mml:mi> <mml:mi>j</mml:mi> <mml:mn>2</mml:mn> </mml:msubsup> <mml:mo>+</mml:mo> <mml:msup> <mml:mi>ε</mml:mi> <mml:mn>2</mml:mn> </mml:msup> </mml:mrow> </mml:mfrac> </mml:mrow> <mml:mo>]</mml:mo> </mml:mrow> <mml:mo>,</mml:mo> <mml:mtext>     </mml:mtext> <mml:mrow> <mml:mo>(</mml:mo> <mml:mn>1</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> <mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGWbaCcqGGOaakcqWG4baEdaWgaaWcbaGaemyAaKMaemOAaOgabeaakiabcMcaPiabg2da9maalaaabaGaeGymaedabaWaaOaaaeaacqaIYaGmiiGacqWFapaCdaqadaqaaiab=n8aZnaaDaaaleaacqWGQbGAaeaacqaIYaGmaaGccqGHRaWkcqWF1oqzdaahaaWcbeqaaiabikdaYaaaaOGaayjkaiaawMcaaaWcbeaaaaGccyGGLbqzcqGG4baEcqGGWbaCdaWadaqaaiabgkHiTmaalaaabaGaeGymaedabaGaeGOmaidaamaalaaabaGaeiikaGIaemiEaG3aaSbaaSqaaiabdMgaPjabdQgaQbqabaGccqGHsislcuWG4baEgaqeamaaBaaaleaacqWGQbGAcqWGRbWAaeqaaOGaeiykaKYaaWbaaSqabeaacqaIYaGmaaGccqGHRaWkcqWF1oqzdaahaaWcbeqaaiabikdaYaaaaOqaaiab=n8aZnaaDaaaleaacqWGQbGAaeaacqaIYaGmaaGccqGHRaWkcqWF1oqzdaahaaWcbeqaaiabikdaYaaaaaaakiaawUfacaGLDbaacqGGSaalcaWLjaGaaCzcamaabmaabaGaeGymaedacaGLOaGaayzkaaaaaa@688B@</mml:annotation> </mml:semantics> </mml:math></inline-formula></p><p>which includes the term <italic>ε </italic>for an unknown systematic error in condition <italic>j</italic>, here assumed to be the same for all <italic>j</italic>. Note that the use of <italic>σ</italic><sub><italic>j </italic></sub>over all genes <bold>I </bold>rather than a <italic>σ</italic><sub><italic>jk </italic></sub>computed over <bold>I</bold><sub><italic>k </italic></sub>results in a lower likelihood <italic>p</italic>(<italic>x</italic><sub><italic>ij</italic></sub>) for those conditions <italic>j </italic>that have a small overall variance, and are therefore more likely to be correlated by random chance. Also, such low-variance conditions could be the result of poor labeling, or other systematic problems.</p><p>The likelihood of the measurements of an arbitrary gene <italic>i </italic>among the conditions in bicluster <italic>k </italic>are <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M6" name="1471-2105-7-280-i6" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:mi>p</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:msub><mml:mi>x</mml:mi><mml:mi>i</mml:mi></mml:msub><mml:mo stretchy="false">)</mml:mo><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:msub><mml:mo>∏</mml:mo><mml:mrow><mml:mi>j</mml:mi><mml:mo>∈</mml:mo><mml:msub><mml:mi>J</mml:mi><mml:mi>k</mml:mi></mml:msub></mml:mrow></mml:msub><mml:mrow><mml:mi>p</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:msub><mml:mi>x</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo stretchy="false">)</mml:mo></mml:mrow></mml:mstyle></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGWbaCcqGGOaakcqWG4baEdaWgaaWcbaGaemyAaKgabeaakiabcMcaPiabg2da9maarababaGaemiCaaNaeiikaGIaemiEaG3aaSbaaSqaaiabdMgaPjabdQgaQbqabaGccqGGPaqkaSqaaiabdQgaQjabgIGioJqabiab=PeaknaaBaaameaacqWGRbWAaeqaaaWcbeqdcqGHpis1aaaa@42C6@</mml:annotation></mml:semantics></mml:math></inline-formula>, and similarly the likelihood of a condition <italic>j</italic>'s measurements are <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M7" name="1471-2105-7-280-i7" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:mi>p</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:msub><mml:mi>x</mml:mi><mml:mi>j</mml:mi></mml:msub><mml:mo stretchy="false">)</mml:mo><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:msub><mml:mo>∏</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>∈</mml:mo><mml:msub><mml:mi>I</mml:mi><mml:mi>k</mml:mi></mml:msub></mml:mrow></mml:msub><mml:mrow><mml:mi>p</mml:mi><mml:mo stretchy="false">(</mml:mo><mml:msub><mml:mi>x</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub><mml:mo stretchy="false">)</mml:mo></mml:mrow></mml:mstyle></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGWbaCcqGGOaakcqWG4baEdaWgaaWcbaGaemOAaOgabeaakiabcMcaPiabg2da9maarababaGaemiCaaNaeiikaGIaemiEaG3aaSbaaSqaaiabdMgaPjabdQgaQbqabaGccqGGPaqkaSqaaiabdMgaPjabgIGioJqabiab=LeajnaaBaaameaacqWGRbWAaeqaaaWcbeqdcqGHpis1aaaa@42C4@</mml:annotation></mml:semantics></mml:math></inline-formula>. We integrate the two tails of the Normal distribution in Eq. 1 to derive <italic>co-expression </italic><italic>p</italic>-values for each gene <italic>i, r</italic><sub><italic>ik</italic></sub>, and for each condition <italic>j, r</italic><sub><italic>jk</italic></sub>, relative to bicluster <italic>k</italic>.</p></sec><sec><title>Sequence component (motif co-occurrence)</title><p>Each gene <italic>i </italic>has an upstream cis-regulatory sequence <italic>S</italic><sub><italic>i </italic></sub>(a string of DNA nucleotides of length <italic>l</italic><sub>S</sub>), and bicluster <italic>k </italic>defines a set of sequences <bold>S</bold><sub><italic>k </italic></sub>for all <italic>S</italic><sub><italic>i'</italic></sub>; <italic>i</italic>' ∈ <bold>I</bold><sub><italic>k</italic></sub>. The decision whether an arbitrary gene's upstream sequence, <italic>S</italic><sub><italic>i</italic></sub>, shares common motif(s) with sequences <bold>S</bold><sub><italic>k</italic></sub>, is determined via a two-step process: (1) identify one or more motif(s) <bold>M</bold><sub><italic>k </italic></sub>that is (are) significantly overrepresented in many (if not all) bicluster sequences <bold>S</bold><sub><italic>k</italic></sub>, and then (2) scan <italic>Si </italic>to see if it also contains <bold>M</bold><sub><italic>k</italic></sub>.</p><p>In this work, we are not advancing the basic methodology for motif detection (step 1), as relatively mature methods exist for finding motifs given a fixed set of sequences [<xref ref-type="bibr" rid="B91">91</xref>]. Instead, we are describing an overall strategy that incorporates previously existing motif finding algorithm(s) into a clustering procedure. As such, the procedure is motif-detection-algorithm agnostic, and the search may be performed using one of many existing methods [<xref ref-type="bibr" rid="B91">91</xref>]. Our only requirements are that (a) significantly overrepresented motifs do not have to exist in <italic>all </italic>sequences <bold>S</bold><sub><italic>k</italic></sub>, and (b) it can produce a score (preferentially a <italic>p</italic>-value) that an arbitrary sequence contains the detected motif(s). The MEME algorithm [<xref ref-type="bibr" rid="B10">10</xref>], which identifies significant sequence motifs using expectation maximization of one or more probabilistic motif models given a fixed set of sequences and a background residue model, is used to perform step (1), as it meets the first criterion (a). MEME's companion algorithm MAST [<xref ref-type="bibr" rid="B10">10</xref>], which computes the <italic>p</italic>-value that an arbitrary sequence matches the set of motifs detected with MEME, is used to perform step (2), as it meets the second criterion (b). During the motif detection step, for any genes in bicluster <italic>k </italic>which are in an operon, we make sure to use only <italic>one copy </italic>of the upstream sequence for that operon (<italic>i.e</italic>. upstream of the first gene in that operon), as described above ("Upstream sequences"). Additional details on the specific parameters passed to these procedures are provided in [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Table Three].</p><p>Thus, using these two algorithms, we can detect a set of motifs <bold>M</bold><sub><italic>k </italic></sub>in sequences <bold>S</bold><sub><italic>k</italic></sub>, and compute a <italic>p</italic>-value that a sequence <italic>S</italic><sub><italic>i </italic></sub>contains those motifs. Note that this <italic>p</italic>-value is computed for <italic>each </italic>upstream sequence in the genome, including those for the genes <italic>within </italic>cluster <italic>k</italic>, to derive the <italic>motif </italic><italic>p</italic>-values, <italic>s</italic><sub><italic>ik</italic></sub>, for each gene <italic>i </italic>relative to bicluster <italic>k</italic>, at each iteration of the MCMC procedure.</p></sec><sec><title>Association network component</title><p>To build up a highly-connected subnetwork among genes that are in a bicluster (given a full set of associations), we aim to add genes preferentially that have a greater number of connections to those currently in the bicluster than one would expect (at random) based upon the overall connectivity in the network. Thus, we compute <italic>p</italic>-values for observing the associations between a gene or experimental condition and the genes or conditions currently in bicluster <italic>k</italic>, given an association network <italic>N </italic>∈ <bold>N</bold>. In the following discussion, genes are the primary consideration, but networks of associations between experimental conditions are conceivable (<italic>e.g</italic>., we might wish to preferentially group conditions that are part of the same time series). The <italic>network association </italic><italic>p</italic>-value, <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M8" name="1471-2105-7-280-i8" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msubsup><mml:mi>q</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>k</mml:mi></mml:mrow><mml:mi>N</mml:mi></mml:msubsup></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGXbqCdaqhaaWcbaGaemyAaKMaem4AaSgabaGaemOta4eaaaaa@3223@</mml:annotation></mml:semantics></mml:math></inline-formula>, is computed based upon the number of edges in network <italic>N </italic>connecting gene <italic>i </italic>to genes <bold>I</bold><sub><italic>k </italic></sub>in bicluster <italic>k</italic>, relative to the total number of edges connected between <italic>i </italic>and the genes <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M9" name="1471-2105-7-280-i9" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msub><mml:msup><mml:mi>I</mml:mi><mml:mo>′</mml:mo></mml:msup><mml:mi>k</mml:mi></mml:msub></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieqacuWFjbqsgaqbamaaBaaaleaacqWGRbWAaeqaaaaa@2F64@</mml:annotation></mml:semantics></mml:math></inline-formula> (that are <italic>not </italic>in cluster <italic>k</italic>), as well as the connections within and between the gene sets <bold>I</bold><sub><italic>k </italic></sub>and <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M10" name="1471-2105-7-280-i9" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msub><mml:msup><mml:mi>I</mml:mi><mml:mo>′</mml:mo></mml:msup><mml:mi>k</mml:mi></mml:msub></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieqacuWFjbqsgaqbamaaBaaaleaacqWGRbWAaeqaaaaa@2F64@</mml:annotation></mml:semantics></mml:math></inline-formula>. The hypergeometric distribution is used to compute the probability of observing such an arrangement of connections by chance:</p><p><inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M11" name="1471-2105-7-280-i10" overflow="scroll"> <mml:semantics definitionURL="" encoding=""> <mml:mrow> <mml:mi>p</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:msup> <mml:mi>I</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>,</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>→</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>,</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>→</mml:mo> <mml:msub> <mml:msup> <mml:mi>I</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>=</mml:mo> <mml:mfrac> <mml:mrow> <mml:mrow> <mml:mo>(</mml:mo> <mml:mrow> <mml:mtable> <mml:mtr> <mml:mtd> <mml:mrow> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>→</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> </mml:mrow> </mml:mtd> </mml:mtr> <mml:mtr> <mml:mtd> <mml:mrow> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> </mml:mrow> </mml:mtd> </mml:mtr> </mml:mtable> </mml:mrow> <mml:mo>)</mml:mo> </mml:mrow> <mml:mrow> <mml:mo>(</mml:mo> <mml:mrow> <mml:mtable> <mml:mtr> <mml:mtd> <mml:mrow> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:msup> <mml:mi>I</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>→</mml:mo> <mml:msub> <mml:msup> <mml:mi>I</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> </mml:mrow> </mml:mtd> </mml:mtr> <mml:mtr> <mml:mtd> <mml:mrow> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:msup> <mml:mi>I</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> </mml:mrow> </mml:mtd> </mml:mtr> </mml:mtable> </mml:mrow> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> <mml:mrow> <mml:mrow> <mml:mo>(</mml:mo> <mml:mrow> <mml:mtable> <mml:mtr> <mml:mtd> <mml:mrow> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>→</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:msup> <mml:mi>I</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>→</mml:mo> <mml:msub> <mml:msup> <mml:mi>I</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> </mml:mrow> </mml:mtd> </mml:mtr> <mml:mtr> <mml:mtd> <mml:mrow> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>→</mml:mo> <mml:msub> <mml:msup> <mml:mi>I</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> </mml:mrow> </mml:mtd> </mml:mtr> </mml:mtable> </mml:mrow> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> </mml:mfrac> <mml:mo>,</mml:mo> <mml:mtext>     </mml:mtext> <mml:mrow> <mml:mo>(</mml:mo> <mml:mn>2</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> <mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGWbaCcqGGOaakcqWGUbGBdaWgaaWcbaGaemyAaKMaeyOKH4kcbeGae8xsaK0aaSbaaWqaaiabdUgaRbqabaaaleqaaOGaeiiFaWNaemOBa42aaSbaaSqaaiabdMgaPjabgkziUkqb=LeajzaafaWaaSbaaWqaaiabdUgaRbqabaaaleqaaOGaeiilaWIaemOBa42aaSbaaSqaaiab=LeajnaaBaaameaacqWGRbWAaeqaaSGaeyOKH4Qae8xsaK0aaSbaaWqaaiabdUgaRbqabaaaleqaaOGaeiilaWIaemOBa42aaSbaaSqaaiab=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LeajzaafaWaaSbaaWqaaiabdUgaRbqabaaaleqaaaGcbaGaemOBa42aaSbaaSqaaiabdMgaPjabgkziUkqb=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@C436@</mml:annotation> </mml:semantics> </mml:math></inline-formula></p><p>where <bold>A </bold>→ <bold>B </bold>represents the set of associations between the elements in gene set <bold>A </bold>with those in set <bold>B</bold>, and <italic>n</italic><sub><bold>A</bold>→<bold>B </bold></sub>is the number of these associations. Expression (2) is analogous to the hypergeometric measure of mutual clustering coefficient described by [<xref ref-type="bibr" rid="B37">37</xref>]. However, it does not account for the global structure of the network; it is only concerned with the local associations, <italic>i.e</italic>. those directly connected to gene <italic>i </italic>and the bicluster's genes, <bold>I</bold><sub><italic>k</italic></sub>. This choice of connectivity measure allows a single value to be directly computed for each gene, relative to each cluster, and gives greater preference to an individual gene <italic>i </italic>being added to cluster <italic>k </italic>if a large fraction of <italic>i's </italic>associations are with the other genes in the cluster (and vice versa), independent of the global distribution of associations in the network. Individual <italic>p</italic>-values, <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M12" name="1471-2105-7-280-i8" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msubsup><mml:mi>q</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>k</mml:mi></mml:mrow><mml:mi>N</mml:mi></mml:msubsup></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacqWGXbqCdaqhaaWcbaGaemyAaKMaem4AaSgabaGaemOta4eaaaaa@3223@</mml:annotation></mml:semantics></mml:math></inline-formula>, for each gene <italic>i </italic>and each network <italic>N </italic>are computed for bicluster <italic>k </italic>by integrating the lower tail of the distribution in Eq. 2.</p></sec><sec><title>The joint cluster membership probability</title><p>The ultimate goal is to decide gene or condition bicluster membership jointly on the basis of the three individual sets of <italic>p</italic>-values <italic>r</italic><sub><italic>ik</italic></sub>, <italic>s</italic><sub><italic>ik</italic></sub>, and <italic>q</italic><sub><italic>ik </italic></sub>computed above (for the remainder of this discussion, we now use <italic>i </italic>to denote a gene <italic>or </italic>experimental condition). A common procedure for combining "scores" such as these into a single joint likelihood is to perform a multi-parametric logistic regression [<xref ref-type="bibr" rid="B41">41</xref>] that treats each <italic>p</italic>-value measure as a random variable and estimates the joint membership likelihood <italic>p</italic>(<italic>y</italic><sub><italic>ik </italic></sub>= 1) using the logistic function,</p><p><inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M13" name="1471-2105-7-280-i11" overflow="scroll"> <mml:semantics definitionURL="" encoding=""> <mml:mrow> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo>≡</mml:mo> <mml:mi>p</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>y</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo>=</mml:mo> <mml:mn>1</mml:mn> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>X</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>,</mml:mo> <mml:msub> <mml:mi>S</mml:mi> <mml:mi>i</mml:mi> </mml:msub> <mml:mo>,</mml:mo> <mml:msub> <mml:mi>M</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>,</mml:mo> <mml:mi>N</mml:mi> <mml:mo stretchy="false">)</mml:mo> <mml:mo>∝</mml:mo> <mml:mi>exp</mml:mi> <mml:mo>⁡</mml:mo> <mml:mrow> <mml:mo>[</mml:mo> <mml:mrow> <mml:msub> <mml:mi>β</mml:mi> <mml:mn>0</mml:mn> </mml:msub> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>r</mml:mi> <mml:mn>0</mml:mn> </mml:msub> <mml:mi>log</mml:mi> <mml:mo>⁡</mml:mo> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>r</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>s</mml:mi> <mml:mn>0</mml:mn> </mml:msub> <mml:mi>log</mml:mi> <mml:mo>⁡</mml:mo> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>s</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>+</mml:mo> <mml:mstyle displaystyle="true"> <mml:munder> <mml:mo>∑</mml:mo> <mml:mrow> <mml:mi>N</mml:mi> <mml:mo>∈</mml:mo> <mml:mi>N</mml:mi> </mml:mrow> </mml:munder> <mml:mrow> <mml:msubsup> <mml:mi>q</mml:mi> <mml:mn>0</mml:mn> <mml:mi>N</mml:mi> </mml:msubsup> </mml:mrow> </mml:mstyle> <mml:mi>log</mml:mi> <mml:mo>⁡</mml:mo> <mml:mo stretchy="false">(</mml:mo> <mml:msubsup> <mml:mi>q</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> <mml:mi>N</mml:mi> </mml:msubsup> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> <mml:mo>]</mml:mo> </mml:mrow> <mml:mo>.</mml:mo> <mml:mtext>     </mml:mtext> <mml:mrow> <mml:mo>(</mml:mo> <mml:mn>3</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> <mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaiiGacqWFapaCdaWgaaWcbaGaemyAaKMaem4AaSgabeaakiabggMi6kabdchaWjabcIcaOiabdMha5naaBaaaleaacqWGPbqAcqWGRbWAaeqaaOGaeyypa0JaeGymaeJaeiiFaWhcbeGae4hwaG1aaSbaaSqaaiabdUgaRbqabaGccqGGSaalcqWGtbWudaWgaaWcbaGaemyAaKgabeaakiabcYcaSiab+1eannaaBaaaleaacqWGRbWAaeqaaOGaeiilaWIae4Nta4KaeiykaKIaeyyhIuRagiyzauMaeiiEaGNaeiiCaa3aamWaaeaacqWFYoGydaWgaaWcbaGaeGimaadabeaakiabgUcaRiabdkhaYnaaBaaaleaacqaIWaamaeqaaOGagiiBaWMaei4Ba8Maei4zaCMaeiikaGIaemOCai3aaSbaaSqaaiabdMgaPjabdUgaRbqabaGccqGGPaqkcqGHRaWkcqWGZbWCdaWgaaWcbaGaeGimaadabeaakiGbcYgaSjabc+gaVjabcEgaNjabcIcaOiabdohaZnaaBaaaleaacqWGPbqAcqWGRbWAaeqaaOGaeiykaKIaey4kaSYaaabuaeaacqWGXbqCdaqhaaWcbaGaeGimaadabaGaemOta4eaaaqaaiabd6eaojabgIGiolab+5eaobqab0GaeyyeIuoakiGbcYgaSjabc+gaVjabcEgaNjabcIcaOiabdghaXnaaDaaaleaacqWGPbqAcqWGRbWAaeaacqWGobGtaaGccqGGPaqkaiaawUfacaGLDbaacqGGUaGlcaWLjaGaaCzcamaabmaabaGaeG4mamdacaGLOaGaayzkaaaaaa@8A43@</mml:annotation> </mml:semantics> </mml:math></inline-formula></p><p>This model approximates a (probabilistic) discriminating hyperplane in the space defined by <italic>r</italic><sub><italic>ik</italic></sub>, <italic>s</italic><sub><italic>ik</italic></sub>, and <italic>q</italic><sub><italic>ik</italic></sub>, parameterized by the four independent variables <italic>β</italic><sub>0 </sub>(the intercept), and <italic>r</italic><sub>0</sub>, <italic>s</italic><sub>0</sub>, and <italic>q</italic><sub>0 </sub>(the slope) that maximally discriminates the genes or conditions within the bicluster (<bold>I</bold><sub><italic>k</italic></sub>) from those outside (<inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M14" name="1471-2105-7-280-i9" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msub><mml:msup><mml:mi>I</mml:mi><mml:mo>′</mml:mo></mml:msup><mml:mi>k</mml:mi></mml:msub></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaaieqacuWFjbqsgaqbamaaBaaaleaacqWGRbWAaeqaaaaa@2F64@</mml:annotation></mml:semantics></mml:math></inline-formula>). Conceptually, the model implies that a gene or condition that poorly matches the bicluster based on one data type can still be added to the bicluster if it matches well to the other data types, analogous to, for example, the explicit "softening" of cluster boundaries performed by [<xref ref-type="bibr" rid="B15">15</xref>]. Note that when element <italic>i </italic>is an experimental condition, <italic>s</italic><sub>0 </sub>(the motif parameter) is zero.</p><p>In practice, during early iterations when the bicluster is not well-discriminated from the background, such an unconstrained regression leads to unstable situations such as unwarranted over-weighting or inversion of one or more variables (<italic>r</italic><sub>0</sub>, <italic>s</italic><sub>0</sub>, or <italic>q</italic><sub>0</sub>). Additionally, depending upon the quality of the data set(s) being used and the predisposition (or prior knowledge) of the researcher, different runs of the algorithm stressing different data types may be desired. Finally, there is good reason to expect that certain data types (<italic>e.g</italic>. sequence motifs) will be less informative early in the procedure when the biclusters are poorly-defined, and only later will it make sense to incorporate them into the bicluster model.</p><p>Therefore, we perform a <italic>constrained </italic>logistic regression by transforming the regression space defined by <italic>r</italic><sub><italic>ik</italic></sub>, <italic>s</italic><sub><italic>ik</italic></sub>, and <italic>q</italic><sub><italic>ik </italic></sub>into one dimension, projecting the log-<italic>p</italic>-values onto a single vector, <italic>g</italic><sub><italic>ik</italic></sub>,</p><p><inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M15" name="1471-2105-7-280-i12" overflow="scroll"> <mml:semantics definitionURL="" encoding=""> <mml:mrow> <mml:msub> <mml:mi>g</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo>=</mml:mo> <mml:msub> <mml:mi>r</mml:mi> <mml:mn>0</mml:mn> </mml:msub> <mml:mi>log</mml:mi> <mml:mo>⁡</mml:mo> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mover accent="true"> <mml:mi>r</mml:mi> <mml:mo>˜</mml:mo> </mml:mover> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>+</mml:mo> <mml:msub> <mml:mi>s</mml:mi> <mml:mn>0</mml:mn> </mml:msub> <mml:mi>log</mml:mi> <mml:mo>⁡</mml:mo> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mover accent="true"> <mml:mi>s</mml:mi> <mml:mo>˜</mml:mo> </mml:mover> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>+</mml:mo> <mml:mstyle displaystyle="true"> <mml:munder> <mml:mo>∑</mml:mo> <mml:mrow> <mml:mi>N</mml:mi> <mml:mo>∈</mml:mo> <mml:mi>N</mml:mi> </mml:mrow> </mml:munder> <mml:mrow> <mml:msubsup> <mml:mi>q</mml:mi> <mml:mn>0</mml:mn> <mml:mi>N</mml:mi> </mml:msubsup> <mml:mi>log</mml:mi> <mml:mo>⁡</mml:mo> <mml:mo stretchy="false">(</mml:mo> <mml:msubsup> <mml:mover accent="true"> <mml:mi>q</mml:mi> <mml:mo>˜</mml:mo> </mml:mover> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> <mml:mi>N</mml:mi> </mml:msubsup> <mml:mo stretchy="false">)</mml:mo> <mml:mo>,</mml:mo> <mml:mtext>     </mml:mtext> <mml:mrow> <mml:mo>(</mml:mo> <mml:mn>4</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> </mml:mstyle> </mml:mrow> <mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=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5eaobqab0GaeyyeIuoaaaa@670A@</mml:annotation> </mml:semantics> </mml:math></inline-formula></p><p>where <italic>r</italic><sub>0</sub>, <italic>s</italic><sub>0</sub>, and <italic>q</italic><sub>0 </sub>are specified for each iteration according to an "annealing schedule," described below. Here, each of the dimensions have been standardized to place the log-<italic>p</italic>-values for each data set on the same scale, with log(<inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M16" name="1471-2105-7-280-i13" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mover accent="true"><mml:mi>χ</mml:mi><mml:mo>˜</mml:mo></mml:mover><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuaHhpWygaacaaaa@2E72@</mml:annotation></mml:semantics></mml:math></inline-formula><sub><italic>ik</italic></sub>) = [log(<italic>χ</italic><sub><italic>ik </italic></sub>) - <italic>μ</italic><sub><italic>k</italic></sub>]/<italic>σ</italic><sub><italic>k</italic></sub>, where <italic>μ</italic><sub><italic>k </italic></sub>and <italic>σ</italic><sub><italic>k </italic></sub>are the mean and standard deviation of the log-<italic>p</italic>-values <italic>log</italic>(<italic>χ</italic><sub><italic>ik</italic></sub>) (<italic>χ</italic><sub><italic>ik </italic></sub>denotes either <italic>r</italic><sub><italic>ik</italic></sub>, <italic>s</italic><sub><italic>ik</italic></sub>, or <italic>q</italic><sub><italic>ik</italic></sub>), only over the genes or conditions in the bicluster (<italic>i ∈ </italic>I<sub><italic>k</italic></sub>). This is necessary because the <italic>p</italic>-values for each component of our model were derived for different types of data, each with widely differing sizes. For example, the <italic>p</italic>-values are likely to be smaller (on average) for the component with the most data (here, the expression data), or for motifs with larger lengths. The projection described in Eq. 4 constrains the regression by fixing the slope of the discriminating hyperplane (via parameters <italic>r</italic><sub>0</sub>, <italic>s</italic><sub>0</sub>, and <italic>q</italic><sub>0</sub>), with only the intercept <italic>β</italic><sub>0 </sub>permitted to vary from cluster to cluster. These parameters may also be interpreted as mixing parameters that control the fractional contribution of each model component to the cluster likelihood, <italic>g</italic><sub><italic>ik</italic></sub>. They may be defined by the user, and/or may be modified throughout the course of cluster optimization. For example, early in the procedure when the bicluster is a poorly-defined seed, co-expression and certain association networks (<italic>e.g</italic>. operon associations, for bacteria and archaea) are extremely informative, whereas a common cis-regulatory motif is less likely exist among the genes in the bicluster. Only later (when the bicluster has been optimized on the basis of expression data and operon associations) does it make sense to incorporate sequence motifs into the bicluster model. Therefore we employ a strategy for slowly varying the relative contribution of each of the regression parameters, as the cluster is optimized, as part of an annealing schedule (described further, below). The constrained binomial regression is now given by</p><p><italic>π</italic><sub><italic>ik </italic></sub>≡ <italic>p</italic>(<italic>y</italic><sub><italic>ik </italic></sub>= 1\|<bold>X</bold><sub><italic>k</italic></sub>, <italic>S</italic><sub><italic>i</italic></sub>, <bold>M</bold><sub><italic>k</italic></sub>, <bold>N</bold>) ∝ exp(<italic>β</italic><sub>0 </sub>+ <italic>β</italic><sub>1</sub>g<sub><italic>ik</italic></sub>),     (5)</p><p>where parameters <bold><italic>β </italic></bold>= [<italic>β</italic><sub>0</sub>, <italic>β</italic><sub>1</sub>] fully determine the conditional probability of membership <italic>p</italic>(<italic>y</italic><sub><italic>ik </italic></sub>= 1) of a gene or condition <italic>i </italic>in bicluster <italic>k</italic>.</p><p>One additional complication arises near the end of a bicluster optimization, that a bicluster may be perfectly discriminated from the background (resulting in an infinite negative log-likelihood and undefined regression). This may be addressed in two ways: the first is to constrain or fix the slope <italic>β</italic><sub>1 </sub>of the regression, allowing only the intercept <italic>β</italic><sub>0 </sub>to vary. We chose a second option, to perform a penalized maximum likelihood estimation described by [<xref ref-type="bibr" rid="B42">42</xref>] and originally proposed by [<xref ref-type="bibr" rid="B35">35</xref>]. This penalized estimate of <italic>β </italic>provides bias reduction in the case of small sample sizes (small biclusters), and solves the separation problem in the context of perfect discrimination and infinite likelihood. <italic>β </italic>can be determined with this penalized likelihood measure using an efficient iterative process [<xref ref-type="bibr" rid="B35">35</xref>].</p><p>We now have a set of probabilities, <italic>π</italic><sub><italic>ik</italic></sub>, that each gene or condition <italic>i </italic>is associated with bicluster <italic>k </italic>given the bicluster's current state. We would now like to perform moves (<italic>i.e</italic>. add or remove genes and conditions) that are most likely to improve the likelihood of the bicluster based upon the model. We do this by sampling moves from <italic>π</italic><sub><italic>ik</italic></sub>. These probabilities may be further adjusted via additional (prior) constraints on the model, as described below.</p></sec></sec><sec><title>The cMonkey iterative procedure</title><sec><title>Seeding the clusters</title><p>The Markov chain process by which a bicluster is optimized requires "seeding" of the bicluster to start the procedure. We experimented with many data-driven methods for generating seed biclusters, including (a) single-gene seeds, (b) random or semi-correlated seeds using a pre-specified distribution of cluster sizes, and (c) seeding on the basis of co-expressed edges in association networks (for example, operon associations). In principle, any seeding method may be used, including the clusters produced by other clustering or biclustering methods. Many <italic>different </italic>seeding methods are used in order to broaden the parameter space which is searched, and depending upon the annealing schedule used, the algorithm can be made more- or less-sensitive to the selected starting points. As biclusters are optimized sequentially, in order to maximize our coverage of the overall (gene) search space, they are seeded only with genes that have not previously been placed into any other biclusters. It should be noted, however, that during subsequent iterations, genes that are already in other biclusters <italic>can </italic>still be added to new biclusters, with additional constraints that are described later.</p><p>Each bicluster is seeded using a random choice from one of a variety of methods, each of which utilizes one or more different types of input data. For each newly-seeded bicluster, <bold>I</bold>' be the set of genes that are currently not in any other biclusters, <italic>i </italic>is a randomly-chosen gene from <bold>I</bold>' and <bold>J</bold><sub><italic>i </italic></sub>is the set of conditions in which <italic>i </italic>has the highest amount of variance. The seeding methods available are:</p><p>1. <italic>A single random gene</italic>: The cluster is seeded with <italic>i </italic>and <bold>J</bold><sub><italic>i</italic></sub>. For the first few iterations of this bicluster's optimization, only gene additions are allowed (forcing the bicluster to grow in size, early on).</p><p>3. <italic>n co-expressed genes from another clustering method</italic>: Clusters are generated using an other clustering or biclustering method, and these are used as seeds for further optimization.</p><p>2. <italic>n semi-co-expressed genes</italic>: Up to <italic>n </italic>- 1 additional genes from <bold>I</bold>' are randomly chosen from those with a high Pearson correlation (<italic>P</italic><sub><italic>c </italic></sub>> 0.8) with <italic>i </italic>in conditions <bold>J</bold><sub><italic>i</italic></sub>. n is chosen randomly from a set of pre-defined cluster seeding sizes, currently 2, 5, 10, 20, <italic>μ</italic>, where <italic>μ</italic>, is described Methods.</p><p>4. <italic>n highly-connected genes</italic>: Up to <italic>n </italic>- 1 random genes from <bold>I</bold>' are added from those with <italic>P</italic><sub><italic>c </italic></sub>> 0 with <italic>i</italic>, and are first neighbors with <italic>i </italic>in a given association network, <italic>n </italic>is chosen as in (2); the association network is chosen randomly from the following (if available for that organism): operons, metabolic assoc., protein-DNA interactions, protein-protein interactions.</p><p>5. <italic>n genes with a common motif</italic>: Up to <italic>n </italic>- 1 genes from <bold>I</bold>' are randomly chosen from those with <italic>P</italic><sub><italic>c </italic></sub>> 0 with <italic>i</italic>, and also have a common <italic>d</italic>-mer with <italic>i </italic>in their upstream sequences, allowing for up to <italic>l </italic>residue differences, <italic>n </italic>is chosen as in (2); defaults of <italic>d </italic>= 9 and <italic>l </italic>= 1 were used.</p></sec><sec><title>Annealing the clusters</title><p>A newly-seeded bicluster <italic>k </italic>is iteratively improved with respect to the joint likelihood derived above. At each iteration, significant motifs are detected (using MEME), and the joint membership probabilities <italic>π</italic><sub><italic>ik </italic></sub>for each gene or condition <italic>i </italic>are computed. We then perform moves using Simulated Annealing (SA) [<xref ref-type="bibr" rid="B51">51</xref>], to preferentially add genes or conditions <italic>i </italic>to bicluster <italic>k </italic>if they have a high probability of membership (<italic>y</italic><sub><italic>ik </italic></sub>= 0 and <italic>π</italic><sub><italic>ik </italic></sub>≈ 1), and to drop genes or conditions from that bicluster if they have a low probability of membership (<italic>y</italic><sub><italic>ik </italic></sub>= 1 and <italic>π</italic><sub><italic>ik </italic></sub>≈ 0). Moves which may decrease the likelihood of the cluster model are permitted, with a frequency that decreases during the course of the procedure, as parameterized by an annealing temperature <italic>T</italic>:</p><p><inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M17" name="1471-2105-7-280-i14" overflow="scroll"> <mml:semantics definitionURL="" encoding=""> <mml:mrow> <mml:mtable> <mml:mtr> <mml:mtd> <mml:mrow> <mml:mi>p</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:mi>a</mml:mi> <mml:mi>d</mml:mi> <mml:mi>d</mml:mi> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>=</mml:mo> <mml:msup> <mml:mi>e</mml:mi> <mml:mrow> <mml:mo>−</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo>/</mml:mo> <mml:mi>T</mml:mi> </mml:mrow> </mml:msup> <mml:mo>;</mml:mo> </mml:mrow> </mml:mtd> <mml:mtd> <mml:mrow> <mml:mi>p</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:mi>d</mml:mi> <mml:mi>r</mml:mi> <mml:mi>o</mml:mi> <mml:mi>p</mml:mi> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>=</mml:mo> <mml:msup> <mml:mi>e</mml:mi> <mml:mrow> <mml:mo>−</mml:mo> <mml:mo stretchy="false">(</mml:mo> <mml:mn>1</mml:mn> <mml:mo>−</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>/</mml:mo> <mml:mi>T</mml:mi> </mml:mrow> </mml:msup> </mml:mrow> </mml:mtd> </mml:mtr> </mml:mtable> <mml:mo>.</mml:mo> <mml:mtext>     </mml:mtext> <mml:mrow> <mml:mo>(</mml:mo> <mml:mn>6</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> <mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=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@6663@</mml:annotation> </mml:semantics> </mml:math></inline-formula></p><p>All moves are performed by sampling them from the probability in Eq. 6. This Simulated Annealing procedure is dampened by restricting the total number of gene/condition moves at each iteration to <italic>n</italic><sub>max </sub>= 5, in order to reduce the chance that a bicluster will change drastically before its model is reevaluated. We find that Simulated Annealing, while not the most efficient search strategy available, improves upon greedy search strategies such as Expectation Maximization, by being able to escape local minima and therefore being able to more completely assign genes and conditions to clusters as appropriate [<xref ref-type="bibr" rid="B24">24</xref>]. Other stochastic or greedy search strategies may be applicable to solving this model, for example if speed is deemed to be a more important consideration than completeness of the solution.</p></sec><sec><title>Additional model constraints: bicluster size and overlap</title><p>The search space for this problem is often dominated by very strong attractors and if we do not restrict the gene/condition move set, biclusters are likely to repeatedly descend into the same set of deep local minima (thereby increasing the bicluster overlap, or redundancy). This is an issue seen in many biclustering algorithms, and a commonly-practiced <italic>ad hoc </italic>remedy is to post-filter the bicluster set to remove redundant ones. We choose a more straightforward, easily-parameterized solution: to constrain the total number of biclusters <italic>z</italic><sub><italic>i </italic></sub>into which each gene <italic>i </italic>may fall (and in effect to reduce the amount of "gene overlap" of the final bicluster set), <italic>z</italic><sub><italic>i </italic></sub>is modeled as a Poisson process with cumulative distribution <italic>F</italic><sub><italic>v</italic></sub>(z<sub><italic>i</italic></sub>) (where <italic>v </italic>is the expected number of biclusters per gene). Then the probability of adding or dropping <italic>i </italic>to/from bicluster <italic>k</italic>, <italic>p</italic>(add\|<italic>π</italic><sub><italic>ik</italic></sub>) and <italic>p</italic>(drop\|<italic>π</italic><sub><italic>ik</italic></sub>) (Eq. 6), is multiplied with this prior probability of observing the gene in that many biclusters (relative to the expected number):</p><p><inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M18" name="1471-2105-7-280-i15" overflow="scroll"> <mml:semantics definitionURL="" encoding=""> <mml:mrow> <mml:mtable> <mml:mtr> <mml:mtd> <mml:mrow> <mml:msup> <mml:mi>p</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mo stretchy="false">(</mml:mo> <mml:mi>a</mml:mi> <mml:mi>d</mml:mi> <mml:mi>d</mml:mi> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> </mml:mtd> <mml:mtd> <mml:mo>=</mml:mo> </mml:mtd> <mml:mtd> <mml:mrow> <mml:msub> <mml:mi>F</mml:mi> <mml:mi>v</mml:mi> </mml:msub> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>z</mml:mi> <mml:mi>i</mml:mi> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>/</mml:mo> <mml:msub> <mml:mi>F</mml:mi> <mml:mi>v</mml:mi> </mml:msub> <mml:mo stretchy="false">(</mml:mo> <mml:mi>v</mml:mi> <mml:mo stretchy="false">)</mml:mo> <mml:msup> <mml:mi>e</mml:mi> <mml:mrow> <mml:mo>−</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo>/</mml:mo> <mml:mi>T</mml:mi> </mml:mrow> </mml:msup> <mml:mo>;</mml:mo> </mml:mrow> </mml:mtd> </mml:mtr> <mml:mtr> <mml:mtd> <mml:mrow> <mml:msup> <mml:mi>p</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mo stretchy="false">(</mml:mo> <mml:mi>d</mml:mi> <mml:mi>r</mml:mi> <mml:mi>o</mml:mi> <mml:mi>p</mml:mi> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> </mml:mtd> <mml:mtd> <mml:mo>=</mml:mo> </mml:mtd> <mml:mtd> <mml:mrow> <mml:mo stretchy="false">[</mml:mo> <mml:mn>1</mml:mn> <mml:mo>−</mml:mo> <mml:msub> <mml:mi>F</mml:mi> <mml:mi>v</mml:mi> </mml:msub> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>z</mml:mi> <mml:mi>i</mml:mi> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo stretchy="false">]</mml:mo> <mml:mo>/</mml:mo> <mml:mo stretchy="false">[</mml:mo> <mml:mn>1</mml:mn> <mml:mo>−</mml:mo> <mml:msub> <mml:mi>F</mml:mi> <mml:mi>v</mml:mi> </mml:msub> <mml:mo stretchy="false">(</mml:mo> <mml:mi>v</mml:mi> <mml:mo stretchy="false">)</mml:mo> <mml:mo stretchy="false">]</mml:mo> <mml:msup> <mml:mi>e</mml:mi> <mml:mrow> <mml:mo>−</mml:mo> <mml:mo stretchy="false">(</mml:mo> <mml:mn>1</mml:mn> <mml:mo>−</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> <mml:mo>/</mml:mo> <mml:mi>T</mml:mi> </mml:mrow> </mml:msup> <mml:mo>.</mml:mo> </mml:mrow> </mml:mtd> </mml:mtr> </mml:mtable> <mml:mtext>     </mml:mtext> <mml:mtable> <mml:mtr> <mml:mtd> <mml:mrow> <mml:mrow> <mml:mo>(</mml:mo> <mml:mn>7</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> </mml:mtd> </mml:mtr> </mml:mtable> </mml:mrow> <mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaafaqadeGadaaabaGafmiCaaNbauaacqGGOaakieaacqWFHbqycqWFKbazcqWFKbazcqGG8baFiiGacqGFapaCdaWgaaWcbaGaemyAaKMaem4AaSgabeaakiabcMcaPaqaaiabg2da9aqaaiabdAeagnaaBaaaleaacqWG2bGDaeqaaOGaeiikaGIaemOEaO3aaSbaaSqaaiabdMgaPbqabaGccqGGPaqkcqGGVaWlcqWGgbGrdaWgaaWcbaGaemODayhabeaakiabcIcaOiabdAha2jabcMcaPiabdwgaLnaaCaaaleqabaGaeyOeI0Iae4hWda3aaSbaaWqaaiabdMgaPjabdUgaRbqabaWccqGGVaWlcqWGubavaaGccqGG7aWoaeaacuWGWbaCgaqbaiabcIcaOiab=rgaKjab=jhaYjab=9gaVjab=bhaWjabcYha8jab+b8aWnaaBaaaleaacqWGPbqAcqWGRbWAaeqaaOGaeiykaKcabaGaeyypa0dabaGaei4waSLaeGymaeJaeyOeI0IaemOray0aaSbaaSqaaiabdAha2bqabaGccqGGOaakcqWG6bGEdaWgaaWcbaGaemyAaKgabeaakiabcMcaPiabc2faDjabc+caViabcUfaBjabigdaXiabgkHiTiabdAeagnaaBaaaleaacqWG2bGDaeqaaOGaeiikaGIaemODayNaeiykaKIaeiyxa0Laemyzau2aaWbaaSqabeaacqGHsislcqGGOaakcqaIXaqmcqGHsislcqGFapaCdaWgaaadbaGaemyAaKMaem4AaSgabeaaliabcMcaPiabc+caViabdsfaubaakiabc6caUaaacaWLjaGaaCzcauaabmqabeaaaeaadaqadaqaaiabiEda3aGaayjkaiaawMcaaaaaaaa@8BE6@</mml:annotation> </mml:semantics> </mml:math></inline-formula></p><p>Thus the solution is constrained to what seems to be a more biologically intuitive model: include each gene in an average of <italic>v </italic>= 2 (the default) clusters. This constraint results in an increased tendency to drop a gene from a bicluster if it is already in more than two biclusters, and a decreased tendency to drop the gene if it is in less than two biclusters.</p><p>Bicluster sizes can also vary widely between biclustering methods; some generate biclusters with only three genes on average [<xref ref-type="bibr" rid="B6">6</xref>], to single biclusters with nearly 3/4 of all genes in the data [<xref ref-type="bibr" rid="B18">18</xref>]. We constrain bicluster <italic>k</italic>'s final size (number of genes, \|<bold>I</bold><sub><italic>k</italic></sub>\|), using a cumulative Normal distribution <italic>N</italic>(\|<bold>I</bold><sub><italic>k</italic></sub>\|, <italic>μ</italic>, <italic>σ</italic>) as a prior constraint on \|<bold>I</bold><sub><italic>k</italic></sub>\|. This conditional distribution is applied by further adjusting the relative ratios of the distributions (Eq. 6) from which the gene moves are sampled:</p><p><inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M19" name="1471-2105-7-280-i16" overflow="scroll"> <mml:semantics definitionURL="" encoding=""> <mml:mrow> <mml:mfrac> <mml:mrow> <mml:mstyle displaystyle="true"> <mml:msub> <mml:mo>∑</mml:mo> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>∈</mml:mo> <mml:msubsup> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> <mml:mo>'</mml:mo> </mml:msubsup> </mml:mrow> </mml:msub> <mml:mrow> <mml:msup> <mml:mi>p</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mo stretchy="false">(</mml:mo> <mml:mi>d</mml:mi> <mml:mi>r</mml:mi> <mml:mi>o</mml:mi> <mml:mi>p</mml:mi> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> </mml:mstyle> </mml:mrow> <mml:mrow> <mml:mstyle displaystyle="true"> <mml:msub> <mml:mo>∑</mml:mo> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>∈</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:msub> <mml:mrow> <mml:msup> <mml:mi>p</mml:mi> <mml:mo>′</mml:mo> </mml:msup> <mml:mo stretchy="false">(</mml:mo> <mml:mi>a</mml:mi> <mml:mi>d</mml:mi> <mml:mi>d</mml:mi> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>π</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> </mml:mstyle> </mml:mrow> </mml:mfrac> <mml:mo>≡</mml:mo> <mml:mfrac> <mml:mrow> <mml:mi>N</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>\|</mml:mo> <mml:mo>,</mml:mo> <mml:mi>μ</mml:mi> <mml:mo>,</mml:mo> <mml:mi>σ</mml:mi> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> <mml:mrow> <mml:mo stretchy="false">[</mml:mo> <mml:mn>1</mml:mn> <mml:mo>−</mml:mo> <mml:mi>N</mml:mi> <mml:mo stretchy="false">(</mml:mo> <mml:mo>\|</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> <mml:mo>\|</mml:mo> <mml:mo>,</mml:mo> <mml:mi>μ</mml:mi> <mml:mo>,</mml:mo> <mml:mi>σ</mml:mi> <mml:mo stretchy="false">)</mml:mo> <mml:mo stretchy="false">]</mml:mo> </mml:mrow> </mml:mfrac> <mml:mo>.</mml:mo> <mml:mtext>     </mml:mtext> <mml:mrow> <mml:mo>(</mml:mo> <mml:mn>8</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> <mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaadaWcaaqaamaaqababaGafmiCaaNbauaacqGGOaakieaacqWFKbazcqWFYbGCcqWFVbWBcqWFWbaCcqGG8baFiiGacqGFapaCdaWgaaWcbaGaemyAaKMaem4AaSgabeaakiabcMcaPaWcbaGaemyAaKMaeyicI4mcbeGae0xsaK0aa0baaWqaaiabdUgaRbqaaiabcEcaNaaaaSqab0GaeyyeIuoaaOqaamaaqababaGafmiCaaNbauaacqGGOaakcqWFHbqycqWFKbazcqWFKbazcqGG8baFcqGFapaCdaWgaaWcbaGaemyAaKMaem4AaSgabeaakiabcMcaPaWcbaGaemyAaKMaeyicI4Sae8xsaK0aaSbaaWqaaiabdUgaRbqabaaaleqaniabggHiLdaaaOGaeyyyIO7aaSaaaeaacqWGobGtcqGGOaakcqGG8baFcqqFjbqsdaWgaaWcbaGaem4AaSgabeaakiabcYha8jabcYcaSiab+X7aTjabcYcaSiab+n8aZjabcMcaPaqaaiabcUfaBjabigdaXiabgkHiTiabd6eaojabcIcaOiabcYha8jab9LeajnaaBaaaleaacqWGRbWAaeqaaOGaeiiFaWNaeiilaWIae4hVd0MaeiilaWIae43WdmNaeiykaKIaeiyxa0faaiabc6caUiaaxMaacaWLjaWaaeWaaeaacqaI4aaoaiaawIcacaGLPaaaaaa@7EE0@</mml:annotation> </mml:semantics> </mml:math></inline-formula></p><p>The result is that if \|<bold>I</bold><sub><italic>k</italic></sub>\| <<italic>μ</italic>, the number of genes sampled to add to the bicluster will tend to be greater than the number sampled to drop, and vice versa if \|<bold>I</bold><sub><italic>k</italic></sub>\| > <italic>μ</italic>. We parameterize our prior expectation of bicluster sizes using <italic>μ </italic>= <italic>σ </italic>= 30, to match previous estimates of regulon sizes for well-studied organisms (<italic>e.g</italic>. Alkema, Lenhard, and Wasserman, 2004). This amounts to a soft constraint, which still allows for considerable variation in final bicluster sizes. A similar constraint may be applied to the biclusters' experiment sizes, which enables the generation of biclusters with a larger number of experiments (on average) than are typically included in biclusters derived by other methods (<italic>e.g</italic>. [<xref ref-type="bibr" rid="B19">19</xref>,<xref ref-type="bibr" rid="B86">86</xref>]).</p></sec><sec><title>The annealing schedule</title><p>To enforce convergence we schedule the annealing temperature <italic>T </italic>to slowly decrease during the procedure, as in standard simulated annealing procedures. We find that varying <italic>T </italic>linearly from 0.15 to 0.05 over 100–150 optimization steps is generally effective for all organisms for which biclustering was performed. As was described previously, there are reasons to vary (in addition to <italic>T</italic>) the three model mixture constraint parameters, <italic>r</italic><sub>0</sub>, <italic>s</italic><sub>0</sub>, and <italic>q</italic><sub>0 </sub>with each iteration. We have found that the most effective schedule up-weights the expression (<italic>r</italic><sub>0</sub>) and certain association networks (<italic>q</italic><sub>0</sub>; <italic>e.g</italic>. operons and metabolic networks) early in a run to build up co-expressed biclusters, and then slowly increases the influence of the sequence motifs (<italic>s</italic><sub>0</sub>) as the biclusters become better-defined (Fig. <xref ref-type="fig" rid="F9">9</xref>). For similar reasons, additional parameters, such as the number of motifs searched for, can also vary (<italic>i.e</italic>. increase or decrease monotonically) with iteration. Details on the default cMonkey parameters used for this work are listed in [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Table Three].</p></sec></sec><sec><title>Implementation</title><p>cMonkey is implemented in the <italic>R </italic>statistical programming language [<xref ref-type="bibr" rid="B5">5</xref>], a highly-flexible cross-platform language widely used in the statistical community. It has been parallelized, using PVM [<xref ref-type="bibr" rid="B84">84</xref>] as implemented in the <underline>S</underline>implified <underline>N</underline>etwork <underline>O</underline>f <underline>W</underline>orkstations (<italic>snow</italic>) <italic>R </italic>library, and runs efficiently on a multi-node Linux cluster; it can be run on a single-processor desktop computer as well. On a typical single-2 GHz processor, the algorithm can generate ~100 biclusters in between 12 and 48 hours, depending on the organism, data size, and motif detection parameters chosen. All parameters relevant to the biclustering procedure that have not been described in the main text are listed in [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Table Three].</p></sec><sec><title>Comparison with other biclustering and clustering methods</title><p>The different bi/clustering algorithms used for the comparative analysis included: Cheng-Church [<xref ref-type="bibr" rid="B25">25</xref>], Order Preserving Submatrix (OPSM [<xref ref-type="bibr" rid="B18">18</xref>]), Iterative Signature (ISA [<xref ref-type="bibr" rid="B19">19</xref>]), xMOTIF [<xref ref-type="bibr" rid="B55">55</xref>], and BIMAX [<xref ref-type="bibr" rid="B6">6</xref>] (all of these algorithms were run using the BICAT implementation [<xref ref-type="bibr" rid="B17">17</xref>]), SAMBA [<xref ref-type="bibr" rid="B86">86</xref>] (as implemented by the authors as part of EXPANDER [<xref ref-type="bibr" rid="B77">77</xref>]), and both hierarchical clustering and <italic>k</italic>-means clustering [<xref ref-type="bibr" rid="B30">30</xref>] with cluster number (<italic>k</italic>) ranging from 10 to 300 (implemented in <italic>R</italic>). None of these methods utilize data integration, and all were run on the same data sets as cMonkey. All biclustering procedures were run using their default parameters and data normalization/discretization schemes (while the effects of varying the parameters for each of these methods would be a worthwhile study, it is beyond the scope of this work). The analysis was performed on the Gasch [<xref ref-type="bibr" rid="B36">36</xref>] subset of the <italic>S. cerevisiae </italic>data containing measurements of 2993 genes over 173 stress-related conditions. <italic>S. cerevisiae </italic>was chosen for these comparisons because of the high-quality data and varied external "references" available for this organism, against which clusters could be compared. In all cases where <italic>p</italic>-values for judging annotation over-representation are listed, they were computed following a procedure similar to [<xref ref-type="bibr" rid="B21">21</xref>]; namely, cumulative hypergeometric <italic>p</italic>-values were corrected for multiple hypothesis testing in an <italic>experiment-wise </italic>manner for each cluster, by computing the fraction of uncorrected <italic>p</italic>-values derived for 1000 randomized instances of the cluster (the null model) that were less than or equal to the best <italic>p</italic>-value obtained for that cluster. To assess the effect of various biases caused by inclusion of different parts of the cMonkey model, we performed these same analyses on cMonkey runs with various model parameters up- and down-weighted, as described in the Results section. In all cases where we compared the motif-detection results of specific biclusters (in the Results section), we used MEME and MAST [<xref ref-type="bibr" rid="B10">10</xref>] (with the same parameters as for cMonkey) to search for motifs <italic>de novo </italic>in the upstream sequences of the clusters' genes.</p><p>Each different biclustering algorithm returned bicluster sets with wide differences in cluster count, cluster size (genes and experiments), amount of overlap/redundancy, expression coherence, and other general characteristics only related to their treatment of the expression data. We therefore computed "bulk" measurements for each bicluster set, such as those listed in [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>, Table Two]. One of these, <italic>f</italic>, is defined as the total fraction of elements in the expression data matrix <bold>X </bold>which fall in at least one bicluster. A measurement that quantifies the degree to which each complete bicluster set recapitulates the variance in <bold>X </bold>is defined as follows:</p><p><inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M20" name="1471-2105-7-280-i17" overflow="scroll"> <mml:semantics definitionURL="" encoding=""> <mml:mrow> <mml:mi>R</mml:mi> <mml:mi>M</mml:mi> <mml:mi>S</mml:mi> <mml:mi>D</mml:mi> <mml:mo>≡</mml:mo> <mml:msqrt> <mml:mrow> <mml:mstyle displaystyle="true"> <mml:munder> <mml:mo>∑</mml:mo> <mml:mi>k</mml:mi> </mml:munder> <mml:mrow> <mml:mstyle displaystyle="true"> <mml:munder> <mml:mo>∑</mml:mo> <mml:mrow> <mml:mi>j</mml:mi> <mml:mo>∈</mml:mo> <mml:msub> <mml:mi>J</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:munder> <mml:mrow> <mml:mstyle displaystyle="true"> <mml:munder> <mml:mo>∑</mml:mo> <mml:mrow> <mml:mi>i</mml:mi> <mml:mo>∈</mml:mo> <mml:msub> <mml:mi>I</mml:mi> <mml:mi>k</mml:mi> </mml:msub> </mml:mrow> </mml:munder> <mml:mrow> <mml:mfrac> <mml:mrow> <mml:msup> <mml:mrow> <mml:mo stretchy="false">(</mml:mo> <mml:msub> <mml:mi>x</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>j</mml:mi> </mml:mrow> </mml:msub> <mml:mo>−</mml:mo> <mml:msub> <mml:mover accent="true"> <mml:mi>x</mml:mi> <mml:mo>¯</mml:mo> </mml:mover> <mml:mrow> <mml:mi>j</mml:mi> <mml:mi>k</mml:mi> </mml:mrow> </mml:msub> <mml:mo stretchy="false">)</mml:mo> </mml:mrow> <mml:mn>2</mml:mn> </mml:msup> </mml:mrow> <mml:mrow> <mml:msub> <mml:mi>n</mml:mi> <mml:mrow> <mml:mi>i</mml:mi> <mml:mi>j</mml:mi> </mml:mrow> </mml:msub> <mml:msup> <mml:mi>σ</mml:mi> <mml:mn>2</mml:mn> </mml:msup> </mml:mrow> </mml:mfrac> </mml:mrow> </mml:mstyle> </mml:mrow> </mml:mstyle> </mml:mrow> </mml:mstyle> </mml:mrow> </mml:msqrt> <mml:mtext>     </mml:mtext> <mml:mrow> <mml:mo>(</mml:mo> <mml:mn>9</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:mrow> <mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=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@5DCE@</mml:annotation> </mml:semantics> </mml:math></inline-formula></p><p>where, as above, <inline-formula><mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" id="M21" name="1471-2105-7-280-i18" overflow="scroll"><mml:semantics definitionURL="" encoding=""><mml:mrow><mml:msub><mml:mover accent="true"><mml:mi>x</mml:mi><mml:mo>¯</mml:mo></mml:mover><mml:mrow><mml:mi>j</mml:mi><mml:mi>k</mml:mi></mml:mrow></mml:msub><mml:mo>=</mml:mo><mml:mstyle displaystyle="true"><mml:msub><mml:mo>∑</mml:mo><mml:mrow><mml:mi>i</mml:mi><mml:mo>∈</mml:mo><mml:msub><mml:mi>I</mml:mi><mml:mi>k</mml:mi></mml:msub></mml:mrow></mml:msub><mml:mrow><mml:msub><mml:mi>x</mml:mi><mml:mrow><mml:mi>i</mml:mi><mml:mi>j</mml:mi></mml:mrow></mml:msub></mml:mrow></mml:mstyle><mml:mo>/</mml:mo><mml:mo>\|</mml:mo><mml:msub><mml:mi>I</mml:mi><mml:mi>k</mml:mi></mml:msub><mml:mo>\|</mml:mo></mml:mrow><mml:annotation encoding="MathType-MTEF"> MathType@MTEF@5@5@+=feaafiart1ev1aaatCvAUfKttLearuWrP9MDH5MBPbIqV92AaeXatLxBI9gBaebbnrfifHhDYfgasaacH8akY=wiFfYdH8Gipec8Eeeu0xXdbba9frFj0=OqFfea0dXdd9vqai=hGuQ8kuc9pgc9s8qqaq=dirpe0xb9q8qiLsFr0=vr0=vr0dc8meaabaqaciaacaGaaeqabaqabeGadaaakeaacuWG4baEgaqeamaaBaaaleaacqWGQbGAcqWGRbWAaeqaaOGaeyypa0ZaaabeaeaacqWG4baEdaWgaaWcbaGaemyAaKMaemOAaOgabeaaaeaacqWGPbqAcqGHiiIZieqacqWFjbqsdaWgaaadbaGaem4AaSgabeaaaSqab0GaeyyeIuoakiabc+caViabcYha8jab=LeajnaaBaaaleaacqWGRbWAaeqaaOGaeiiFaWhaaa@449D@</mml:annotation></mml:semantics></mml:math></inline-formula> is the mean expression profile of bicluster <italic>k</italic>, and <italic>n</italic><sub><italic>ij </italic></sub>= ∑<sub><italic>k</italic></sub><italic>i </italic>∈ <bold>I</bold><sub><italic>k </italic></sub>∧ <italic>j </italic>∈ <bold>J</bold><sub><italic>k </italic></sub>is the number of biclusters containing element <italic>x</italic><sub><italic>ij</italic></sub>. This measure is dependent upon the fractional coverage <italic>f </italic>of the expression data matrix by the bicluster set (better coverage will generally lead to better RMSD) as well as the average bicluster residual (better residual leads to better RMSD), but is nearly independent of bicluster redundancy. It therefore is a good measure of the tradeoff that each bi/clustering method chooses between data coverage and bicluster co-expression. Because the expression data set has been variance-normalized (see Methods), RMSD ranges between 0–1, where a smaller RMSD implies that the mean expression profiles of the biclusters more accurately "generate" the original data matrix <bold>X</bold>.</p><p>In an attempt to remove <italic>some </italic>overlap and size bias related to these quality measurements (see Discussion), we also performed tests on a "filtered" set of biclusters, in which we greedily identified the largest 100 clusters with a volume-overlap (genes × conditions) of < 0.5 [<xref ref-type="bibr" rid="B6">6</xref>], excluding any with > 200 genes. For methods such as cMonkey, this filtering step removes a large number of non-redundant (but smaller) clusters, while for other methods (<italic>e.g</italic>. OPSM), it removes a large fraction of derived clusters and for others (such as SAMBA) it has little effect. Finally, in a further attempt to disentangle the cluster size bias inherent in these comparisons, we performed the same analyses on a set of evenly divided cluster sets ("big" and "small" halves; results shown in [<xref ref-type="supplementary-material" rid="S1">Additional File 1</xref>]).</p></sec></sec><sec><title>Abbreviations</title><p>Markov chain Monte Carlo (MCMC), cumulative distribution function (CDF), Position specific scoring matrix (PSSM), Gene Ontology (GO)</p></sec><sec><title>Authors' contributions</title><p><bold>DJR </bold>Developed and implemented the cMonkey algorithm; ran the procedure and analyzed the results; wrote this manuscript.</p><p><bold>NSB </bold>Contributed to the inception of this project; provided important feedback on the results; assisted with the writing of this manuscript.</p><p><bold>RB </bold>Conceived and initiated this project; contributed to the development and implementation of cMonkey; wrote this manuscript.</p></sec><sec sec-type="supplementary-material"><title>Supplementary Material</title><supplementary-material content-type="local-data" id="S1"><caption><title>Additional File 1</title><p>Supplementary file containing additional figures and tables, with captions, referenced in the main manuscript.</p></caption><media xlink:href="1471-2105-7-280-S1.pdf" mimetype="application" mime-subtype="pdf"><caption><p>Click here for file</p></caption></media></supplementary-material></sec>
Repression of Smoothened by Patched-Dependent (Pro-)Vitamin D3 Secretion	<p>The developmentally important hedgehog (Hh) pathway is activated by binding of Hh to patched (Ptch1), releasing smoothened (Smo) and the downstream transcription factor glioma associated (Gli) from inhibition. The mechanism behind Ptch1-dependent Smo inhibition remains unresolved. We now show that by mixing Ptch1-transfected and Ptch1 small interfering RNA–transfected cells with Gli reporter cells, Ptch1 is capable of non–cell autonomous repression of Smo. The magnitude of this non–cell autonomous repression of Smo activity was comparable to the fusion of Ptch1-transfected cell lines and Gli reporter cell lines, suggesting that it is the predominant mode of action. CHOD-PAP analysis of medium conditioned by Ptch1-transfected cells showed an elevated 3β-hydroxysteroid content, which we hypothesized to mediate the Smo inhibition. Indeed, the inhibition of 3β-hydroxysteroid synthesis impaired Ptch1 action on Smo, whereas adding the 3β-hydroxysteroid (pro-)vitamin D3 to the medium effectively inhibited Gli activity. Vitamin D3 bound to Smo with high affinity in a cyclopamine-sensitive manner. Treating zebrafish embryos with vitamin D3 mimicked the <italic>smo</italic> <sup>–/–</sup> phenotype, confirming the inhibitory action in vivo. Hh activates its signalling cascade by inhibiting Ptch1-dependent secretion of the 3β-hydroxysteroid (pro-)vitamin D3. This action not only explains the seemingly contradictory cause of Smith-Lemli-Opitz syndrome (SLOS), but also establishes Hh as a unique morphogen, because binding of Hh on one cell is capable of activating Hh-dependent signalling cascades on other cells. </p>	<contrib contrib-type="author" corresp="yes"><name><surname>Bijlsma</surname><given-names>Maarten F</given-names></name><email>[email protected]</email><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Spek</surname><given-names>C. Arnold</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Zivkovic</surname><given-names>Danica</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>van de Water</surname><given-names>Sandra</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Rezaee</surname><given-names>Farhad</given-names></name><xref ref-type="aff" rid="aff3"> <sup>3</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Peppelenbosch</surname><given-names>Maikel P</given-names></name><xref ref-type="aff" rid="aff4"> <sup>4</sup> </xref></contrib>	PLoS Biology	<sec id="s1"><title>Introduction</title><p>Signal transduction of the morphogen hedgehog (Hh) is highly unusual, with many features unique to this signalling system, many of which are only partly understood [ <xref rid="pbio-0040232-b001" ref-type="bibr">1</xref>]. For instance, Hh signalling is mediated by two membrane proteins: patched (Ptch1) and smoothened (Smo). The former is a 12-pass transmembrane protein [ <xref rid="pbio-0040232-b002" ref-type="bibr">2</xref>, <xref rid="pbio-0040232-b003" ref-type="bibr">3</xref>] resembling the Niemann-Pick disease type C1 (NPC1) protein (which is involved in cholesterol trafficking and has a pump function); the latter is a seven-pass transmembrane protein that resembles a G protein–coupled receptor [ <xref rid="pbio-0040232-b004" ref-type="bibr">4</xref>]. In the absence of the inhibitory receptor Ptch1 (or in the case of a specific mutation rendering Ptch1 dysfunctional), Smo is constitutively active and leads to the activation and nuclear translocation of its downstream transcription factor glioma associated (Gli). Under normal physiological circumstances, the activation of the signalling pathway is caused by binding of Hh to Ptch1, resulting in the internalisation of Ptch1 and, consequently, the alleviation of the inhibitory effect of Ptch1 on Smo [ <xref rid="pbio-0040232-b005" ref-type="bibr">5</xref>, <xref rid="pbio-0040232-b006" ref-type="bibr">6</xref>]. The exact mechanism behind Ptch1 inhibition of Smo, however, remains unclear [ <xref rid="pbio-0040232-b007" ref-type="bibr">7</xref>]. </p><p>Similarities in the phenotypes of humans with inherited disorders of sterol biosynthesis ( <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>A, lathosterolosis and Smith-Lemli-Opitz syndrome [SLOS]) and the phenotypes seen with mutations in the Hh signalling pathway have led to the suggestion that cholesterol-synthesizing enzymes may somehow be involved in Ptch1-dependent Smo inhibition [ <xref rid="pbio-0040232-b001" ref-type="bibr">1</xref>]. Sterol content has been found to be crucial to a cell's responsiveness to Hh, independent of proper Hh sterolation [ <xref rid="pbio-0040232-b008" ref-type="bibr">8</xref>], suggesting that sterol—or more specifically cholesterol—levels are pivotal to a cell's ability to respond to Hh, rather than to the accumulation of a specific inhibitory metabolite. <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>B depicts the proposed possible modes of Ptch1 action on Smo. Because Ptch1 and Smo do not unambiguously show physical association and because Ptch1 can inhibit Smo substoichiometrically, direct binding of Ptch1 to Smo as shown in <xref ref-type="fig" rid="pbio-0040232-g001">Figures 1</xref>B, columns 1 and 2, seems an unlikely mechanism of inhibition [ <xref rid="pbio-0040232-b009" ref-type="bibr">9</xref>]. Ptch1 shares high homology (for instance, a sterol-sensing domain) with NPC1, a protein involved in cholesterol trafficking [ <xref rid="pbio-0040232-b010" ref-type="bibr">10</xref>], as well as with various (prokaryotic) pump proteins. Also, several small molecules with homology to cholesterol that act as Hh pathway antagonists have been identified [ <xref rid="pbio-0040232-b011" ref-type="bibr">11</xref>– <xref rid="pbio-0040232-b013" ref-type="bibr">13</xref>]. Cyclopamine, a well-known antagonist of Smo, has been widely used for assessing the effects of Hh pathway inhibition in several model systems (for example, in [ <xref rid="pbio-0040232-b014" ref-type="bibr">14</xref>]). Consequently, it has been anticipated that in the absence of Hh, Ptch1 translocates a small molecule resembling cholesterol across the membrane, acting as a Smo antagonist. This hypothesis is summarized in <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>B, column 3. </p><fig id="pbio-0040232-g001" position="float"><label>Figure 1</label><caption><title>Cholesterol Synthesis and Possible Modes of Action of Ptch1</title><p>(A) Shown is a schematic representation of cholesterol synthesis. Boxed are lathosterol and 7-DHC, cholesterol precursors known to accumulate in lathosterolosis and SLOS patients, respectively. Mutations in or genetic loss of sterol C5-desaturase <italic>(Sc5d)</italic> cause stacking of lathosterol, whereas dysfunction of 7-DHC reductase <italic>(Dchr7)</italic> or the addition of its synthetic inhibitor AY-9944 causes accumulation of 7-DHC. The conversion of 7-DHC to vitamin D3 (cholecalciferol) is mediated by UV light. Statins, like pravastatin, inhibit HMG-CoA reductase, the enzyme that forms mevalonate. </p><p>(B) Shown are three possible models for the inhibitory action of the Hh receptor Ptch1 on Smo. (1) A cell-autonomous mode of action, in which direct binding of Ptch1 inhibits Smo. (2) An intracellular inhibitory action, mediated by direct binding of Ptch1 to Smo. (3) The model in which Ptch1 pumps an inhibitory small molecule that is capable of Smo-repression intercellularly (as well as intracellularly; not shown).</p></caption><graphic xlink:href="pbio.0040232.g001"/></fig><p>A consequence of the proposed “pump” model for Ptch1 is that the action of Ptch1 should be non–cell autonomous. We set out to test this prediction, and we established that Hh activates its signalling cascade by inhibiting Ptch1-dependent secretion of the 3β-hydroxysteroid (pro-)vitamin D3. These data provide fundamental new insight into the molecular mechanisms by which Hh exerts its action in pathophysiology.</p></sec><sec id="s2"><title>Results</title><sec id="s2a"><title>A Model System for Measuring Ptch1-Dependent Smo Inhibition</title><p>An experimental system allowing the study of intercellular inhibitory actions of Ptch1 on Smo should fulfill four requirements: (1) cells must be capable of sustaining Ptch1 expression, (2) expressed Ptch1 must be functionally active, (3) the inhibition of Gli activity should be Smo-mediated, and (4) endogenous Hh should not be a contributing factor. To set up a model system for studying Ptch1-dependent inhibition, C3H/10T1/2 fibroblasts, which are cells extensively used for Hh research, were transfected with an 8× Gli–binding site luciferase construct [ <xref rid="pbio-0040232-b015" ref-type="bibr">15</xref>] together with Ptch1, Smo, Smo and Ptch1, or Gli1. </p><p>Overexpression of Ptch1 drove cells into apoptosis (unpublished data, consistent with [ <xref rid="pbio-0040232-b016" ref-type="bibr">16</xref>]), possibly a reflection of the function of Ptch1 as a so-called dependence receptor. To overcome this problem, we performed experiments in the presence of 20 μM caspase inhibitor zVADfmk. Under these conditions, transfection with a Ptch1 expression construct led to efficient overexpression of Ptch1 as detected by Western blot ( <xref ref-type="fig" rid="pbio-0040232-g002">Figure 2</xref>A). Transfection of Ptch1 effectively inhibited transactivation of the Gli reporter in the presence of overexpressed Smo ( <xref ref-type="fig" rid="pbio-0040232-g002">Figure 2</xref>B). The Ptch1-insensitive mutant SmoM2 [ <xref rid="pbio-0040232-b011" ref-type="bibr">11</xref>, <xref rid="pbio-0040232-b017" ref-type="bibr">17</xref>] was not inhibited by Ptch1 cotransfection and showed a high basal Gli activity. Ptch1 transfection in the absence of Smo overexpression yielded no inhibition below control values, which has been described previously [ <xref rid="pbio-0040232-b018" ref-type="bibr">18</xref>]. This finding suggests the presence of high basal Ptch1 levels and indicates the specificity of Ptch1 inhibition acting through Smo in our system. The inhibitory effect of the high basal Ptch1 levels could be overcome by the addition of 1 μg/ml recombinant N-terminal Sonic hedgehog (Shh) for 6 h, as can be seen from the high reporter activity upon stimulation ( <xref ref-type="fig" rid="pbio-0040232-g002">Figure 2</xref>B). Addition of 1 μg/ml 5E1 Shh-blocking antibody [ <xref rid="pbio-0040232-b019" ref-type="bibr">19</xref>] could counteract the stimulation by Shh. Because Smo and Gli overexpression were capable of increasing transactivation of the Gli reporter despite the presence of a caspase inhibitor, our experimental system is a valid readout for Smo-mediated Gli activity. </p><fig id="pbio-0040232-g002" position="float"><label>Figure 2</label><caption><title>Confirmation of Functionality of Constructs and Model System Used</title><p>(A) Transfection of Ptch1 expression construct in C3H/10T1/2 fibroblasts increased Ptch1 expression over basal expression, as seen on Western blot. Actin levels remained unaltered; cells were lysed 24 h post transfection.</p><p>(B) C3H/10T1/2 cells are sensitive to Hh pathway components as indicated by Gli reporter activity when pathway components are expressed: Smo increased Hh pathway activity as determined by Gli reporter luciferase assay. Cotransfection of Ptch1 suppressed Smo-induced Gli activation. Transfection of Ptch1 in the absence of Smo overexpression did not decrease Gli activity below control levels. Shh stimulation (1 μg/ml for 6 h, 16 h post transfection) and transfection of a Gli1 expression construct showed highest reporter activity, as expected. Addition of 1 μg/ml Shh-blocking antibody 5E1 reduced Shh-mediated activation of Gli reporter activity. The Ptch1-insensitive mutant SmoM2 showed a high basal activity that could not be diminished by cotransfecting Ptch1, as expected. Cells were transfected and lysed after 24 h. Values shown are relative light unit (RLU) values corrected for an internal CMV Renilla standard, expressed as percent increase relative to vector (pcDNA 3.1–) transfection or control stimulation. Depicted is the mean ± SEM. ( <italic>n</italic> = 4; , <italic>p</italic> < 0.05; , <italic>p</italic> < 0.01). </p><p>(C) Shh concentration in medium is below the detection limit (5 ng/ml) of Western blotting. Medium was spiked with decreasing concentrations of recombinant Shh, and blotted along with a 4× concentrated medium sample obtained from C3H/10T1/2 fibroblasts (incubated for 16 h at a volume-to-surface ratio identical to the mix-and-match and medium transfer experiments).</p></caption><graphic xlink:href="pbio.0040232.g002"/></fig><p>The requirement for Smo overexpression to respond to Ptch1 inhibition and the high responsiveness to exogenously added Shh suggest that endogenous Hh production is not a major factor in our setup, although others have shown mRNA for Indian hedgehog (Ihh) and Shh in C3H/10T1/2 cells [ <xref rid="pbio-0040232-b020" ref-type="bibr">20</xref>, <xref rid="pbio-0040232-b021" ref-type="bibr">21</xref>]. To exclude Hh protein excretion by our C3H/10T1/2 cells as a contributing factor in our model system, we performed Western blot analysis on Shh-spiked medium and medium from fibroblast cell culture. The cells did not excrete Hh protein ( <xref ref-type="fig" rid="pbio-0040232-g002">Figure 2</xref>C). As can be seen, the detection limit was about 5 ng/ml, which was not reached in the (4× concentrated) medium, indicating that the Hh concentration in the medium was less than 1.25 ng/ml, a concentration too low to evoke a response [ <xref rid="pbio-0040232-b022" ref-type="bibr">22</xref>]. Reprobing the blot for Ihh did not yield a signal for the medium either. </p></sec><sec id="s2b"><title>Ptch1 Can Inhibit Smo in a Non–Cell Autonomous Fashion</title><p>To assess the ability of Ptch1 to act upon Smo on another cell, we used a “mix-and-match” approach (as schematically shown in <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>A), in which two populations of cells are mixed. The donor population expresses Ptch1 at different levels (using an overexpression construct or small interfering RNA [siRNA]) that could act on reporter cells with a constitutively active Hh pathway (reporter cells: Smo receptor overexpressed and a Gli-luciferase reporter). We mixed reporter cells with vector, Ptch1, Ptch1 siRNA and scrambled control siRNA transfectants. Cell populations were detached by EDTA treatment, mixed, and replated. By using two fluorescent tracers for labelling the cell populations, the capability of this procedure to obtain a homogenous mixture could be assessed. <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>A shows that combining two cell populations yields evenly mixed cells with intimate cell-cell contacts (micrograph). When Gli reporter cells were mixed with Ptch1- overexpressing cells in these mix-and-match experiments, a significant reduction in Smo-mediated Gli activation could be seen compared with control. This inhibition could not be diminished by addition of 1 μg/ml 5E1 Hh-blocking antibody ( <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>A), again demonstrating that endogenous Hh is not a contributing factor in our experimental setup. This is important, specifically because the presence of Hh in the medium would increase reporter activity. As Ptch1 scavenges Hh from the medium, medium of Ptch1-overexpressing cells would contain lower Hh levels than would control or Ptch1 siRNA-transfected cells. The reduction in reporter activity as consequence of such scavenging would mask the action of a potential inhibitory molecule. The addition of 1 μg/ml recombinant Shh, which should inhibit Ptch1 present in these experiments, was able to abolish the inhibition conferred by Ptch1. Mixing reporter cells with Ptch1 siRNA-transfected cells increased Smo-dependent reporter activity ( <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>A, blue bars), demonstrating that the inhibition is Ptch1 dependent and again suggests that C3H/10T1/2 cells have a significant basal level of Ptch1 activity. Showing opposite effects of Ptch1 siRNA versus DNA excludes any overexpression artefacts from responsibility for the observed non–cell autonomous effect. To exclude cell-specific artefacts, the procedures were also performed with MDA-MB-231 breast tumour cells, which had even higher levels of endogenous PTCH1 and SMO than the C3H/10T1/2 fibroblasts (Western blot, unpublished data). Indeed, using these cells yielded similar but more pronounced effects, as did the C3H/10T1/2 cells ( <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>A, hatched bars). </p><fig id="pbio-0040232-g003" position="float"><label>Figure 3</label><caption><title>Ptch1 Inhibits Smo Intercellularly</title><p>(A) Detaching and mixing of two populations of fluorescently labelled cells resulted in a homogenous mixture with intimate cell–cell contacts. The procedure is summarized in the left panel. Original magnification is 400×. Intercellular capacity of Ptch1 to inhibit Smo was assessed by mixing reporter cells overexpressing Smo and a Gli reporter with Ptch1-overexpressing donor cells. The observed decrease in Gli reporter activity could not be inhibited by the addition of 1 μg/ml 5E1 Shh-blocking antibody. Exogenously added Shh (1 μg/ml) was capable of abolishing the Ptch1-mediated inhibition. Transfecting donor cells with Ptch1 siRNA reversed the inhibitory effect (blue bars). MDA-MB-231 breast carcinoma cells (hatched bars) showed a response similar to the C3H/10T1/2 fibroblasts'. Donor cells were transfected, washed (16 h post transfection), mixed with reporter cells, and lysed after 6 to 8 h. Shown are RLU values corrected for an internal CMV Renilla luciferase control, expressed as percent differences relative to control-transfected donor cells (vector DNA or scrambled control siRNA). ( <italic>n</italic> ≥ 4; , <italic>p</italic> < 0.05; **, <italic>p</italic> < 0.005). Depicted is the mean ± SEM. </p><p>(B) Fusion of reporter cells with Ptch1 overexpressing donor cells showed a decrease in Gli reporter activity comparable in magnitude to the inhibition observed in mixed cells. Right panel: treating fluorescently labelled cells with PEG1500 4 h after mixing resulted in multiple nuclei per cell and composite cell colours, indicating successful cell fusion. Magnification is 1,000×. Statistics and incubation times are as in (A).</p><p>(C) To assess the specificity of Ptch1 acting on Smo, reporter cells expressing the Ptch1-insensitive mutant SmoM2 (as indicated on the <italic>y</italic>-axis) were mixed with donor cells. The loss of inhibition indicates the specificity of Ptch1 acting through Smo in the mix-and-match setup. Statistics and incubation times are as in (A). </p><p>(D) Shown is a model for the proposed intercellular action of Ptch1 inhibition following the results obtained in the mix-and-match experiments.</p></caption><graphic xlink:href="pbio.0040232.g003"/></fig><p>To determine the maximal inhibition Ptch1 confers in a cell-autonomous fashion—that is, the extent to which Ptch1 can inhibit Smo on the same cell—we mixed and fused the two previously described cell populations. PEG1500 induced efficient cell fusion, which was evident from blended staining of PEG1500-treated cells and the multinucleated cells (procedure and micrographs shown in <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>B, left panel). Fusion of reporter cells with Ptch1-overexpressing cells again showed a reduction in Gli activation, similar to that observed in the mix-and-match experiments ( <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>A). The reduction in reporter activity was of the same magnitude as that obtained in the mixing experiments. Therefore, we conclude that Ptch1 inhibition of Gli reporter activity in our system is mediated mainly intercellularly. </p><p>The specificity of the observed inhibitory action of Ptch1 acting through Smo was assessed by using the Ptch1-insensitive SmoM2 mutant, as previously shown in <xref ref-type="fig" rid="pbio-0040232-g002">Figure 2</xref>B. As can be seen in <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>C, SmoM2-transfected reporter cells are no longer sensitive to mixing with Ptch1-overexpressing donor cells, even when cotransfected with wild-type Smo. These findings led us to suggest the model for Ptch1 action as depicted in <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>B, column 3, and <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>D, in which Ptch1 on one cell is capable of specifically inhibiting Smo on another cell, to act non–cell autonomously. </p></sec><sec id="s2c"><title>Non–Cell Autonomous Ptch1-Dependent Smo Inhibition Is Carried Out by a Medium-Borne Factor</title><p>The intercellular Ptch1-mediated inhibition of Smo as identified in the mix-and-match experiments may be conferred by either an inhibitory molecule secreted or translocated by Ptch1 or by diffusion of a cytosolic mediator through tight junctions. The latter possibility is less likely, because it would result in a merged fluorophore signal in the mixed cell population, which was not observed ( <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>A). Indeed, important support for the possibility that Ptch1 secretes a Smo inhibitory molecule into the medium was obtained from experiments in which medium conditioned on Ptch1 <italic>-</italic>transfected cells was transferred to reporter cells in which Gli activity was assayed (procedure schematically shown in <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>A, left panel). </p><fig id="pbio-0040232-g004" position="float"><label>Figure 4</label><caption><title>Ptch1 Secretes Smo-Inhibitory 3β-Hydroxysteroids</title><p>(A) The left panel shows a schematic representation of the medium transfer experiments, in which medium was incubated on (transfected) donor cells and subsequently transferred to reporter cells. Right panel: intercellular Smo inhibition by Ptch1 is carried by the medium. Medium conditioned by control-, Ptch1-, or Ptch1 siRNA-transfected (C3H/10T1/2, solid bars, and MDA-MB-231, hatched bars) cells was transferred to Gli reporter cells. The same inhibitory action for Ptch1 was found as in the mix-and-match experiments. Neither 5E1 blocking antibody nor recombinant Shh addition to the reporter cells could diminish the inhibitory potential of the conditioned media. Medium conditioned on wild-type <italic>(Ptch1 <sup>+/+</sup>) </italic> MEFs showed a pronounced inhibitory effect on reporter cells as compared with <italic>Ptch1</italic> knockout <italic>(Ptch1 <sup>–/–</sup>) </italic> MEF-conditioned medium. Gli activity in reporter cells transfected with SmoM2 was not inhibited by <italic>Ptch1 <sup>+/+</sup></italic> MEF-conditioned medium. In the absence of serum, Ptch1 transfectant conditioned medium did not contain the inhibitory activity. Incubation times are as in <xref ref-type="fig" rid="pbio-0040232-g003">Figure 3</xref>B. Values shown are RLU values corrected for an internal CMV Renilla standard and expressed as percent difference to control-transfected donor cells. Depicted is the mean ± SEM. ( <italic>n</italic> ≥ 4; , <italic>p</italic> < 0.05; *, <italic>p</italic> < 0.01; medium Ptch1-transfected compared with medium Ptch1 siRNA-transfected). </p><p>(B) Vector, Ptch1, and Ptch1 siRNA transfectant-conditioned media investigated with FPLC-coupled CHOD-PAP analysis shows loading of 3β-hydroxysteroids on lipoproteins by control cells. Medium conditioned by Ptch1-transfected cells shows a strong increase in 3β-hydroxysteroids, mainly in the LDL fraction. Ptch1 siRNA transfection abolishes 3β-hydroxysteroid loading on LDL. The inset shows the lipid standard FCS containing VLDL, LDL, and HDL. Medium was incubated on cells for 16 h. Shown are typical profiles.</p><p>(C) Shown is quantification of the LDL peaks, expressed as mM. LDL-C shows the Ptch1-induced increase, a reduction by HMG-CoA reductase inhibitor pravastatin treatment, and the abolition by Ptch1 siRNA transfection of 3β-hydroxysteroid loading on LDL ( <italic>n</italic> = 3; , <italic>p</italic> < 0.05; *, <italic>p</italic> < 0.01). </p><p>(D) Pravastatin-inhibited Ptch1 action on Smo-driven Gli reporter activity, whereas the addition of the cholesterol precursor mevalonate enhanced this inhibition. Cells were transfected, and, after 16 h, pravastatin or mevalonate was added for 6 to 8 h. Depicted is the mean ± SEM. ( <italic>n</italic> = 4; , <italic>p</italic> < 0.05; , <italic>p</italic> < 0.01). </p><p>(E) Hh-induced endocytosis is not inhibited by 1 mM pravastatin ( <italic>n</italic> = 24; , <italic>p</italic> < 0.005). Cells were stimulated with 1 μg/ml Shh for 1 h, and preincubated with 1 mM pravastatin or control for 6 h. </p></caption><graphic xlink:href="pbio.0040232.g004"/></fig><p>By using this experimental setup, we found that medium conditioned on Ptch1 <italic>-</italic>transfected cells exerted a strong inhibitory effect on Gli activation as compared with medium conditioned on control-transfected cells ( <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>A). This inhibition could not be blocked by the addition of 1 μg/ml Shh or 5E1 to the reporter cells. The inability of Shh to prevent Ptch1-conditioned, medium-dependent inhibition of Smo activity is expected, because the transferred medium does not contain donor cells expressing Ptch1 on which Shh should exert its action, and the presence of the medium-borne factor in the Ptch1-conditioned medium shortcuts the function of endogenous Ptch1 in the reporter cells. Using MDA-MB-231 cells in this setup, similar results were obtained ( <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>A, hatched bars). As for the mix-and-match experiments, medium conditioned on Ptch1 siRNA-transfected cells increased Smo-dependent reporter activity in both cell lines. Medium conditioned on mouse embryonic fibroblasts (MEFs) from <italic>Ptch1</italic> knockout mice [ <xref rid="pbio-0040232-b023" ref-type="bibr">23</xref>] showed a distinct deficiency in their capacity to inhibit Smo compared with medium conditioned on wild-type ( <italic>Ptch1 <sup>+/+</sup></italic>) MEFs ( <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>, pink bars). This inhibitory effect could again not be conferred to reporter cells transfected with the Ptch1-insensitive SmoM2. The opposite effects, as observed with Ptch1 DNA transfection versus either Ptch1 siRNA transfection (blue bars) or genetic Ptch1-deficient MEFs, exclude transfection artefacts on the donor cells as the potential cause of the observed effects. Overall, these findings suggest that Ptch1 transfers a molecule to the medium that inhibits Smo activity. Medium conditioned on Ptch1-overexpressing MDA-MD-231 cells could inhibit Gli activity in C3H/10T1/2 cells and vice versa (unpublished data), suggesting that the inhibitory factor is not species-specific, at least not among mammals. </p><p>Using serum-free medium, we could confer no inhibitory action of Ptch1-conditioned medium on reporter cells ( <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>A). Serum-free medium contains no lipoproteins, which implies an important role for lipoproteins in conveying the inhibitory signal from cell to cell. Alternatively, serum-free medium also lacks lipids, which could stress the necessity for lipids supplied in the medium for Hh responsiveness [ <xref rid="pbio-0040232-b008" ref-type="bibr">8</xref>]. However, serum-free conditions for 8 h (the incubation time in our experiments) do not deplete a cell's cholesterol metabolism, arguing in favour of a role for lipoproteins in communicating the inhibitory signal. </p></sec><sec id="s2d"><title>Ptch1-Dependent Secretion of 3β-Hydroxysteroids</title><p>The conditioned media were subjected to fast performance liquid chromatography (FPLC)-coupled cholesterol oxidase peroxidase-amidopyrine (CHOD-PAP) analysis, a technique that analyzes 3β-hydroxysteroid content. The Ptch1-conditioned medium was found to be enriched with 3β-hydroxysteroids relative to medium conditioned by control cells ( <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>B and <xref ref-type="fig" rid="pbio-0040232-g004">4</xref>C; quantification of FPLC profiles as mM total 3β-hydroxysteroid). Judging from the retention profile, these 3β-hydroxysteroids were present on low-density lipoprotein (LDL) particles and, to a lesser extent, on very-low-density lipoprotein (VLDL). High-density lipoprotein (HDL) 3β-hydroxysteroid levels remained constant in all conditions. Medium incubated on cells transfected with siRNA for Ptch1 was virtually devoid of LDL-associated 3β-hydroxysteroids. This correlation between modified Ptch1 levels and the accumulation of 3β-hydroxysteroid molecules in the medium suggests that Ptch1 translocates or pumps these 3β-hydroxysteroids into the medium and thereby transfers the Smo inhibitory activity. </p></sec><sec id="s2e"><title>3β-Hydroxysteroids Are Necessary for Ptch1-Dependent Smo Inhibition</title><p>In the presence of pravastatin, a 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitor that abrogates the biosynthesis of the 3β-hydroxysteroid precursor mevalonate ( <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>A), LDL-associated 3β-hydroxysteroid (named LDL-C on axis) levels were substantially reduced ( <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>C), and, correspondingly, reduced Gli inhibition was observed in Ptch1-overexpressing reporter cells as shown in <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>D. Conversely, the 3β-hydroxysteroid precursor mevalonate strongly increased LDL-associated 3β-hydroxysteroids and Ptch1-mediated inhibition of Gli transactivation. </p><p>If the observed effects of steroid synthesis modifiers on Gli activation are Smo-specific, Hh responses mediated by Ptch1 (but not those by Smo) should remain unaffected. The endocytosis of Ptch1 is such a Smo-independent response to Hh [ <xref rid="pbio-0040232-b003" ref-type="bibr">3</xref>, <xref rid="pbio-0040232-b004" ref-type="bibr">4</xref>]. By using [3H]-sucrose as a tracer [ <xref rid="pbio-0040232-b024" ref-type="bibr">24</xref>], we observed no effect of 1 mM pravastatin on Hh-mediated endocytosis ( <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>E), and thus the effects of steroid-synthesis inhibitors are Smo-specific. The above-mentioned reporter data, however, are not from medium transfer experiments, and the inhibition or stimulation of cholesterol synthesis and the concomitant effects on Gli activity do not unequivocally prove the contribution of sterol biosynthesis to the intercellular action of Ptch1. </p></sec><sec id="s2f"><title>Using <italic>Dhcr7 <sup>–/–</sup></italic> and <italic>Sc5d <sup>–/–</sup></italic> MEFs to Identify the Inhibitory Compound </title><p>The paradoxical phenotype of SLOS patients, in which diminished Hh signalling is accompanied by an accumulation of the sterol 7-dehydrocholesterol (7-DHC; see <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>A), led us to hypothesize that 7-DHC might be a Smo inhibitor. To test this hypothesis, we used MEFs [ <xref rid="pbio-0040232-b008" ref-type="bibr">8</xref>] from mice genetically deficient for 7-DHC reductase ( <italic>Dhcr7 <sup>–/–</sup></italic>), because these cells stack 7-DHC. To confirm the specificity of 7-DHC as Smo inhibitor, we also used sterol C5-desaturase ( <italic>Sc5d <sup>–/–</sup></italic>) MEFs that stack the precursor of 7-DHC (lathosterol) and contain little or no 7-DHC. Indeed, as shown in <xref ref-type="fig" rid="pbio-0040232-g005">Figure 5</xref>A, <italic>Dhcr7 <sup>–/–</sup></italic> MEFs had a significantly reduced Gli activity as compared with <italic>Sc5d <sup>–/–</sup></italic> MEFs. In addition, the Smo-inhibitory potential of <italic>Dhcr7 <sup>–/–</sup></italic> MEF-conditioned medium was much higher than that of <italic>Sc5d <sup>–/–</sup></italic> MEF-conditioned medium, as shown in the medium transfer experiment depicted in <xref ref-type="fig" rid="pbio-0040232-g005">Figure 5</xref>A. In addition to stacking a specific metabolite, both MEFs are equally incapable of sterol synthesis. Our data therefore argue against reduced sterol levels as being responsible for the observed Smo inhibition. Overall, these data strongly suggest that 7-DHC or a Dhcr7-independent metabolite of 7-DHC have an inhibitory action on Smo. </p><fig id="pbio-0040232-g005" position="float"><label>Figure 5</label><caption><title>Differentially Modulated Ptch1 Action in <italic>Dhcr7 <sup>–/–</sup></italic> and <italic>Sc5d <sup>–/–</sup></italic> MEFs </title><p>(A) Basal Gli reporter activity measured in MEFs either deficient for Sc5d or Dhcr7 shows that the stacking of 7-DHC (in the <italic>Dhcr7 <sup>–/–</sup></italic> MEFs) lowers Gli activity in the mutant cells compared with lathosterol stacking in the <italic>Sc5d <sup>–/–</sup></italic> MEFs. Transfer of the conditioned media to reporter cells (indicated as “MEFs as medium donors”) shows that the inhibitory potential of the <italic>Dhcr7 <sup>–/–</sup></italic> MEFs is medium-borne. Media were incubated on donor MEFs for 16 h and transferred to reporter cells for 6 to 8 h. ( <italic>n</italic> ≥ 4; , <italic>p</italic> < 0.05; ***, <italic>p</italic> < 0.005). </p><p>(B) Medium transfer of Ptch1, Ptch1 siRNA, or control-transfected <italic>Sc5d <sup>–/–</sup></italic> and <italic>Dhcr7 <sup>–/–</sup></italic> MEFs shows that cells stacking lathosterol but lacking 7-DHC ( <italic>Sc5d <sup>–/–</sup></italic> MEFs) are not able to translate different Ptch1 levels to an inhibitory action on reporter cells. The <italic>Dhcr7 <sup>–/–</sup></italic> cells were able to properly convey an inhibitory signal when transfected with Ptch1 or, inversely, show a diminished inhibitory potential upon <italic>Ptch1</italic> siRNA transfection. UVB treatment of Ptch1 transfectant medium (2 h) amplified the inhibitory action. Media incubations and statistics are as in (A). </p></caption><graphic xlink:href="pbio.0040232.g005"/></fig><p>To assess whether Ptch1 uses 7-DHC to inhibit Smo, we performed medium transfer experiments with Ptch1 (or Ptch1 siRNA)-transfected <italic>Dhcr7 <sup>–/–</sup></italic> and <italic>Sc5d <sup>–/–</sup></italic> MEFs as donor cells. If Ptch1 would indeed pump 7-DHC, Ptch1 overexpression or knockdown in the <italic>Sc5d <sup>–/–</sup></italic> MEFs should show no effect on Smo inhibition. As shown in <xref ref-type="fig" rid="pbio-0040232-g005">Figure 5</xref>B, the <italic>Sc5d <sup>–/–</sup></italic> MEFs were severely hampered in their ability to transfer Ptch1 action to the medium, because neither Ptch1 DNA nor siRNA transfectants differed from control transfectants in their ability to inhibit Smo on reporter cells. The <italic>Dhcr7 <sup>–/–</sup></italic> MEFs, however, were well capable of translating Ptch1 expression levels to differential inhibitory action on reporter cells. UVB treatment of <italic>Dhcr7 <sup>–/–</sup></italic> MEF-conditioned media, which catalyzes the reaction from 7-DHC to vitamin D3, increased the Ptch1 effect on reporter cells, raising the tantalising option that Ptch1 uses vitamin D3 to inhibit Smo. </p></sec><sec id="s2g"><title>Vitamin D3 Is Sufficient for Smo Inhibition</title><p>From the experiments described above, we hypothesized that the addition of synthetic 7-DHC or vitamin D3 would inhibit Gli activity in reporter cells as well. Indeed, as can be seen from <xref ref-type="fig" rid="pbio-0040232-g006">Figure 6</xref>A, 7-DHC was capable of inhibiting Smo, but was not nearly as potent as its derivative, vitamin D3. This fits the observation that UV treatment enhanced the inhibitory potential of Ptch1-conditioned medium ( <xref ref-type="fig" rid="pbio-0040232-g005">Figure 5</xref>B). The addition of the 7-DHC reductase inhibitor AY-9944 [ <xref rid="pbio-0040232-b025" ref-type="bibr">25</xref>] successfully enhanced the effect of vitamin D3 treatment but was also capable of inhibiting Smo by itself, possibly by causing accumulation of endogenously synthesized 7-DHC or by acting as a 7-DHC mimetic. The magnitude of inhibition conveyed by either the transfer of Ptch1 transfectant–conditioned medium or Ptch1 cotransfection were both smaller than that of vitamin D3. In addition, inhibition conferred by vitamin D3 was stronger than that of 10 μM cyclopamine. The finding that AY-9944 was not a necessity for inhibitory action excludes a role for sterol deprivation in this model ( <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>A), because the exogenously added 7-DHC or vitamin D3 can be readily metabolized by these (wild-type) fibroblasts to produce downstream sterols. </p><fig id="pbio-0040232-g006" position="float"><label>Figure 6</label><caption><title>Analysis of Vitamin D3 as a Specific Smo Antagonist</title><p>(A) Shown are the Gli reporter inhibitions by 7-DHC, AY-9944, a Dhcr7 inhibitor, vitamin D3, and a combination of the latter two. Also shown is the inhibition that can be conferred by the well-known Smo antagonist cyclopamine and the previously observed effects for Ptch1 transfectant conditioned medium as well as Ptch1 cotransfection (data taken from <xref ref-type="fig" rid="pbio-0040232-g004">Figure 4</xref>B and <xref ref-type="fig" rid="pbio-0040232-g002">Figure 2</xref>B, respectively). Cells were stimulated with the various compounds for 6 h (16 h past transfection), except for the cotransfection, which was lysed after 24 h ( <italic>n</italic> ≥ 4). </p><p>(B) A range of concentrations of vitamin D3 was tested for inhibition of Gli reporter activity (also in SmoM2-transfected cells), MTT viability assay, and N-terminal repressor form Gli3 levels. Cells were stimulated with vitamin D3 for 6 h before lysis. For the reporter assays, RLU values were corrected for an internal CMV Renilla standard and expressed as percent difference to control stimulated cells ( <italic>n</italic> = 4). The MTT assays shown were raw absorption data expressed as percent difference to control stimulated ( <italic>n</italic> = 8); MTT agent was added for 3 h. N-terminal repressor form Gli3 was determined by quantifying ECL signal from Western blot and corrected for β-actin ( <italic>n</italic> = 3). Depicted is the mean ± SEM. </p><p>(C) Gli specificity of the inhibitory effect of 10 μM vitamin D3 was assayed by using a panel of luciferase reporter constructs, one of which was a mGli reporter (mGli-LUC). Cells were stimulated for 6 h with vitamin D3, and the luciferase activity was assayed; values were calculated from RLU values corrected for an internal CMV Renilla standard and expressed as percent difference to control stimulated cells ( <italic>n</italic> ≥ 4). Depicted is the mean ± SEM. </p><p>(D) Different cell types share the inhibitory response to vitamin D3 on Gli activity. C3H/10T1/2 and MDA-MD-231 cells were used earlier in the mix-and-match and medium transfer experiments. <italic>Ptch1 <sup>–/–</sup></italic> MEFs served as a control for any possible effects of vitamin D3 on Ptch1 function. Cells were stimulated for 6 h, and the luciferase activity was assayed,; values are RLU values corrected for an internal CMV Renilla standard and expressed as percent difference to control stimulated cells ( <italic>n</italic> = 4). Depicted is the mean ± SEM. </p><p>(E) Proposed mechanism of vitamin D3 action. First panel: in the presence of Hh, Ptch1 is inactive and does not inhibit Smo. The Hh pathway is active, and Gli activity can be measured in a reporter assay. Second panel: in the absence of Hh, Ptch1 is active and uses vitamin D3 to inhibit Smo. The pathway is inactive, and low Gli activity is measured. Third panel: in the presence of Hh as well as exogenous D3, Smo is inhibited independently of Ptch1, and Hh can no longer elicit a Gli response.</p><p>(F) Confluent Shh-LIGHT II stable Gli reporter transfectants were stimulated with 10 μM vitamin D3 and/or 200 ng/ml Shh overnight in the presence of 0.5% FCS. Luciferase activity was assayed; values are RLU values corrected for an internal CMV Renilla standard and expressed as percent difference to control stimulated cells (0 μM vitamin D3; 0 ng/ml Shh). In the presence of vitamin D3, Shh is no longer able to induce reporter activity. <italic>n</italic> = 4; depicted is the mean ± SEM. </p></caption><graphic xlink:href="pbio.0040232.g006"/></fig><p>Shown in <xref ref-type="fig" rid="pbio-0040232-g006">Figure 6</xref>B is a dose-dependent response of reporter cells to vitamin D3 for 6 h. In agreement, the level of inhibitory N-terminal Gli3 protein increased accordingly, as quantified from Western blot. This digestion product of Gli3 originates from proteolysis in the SuFu/Fu complex present in Hh pathway inactive cells and is considered the repressor form [ <xref rid="pbio-0040232-b026" ref-type="bibr">26</xref>]. To exclude cytotoxic artefacts of vitamin D3, we measured cell viability by MTT reduction. Only at very high (1 mM) concentrations of vitamin D3 could we could observe a slight decrease in cell viability. Using SmoM2-transfected reporter cells, Gli reporter inhibition occurred only at 100 μM vitamin D3, and below that concentration, no inhibition could be observed. This is very similar to the impaired but not absent response of SmoM2 to cyclopamine [ <xref rid="pbio-0040232-b011" ref-type="bibr">11</xref>]. </p><p>To exclude Gli-independent effects of vitamin D3 from being responsible for the observed reporter inhibition, we used a panel of various luciferase reporter constructs. As can be seen in <xref ref-type="fig" rid="pbio-0040232-g006">Figure 6</xref>C, interferon-stimulated response element (pISRE), TATA-like promoter (pTAL), nuclear factor–kappa B (pNF-κB), and mutant Gli binding site (mGli) luciferase reporter constructs were not significantly inhibited by the addition of 10 μM vitamin D3 for 6 h, whereas reporter activity of a TATA-like promoter–luciferase construct was increased. Thus, the inhibitory effect of vitamin D3 on Gli reporter activity appears to be a genuine effect of Smo-dependent signalling. </p><p> <xref ref-type="fig" rid="pbio-0040232-g006">Figure 6</xref>D shows Gli reporter inhibition by vitamin D3 in the cell lines C3H/10T1/2 and MDA-MB-231 previously used in the mix-and-match and medium transfer experiments. Both showed a marked inhibition in reporter activity after treatment with 10 μM vitamin D3. <italic>Ptch1</italic> <sup>–/–</sup> MEFs also responded to vitamin D3. Because these cells have no Ptch1, any artefacts of vitamin D3 by interference with Ptch1 action rather than Smo activity can be excluded. </p><p>A consequence of our model in which vitamin D3 or a very similar molecule mediates Ptch1 action on Smo ( <xref ref-type="fig" rid="pbio-0040232-g006">Figure 6</xref>E) is that exogenously added vitamin D3 should overrule any effect of Hh on Ptch1. To confirm this, we stimulated stable Gli-luciferase transfectants (Shh-LIGHT II) overnight with 10 μM vitamin D3 and/or 200 ng/ml Shh. What is apparent from <xref ref-type="fig" rid="pbio-0040232-g006">Figure 6</xref>F is that in the presence of vitamin D3 reporter activity could not be increased by the addition of Shh, confirming our hypothesis. </p></sec><sec id="s2h"><title>Scatchard Analysis Reveals Vitamin D3 Binding to Smo</title><p>To assess possible binding of vitamin D3 to Smo, we used a yeast strain <italic>(Pichia pastoris)</italic> transformed with Smo. The expression of Smo could be induced by the addition of methanol to the growth medium (kind gift of Dr. I. Mus-Veteau [ <xref rid="pbio-0040232-b027" ref-type="bibr">27</xref>]). We chose this approach for two reasons: first, to our knowledge there is no vitamin D3 receptor in <italic>P. pastoris;</italic> second, baseline (or background) levels of Smo in noninduced <named-content content-type="genus-species">P. pastoris</named-content> are negligible. </p><p>First we confirmed the successful induction of Smo expression by Western blot using a Smo antibody as shown in <xref ref-type="fig" rid="pbio-0040232-g007">Figure 7</xref>A. We then set out to perform a Scatchard analysis [ <xref rid="pbio-0040232-b024" ref-type="bibr">24</xref>] with heterologous competition by cyclopamine. Cyclopamine is a ligand for Smo, and this enabled us to distinguish the observed competition from any background by nonspecific binding or specific binding to a vitamin D3 receptor (although it is not known to exist in these cells). As can be seen from <xref ref-type="fig" rid="pbio-0040232-g007">Figure 7</xref>B, [3H]-D3 was not capable of specific binding to non–methanol induced <named-content content-type="genus-species">P. pastoris</named-content> (BMGY), whereas after methanol induction (BMMY), specific binding of [3H]-D3 was observed that was subject to competition by cyclopamine. Cyclopamine replaced [3H]-D3 with an apparent dissociation constant <italic>K</italic> <sub>d</sub> of 2 nM in this heterologous assay, a higher affinity than the earlier reported 20-nM affinity of cyclopamine [ <xref rid="pbio-0040232-b028" ref-type="bibr">28</xref>]. Nevertheless, these results show that vitamin D3 has the potential to bind Smo at the same site as cyclopamine, indicating that vitamin D3 is indeed a physiological ligand for Smo. </p><fig id="pbio-0040232-g007" position="float"><label>Figure 7</label><caption><title>Scatchard Analysis of Vitamin D3 Using Smo-Expressing <named-content content-type="genus-species">P. pastoris</named-content> </title><p>(A) 24 h methanol induction (buffered methanol-complex medium; BMMY) of Smo-transformed <named-content content-type="genus-species">P. pastoris</named-content> yielded successful Smo expression as shown by Western blot analysis with an antibody directed against Smo. </p><p>(B) Scatchard analysis of heterologous competition with cyclopamine representing the binding of [3H]-D3 to Smo-expressing (solid circles, MeOH induced) and nonexpressing (open circles, noninduced) yeast strains. The solid line indicates the induced high-affinity binding, whereas the dotted line indicates nonspecific low-affinity binding. Scatchard analysis and fitting was performed as described earlier [ <xref rid="pbio-0040232-b053" ref-type="bibr">53</xref>]. </p></caption><graphic xlink:href="pbio.0040232.g007"/></fig></sec><sec id="s2i"><title>Vitamin D3 Treatment Mimics the <italic>smo <sup>–/–</sup></italic> Phenotype in Zebrafish </title><p>To test the action of vitamin D3 in a developmental in vivo model system, we continuously incubated zebrafish embryos starting between 32- and 64-cell stage in 1.2 mg/ml sonicated vitamin D3. We used sonication as described for 7-DHC [ <xref rid="pbio-0040232-b029" ref-type="bibr">29</xref>], but did not use a stabilizing reagent. The vitamin D3 treatment phenocopied to a great extent the <italic>slow muscle omitted (smu)</italic> mutant phenotype [ <xref rid="pbio-0040232-b030" ref-type="bibr">30</xref>], which has been identified as the zebrafish ortholog of <italic>smo</italic> [ <xref rid="pbio-0040232-b031" ref-type="bibr">31</xref>, <xref rid="pbio-0040232-b032" ref-type="bibr">32</xref>]. </p><p>One of the most striking features of the <italic>smo <sup>–/–</sup></italic> phenotype is the U-shaped somites in contrast to the sharply defined chevron (V)-shaped somites of wild-type embryos. The vitamin D3–treated embryos showed these disrupted somites (as indicated by arrow 1 in <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>A) at the developmental stages shown. Of the treated animals, approximately 75% ( <italic>n</italic> = 30 embryos in six individual experiments) showed this somite phenotype as well as the typical responses as listed below. Quantification of the somite angle was performed, measuring the angle between the dorsal and ventral portions of the individual somites while positioning the apex of the angle on the horizontal myoseptum ( <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>B, top panel). In properly shaped somites, this angle was relatively sharp, whereas in vitamin D3–treated embryos, somites were kidney-shaped, displaying a shallow angle ( <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>B). The ventrally curved body that is characteristic of <italic>smo</italic> <sup>–/–</sup> zebrafish embryos was also observed upon vitamin D3 treatment. The horizontal myoseptum was malformed ( <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>A; see also Figure 1 B of [ <xref rid="pbio-0040232-b032" ref-type="bibr">32</xref>]). Another aspect of the <italic>smo <sup>–/–</sup></italic> phenotype is the aberrant extension of the yolk tube, resulting in a shorter and thicker yolk tube; this was also clearly noticeable in the vitamin D3–treated embryos (indicated by arrow 2 in <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>A; compare to Figure 1 A of [ <xref rid="pbio-0040232-b032" ref-type="bibr">32</xref>] and Figure 4 of [ <xref rid="pbio-0040232-b031" ref-type="bibr">31</xref>]). Although motoneurons have been described to be missing in <italic>smo <sup>–/–</sup></italic> animals [ <xref rid="pbio-0040232-b032" ref-type="bibr">32</xref>], we observed no apparent difference in touch response between control and vitamin D3 animals. To our knowledge not previously reported, it appeared that the dorsal midbrain was more prominent in the treated embryos versus controls, resembling the prominent dorsal midbrain in <italic>smo <sup>–/–</sup></italic> embryos (indicated by arrow 3 in <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>A; compare to Figure 1 of [ <xref rid="pbio-0040232-b032" ref-type="bibr">32</xref>] and Figure 4 of [ <xref rid="pbio-0040232-b031" ref-type="bibr">31</xref>]). Also, note the similarity of the treated embryo shown at 18 h with regard to a reduced trunk, possibly caused by increased cell death as shown for <italic>smo</italic> mutants (indicated by arrow 4 in <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>A; compare to Figure 8 B of [ <xref rid="pbio-0040232-b033" ref-type="bibr">33</xref>]). </p><fig id="pbio-0040232-g008" position="float"><label>Figure 8</label><caption><title>Effects of Vitamin D3 on Developing Zebrafish Embryos</title><p>(A) Fertilized zebrafish eggs were dechorionated and placed in buffer containing 1.2 mg/ml sonicated vitamin D3, starting the continuous incubation between the 32- and 64-cell stages. At the different developmental stages (as indicated), the embryos were analysed, revealing persistent somite malformations, a ventrally curved body, poorly developed myoseptum, an aberrant extension of the yolk tube, a prominent dorsal midbrain, and a reduced trunk. The wild-type Engrailed staining (4D9) of muscle pioneers and surrounding cells in control-treated animals was not observed upon vitamin D3 treatment. ISH for <italic>ptc1</italic> mRNA showed a strongly reduced signal in vitamin D3–treated animals compared with controls. Staining with the F59 antibody revealed that slow muscle fibres were disturbed in number and orientation in the treated animals. </p><p>(B) At 18 hpf, embryos were photographed in detail and somite angles were determined (as shown in top panel). The measurement was taken of an individual somite between dorsal and ventral portions of the vertical myoseptum for corresponding somites between the wild-type and vitamin D3–treated embryos. Shown are the mean angles (±SEM; <italic>n</italic> = 3) of wild-type control- and vitamin D3–treated embryos. </p><p>C) Detail of 4-dpf embryo. Note the slightly enlarged pericardial cavity in vitamin D3–treated animals. The orientation of all images is a lateral view, anterior to the left.</p></caption><graphic xlink:href="pbio.0040232.g008"/></fig><p>To assess the specificity of vitamin D3 action on the Hh pathway, we used molecular markers specific for Hh pathway activity in zebrafish. The protein Engrailed is a marker for muscle pioneers in normal development, and as development advances, it is also expressed in the cells surrounding the muscle pioneers [ <xref rid="pbio-0040232-b034" ref-type="bibr">34</xref>]. In zebrafish embryos with an abrogated Hh signalling, muscle pioneers fail to differentiate and Engrailed is totally absent [ <xref rid="pbio-0040232-b031" ref-type="bibr">31</xref>, <xref rid="pbio-0040232-b035" ref-type="bibr">35</xref>– <xref rid="pbio-0040232-b037" ref-type="bibr">37</xref>]. No Engrailed staining of muscle pioneers (indicated by arrow in <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>, panel labelled 4D9) or surrounding cells was observed in vitamin D3–treated embryos, suggestive of inhibition of Hh signalling by vitamin D3. In the control-treated animals, staining of muscle pioneers as well as ventrally located cells in the somites was observed. Expression of <italic>ptc1</italic> as determined by in situ hybridisation (ISH) ( <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>A, <italic>ptc1</italic>) showed a marked reduction of transcript levels in the vitamin D3–treated animals. Because ptc1 is not only the primary receptor for the Hh pathway but also a transcriptional readout of Hh activity [ <xref rid="pbio-0040232-b032" ref-type="bibr">32</xref>, <xref rid="pbio-0040232-b037" ref-type="bibr">37</xref>], the attenuated expression seen in the treated animals confirms that vitamin D3 specifically inhibits the Hh pathway. Another marker previously described to be indicative of Hh pathway activity is slow muscle fibre orientation and number in the somites. In <italic>smo <sup>–/–</sup></italic> zebrafish, slow muscle fibres are fewer and malformed [ <xref rid="pbio-0040232-b030" ref-type="bibr">30</xref>, <xref rid="pbio-0040232-b031" ref-type="bibr">31</xref>]. Shown in <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref> (panel marked F59) is the disturbed pattern and number of slow muscle fibres after vitamin D3 treatment, reminiscent of <italic>smo <sup>–/–</sup></italic> zebrafish embryos. The use of the abovementioned molecular markers confirms that the effects of vitamin D3 on zebrafish embryo morphology are indeed due to reduced Hh pathway activity. </p><p>At later stages, we observed a subtle cardiac edema, and all treated embryos died after 5 d of vitamin D3 treatment, a feature again similar to the <italic>smo <sup>–/–</sup></italic> phenotype ( <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>C). Although we found the eyes to be relatively ventrally positioned throughout various developmental stages ( <xref ref-type="fig" rid="pbio-0040232-g008">Figure 8</xref>C), no absolute cyclopia was observed. However, cyclopia is not an unambiguous hallmark of the <italic>smo</italic>-deficient phenotype [ <xref rid="pbio-0040232-b031" ref-type="bibr">31</xref>– <xref rid="pbio-0040232-b033" ref-type="bibr">33</xref>]. </p><p>Because nearly all of the typical <italic>smo <sup>–/–</sup></italic> phenotypic features were also prominently present in vitamin D3–treated zebrafish embryos and the molecular markers used all pointed to a defective Hh signalling, we conclude that the in vivo action of vitamin D3 on embryonic development is due to a specific inhibitory effect on Smo. </p></sec></sec><sec id="s3"><title>Discussion</title><p>In the present study, we investigated the molecular mode by which Ptch1 represses Smo activity. By mixing Ptch1-expressing cells with cells expressing Smo and a Gli reporter, we were able to show that Ptch1 is capable of non–cell autonomous repression of Smo. This non–cell autonomous repression is carried by the extracellular medium, because medium conditioned by Ptch1-transfected cells was capable of inhibiting Gli reporter activity as well. Conversely, medium conditioned by cells transfected with Ptch1 siRNA (so lacking endogenous Ptch1 expression) elevated Gli reporter activity significantly above medium that was conditioned by control cells, demonstrating that Ptch1 overexpression is not a prerequisite for observing this cell non-autonomous repression of Smo by Ptch1. Mixing Ptch1-overexpression cells with cells expressing Smo and Gli reporter produced inhibition of Gli reporter activity to approximately the same extent as fusing these cells, showing that this non–cell autonomous repression of Smo by Ptch1 is of roughly similar potency compared to intracellular Smo repression by Ptch1. This result suggests that this intercellular repression constitutes an important physiological mechanism. In this context, it is important to note that Hh quenching was not a major contributing factor in our experimental system, because Hh-neutralising antibodies do not interfere with Ptch1-dependent intercellular repression of Smo and we were not able to detect Hh release by the cells.</p><p>Conditioning medium with Ptch1-transfected cells led to the appearance of a 3β-hydroxysteroid in this medium, suggesting that this compound is increasingly translocated into the medium as a consequence of elevated Ptch1 expression. These results are concordant with the observation that Smo antagonists bear structural similarities to steroids [ <xref rid="pbio-0040232-b012" ref-type="bibr">12</xref>] and the high homology of Ptch1 to the NPC1 protein, which has cholesterol trafficking and pump functions [ <xref rid="pbio-0040232-b010" ref-type="bibr">10</xref>], as well as to various prokaryotic pump proteins. These homologies gave rise to the suggestion that, in the absence of Hh, Ptch1 pumps a compound out of the cell that binds to and inhibits Smo, and the data from this study confirm this model. Transfection of Ptch1 siRNA abolished deposition of this 3β-hydroxysteroid into the medium when compared to control-transfected cells, showing that 3β-hydroxysteroid translocation is not an artefact of Ptch1 overexpression, but that endogenous Ptch1 expression levels are sufficient to sustain detectable 3β-hydroxysteroid translocation. This 3β-hydroxysteroid translocation seems important for the intercellular repression of Smo by Ptch1, because in the presence of 3β-hydroxysteroid synthesis inhibitors Ptch1 was no longer capable of repressing Smo, whereas artificial stimulation of 3β-hydroxysteroid biosynthesis by using mevalonate supplementation enhanced the inhibitory potential of Ptch1 on Smo. </p><p>The 3β-hydroxysteroid responsible for intercellular repression of Smo is most likely the (pro-)vitamin D3 (that is, either 7-DHC or its metabolite vitamin D3) or a highly similar molecule. Fibroblasts derived from animals genetically incapable of synthesizing these 3β-hydroxysteroids <italic>(Sc5d <sup>–/–</sup>) </italic> do not sustain intercellular repression of Smo by Ptch1, whereas fibroblasts that have a genetic defect leading to increased (pro-)vitamin D3 production <italic>(Dhcr7 <sup>–/–</sup>) </italic> show exaggerated intercellular repression of Smo by Ptch1. Exogenous addition of vitamin D3 was more efficient in inhibiting Gli reporter activity than the established Smo antagonist cyclopamine. SmoM2, which has a reduced sensitivity towards cyclopamine, showed diminished sensitivity to vitamin D3 as well. In Scatchard assays, cyclopamine and vitamin D3 competed for Smo binding, suggesting that both compounds have an identical binding site on Smo. Scatchard analysis using the precursor 7-DHC as a competitor revealed a 10-fold lower affinity for Smo than did vitamin D3 or the hydroxylated (active) form of vitamin D3 (unpublished data). Treatment of <named-content content-type="genus-species">Danio rerio</named-content> embryos with vitamin D3 phenocopied most of the aspects of the <italic>smo <sup>–/–</sup></italic> phenotype in zebrafish. These findings strongly suggest that Ptch1 translocates either vitamin D3, a related molecule, or possibly a precursor that is subsequently metabolised in the extracellular medium that acts as a Smo antagonist. </p><p>Our findings shed light on the puzzling biochemical basis of SLOS, in which accumulation of 7-DHC ( <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>A) accompanies a reduced cholesterol content and a seemingly decreased Hh signalling [ <xref rid="pbio-0040232-b038" ref-type="bibr">38</xref>]. Previously, the SLOS phenotype has been mimicked by using AY-9944 [ <xref rid="pbio-0040232-b025" ref-type="bibr">25</xref>], which inhibits 7-DHC reductase ( <xref ref-type="fig" rid="pbio-0040232-g001">Figure 1</xref>A), leading to the accumulation of its substrate. This result suggests that this molecule mediates Hh inhibition. Treatment of patients with a cholesterol-rich diet, however, has been successful, and statin treatment does not seem to alleviate the symptoms, suggesting that the lack of cholesterol is the culprit [ <xref rid="pbio-0040232-b038" ref-type="bibr">38</xref>, <xref rid="pbio-0040232-b039" ref-type="bibr">39</xref>]. Others have been able to counteract the effect of 7-DHC stacking in experimental settings by exogenously adding cholesterol [ <xref rid="pbio-0040232-b040" ref-type="bibr">40</xref>], confirming that AY-9944 merely inhibits synthesis of cholesterol needed for proper Hh processing. Although these findings seem to preclude the stacking of a specific metabolite to be responsible for the SLOS phenotype, it has become clear that the clinical phenotype correlates best with the ratio of 7-DHC to cholesterol levels in patient plasma rather than the levels of cholesterol alone. Also, the treatment of infants with cholesterol is arguably too late an intervention, because most of the Hh patterning events are completed in utero. Our experiments show that even in the absence of AY-9944, addition of vitamin D3 strongly inhibits Hh signalling. This inhibition is stronger than Ptch1 cotransfection and Ptch1-transfectant conditioned medium, and we therefore conclude that the stacking of 7-DHC is responsible for the SLOS phenotype. </p><p>Very relevant to this manuscript and the abovementioned studies are the findings of Cooper et al., who found that reduced sterol levels impair a cell's responsiveness to Hh [ <xref rid="pbio-0040232-b008" ref-type="bibr">8</xref>]. By using fibroblasts from mice genetically deficient for the enzymes involved in lathosterolosis (Sc5d) and SLOS (Dhcr7) and by depleting lipids and sterols from the medium, they showed that sterol levels were crucial to a cell's responsiveness to Hh. By analyzing the Hh protein in these cells under the various conditions, they could demonstrate proper Hh maturation, thereby excluding improper Hh sterolation as the cause of SLOS. They did not, however, hypothesize an inhibitory role for excess cholesterol precursors. We explain our hypothesis and findings in light of their results: by genetic and/or medium condition causes, steroid synthesis is seriously impaired in cells in their experiments. As a result, 7-DHC levels will drop, leaving Ptch1 devoid of substrate. Without substrate to pump or translocate, Ptch1 is “paralysed”, and addition of Hh will not exert an effect, explaining the reduced responsiveness. Hh responsiveness does not equal pathway activity, and the absence of Ptch1 substrate will diminish responsiveness but not pathway activity. Because both cell lines share the inability to properly synthesize cholesterol, and because we find a differential inhibitory potential on Smo, we conclude that sterol content alone cannot account for Ptch1-mediated Smo inhibition, and we suggest that the effects found by other groups take place elsewhere in the Hh signalling pathway. In our system, LDL seems to be a prerequisite for the diffusion of the inhibitory molecule, and, recently, a similar requirement has also been found for the distribution of Hh itself [ <xref rid="pbio-0040232-b041" ref-type="bibr">41</xref>]. </p><p>Although the present study clearly demonstrates that Ptch1 can inhibit Smo in an intercellular fashion, we have not addressed whether such intercellular action of Ptch1 on Smo is relevant for Hh action in vivo. Vitamin D3 is highly hydrophobic and is not expected to diffuse very far in the aqueous environment that surrounds cells. We observed that in the absence of LDL as a carrier for the hydrophobic Smo inhibitor, intercellular repression of Smo activity was undetectable. Hence, although Ptch1 exports (pro-)vitamin D3, it might act on Smo on the same cell (and thus function cell autonomously) or, depending on the cell's surroundings, on Smo on adjacent cells (and thus function non–cell autonomously). Nevertheless, the expression of Ptch1 and Smo is often discordant, with Ptch1 expression being in the vicinity of Smo-expressing cells, without complete colocalisation [ <xref rid="pbio-0040232-b042" ref-type="bibr">42</xref>, <xref rid="pbio-0040232-b043" ref-type="bibr">43</xref>]. We observed that in cardiac nerve branch ganglia, Smo and Ptch1 colocalise in the bronchial cartilage and epithelium, but not in the ganglion itself. In the ganglion, a pronounced staining for Ptch1 can be seen, whereas Smo staining is observed in only the adjacent cells. These findings are suggestive for intercellular signalling actually occurring in vivo (MF Bijlsma, unpublished data). In contrast with these findings are earlier reported experiments by Briscoe et al. [ <xref rid="pbio-0040232-b044" ref-type="bibr">44</xref>] in which a Hh uninhabitable form of Ptch1 (Ptch1 <sup>Δloop2</sup>) was evaluated with respect to its effects on Hh signalling in <named-content content-type="genus-species">Drosophila melanogaster</named-content> and <named-content content-type="genus-species">Gallus gallus</named-content> embryogenesis. The results obtained in this study suggest that the border of Ptch1 <sup>Δloop2</sup> expression more or less corresponds with the boundary of Hh-induced gene expression. Thus, it would appear that in these experimental systems, a Smo-inhibitory, vitamin D3–like ligand has not much capacity to diffuse over multiple cell layers. We therefore suggest that the action of Ptch1 can be cell autonomous or non–cell autonomous in various in vivo situations, perhaps dependent on the presence or absence of molecules that can carry the hydrophobic ligand over larger distances and subsequently present this molecule to Smo. </p><p>Recently it was demonstrated that Hh-bound Ptch1 could titrate the inhibitory action of the abovementioned Ptch1 <sup>Δloop2</sup>. Furthermore, Hh-bound Ptch1 allowed Smo signalling to occur even in the presence of a Ptch1 form that is incapable of binding Hh [ <xref rid="pbio-0040232-b045" ref-type="bibr">45</xref>]. This observation is inconsistent with a model in which Ptch1 serves to translocate a Smo-inhibitory molecule out of the cell, because in this model, Ptch1 action would depend on the amount of unliganded Ptch1, not the amount of liganded Ptch1. Casali and Struhl provide as a possible explanation for the titration effect that liganded and unliganded Ptch1 compete for access to Smo: Ptch1 acts as a multimer in which binding of Hh to any of the subunits blocks the action of the complex, or liganded and unliganded Ptch1 exert counteracting catalytic activities. Only the latter explanation seems consistent with the data presented in this study; the other explanation requires Ptch1 to function in a cell-autonomous fashion. Hence, it is possible that unliganded Ptch1 translocates a Smo inhibitor to extracellular medium, whereas Hh-bound Ptch1 would transport this inhibitor to the cytosol. Alternatively, Hh-bound Ptch1 is internalized, and it is possible that in this process it cointernalizes the Ptch1 <sup>Δloop2</sup> protein, especially if both proteins would localize to the same plasma membrane compartment—for example, lipid rafts. Support for this notion might be deduced from the strong Hh-dependent endocytosis we observed in the present study, suggesting that endocytosis is indeed a major cellular response to Hh binding to Ptch1. But further studies are needed to address the exact mode by which wild-type Ptch1 impairs Ptch1 <sup>Δloop2</sup> function. </p><p>Activation of the Hh pathway is, under certain conditions, associated with the development of cancer, especially in the upper digestive tract [ <xref rid="pbio-0040232-b046" ref-type="bibr">46</xref>, <xref rid="pbio-0040232-b047" ref-type="bibr">47</xref>] and in basal cell carcinoma in the skin, where this disease is linked to mutations in Ptch1 [ <xref rid="pbio-0040232-b048" ref-type="bibr">48</xref>– <xref rid="pbio-0040232-b051" ref-type="bibr">51</xref>]. Intercellular action of Ptch1 will allow this molecule to exert its action as a tumour suppressor, on not only the Ptch1-expressing cell but also neighbouring cells that may have acquired inactivating mutations of Ptch1, making this mode of Ptch1-dependent Smo inhibition especially attractive with respect to tumour suppression. Thus, it will be interesting to investigate whether intercellular repression of Smo by Ptch1 actually contributes to tumour suppression in vivo. Experiments addressing this possibility are currently in progress. </p></sec><sec id="s4"><title>Materials and Methods</title><sec id="s4a"><title>Constructs/siRNA used.</title><p>Mouse Ptch1 (a generous gift of Dr. M. P. Scott) was cloned into the pcDNA 3.1(–) vector (Invitrogen, Carlsbad, California, United States). Human SMO (referred to as Smo) cDNA (Image Consortium/RZPD, Berlin, Germany) was cloned into the pcDNA 3.1(+) vector. The Gli1 cDNA is in pcDNA1 (a generous gift of Dr. A. Ruiz i Altaba). The Gli-reporter δ51-LucII was kindly provided by Dr. H. Sasaki, as was the mutant binding site variant (mGli). The NF-κB, ISRE, and TAL luciferase constructs were from Clontech Laboratories (Mountain View, California, United States). The internal control cytomegalovirus (CMV) promoter-driven Renilla luciferase vector was from Promega (Madison, Wisconsin, United States). SmoM2 in pRK7 was obtained from Genentech (South San Francisco, California, United States). The Ptch1 (#63908) and scrambled control siRNA were from Ambion (Austin, Texas, United States).</p></sec><sec id="s4b"><title>Western blotting.</title><p>Transfected cells were lysed in Laemmli buffer and brought onto SDS-PAGE gels. After electrophoresis, protein was transferred onto Immobilon-PVDF membranes (Millipore, Billerica, Massachusetts, United States). Membranes were blocked in 5% bovine serum albumin (BSA) (Sigma-Aldrich, St Louis, Missouri, United States) in TBS/0.1% Tween-20 (TBST) for 1 h. Goat polyclonal α-Ptch1 antibody G-19, α-Smo C-17, α-Gli3 N-terminal N-19, α-Shh N-19, or α-Ihh I-19 (Santa Cruz Biotechnology, Santa Cruz, California, United States) were diluted to 1:500 in 3% BSA in TBST, and membranes were incubated overnight. Goat polyclonal α-actin I-19 (Santa Cruz Biotechnology) was diluted to 1:1000 in 3% BSA in TBST. After 1 h incubation in 1:1000 α-goat HRP-conjugated secondary antibody (DakoCytomation, Glostrup, Denmark), blots were imaged using LumiLight Plus enhanced chemiluminescence substrate (ECL) (Roche, Basel, Switzerland) on a GeneGnome chemiluminescence imager (Syngene, Cambridge, United Kingdom).</p></sec><sec id="s4c"><title>Mix-and-match procedure.</title><p>Mouse mesenchymal fibroblasts (C3H/10T1/2 from American Type Culture Collection CCL-226, Manassas, Virginia, United States) were grown in Dulbecco's Modified Eagle Medium (DMEM) (Cambrex, East Rutherford, New Jersey, United States) containing 10% fetal calf serum (FCS) (Cambrex). Human MDA-MB-231 breast carcinoma cells (American Type Culture Collection HTB-26) were grown in L-15 Leibovitz medium (Cambrex) with 10% FCS. Cells were grown to approximately 70% and transfected with either Ptch1 or Smo and the Gli luciferase reporter in combination with a CMV-Renilla luciferase internal standard using Effectene (Qiagen, Hilden, Germany) according to routine procedures. Ptch1 and scrambled control siRNA were transfected using RNAiFect (Qiagen) following the supplied protocol. Transfected cells were treated with o-methylated 20 μM zVADfmk (Sigma-Aldrich) to neutralize the apoptotic effects of Ptch1. After 16 h, cells were washed three times with phosphate buffered saline (PBS), detached by 2 mM EDTA (Sigma-Aldrich) in PBS, resuspended in DMEM with 10% FCS, and pipetted three times through a 40-μm cell strainer (BD Biosciences, San Jose, California, United States) into a tube to allow a homogenous mixture. Equal volumes of reporter and donor cells were mixed thoroughly and subsequently transferred to 24-well plates. After 6 h (4 + 2 h for the fusion protocol), cells were lysed with passive lysis buffer as provided by Promega, and luciferase activity was assayed according to the Promega Dual-Glo Luciferase Assay System (Promega) protocol on a Lumat Berthold LB 9501 Luminometer (Berthold Technologies, Bad Wildbad, Germany). Each firefly luciferase value was corrected for its cotransfected CMV-driven Renilla luciferase standard to correct for transfection efficiency or dilution effects. Recombinant N-terminal Shh was obtained from R&D Systems (Minneapolis, Minnesota, United States) and dissolved in PBS with 0.1% BSA. The 5E1 Shh-blocking antibody was from the Developmental Hybridoma Bank (Iowa City, Iowa, United States).</p></sec><sec id="s4d"><title>Fluorescent tracking of cells.</title><p>Cells were grown to approximately 70% confluence in DMEM containing 10% FCS and washed with PBS to remove serum. CellTracker Green CMFDA or Orange CMTMR (Molecular Probes, Eugene, Oregon, United States) were diluted to 5 μM in serum-free DMEM and added to cells for 45 min, the medium was aspirated, and fresh medium containing 10% FCS was added. Cells were detached and mixed as described above. After 16 h, cells were fixed in 3.7% formaldehyde overnight or mixed as described above and subsequently fixed. For fluorescent staining of nuclei, cells were washed with PBS/0.1% Triton X-100 (PBST) and incubated in 200 ng/ml DAPI (Roche) in PBST for 1 h. Before microscopy, cells on cover slips were washed briefly in PBS, placed upside-down on another cover slip if imaged on a Leica TCS SP II confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany), or placed on an object glass when imaged on a Zeiss Axioskop (Carl Zeiss, Oberkochen, Germany).</p></sec><sec id="s4e"><title>Cell fusion.</title><p>After fluorescent labelling (1 h) or transfection (16 h), cells were mixed, medium was aspirated, and cells were washed with PBS. Prewarmed PEG 1500/HEPES (pH 7.4) 50% (Roche) was added to the cells for 90 s, after which the cells were washed three times with PBS. Fresh medium was added, and after 2 h, cells were either lysed with passive lysis buffer and assayed for luciferase activity or fixed for microscopy after 8 h.</p></sec><sec id="s4f"><title>Medium transfer experiments.</title><p>Donor and reporter cells were transfected for 16 h, after which the donor cells were washed extensively to wash out any remaining transfection complexes and supplied with fresh medium for 6 h. Medium was transferred to reporter cells for 6 to 8 h, after which cells were lysed and assayed for luciferase activity (according to manufacturer's protocol). The mutant MEFs were grown in complete DMEM supplemented with 15% FCS and nonessential amino acids (Sigma).</p></sec><sec id="s4g"><title>CHOD-PAP–coupled FPLC.</title><p>Steroid levels in the main lipoprotein classes (VLDL, LDL, and HDL) were determined using high-performance gel filtration chromatography. The system contained a PU-980 ternary pump with an LG-980–02 linear degasser, FP-920 fluorescence, and UV-975 UV/VIS detectors (Jasco, Tokyo, Japan). An extra P-50 pump (Pharmacia Biotech, Uppsala, Sweden) was used for in-line CHOD-PAP enzymatic reagent (Biomerieux, Marcy l'Etoile, France) addition at 0.1 ml/min. 60 μl from each sample was subjected to size-exclusion chromatography to determine whether there is a relationship between Ptch1 expression and medium 3β-hydroxysteroid levels using a Superose 6 HR 10/30 (Pharmacia Biotech) column with Tris-buffered saline (TBS) (pH 7.4) at a flow rate of 0.31 ml/min with in-line fluorescence and UV detection on the Jasco system with CHOD-PAP assay described above. Commercially available lipid plasma standards were used for total cholesterol pattern analysis (SKZL, Nijmegen, Netherlands).</p></sec><sec id="s4h"><title>Endocytosis assay.</title><p>Cells on 24-well plates were grown to 70% confluence and treated with no or 1 mM pravastatin for 6 h. Subsequently, 200 nCi of [3H]-labelled sucrose (Amersham Pharmacia Biotech, Freiburg, Germany) was added per well. Cells were stimulated with either 1 μg/ml Shh or solvent control (0.1% BSA/PBS) for 1 h. After washing, cells were lysed in 1% Nonidet P-40 and the lysate was transferred to 4 ml of scintillation fluid and activity was determined on a Packard Tri-Carb scintillation counter (PerkinElmer, Wellesley, Massachusetts, United States). Values were corrected for solvent control treated cells on ice.</p></sec><sec id="s4i"><title>Stimulation experiments.</title><p>Cells transfected with the Gli-luciferase reporter were stimulated for the indicated times and concentrations with the following: cyclopamine (Biomol, Plymouth Meeting, Pennsylvania, United States), AY-9944 (Calbiochem, San Diego, California, United States), vitamin D3, or 7-DHC (Sigma).</p></sec><sec id="s4j"><title>MTT assay.</title><p>Cells were seeded in flat-bottom 96-well plates and treated with the indicated concentrations of vitamin D3 for 6 h. During the last 3 h, 0.5 mg/ml thiazolyl blue tetrazolium bromide (MTT) was added. After incubation, supernatant was discarded, cells were lysed in 50 μl of 40 mM HCl in isopropanol, and absorbance was measured at 570 nm in a Benchmark Plus Microplate Spectrophotometer (Bio-Rad, Hercules, California, United States).</p></sec><sec id="s4k"><title> <named-content content-type="genus-species">P. pastoris</named-content> culture. </title><p>The Smo-transformed <named-content content-type="genus-species">P. pastoris</named-content> strain was a kind gift from Dr. I. Mus-Veteau. Culture of yeast and induction of Smo expression was performed as described [ <xref rid="pbio-0040232-b027" ref-type="bibr">27</xref>]. </p></sec><sec id="s4l"><title>Scatchard analysis.</title><p>Scatchard analysis was performed on intact cells as described earlier for [ <xref rid="pbio-0040232-b052" ref-type="bibr">52</xref>, <xref rid="pbio-0040232-b053" ref-type="bibr">53</xref>]. After 24 h growth in methanol- or glycerol-complex medium, <named-content content-type="genus-species">P. pastoris</named-content> strains were washed with PBS and diluted to the same OD <sub>600</sub>. Subsequently, aliquots of cells were labelled for 1.5 h at 4 °C in PBS containing 1.66 nM of [3H]-labelled vitamin D3 (Amersham Pharmacia Biotech) and eight different concentrations of unlabelled cyclopamine, 7-DHC, vitamin D3, or hydroxylated vitamin D3 (Sigma-Aldrich). The reaction was stopped by washing four times with ice-cold PBS. The bound radioactivity was determined by transferring the washed pellets to 4 ml of scintillation fluid, and activity was determined on a Packard Tri-Carb scintillation counter (PerkinElmer). In each experiment, each condition was performed in duplicate. In general, Scatchard plots on intact cells show considerable nonspecific low-affinity binding of [3H]-vitamin D3. Therefore, Scatchard plots were fitted according to a one- or two-site model, as appropriate. The observed points of noninduced yeast were a satisfactory fit with a one-site (low affinity) model, whereas two affinity sites could be distinguished in the induced <named-content content-type="genus-species">P. pastoris</named-content>. </p></sec><sec id="s4m"><title>Animals.</title><p>Wild-type zebrafish <italic>(D. rerio)</italic> were kept and bred according to standard protocols. 1- to 1.5-h post fertilisation (hpf) embryos were dechorionated following 2 mg/ml pronase treatment for 1 min followed by extensive rinsing. 1.2 mg/ml vitamin D3 was added to HEPES buffered Ringer's solution and sonicated on ice for 3 min; control buffer was also sonicated. Suspension temperatures were equilibrated and added to the embryos. </p></sec><sec id="s4n"><title>ISH and immunofluorescence.</title><p>After treatment, embryos were fixed in 4% paraformaldehyde/PBS, and ISH was carried out as described [ <xref rid="pbio-0040232-b054" ref-type="bibr">54</xref>] using a <italic>ptc1</italic> probe synthesized according to [ <xref rid="pbio-0040232-b055" ref-type="bibr">55</xref>]. The 4D9 and F59 antibodies were obtained from the Developmental Hybridoma Bank and used as previously described [ <xref rid="pbio-0040232-b056" ref-type="bibr">56</xref>]. </p></sec></sec><sec sec-type="supplementary-material" id="s5"><title>Supporting Information</title><p> <bold>Accession Numbers</bold> </p><p>The Entrez Protein ( <ext-link ext-link-type="uri" xlink:href="http://www.pubmedexpress.nih.gov/entrez/query.fcgi?db=Protein">http://www.pubmedexpress.nih.gov/entrez/query.fcgi?db=Protein</ext-link>) accession numbers for the genes and gene products discussed in this paper are Gli1 (BAA85004), NPC1 (AAB63373), Ptch1 (NP_032983), and Smo (AAH48091). </p></sec>
Inhibiting Hedgehog: New Insights into a Developmentally Important Signaling Pathway	Could not extract abstract	<contrib contrib-type="author"><name><surname>Hoff</surname><given-names>Mary</given-names></name></contrib>	PLoS Biology	<p>What transforms a ball of undifferentiated cells into an organism with a nervous system, digestive tract, and other specialized body parts? Among the proteins that play an important role is one with the unlikely name of hedgehog (Hh). When Hh attaches to a transmembrane protein known as patched (Ptch1), it initiates a series of molecular interactions that lead to activation of the transcription factor Gli (for “glioma associated”) and the onset of key events in embryonic differentiation.</p><p>We know this process involves freeing a second transmembrane protein, smoothened (Smo), from inhibition. But how do Hh and Ptch1 accomplish this? Scientists would like to know because the ability of this signaling pathway to function properly makes the difference between normal development and devastating abnormalities—and because the pathway is also implicated in tumor growth.</p><p>Maarten F. Bijlsma, Maikel P. Peppelenbosch, and colleagues began their attempt to find out by noting that previous studies show that enzymes used to make cholesterol are involved in the pathway; that Ptch1 and Smo don’t necessarily bond to each other; that Ptch1 looks like other proteins that pump molecules from one side of the cell membrane to the other; and that cholesterol-like molecules can inhibit the pathway. Based on that information, the researchers hypothesized that when—and only when—Ptch1 is unencumbered by Hh, it pumps a cholesterol-like molecule into the extracellular space, where it inhibits Smo.</p><p>To test this, the researchers developed an experimental system made up of fibroblasts modified to luminesce when Gli is active (called reporter cells), and to overexpress various combinations of Ptch1, Smo, and Gli. When they mixed reporter cells overexpressing Smo with cells overexpressing Ptch1, Gli activation in the reporter cells was reduced. When they mixed them instead with cells in which Ptch1-producing genes were silenced, Gli activation increased. After performing additional tests to eliminate alternative explanations, the team concluded that Ptch1 inhibits Smo through an intermediary, and that the intermediary molecule can exert its influence between individual cells.</p><p>The researchers next exposed reporter cells to a medium that had contained Ptch1-overexpressing cells, and found Gli activation to be strongly inhibited. However, when they exposed reporter cells to a serum-free, Ptch-conditioned medium, they found no inhibition. Since serum-free medium doesn’t contain lipoproteins, they concluded that a lipoprotein is involved. Further tests suggested that the lipoprotein acts by helping transport a 3β-hydroxysteroid involved in the pathway inhibition.</p><p>Interestingly, the researchers noted that certain people with Hh signaling problems have elevated levels of a particular hydroxysteroid, 7-dehydrocholesterol (7-DHC). This led them to test the link between 7-DHC and Gli activity in mouse cells, which in turn led to the conclusion that 7-DHC indeed participates in Ptch1’s inhibition of Smo.</p><p>But is the Smo-inhibiting molecule actually 7-DHC or a compound derived from 7-DHC? When the researchers exposed medium from 7-DHC-producing mouse cells to ultraviolet radiation—which changes 7-DHC into vitamin D3—or used vitamin D3 in place of 7-DHC, Hh pathway inhibition was even stronger. They also showed that vitamin D3 binds to Smo, and that it inhibits the Hh pathway in live zebrafish embryos.</p><p>Putting it all together, the researchers concluded that when Hh isn’t present, Ptch1 pumps (pro)-vitamin D3 (i.e., either 7-DHC or vitamin D3) into the extracellular space, where the hydroxysteroid grabs onto Smo, inhibiting Gli activation. When Hh binds to Ptch1, the pump grinds to a halt, Smo is freed of inhibition, and transcription of developmentally important genes kicks in.</p><p>This new knowledge of the Hh signaling pathway shows how some cells can affect neighboring cells’ development, and it helps explain some of the problems associated with mutations affecting cholesterol biosynthesis. Because the Hh pathway is linked with certain cancers, it also has implications for tumor development. Additional work is now underway to see whether this new understanding of Ptch1’s intercellular inhibition of Smo can be applied to help suppress tumor growth.</p><sec><title/><fig id="pbio-0040258-g001" position="float"><caption><p>Structure of the vitamin D3 molecule overlaying a photograph of a zebrafish embryo stained for hedgehog pathway activity.</p></caption><graphic xlink:href="pbio.0040258.g001"/></fig></sec>
Wnt Sets the Stage for Spinal Cord Patterning in the Chick	Could not extract abstract	<contrib contrib-type="author"><name><surname>Chanut</surname><given-names>Françoise</given-names></name></contrib>	PLoS Biology	<p>The spinal cord and brain harbor a variety of motor neurons that innervate limbs, thorax, neck, or face. Different classes of motor neurons reside at different locations along the head-to-tail (rostrocaudal) axis of the nervous system, reflecting a patterning that arises early in embryonic development, when the nervous system is but a simple tube. Indeed, different sections of the neural tube begin expressing distinct subsets of genes thought to govern cell fate, such as members of the <italic>Hox</italic> gene family, shortly after the tube forms. The question is how this patterned gene expression arises. Ulrika Nordström, Thomas Edlund, and their colleagues have tackled this problem in the chick embryo. They report that rostrocaudal patterning of the spinal cord and hindbrain is under the influence of the signaling molecule Wnt. Other signals refine Wnt patterning and create distinct sections in which different types of motor neurons eventually differentiate. </p><p>The embryonic nervous system of vertebrates is shaped by signals coming from surrounding tissues, in particular from the mesoderm, which gives rise to skeletal bones and muscles. The mesodermal signal fibroblast growth factor (FGF), for instance, imposes caudal fates on the developing nervous system, whereas Wnt and FGF in combination pattern the middle portions of the prospective brain. Since Wnt remains abundant in regions where the spinal cord will start to grow behind the nascent brain, Nordström and her colleagues reasoned that it might also pattern the posterior regions of the embryo’s nervous system.</p><p>The researchers cut out small fragments (called “explants”) of prospective neural tissue from chick embryos that had reached different developmental stages and observed the explants’ development in culture. At developmental stage 4, the chick embryo is a pin-head-size disc of cells that lies atop the yolk. Its most prominent feature is the primitive streak, a thin furrow that starts as an anterior dimple known as Hensen’s node and runs all the way to the embryo’s posterior end. Cells destined to become neural occupy a horseshoe-shaped halo around Hensen’s node. The anterior and middle portions of the brain derive from the horseshoe’s arch, whereas hindbrain and spinal cord come from its branches.</p><p>The researchers found that stage 4 explants taken from the horseshoe’s branch expressed <italic>Hoxc9</italic>, <italic>Hoxb8</italic> and <italic>Hoxb4</italic> in its caudal half and <italic>Krox20</italic> in its rostral half after a day in culture. The <italic>Hox</italic> gene combination is characteristic of the lumbar and tail region of the spinal cord, whereas expression of <italic>Krox20</italic> marks the anterior hindbrain. When cultured in the presence of an inhibitor of Wnt signaling, the explants failed to turn on <italic>Krox20</italic> and the <italic>Hox</italic> genes, and instead expressed <italic>Otx2</italic>, a marker of the forebrain. Therefore, Wnt signaling appears necessary for the neural tube to adopt the posterior characteristics of hindbrain and spinal cord. </p><p>Missing from the cultured explants, however, were cells expressing a combination of <italic>Hoxb8</italic> and <italic>Hoxb4</italic>, which mark the thoracic and neck regions of the spinal cord, or <italic>Hoxb4</italic> alone, which marks the posterior hindbrain. By contrast, these cells appeared readily in cultured explants taken from stage 8 embryos. By this stage, mesoderm has crept through the backward-sliding Hensen’s node to lie under the elongating neural tube, where it begins to organize into somites, the precursors of vertebrae and ribs. The researchers reasoned that retinoic acid, a signal secreted by somites, might combine with Wnt to specify hindbrain and anterior spinal cord fates. </p><p>Adding retinoic acid to cultures of stage 4 explants confirmed this hypothesis. retinoic acid suppressed the expression of <italic>Krox20</italic> and <italic>Hoxc9</italic>, while inducing the appearance of cells expressing <italic>Hoxb8</italic> and <italic>Hoxb4</italic>, or <italic>Hoxb4</italic> alone. If the cultures also contained FGF, which normally diffuses from the primitive streak, the explants turned mostly into <italic>Hoxb8/b4</italic>-expressing cells. Culturing stage 4 explants from the forebrain region in the presence of various combinations of Wnt with retinoic acid or FGF also led to the appearance of distinct rostrocaudal subsets of spinal cord markers. </p><p>The researchers propose that, during normal development, early Wnt signaling steers the growing neural tube toward posterior fates, while retinoic acid and FGF later subdivide it into smaller rostrocaudal domains from which distinct motor neurons emerge.</p><sec><title/><fig id="pbio-0040280-g001" position="float"><caption><p>Combinatorial Wnt, retinoic acid, and FGF signals specify progenitor cell identity that prefigure motor neuron subtype in the developing hindbrain and spinal cord.</p></caption><graphic xlink:href="pbio.0040280.g001"/></fig></sec>
An Early Role for Wnt Signaling in Specifying Neural Patterns of <italic>Cdx</italic> and <italic>Hox</italic> Gene Expression and Motor Neuron Subtype Identity	<p>The link between extrinsic signaling, progenitor cell specification and neuronal subtype identity is central to the developmental organization of the vertebrate central nervous system. In the hindbrain and spinal cord, distinctions in the rostrocaudal identity of progenitor cells are associated with the generation of different motor neuron subtypes. Two fundamental classes of motor neurons, those with dorsal (dMN) and ventral (vMN) exit points, are generated over largely non-overlapping rostrocaudal domains of the caudal neural tube. <italic>Cdx</italic> and <italic>Hox</italic> genes are important determinants of the rostrocaudal identity of neural progenitor cells, but the link between early patterning signals, neural <italic>Cdx</italic> and <italic>Hox</italic> gene expression, and the generation of dMN and vMN subtypes, is unclear. Using an in vitro assay of neural differentiation, we provide evidence that an early Wnt-based program is required to interact with a later retinoic acid- and fibroblast growth factor–mediated mechanism to generate a pattern of <italic>Cdx</italic> and <italic>Hox</italic> profiles characteristic of hindbrain and spinal cord progenitor cells that prefigure the generation of vMNs and dMNs. </p>	<contrib contrib-type="author"><name><surname>Nordström</surname><given-names>Ulrika</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Maier</surname><given-names>Esther</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Jessell</surname><given-names>Thomas M</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author" corresp="yes"><name><surname>Edlund</surname><given-names>Thomas</given-names></name><email>[email protected]</email><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib>	PLoS Biology	<sec id="s1"><title>Introduction</title><p>During the early development of the vertebrate central nervous system, the position of generation of post-mitotic neurons depends on the patterning of progenitor cells along the dorsoventral and rostrocaudal axes of the neural tube [ <xref rid="pbio-0040252-b001" ref-type="bibr">1</xref>– <xref rid="pbio-0040252-b003" ref-type="bibr">3</xref>]. At many levels of the neuraxis, the dorsoventral pattern of progenitor cells, which later gives rise to motor, sensory, and local circuit neurons, is initiated by the opponent signaling activities of Sonic hedgehog (Shh) and bone morphogenetic proteins [ <xref rid="pbio-0040252-b002" ref-type="bibr">2</xref>, <xref rid="pbio-0040252-b004" ref-type="bibr">4</xref>, <xref rid="pbio-0040252-b005" ref-type="bibr">5</xref>]. In contrast, the rostrocaudal pattern of neural progenitor cells that differentiate into distinct neuronal subtypes is imposed, in part, by opponent retinoid and fibroblast growth factor (FGF) signals [ <xref rid="pbio-0040252-b006" ref-type="bibr">6</xref>– <xref rid="pbio-0040252-b009" ref-type="bibr">9</xref>]. Within the hindbrain and spinal cord, the rostrocaudal positional identity of neurons is reflected most clearly by the generation of different motor neuron (MN) subtypes. One fundamental distinction in MN subtype identity is the emergence of two major classes of MNs that exhibit distinctive axonal trajections, ventral exiting motor neurons (vMNs) and dorsal exiting motor neurons (dMNs) [ <xref rid="pbio-0040252-b010" ref-type="bibr">10</xref>]. vMNs include most spinal MNs as well as hypoglossal and abducens MNs of the caudal hindbrain [ <xref rid="pbio-0040252-b001" ref-type="bibr">1</xref>, <xref rid="pbio-0040252-b010" ref-type="bibr">10</xref>, <xref rid="pbio-0040252-b011" ref-type="bibr">11</xref>], whereas dMNs are found throughout the hindbrain and at cervical levels of the spinal cord [ <xref rid="pbio-0040252-b010" ref-type="bibr">10</xref>]. Each of the many subsequent distinctions in MN subtype identity emerge through the diversification of these two basic neuronal classes [ <xref rid="pbio-0040252-b002" ref-type="bibr">2</xref>]. </p><p>Despite many advances in defining the mechanisms of MN diversification [ <xref rid="pbio-0040252-b007" ref-type="bibr">7</xref>, <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>, <xref rid="pbio-0040252-b012" ref-type="bibr">12</xref>– <xref rid="pbio-0040252-b014" ref-type="bibr">14</xref>], it remains unclear how neural progenitors in the hindbrain and spinal cord acquire a rostrocaudal positional character that results in the generation of dMN and vMN classes. At both hindbrain and spinal levels, <italic>Hox</italic> genes are informative markers of the rostrocaudal positional identity of progenitor cells. Within the hindbrain, distinct rhombomeres are delineated by the nested expression of 3′ <italic>Hox</italic> genes [ <xref rid="pbio-0040252-b003" ref-type="bibr">3</xref>, <xref rid="pbio-0040252-b015" ref-type="bibr">15</xref>], whereas the spinal expression of 5′ <italic>Hox</italic> genes distinguishes progenitor cells and post-mitotic neurons at cervical, brachial, thoracic, and lumbar levels [ <xref rid="pbio-0040252-b001" ref-type="bibr">1</xref>, <xref rid="pbio-0040252-b006" ref-type="bibr">6</xref>– <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>, <xref rid="pbio-0040252-b013" ref-type="bibr">13</xref>, <xref rid="pbio-0040252-b016" ref-type="bibr">16</xref>]. Moreover, <italic>Hox</italic> genes are determinants of MN subtype identity in both hindbrain and spinal cord. In the hindbrain, for example, the restricted expression of Hoxb1 helps to determine the identity of facial MNs [ <xref rid="pbio-0040252-b001" ref-type="bibr">1</xref>, <xref rid="pbio-0040252-b017" ref-type="bibr">17</xref>– <xref rid="pbio-0040252-b020" ref-type="bibr">20</xref>], and in the spinal cord the restricted expression of Hox6, Hox9, and Hox10 proteins establishes MN columnar subtype [ <xref rid="pbio-0040252-b007" ref-type="bibr">7</xref>, <xref rid="pbio-0040252-b016" ref-type="bibr">16</xref>]. In addition, a more complex Hox transcriptional regulatory network specifies spinal MN pool identity and connectivity [ <xref rid="pbio-0040252-b021" ref-type="bibr">21</xref>]. The neural pattern of <italic>Hox</italic> expression is, in turn, regulated by members of the <italic>Cdx</italic> homeobox gene family [ <xref rid="pbio-0040252-b006" ref-type="bibr">6</xref>, <xref rid="pbio-0040252-b022" ref-type="bibr">22</xref>– <xref rid="pbio-0040252-b025" ref-type="bibr">25</xref>]. <italic>Cdx</italic> genes are transiently expressed in the caudal-most region of the neural plate prior to the onset of 5′ <italic>Hox</italic> gene expression [ <xref rid="pbio-0040252-b026" ref-type="bibr">26</xref>– <xref rid="pbio-0040252-b028" ref-type="bibr">28</xref>] and appear to be direct regulators of the expression of 5′ <italic>Hoxb</italic> genes [ <xref rid="pbio-0040252-b006" ref-type="bibr">6</xref>, <xref rid="pbio-0040252-b023" ref-type="bibr">23</xref>, <xref rid="pbio-0040252-b024" ref-type="bibr">24</xref>, <xref rid="pbio-0040252-b027" ref-type="bibr">27</xref>]. Thus, analysis of spatial profiles of <italic>Cdx</italic> and <italic>Hox</italic> gene expression may provide clues about the identity of signals that pattern MN subtypes in the hindbrain and spinal cord. </p><p>Several recent studies have provided insight into the signals that impose rostrocaudally-restricted patterns of neural <italic>Cdx</italic> and <italic>Hox</italic> expression. Retinoic acid (RA) and FGF signals appear to have opponent roles in the rostrocaudal patterning of <italic>Hox</italic> gene expression in the caudal hindbrain (cHB) and spinal cord [ <xref rid="pbio-0040252-b006" ref-type="bibr">6</xref>, <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>]. Mesodermal-derived RA signals promote the expression of <italic>Hox</italic> genes characteristic of the cHB and rostral spinal cord (rSC) [ <xref rid="pbio-0040252-b011" ref-type="bibr">11</xref>, <xref rid="pbio-0040252-b029" ref-type="bibr">29</xref>, <xref rid="pbio-0040252-b030" ref-type="bibr">30</xref>], whereas FGF signals pattern the expression of <italic>Hox</italic> genes at more caudal levels of the spinal cord. At an earlier developmental stage, neural progenitors have been shown to acquire caudal forebrain, midbrain, and rostral hindbrain positional identities in response to graded Wnt signaling at the gastrula stage [ <xref rid="pbio-0040252-b031" ref-type="bibr">31</xref>, <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>]. It is unclear, however, whether an early phase of Wnt signaling is also required to establish <italic>Cdx</italic> and <italic>Hox</italic> gene expression profiles characteristic of the cHB and spinal cord, in turn specifying the generation of dMN and vMN subtypes. </p><p>This study uses in vitro assays of neural cell differentiation to obtain evidence that early Wnt signaling does indeed have a crucial role in specifying the identity of hindbrain and spinal cord progenitor cells as revealed by profiles of <italic>Cdx</italic> and <italic>Hox</italic> gene expression. This early influence of Wnt signaling is later refined by retinoid and FGF signals to impart additional rostrocaudal distinctions in <italic>Hox</italic> expression that correlate tightly with the generation of dMNs and vMNs. Our findings therefore define a crucial early role for Wnt signaling in inducing profiles of <italic>Cdx</italic> and <italic>Hox</italic> expression that prefigure the differentiation of dMN and vMN subtypes in the developing hindbrain and spinal cord. </p></sec><sec id="s2"><title>Results</title><sec id="s2a"><title>Transcriptional Markers of Progenitor Cell Position and MN Subtype</title><p>To explore how progenitor cells of different rostrocaudal regional identity differentiate into dMNs and vMNs, we analyzed a panel of transcription factors that are expressed in different temporal and spatial patterns in the developing hindbrain and spinal cord.</p><p>In the hindbrain, progenitor cells do not express <italic>Cdx</italic> genes [ <xref rid="pbio-0040252-b022" ref-type="bibr">22</xref>, <xref rid="pbio-0040252-b026" ref-type="bibr">26</xref>]. Cells in rhombomeres (r)3 and r5—here defined as the rostral hindbrain (rHB)—express Krox20 (Figure1D, [ <xref rid="pbio-0040252-b033" ref-type="bibr">33</xref>]) and generate both dMNs defined by the expression of <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> and vMNs defined by the expression of Hb9 <sup>+</sup>/Isl <sup>+</sup> ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>D, [ <xref rid="pbio-0040252-b034" ref-type="bibr">34</xref>– <xref rid="pbio-0040252-b036" ref-type="bibr">36</xref>]). In contrast, progenitor cells in r7 and r8—here defined as the cHB—express <italic>Hoxb4</italic> but not <italic>Hoxb8</italic> or <italic>Hoxc9</italic> (termed <bold><italic>Hoxb4 <sup>+</sup></italic></bold> <italic>/b8</italic> <sup>−</sup> <italic>/c9 <sup>−</sup></italic> cells and all monitored by in situ hybridization) ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>A, <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>B, and <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>D) and generate <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> dMNs. ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>C, and <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>D, [ <xref rid="pbio-0040252-b034" ref-type="bibr">34</xref>– <xref rid="pbio-0040252-b036" ref-type="bibr">36</xref>]). </p><fig id="pbio-0040252-g001" position="float"><label>Figure 1</label><caption><title>Profiles of <italic>Hox</italic> Gene Expression, dMNs, and vMNs in the Hindbrain and Spinal Cord </title><p>(A) Schematic figure of the caudal embryonic neural tube divided into four distinct regions along the rostrocaudal axis: rHB; r7 and r8 defined as cHB; the region of the spinal cord located at the level of somite 6–19 defined as the rSC; and the region located caudal to somite 19 defined as cSC.</p><p>(B) In a stage 17 (30-somite) chick embryo, the rostral borders of <italic>Hoxb4, Hoxb8,</italic> and <italic>Hoxc9</italic> expression are located just caudal to the otic vesicle (OV) (at the level of the r6/7 border), at the level of somite 5/6, and at the level of somite 19/20, respectively. <italic>Hoxb4</italic> expression in the absence of <italic>Hoxb8</italic> and <italic>Hoxc9</italic> is characteristic of cells in the cHB (r7 and r8). Expression of <italic>Hoxb4</italic> and <italic>Hoxb8</italic> in the absence of <italic>Hoxc9</italic> is characteristic of the rSC domain, and expression of <italic>Hoxb4, Hoxb8, and Hoxc9</italic> is characteristic of cells in the cSC domain. </p><p>(C) In stage 20 (42-somite) chick embryos, the rostral boundaries of expression of <italic>Hoxb4, Hoxb8,</italic> and <italic>Hoxc9</italic> in the neural tube are maintained as in a stage 17 embryo (± 1-somite). <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> dMNs are present at high numbers in the cHB and at lower numbers in the rSC. No <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> dMNs are found in the cSC. In contrast, Hb9 <sup>+</sup>/Isl <sup>+</sup> vMNs are present at high numbers in both rSC and cSC, and at lower numbers in r8 of the cHB. Hoxc9 protein is expressed in a subset of Hb9 <sup>+</sup>/Isl <sup>+</sup> vMNs in the cSC and thus, distinguishes vMNs in the cSC from vMNs in the rSC. </p><p>(D) Horizontal bars represent rostrocaudal restrictions (applied to <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>A) of marker genes expressed by neural progenitor cells in the stage 17 neural tube, and by MNs in stage 20 embryos. </p></caption><graphic xlink:href="pbio.0040252.g001"/></fig><p>Cells that give rise to the spinal cord can, at Hamburger and Hamilton (HH) stages 6–8, be defined by their profile of <italic>CdxB</italic> and <italic>CdxC</italic> expression ( <xref ref-type="fig" rid="pbio-0040252-g002">Figure 2</xref> and [ <xref rid="pbio-0040252-b022" ref-type="bibr">22</xref>, <xref rid="pbio-0040252-b026" ref-type="bibr">26</xref>]). Later, at stage 17, a complex spatial pattern of <italic>Hox</italic> gene expression defines progenitor cells of different regional identities ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>). Spinal progenitor cells at prospective cervical/brachial levels adjacent to somites 6–19 (here termed “rostral” spinal cord [rSC])—express <italic>Hoxb4</italic> and <italic>Hoxb8,</italic> but do not express <italic>Hoxc9</italic> (termed <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells) ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>A, <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>B, and <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>D). MN progenitor cells at this level generate Hb9 <sup>+</sup>/Isl1 <sup>+</sup> vMNs, but only a few <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> dMNs ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>C, <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>D and [ <xref rid="pbio-0040252-b034" ref-type="bibr">34</xref>– <xref rid="pbio-0040252-b036" ref-type="bibr">36</xref>]). Progenitor cells at thoracic levels of the spinal cord—adjacent to somite 20–30 (here termed “caudal” spinal cord [cSC])—express <italic>Hoxb4, Hoxb8,</italic> and <italic>Hoxc9</italic> (termed <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cells) ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>A, <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>B, and 1D) and generate Hb9 <sup>+</sup>/Isl <sup>+</sup> vMNs that can be distinguished from those found at more rostral levels by their expression of Hoxc9 (Hb9 <sup>+</sup>/Isl <sup>+</sup>/Hoxc9 <sup>+</sup> MNs) ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>C and <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>D, [ <xref rid="pbio-0040252-b007" ref-type="bibr">7</xref>]). Since most of these markers are also expressed by non-neural cells, we used Sox proteins as general neural markers: Sox1 at stage 17 [ <xref rid="pbio-0040252-b037" ref-type="bibr">37</xref>, <xref rid="pbio-0040252-b038" ref-type="bibr">38</xref>] and Sox2 in combination with Sox3 at stage 8 as presumptive neural markers [ <xref rid="pbio-0040252-b039" ref-type="bibr">39</xref>, <xref rid="pbio-0040252-b040" ref-type="bibr">40</xref>]. In addition, we used Otx2 as a marker for neural cells located rostral to the midbrain/hindbrain boundary [ <xref rid="pbio-0040252-b041" ref-type="bibr">41</xref>, <xref rid="pbio-0040252-b042" ref-type="bibr">42</xref>]. </p><fig id="pbio-0040252-g002" position="float"><label>Figure 2</label><caption><title>Expression Pattern of <italic>CdxB</italic> and <italic>CdxC</italic> </title><p>Whole mount in situ hybridization of <italic>CdxC</italic> and <italic>CdxB</italic> in HH stage 8 embryos (4-somite). Expression of <italic>CdxC</italic> and <italic>CdxB</italic> is limited to levels adjacent and caudal to the node. The node is indicated by the red arrowhead. </p></caption><graphic xlink:href="pbio.0040252.g002"/></fig></sec><sec id="s2b"><title>Hindbrain and Spinal Cord Progenitor Cells Acquire Rostrocaudal Regional Identity at Early Somite Stages</title><p>To study the patterning of <italic>Cdx</italic> and <italic>Hox</italic> genes by neural progenitor cells and its link to the differentiation of MN subtypes, we used in vitro differentiation assays that employed stage 4–8 prospective hindbrain and spinal cord explants, and stage 4 prospective forebrain (FB) explants. In stage 4 caudal (C) explants, cells have been exposed to caudalizing signals at the time of their isolation [ <xref rid="pbio-0040252-b031" ref-type="bibr">31</xref>, <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>], and these explants were used to examine the signals that specify hindbrain and spinal cord character. Cells in stage 4 FB explants have not been exposed to caudalizing signals at the time of their isolation [ <xref rid="pbio-0040252-b031" ref-type="bibr">31</xref>, <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>], and these explants were used in attempts to reconstitute more completely the events that direct the generation of dMNs and vMNs from “naive” neural cells. </p><p>Prior to stage 8, the caudal neural plate is either specified as rSC or cSC (HH 5–7), but no explants generating cells of cHB character can be isolated (unpublished data). By stage 8, however, cells in explants isolated at a position just rostral to the regressing Hensen's node (RN explants) did not express <italic>CdxB</italic> and <italic>CdxC</italic> ( <xref ref-type="fig" rid="pbio-0040252-g002">Figure 2</xref>) and generated cells characteristic of the cHB that expressed <italic>Hoxb4</italic> alone ( <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cells (85 ± 10% of total cell number) after 40 h in culture ( <xref ref-type="fig" rid="pbio-0040252-g003">Figure 3</xref>B). In contrast, cells in stage 8 explants isolated at the level of the node (NL explants) expressed <italic>CdxB</italic> and <italic>CdxC</italic> ( <xref ref-type="fig" rid="pbio-0040252-g002">Figure 2</xref>). These explants also generated cells that expressed <italic>Hoxb4</italic> and <italic>Hoxb8</italic> in the absence of <italic>Hoxc9</italic> ( <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells; 96 ± 2% of total cell number), a marker profile characteristic of the rSC ( <xref ref-type="fig" rid="pbio-0040252-g003">Figure 3</xref>C). Cells in stage 8 explants isolated caudal to the node (CN explants) also expressed <italic>CdxB</italic> and <italic>CdxC</italic> ( <xref ref-type="fig" rid="pbio-0040252-g002">Figure 2</xref>), and generated <bold><italic>Hoxb4 <sup>+</sup></italic>/ <italic>b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold>cells (88 ± 7% of total cell number), a profile characteristic of the cSC ( <xref ref-type="fig" rid="pbio-0040252-g003">Figure 3</xref>D). Thus by stage 8, prospective hindbrain and spinal cord progenitor cells appear to have acquired a coarse rostrocaudal regional identity. </p><fig id="pbio-0040252-g003" position="float"><label>Figure 3</label><caption><title>Hindbrain and Spinal Cord Progenitor Cells Acquire Rostrocaudal Regional Identity at the Early Somite Stage</title><p>(A) Schematic drawing of a stage 8 chick embryo. The boxed regions indicate neural plate explants cultured in vitro for 40 h.</p><p>(B–D) Sox1 was used as a general neural marker. Bars represent mean ± s.e.m. number of cells in <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup>, </italic> <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup>, </italic> and <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> domains, respectively, as percentage of total cell number. Each row represents consecutive sections from a single explant. </p><p>(B) Explants isolated from the RN generated <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cells and few <italic><bold>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </bold>c9 <sup>−</sup></italic> cells ( <italic>n</italic> = 8 explants). </p><p>(C) Explants isolated at the NL generated <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells ( <italic>n</italic> = 15 explants). </p><p>(D) Explants isolated from the CN generated <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> and only a few <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells ( <italic>n</italic> = 12 explants). Scale bar represents 100 μm. </p></caption><graphic xlink:href="pbio.0040252.g003"/></fig><p>To determine whether the <italic>Hox</italic> gene profiles generated in stage 8 neural plate explants were correlated with the emergence of dMNs and vMNs, we exposed explants to the diffusible N-terminal fragment of the Shh protein (Shh-N) that exhibits MN-inducing activity [ <xref rid="pbio-0040252-b043" ref-type="bibr">43</xref>, <xref rid="pbio-0040252-b044" ref-type="bibr">44</xref>]. In the presence of Shh-N (15 nM) for ˜50 h, stage 8 RN explants generated many <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> dMNs (22 ± 7% of total cell number) and very few Hb9 <sup>+</sup>/Isl <sup>+</sup> vMNs (1 ± 0.5% of total cell or number), a profile of MNs characteristic of the cHB (r7 and rostral r8) ( <xref ref-type="fig" rid="pbio-0040252-g004">Figure 4</xref>B). Under these conditions, NL explants generated very few <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> dMNs (1.5 ± 1% of total cell number) and many Hb9 <sup>+</sup>/Isl <sup>+</sup> vMNs (25 ± 4% of total cell number). Few, if any, of the induced Hb9 <sup>+</sup>/Isl <sup>+</sup> vMNs co-expressed Hoxc9 (0.3 ± 0.3% of total cell number), indicative of an rSC positional character ( <xref ref-type="fig" rid="pbio-0040252-g004">Figure 4</xref>C). CN explants did not generate <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> dMNs but did generate Hb9 <sup>+</sup>/Isl <sup>+</sup> vMNs (27 ± 7% of total cell number), most of which expressed Hoxc9 protein (80 ± 6% of Hb9 <sup>+</sup> cells)—a profile characteristic of the thoracic spinal cord ( <xref ref-type="fig" rid="pbio-0040252-g004">Figure 4</xref>D). Thus, by stage 8, progenitor cells that occupy different rostrocaudal positions within the caudal neural plate, defined in part by their <italic>Cdx</italic> profiles, are specified as hindbrain and spinal cord cells of either rostral or caudal regional character, and have acquired sufficient positional information to differentiate into dMNs and vMNs in a position-appropriate manner. </p><fig id="pbio-0040252-g004" position="float"><label>Figure 4</label><caption><title>Hindbrain and Spinal Cord Progenitor Cells Generate dMNs and vMNs as Predicted by Their Respective <italic>Hox</italic> Expression Profile When Exposed to Shh-N </title><p>(A) Schematic of a stage 8 chick embryo. The boxed regions indicate isolated neural plate explants cultivated alone for 4 h and then exposed to Shh-N (15 nM) for an additional ˜50 h.</p><p>(B–D) Bars represent mean ± s.e.m. number of <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup>, Hb9 <sup>+</sup>/Isl <sup>+</sup>, and Hb9 <sup>+</sup>/Hoxc9 <sup>+</sup> cells, respectively, as percentage of total cell number. Each row represents consecutive sections from a single explant. </p><p>(B) Explants isolated from the RN generated <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> cells and a few Hb9 <sup>+</sup>/Isl <sup>+</sup> cells ( <italic>n</italic> = 9 explants). </p><p>(C) Explants isolated at the NL generated Hb9 <sup>+</sup>/Isl <sup>+</sup> cells, only a few <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> cells, and no Hb9 <sup>+</sup>/Isl <sup>+/</sup>Hoxc9 <sup>+</sup> cells ( <italic>n</italic> = 12 explants). </p><p>(D) Explants isolated from the CN generated Hb9 <sup>+</sup>/Isl <sup>+</sup> cells of which 80 ± 6% expressed Hoxc9 but no <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> cells appeared ( <italic>n</italic> = 10 explants). Scale bars represent 100 μm (Isl) and 50 μm (double labels), respectively. </p></caption><graphic xlink:href="pbio.0040252.g004"/></fig></sec><sec id="s2c"><title>Spinal Cord Cells of Caudal Character Are Specified at the Late Gastrula Stage</title><p>To examine when the early pattern of <italic>Cdx</italic> and <italic>Hox</italic> profiles characteristic of hindbrain and spinal cord progenitor cells is established, we first monitored the generation of cells expressing <italic>CdxB</italic> and <italic>CdxC</italic> by in situ hybridization in stage 4 C explants cultured for 15 h in vitro, corresponding to a stage 8 embryo. We tracked the rostrocaudal orientation of these explants by labeling cells at the caudal margin with DiI crystals. Under these conditions, <italic>CdxB <sup>+</sup></italic> and <italic>CdxC <sup>+</sup></italic> cells (36 ± 11% of total cell number) were generated in a caudal domain of the explants ( <xref ref-type="fig" rid="pbio-0040252-g005">Figure 5</xref>D), adjacent to the DiI labeling. To examine the rostrocaudal identity of the <italic>Cdx <sup>+</sup></italic> cells, we monitored the generation of <italic>Hoxb4 <sup>+</sup>, Hoxb8 <sup>+</sup>, </italic> and <italic>Hoxc9 <sup>+</sup></italic> cells by in situ hybridization in stage 4 C explants cultured for 44 h. Stage 4 C explants cultured for 44 h generated Krox20 <sup>+</sup> cells (45 ± 15% of total cell number), characteristic of the rHB away from the DiI label and, in a separate caudal domain (adjacent to the DiI label) <bold><italic>Hoxb4 <sup>+</sup></italic>/ <italic>b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold> cells (52 ± 12% of total cell number) characteristic of the cSC ( <xref ref-type="fig" rid="pbio-0040252-g007">Figures 7</xref>B and <xref ref-type="fig" rid="pbio-0040252-g005">5</xref>E). Thus, the generation of <italic>CdxB <sup>+</sup>/C <sup>+</sup></italic> cells and <bold><italic>Hoxb4 <sup>+</sup></italic>/ <italic>b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold> cells in a similar caudal domain of the explants supports the view that the expression of <italic>Cdx</italic> genes in neural plate cells is restricted to prospective spinal cord cells. In contrast, only a small domain of the explants generated <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells (12 ± 8% of total cell number), a marker profile characteristic of the rSC. Moreover, these explants lacked cells that expressed <italic>Hoxb4</italic> alone ( <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cells) ( <xref ref-type="fig" rid="pbio-0040252-g005">Figures 5</xref>E and <xref ref-type="fig" rid="pbio-0040252-g007">7</xref>B), characteristic of the cHB. These results provide evidence that, at stage 4, prospective caudal neural plate cells are specified as cells of rHB and cSC character, and only later acquire cHB or rSC character. </p><fig id="pbio-0040252-g005" position="float"><label>Figure 5</label><caption><title>Cells of Spinal Cord Character Are Induced by Combinatorial Wnt and FGF Signaling at the Late Gastrula Stage</title><p>(A and B) Caudal (C) neural plate tissue explants (black box) were isolated from HH stage 4 embryos and embedded in collagen matrix where their rostrocaudal orientation was maintained during in vitro cultivation for 15 h, corresponding to a stage 8 embryo (A, D, F, H, and J) or for 44 h, corresponding to a stage 17 ˜30-somite embryo (B, E, G, I, and K).</p><p>(C) Schematic drawing indicating the expression pattern of <italic>Wnt</italic> (red) and <italic>Fgf</italic> genes (green) in the primitive streak and caudal ectoderm, and in the node and primitive streak, respectively. Black dotted line indicates the presumptive neural plate. </p><p>(D–K) Each row represents consecutive sections from a single explant.</p><p>(D, F, H, and J) Sox2/3 was used as a presumptive neural marker.</p><p>(D) Stage 4 C explants were DiI-labeled and cultured alone. Cells in the caudal domain of the explants, close to the DiI-labeled cells, expressed <italic>CdxB</italic> and <italic>CdxC</italic> ( <italic>n</italic> = 12 explants). White arrowhead indicates the DiI-labeled cells. </p><p>(E, G, I, and K) Sox1 was used as a general neural marker.</p><p>(E) Cells in the rostral domain of 4 C explants cultured alone expressed Krox20, whereas <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cells appeared in the caudal domain of the explants. A small domain of <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells, but no domain of cells expressing <italic>Hoxb4</italic> alone, were generated ( <italic>n</italic> = 30 explants). </p><p>(F) Explants cultured in the presence of mFrz8CRD-IgG conditioned medium (300 μl/ml culture medium) generated Sox2/3 <sup>+</sup> neural cells but no <italic>CdxB</italic> or <italic>CdxC</italic> positive caudal neural cells ( <italic>n</italic> = 11 explants). </p><p>(G) Explants cultured in the presence of mFrz8CRD-IgG conditioned medium (300 μl/ml culture medium) generated Otx2 <sup>+</sup> but no, or few, caudal neural cells ( <italic>n</italic> = 12 explants). </p><p>(H) Explants cultured in the presence of SU5402 (5 μM), an inhibitor of FGF signaling, generated Sox2/3 <sup>+</sup> neural cells but no <italic>CdxB</italic> or <italic>CdxC</italic> positive caudal neural cells ( <italic>n</italic> = 8 explants). </p><p>(I) Explants cultured in the presence of SU5402 (5 μM), an inhibitor of FGF signaling, generated Otx2 <sup>+</sup> but no, or few, caudal neural cells ( <italic>n</italic> = 9 explants). </p><p>(J) Simultaneous exposure to Wnt3A (˜75 ng/ml) and FGF4 (60 ng/ml) resulted in the generation of cells that expressed <italic>CdxB</italic> and <italic>CdxC</italic> in the entire explant ( <italic>n</italic> = 17 explants). Scale bar represents 100 μm. </p><p>(K) Exposure to Wnt3A (˜75 ng/ml) and FGF4 (60 ng/ml), in combination, almost completely blocked the generation of Krox20 <sup>+</sup> rHB cells, and only <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> spinal cord cells were generated ( <italic>n</italic> = 17 explants). Scale bar represents 100 μm. </p></caption><graphic xlink:href="pbio.0040252.g005"/></fig><p>We next examined whether prospective rHB and cSC cells have acquired sufficient rostrocaudal positional information by stage 4 to permit them to differentiate into dMNs or vMNs when exposed to Shh-N. To test this possibility, we cultured stage 4 C explants for 28 h to allow cells to acquire a stable rostrocaudal positional identity, and then for an additional 38 h in the presence of Shh-N (15 nM) (combined culture time corresponding to a stage 20, ˜40-somite embryo). In the presence of Shh-N, <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> dMNs (1 ± 1% of total cell number) and Hb9 <sup>+</sup>/Isl <sup>+</sup> vMNs (16 ± 2% of total cell number) were generated in separate domains of the explants ( <xref ref-type="fig" rid="pbio-0040252-g006">Figure 6</xref>B). A majority of the vMNs expressed Hoxc9 (10 ± 2% of total cell number)—a profile characteristic of vMNs at thoracic levels of the spinal cord ( <xref ref-type="fig" rid="pbio-0040252-g006">Figure 6</xref>B). These results indicate that by stage 4, cells in the prospective caudal neural plate have acquired a positional character that permits them to differentiate into dMNs and vMNs. </p><fig id="pbio-0040252-g006" position="float"><label>Figure 6</label><caption><title>Wnt Signaling Is Required for the Generation of dMNs and vMNs in the Hindbrain and Spinal Cord</title><p>(A) Caudal (C) neural plate tissue explants (black box) were isolated from HH stage 4 embryos. Explants were cultured alone or in the presence of mFrz8CRD-IgG for 28 h and then exposed to Shh-N (15 nM) for an additional 38 h.</p><p>(B–C) Bars represent mean ± s.e.m. number of <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup>, Hb9 <sup>+</sup>/Isl <sup>+</sup>, Hb9 <sup>+</sup>/Hoxc9 <sup>+</sup>, and Isl1 <sup>+</sup> cells, respectively, as percentage of total cell number. Each row represents consecutive sections from a single explant. </p><p>(B) Stage 4 C explants cultured with Shh-N alone generated <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> cells in the rostral domain of the explant and Hb9 <sup>+</sup>/Isl <sup>+</sup> cells and Hb9 <sup>+</sup>/Hoxc9 <sup>+</sup> cells in the caudal domain of the explant ( <italic>n</italic> = 18 explants). </p><p>(C) Explants cultured in the presence of mFrz8CRD-IgG conditioned medium (500 μl/ml culture medium) and Shh-N generated Isl1/2 <sup>+</sup> cells but no <italic>Tbx20 <sup>+</sup>, </italic> Hb9 <sup>+</sup>, or Hoxc9 <sup>+</sup> cells ( <italic>n</italic> = 7 explants). Scale bars represent 100 μm (Isl1/2) and 50 μm (double labels), respectively. </p></caption><graphic xlink:href="pbio.0040252.g006"/></fig><fig id="pbio-0040252-g007" position="float"><label>Figure 7</label><caption><title>RA Induces cHB Cells, and RA and FGF, in Combination, Induce rSC Cells</title><p>(A) Schematic drawing of a stage 4 embryo. Dotted line indicates the presumptive neural plate. Black box indicates caudal (C) neural plate explant isolated and cultured in vitro for 44 h. The red mark indicates the caudal margin of the explant labeled with DiI. Bars represent mean ± s.e.m. number of cells in Krox20 <sup>+</sup> <bold><italic>Hoxb4 <sup>+</sup></italic></bold> <italic>/b8 <sup>−</sup>/c9 <sup>−</sup>, </italic> <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup>, </italic> and <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> domains, respectively, as percentage of total cell number. </p><p>(B–D) Each row represents consecutive sections from a single explant.</p><p>(B) Control stage 4 C explants generated <bold>Krox20 <sup>+</sup></bold> cells in the rostral, and <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold> cells were generated in the caudal region of the explant, adjacent to the DiI-labeled cells. A small domain of <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells was generated in the medial region but no domain of cells expressing <italic>Hoxb4</italic> alone was generated ( <italic>n</italic> = 7 explants). </p><p>(C) RA (10 nM) blocked the generation of Krox20 <sup>+</sup> and induced <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic>cells in the rostral region. Adjacent to the DiI-labeled cells in the caudal region of the explant, <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells, but no <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cells, were generated ( <italic>n</italic> = 6 explants). </p><p>(D) RA (10 nM) and FGF4 (30 ng/ml), in combination, generated <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup></italic></bold> <italic>/c9 <sup>−</sup></italic> cells in both the rostral and caudal regions. No, or a few, <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold> cells appeared, and no Krox20 <sup>+</sup> or <bold><italic>Hoxb4 <sup>+</sup></italic></bold> <italic>/b8 <sup>−</sup>/c9 <sup>−</sup></italic> cells were generated ( <italic>n</italic> = 5 explants). Scale bar represents 100 μm. </p></caption><graphic xlink:href="pbio.0040252.g007"/></fig></sec><sec id="s2d"><title>Joint Wnt and FGF Signaling at Late Gastrula Stages Specifies Spinal Cord Character</title><p>We next addressed the identity of the secreted signals that might be involved in the early specification of cells of spinal cord character. Several <italic>Wnt</italic> and <italic>Fgf</italic> genes are expressed in and around the prospective caudal neural plate [ <xref rid="pbio-0040252-b045" ref-type="bibr">45</xref>– <xref rid="pbio-0040252-b047" ref-type="bibr">47</xref>], and the specification of rHB cells at stage 4 requires convergent Wnt and FGF signaling [ <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>, <xref rid="pbio-0040252-b048" ref-type="bibr">48</xref>]. To examine whether combined Wnt and FGF signaling is required for the specification of cells of spinal cord positional character, we cultured stage 4 C explants in the presence of a soluble fragment of the Frizzled receptor 8 protein (Frz8CRD-IgG), an antagonist of Wnt signals [ <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>, <xref rid="pbio-0040252-b049" ref-type="bibr">49</xref>, <xref rid="pbio-0040252-b050" ref-type="bibr">50</xref>], or with SU5402, an antagonist of FGF receptor signaling [ <xref rid="pbio-0040252-b049" ref-type="bibr">49</xref>, <xref rid="pbio-0040252-b051" ref-type="bibr">51</xref>, <xref rid="pbio-0040252-b052" ref-type="bibr">52</xref>] and monitored the expression of <italic>Cdx</italic> and <italic>Hox</italic> genes. </p><p>In the presence of Frz8CRD-IgG or SU5402 (5 μM) for 15 h, the expression of both <italic>CdxB</italic> and <italic>CdxC</italic> was blocked ( <xref ref-type="fig" rid="pbio-0040252-g005">Figure 5</xref>F and <xref ref-type="fig" rid="pbio-0040252-g005">5</xref>H). After 44 h under these conditions, the generation of cells of both rHB (Krox20 <sup>+</sup>) and of spinal cord <italic>(</italic> <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> and <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> <italic>)</italic> character was almost completely blocked (3 ± 3% caudal cells remaining versus 64 ± 10% in the controls) ( <xref ref-type="fig" rid="pbio-0040252-g005">Figure 5</xref>G and <xref ref-type="fig" rid="pbio-0040252-g005">5</xref>I). Instead, Otx2 <sup>+</sup> forebrain-like cells were generated (79 ± 9% of total cell number versus 0% in the controls) ( <xref ref-type="fig" rid="pbio-0040252-g005">Figure 5</xref>G and <xref ref-type="fig" rid="pbio-0040252-g005">5</xref>I). These results support the idea that the specification of cells of spinal cord positional character also involves convergent Wnt and FGF signaling. </p><p>To test whether Wnt signaling in prospective rHB and cSC cells is required for the generation of vMNs and dMNs, stage 4 C explants were cultured in the presence of mFrz8CRD-IgG, and after 28 h Shh-N (15 nM) was added for a further 38 h. Under these conditions, the generation of <italic>Tbx20</italic> <sup>+</sup>/Isl <sup>+</sup> dMNs and Hb9 <sup>+</sup>/Isl <sup>+</sup> and Hoxc9 <sup>+</sup>/Hb9 <sup>+</sup> vMNs was blocked (0% of total cell number), and instead <bold>Isl <sup>+</sup></bold>/ <italic>Tbx20</italic> <sup>−</sup>/Hb9 <sup>−</sup>/Hoxc9 <sup>−</sup> neurons, characteristic of the ventral forebrain [ <xref rid="pbio-0040252-b053" ref-type="bibr">53</xref>, <xref rid="pbio-0040252-b054" ref-type="bibr">54</xref>], were generated (18 ± 3% of total cell number) ( <xref ref-type="fig" rid="pbio-0040252-g006">Figure 6</xref>C). Thus, exposure of prospective caudal neural plate cells to Wnt signals is required for the generation of vMNs and dMNs. </p></sec><sec id="s2e"><title>High-Level Wnt Signaling Promotes Spinal Cord Character</title><p>We next examined whether differences in the level or duration of exposure to Wnt and FGF signals contribute to the early distinction in hindbrain and spinal cord character. To test this possibility, we exposed stage 4 C explants to exogenous Wnt and FGF signals for 15 h or 44 h in vitro. In explants exposed to Wnt3A (75 ng/ml) and FGF4 (60 ng/ml) simultaneously for 15 h, <italic>CdxB</italic> <sup>+</sup> and <italic>CdxC <sup>+</sup></italic> cells were generated throughout the entire explant ( <xref ref-type="fig" rid="pbio-0040252-g005">Figure 5</xref>J). After 44 h of culture under these conditions, the generation of Krox20 <sup>+</sup> cells was largely suppressed (3 ± 2% of total cell number versus 45 ± 15% in the controls) and most cells acquired a <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cSC character (96 ± 2% of total cell number versus 52 ± 12% in the controls) ( <xref ref-type="fig" rid="pbio-0040252-g005">Figure 5</xref>K). To examine whether mesodermal cells were generated under these conditions, we monitored the expression of <italic>Mox1</italic>[ <xref rid="pbio-0040252-b055" ref-type="bibr">55</xref>], which is expressed in caudal paraxial mesoderm and of <italic>Brachyury (Bra)</italic> [ <xref rid="pbio-0040252-b056" ref-type="bibr">56</xref>], which at caudal levels is expressed in both the mesoderm and in cells of the forming caudal neural plate ( <xref ref-type="supplementary-material" rid="sg003">Figure S3</xref>G). No <italic>Mox1</italic> cells and no, or very few, <italic>Bra <sup>+</sup></italic> cells were induced ( <xref ref-type="supplementary-material" rid="sg003">Figure S3</xref>D), indicating that the few <italic>Bra <sup>+</sup></italic> cells represent caudal neural cells and not mesodermal cells. Exposure to FGF4 (60–120 ng/ml), or Wnt3A (150 ng/ml) alone did not increase the number of cells expressing <italic>Cdx</italic> genes, nor did it change the ratio of Krox20 <sup>+</sup> and <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cells ( <xref ref-type="supplementary-material" rid="sg001">Figure S1</xref> and unpublished data). These results are consistent with the view that exposure of cells to prolonged or higher level Wnt and FGF signaling promotes the specification of cells of spinal cord rather than midbrain or hindbrain character. </p></sec><sec id="s2f"><title>RA Imparts Caudal Character to Hindbrain Cells and Rostral Character to Spinal Cord Cells</title><p>How then, are rostral and caudal sub-domains of the hindbrain and spinal cord established? Since Wnt signaling contributes to the distinction in specification of prospective rHB and cSC cells at gastrulation stages, we examined first whether exposure of prospective hindbrain and spinal cord cells to Wnt signals beyond stage 8 was required for the later specification of cHB and rSC cells. Stage 8 caudal neural plate explants exposed to Frz8CRD-IgG still generated cells of cHB and spinal cord character (unpublished data); indicating that prolonged exposure to Wnt signals is not required for the generation of these two sub-domains of the caudal neural tube.</p><p>By stage 8, the retinoid acid synthesizing enzyme retinaldehyde dehydrogenase 2 (RALDH2) is expressed in the paraxial mesoderm adjacent to the prospective cHB and rSC [ <xref rid="pbio-0040252-b057" ref-type="bibr">57</xref>– <xref rid="pbio-0040252-b059" ref-type="bibr">59</xref>]. RA signaling might therefore promote the generation of cells of cHB and rSC character by acting on neural cells that have already acquired an initial caudal character, through convergent Wnt and FGF signaling. To test this idea, we exposed stage 4 C explants to RA, and used the combinatorial expression of <italic>Hoxb4, Hoxb8,</italic> and <italic>Hoxc9</italic> to distinguish cells of cHB, and rSC or cSC character. Since <italic>Cdx</italic> expression does not distinguish between cells of rSC and cSC character, we did not monitor <italic>Cdx</italic> expression in these experiments. To map prospective rHB and cSC cells in these explants, we tracked their rostrocaudal orientation by labeling cells at the caudal margin with DiI crystals. We found that <bold><italic>Hoxb4 <sup>+</sup></italic>/ <italic>b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold> cells derive from the caudal-most region of the explants, whereas Krox20 <sup>+</sup> cells derive from a rostral domain of the explants ( <xref ref-type="fig" rid="pbio-0040252-g007">Figure 7</xref>B). </p><p>In the presence of RA (10 nM), no Krox20 <sup>+</sup> cells of rHB character were generated in the rostral domain of the explants—the domain lacking DiI-labeled cells. Instead, <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cells characteristic of the cHB (47 ± 10% of total cell number versus 0% in the controls) were generated in this domain ( <xref ref-type="fig" rid="pbio-0040252-g007">Figure 7</xref>C). In the caudal domain of the explants—the domain adjacent to DiI-labeled cells—the presence of RA (10 nM) blocked the generation of <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cells characteristic of the cSC (0% of total cell number versus 52 ± 12% in controls), and instead, <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells characteristic of the rSC (55 ± 10% of total cell number versus 12 ± 8% in controls) were generated ( <xref ref-type="fig" rid="pbio-0040252-g007">Figure 7</xref>C). Furthermore, RA was unable to induce caudal character in stage 4 prospective FB cells that had not been exposed to caudalizing signals [ <xref rid="pbio-0040252-b031" ref-type="bibr">31</xref>] ( <xref ref-type="supplementary-material" rid="sg002">Figure S2</xref>E). These results provide evidence that RA promotes the generation of <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cHB cells by blocking Krox20 and inducing <italic>Hoxb4</italic> expression in caudalized prospective Krox20 <sup>+</sup> rHB cells. In addition, they show that RA induces <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> rSC cells by preventing <italic>Hoxc9</italic> expression in prospective cSC cells. </p></sec><sec id="s2g"><title>FGF Signaling Contributes to the Distinction between Cells of cHB and rSC Character</title><p>We next examined how the distinction between cHB and rSC cells is established. RA promotes the generation of cHB cells, and FGFs promote the expression of <italic>Hox</italic> genes characteristic of the caudal region of the spinal cord [ <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>, <xref rid="pbio-0040252-b046" ref-type="bibr">46</xref>]. Moreover, RA and FGF signals act in an opponent manner during rostrocaudal patterning of <italic>Hox</italic> gene expression and MN progenitor cell specification [ <xref rid="pbio-0040252-b006" ref-type="bibr">6</xref>, <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>]. These observations led us to examine the possibility that FGF and RA signals converge during the initial assignment of cHB and rSC positional character. </p><p>We exposed stage 4 C explants to both RA (10 nM) and FGF4 (30 ng/ml) and assayed their <italic>Hox</italic> profile after 44 h. The generation of <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cHB cells was suppressed (1 ± 1% of total cell number versus 47 ± 10% in explants cultured with RA alone), and most cells (92 ± 3% of total cell number versus 55 ± 10% in explants cultured with RA alone) expressed <italic>Hoxb4</italic> and <italic>Hoxb8,</italic> a profile indicative of rSC character. A few <bold><italic>Hoxb4 <sup>+</sup></italic>/ <italic>b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold> cells (7 ± 4% of total cell number versus 0% in explants cultured with RA alone), characteristic of cSC character were also detected ( <xref ref-type="fig" rid="pbio-0040252-g007">Figure 7</xref>D). Thus, in the presence of RA, prolonged exposure to FGF signals promotes the generation of cells of rSC at the expense of cHB character. This finding supports the idea that the status of FGF signaling biases whether RA exposure induces <italic>Hox</italic> gene profiles characteristic of cHB or rSC. </p></sec><sec id="s2h"><title>Combinatorial Wnt, RA, and FGF Signals Impose Hindbrain and Spinal Cord Character in Naive Neural Cells</title><p>As a further test of the sufficiency of Wnt, FGF, and RA signals in establishing hindbrain and spinal cord pattern, we examined whether a combination of these factors can establish appropriate <italic>Cdx</italic> and <italic>Hox</italic> gene profiles in naive rostral forebrain cells that appear not to have been exposed to caudalizing signals in ovo. </p><p>Stage 4 FB explants cultured alone for 44 h generated Sox1 <sup>+</sup>/Otx2 <sup>+</sup> cells (94 ± 4% of total cell number) ( <xref ref-type="fig" rid="pbio-0040252-g009">Figures 9</xref>B and S2B). Wnt3A, FGF4, or RA added separately, or FGF4 and RA added in combination, did not induce cells of caudal neural character ( <xref ref-type="supplementary-material" rid="sg002">Figure S2</xref>; see also [ <xref rid="pbio-0040252-b031" ref-type="bibr">31</xref>, <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>]). Stage 4 FB explants exposed to Wnt3A and FGF4 in combination for 15 h, corresponding to stage 8, generated <italic>CdxB <sup>+</sup></italic> and <italic>CdxC <sup>+</sup></italic> cells (95 ± 4 % of total cell number) ( <xref ref-type="fig" rid="pbio-0040252-g008">Figure 8</xref>C). After 44 h of culture, <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cSC cells (96 ± 2% of total cell number) and only a few Krox20 <sup>+</sup> cells (4 ± 1% of total cell number) were generated ( <xref ref-type="fig" rid="pbio-0040252-g009">Figure 9</xref>C). Thus, combined Wnt and FGF signals establish <italic>Cdx</italic> and <italic>Hox</italic> gene profiles indicative of cSC identity in prospective FB cells. </p><fig id="pbio-0040252-g008" position="float"><label>Figure 8</label><caption><title>Wnt and FGF, in Combination, Induce <italic>Cdx</italic> Gene Expression in Prospective FB Cells </title><p>(A) Schematic drawing of a stage 4 embryo. Dotted line indicates the presumptive neural plate. Red box indicates prospective FB explants used for in vitro studies.</p><p>(B–D) Sox2 and Sox3, in combination, were used as general presumptive neural markers. Each row represents consecutive sections from a single explant.</p><p>(B) Control stage 4 FB explants generated Sox2 <sup>+</sup>/3 <sup>+</sup> but no <italic>CdxC <sup>+</sup>/CdxB </italic> <sup>+</sup> presumptive neural cells ( <italic>n</italic> = 24 explants). </p><p>(C) Stage 4 FB explants cultured in the presence of Wnt (˜150 ng/ml) and FGF (60 ng/ml) simultaneously generated <italic>CdxC <sup>+</sup></italic> <italic>/CdxB <sup>+</sup></italic> presumptive caudal neural cells ( <italic>n</italic> = 26 explants). </p><p>(D) 4 FB explants cultivated in the presence of Wnt3A (˜150 ng/ml), RA (10 nM), and FGF (30 ng/ml) generated <italic>CdxC <sup>+</sup></italic> <italic>/CdxB <sup>+</sup></italic> presumptive caudal neural cells ( <italic>n</italic> = 18 explants). </p><p>(E) 4 FB explants cultivated in the presence of Wnt3A (˜150 ng/ml) and RA (10 nM) did not generate <italic>CdxC <sup>+</sup></italic> <italic>/CdxB <sup>+</sup></italic> presumptive caudal neural cells ( <italic>n</italic> =14 explants). </p></caption><graphic xlink:href="pbio.0040252.g008"/></fig><fig id="pbio-0040252-g009" position="float"><label>Figure 9</label><caption><title>Combinatorial Wnt, RA, and FGF Signaling Reconstruct <italic>Hox</italic> Gene Profiles Characteristic of the cHB and Spinal Cord </title><p>(A) Schematic drawing of a stage 4 chick embryo. Dotted line indicates the presumptive neural plate. Red box indicates FB explants isolated and cultured in vitro for 44 h.</p><p>(B–F) Sox1 was used as a general neural marker. Bars represent mean ± s.e.m. number of cells in Otx2 <sup>+</sup>, <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup>, </italic> <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup>, </italic> and <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> domains, respectively, as percentage of total cell number. Each row represents consecutive sections from a single explant. </p><p>(B) Control stage 4 FB explants generated Sox1 <sup>+</sup>/Otx2 <sup>+</sup> but no caudal neural cells ( <italic>n</italic> = 24 explants). </p><p>(C) Stage 4 FB explants cultured in the presence of Wnt (˜150 ng/ml) and FGF (60 ng/ml) generated <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cells and only a few Krox20 <sup>+</sup> cells ( <italic>n</italic> = 24 explants). </p><p>(D) Cultivation in the presence of Wnt3A (˜150 ng/ml), RA (10 nM), and FGF (30 ng/ml) generated <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells and a few <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cells ( <italic>n</italic> = 18 explants). </p><p>(E) Cultivation in the presence of Wnt3A (˜150 ng/ml) and RA (10 nM) generated <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cells and no, or only a few, <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells ( <italic>n</italic> = 28 explants). </p><p>(F) Exposure to Wnt3A (˜150 ng/ml) and RA (10 nM) in the presence of SU5402 (3 μM), an inhibitor of FGF signaling, generated <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cells ( <italic>n</italic> = 12 explants). Scale bar represents 100 μm. </p></caption><graphic xlink:href="pbio.0040252.g009"/></fig><p>We next examined whether, in the presence of Wnt and FGF signals, RA can induce the generation of cHB or rSC cells. In the combined presence of Wnt3A (150 ng/ml), FGF4 (30 ng/ml), and RA (10 nM) for 15 h, stage 4 FB explants generated <italic>CdxB <sup>+</sup></italic> and <italic>CdxC <sup>+</sup></italic> cells (97 ± 3% of total cell number) ( <xref ref-type="fig" rid="pbio-0040252-g008">Figure 8</xref>D). After 44 h of culture, <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> rSC cells (91 ± 5% of total cell number) and some <bold><italic>Hoxb4 <sup>+</sup></italic>/ <italic>b8 <sup>+</sup>/ </italic>c9 <italic><sup>+</sup></italic></bold> cSC cells (9 ± 5% of total cell number), but no <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cells of cHB character, were generated ( <xref ref-type="fig" rid="pbio-0040252-g009">Figure 9</xref>D). Thus, Wnt signals in combination with RA and FGF signals predominantly induce a <italic>Hox</italic> gene profile characteristic of the rSC. </p><p>Next, we tested whether combined exposure to Wnt3A (150 ng/ml) and RA (10 nM) induces cHB cells. No <italic>CdxB <sup>+</sup></italic> and <italic>CdxC <sup>+</sup></italic> cells were generated after 15 h of culture ( <xref ref-type="fig" rid="pbio-0040252-g008">Figure 8</xref>E). After 44 h of culture, many <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cHB cells (92 ± 3% of total cell number versus 0% in control explants), few (3 ± 3% of total cell number) <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup>, </italic> and no <bold><italic>Hoxb4 <sup>+</sup></italic>/ <italic>b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold> cells characteristic of the rSC and cSC, were generated ( <xref ref-type="fig" rid="pbio-0040252-g009">Figure 9</xref>E). Moreover, in the presence of SU5402 (3 μM), Wnt3A (150 ng/ml) and RA (10 nM) still induced <bold><italic>Hoxb4 <sup>+</sup>/ </italic></bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup>/ </italic> cHB cells (97 ± 2% of total cell number) but no <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> cells, suggesting that the convergent activities of Wnt and RA signaling are sufficient to induce cHB cells ( <xref ref-type="fig" rid="pbio-0040252-g009">Figure 9</xref>F). No combination of Wnt3A-, RA-, and FGF4-induced <italic>Mox1 <sup>+</sup></italic> cells characteristic of caudal paraxial mesoderm, and only Wnt3A and FGF4 in combination, induced a few <italic>Bra <sup>+</sup></italic> cells ( <xref ref-type="supplementary-material" rid="sg003">Figure S3</xref>B and <xref ref-type="supplementary-material" rid="sg003">S3</xref>C, unpublished data). These results provide evidence that Wnt signals, in combination with RA and/or FGF, are sufficient to reconstruct <italic>Cdx</italic> and <italic>Hox</italic> gene profiles indicative of cHB and both rSC and cSC in naive prospective FB explants. </p><p>We also addressed whether <italic>Hox</italic> gene profiles induced by Wnt, RA, and/or FGF signals in naive prospective FB cells predict the later generation of dMNs and vMNs. We cultivated stage 4 FB explants in the presence of Wnt, RA, and/or FGF signals for 44 h, to allow hindbrain and spinal cord progenitor cells to acquire their rostrocaudal positional identity, and for an additional 22 h in the absence or presence of Shh-N (15 nM) to induce MN differentiation. Stage 4 FB explants, grown both in the absence or presence of Shh-N (15 nM), generated Otx2 <sup>+</sup> neural progenitor cells and <bold>Isl <sup>+</sup></bold>/ <italic>Tbx20</italic> <sup>−</sup>/Hb9 <sup>−</sup>/Hoxc9 <sup>−</sup> ventral FB neurons ( <xref ref-type="fig" rid="pbio-0040252-g010">Figure 10</xref>B and <xref ref-type="table" rid="pbio-0040252-t001">Table 1</xref>, [ <xref rid="pbio-0040252-b054" ref-type="bibr">54</xref>]). Stage 4 FB cells exposed to FGF4 (60 ng/ml) and Wnt3A (150 ng/ml), which generate <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/c9 <sup>+</sup></italic></bold> cSC progenitor cells, generated Hb9 <sup>+</sup>/Isl <sup>+</sup> and Hoxc9 <sup>+</sup>/Hb9 <sup>+</sup> vMNs in the presence of Shh-N (15nM) ( <xref ref-type="fig" rid="pbio-0040252-g010">Figure 10</xref>E and <xref ref-type="table" rid="pbio-0040252-t001">Table 1</xref>). We also determined which MN subtypes are generated from the <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup>/ </italic></bold> <italic>c9 <sup>−</sup></italic> rSC progenitor cells, induced by Wnt3A (150 ng/ml), FGF4 (30 ng/ml), and RA (10 nM). Addition of Shh-N (15 nM) induced a large number of Hb9 <sup>+</sup>/Isl <sup>+</sup> MNs, but few Hb9 <sup>+</sup>/Hoxc9 <sup>+</sup> or <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> MNs—a MN profile characteristic of the rSC ( <xref ref-type="fig" rid="pbio-0040252-g010">Figure 10</xref>D and <xref ref-type="table" rid="pbio-0040252-t001">Table 1</xref>). Finally, we assessed the MN subtype generated from stage 4 FB cells which generate <bold><italic>Hoxb4 <sup>+</sup></italic>/ </bold> <italic>b8 <sup>−</sup>/c9 <sup>−</sup></italic> cHB progenitor cells, upon Wnt3A (150 ng/ml) and RA (10 nM) exposure. Addition of Shh-N (15 nM) induced <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup>, but only a few Hb9 <sup>+</sup>/Isl <sup>+</sup> MNs and no Hb9 <sup>+</sup>/Hoxc9 <sup>+</sup> MNs ( <xref ref-type="fig" rid="pbio-0040252-g010">Figure 10</xref>C and <xref ref-type="table" rid="pbio-0040252-t001">Table 1</xref>)—a profile characteristic of cHB (r7 and r8) MNs ( <xref ref-type="fig" rid="pbio-0040252-g001">Figure 1</xref>C and <xref ref-type="fig" rid="pbio-0040252-g001">1</xref>D). </p><fig id="pbio-0040252-g010" position="float"><label>Figure 10</label><caption><title>Hox Gene Profiles Induced by Wnt, RA, and/or FGF Signals Predict Later MN Subtype</title><p>(A) Schematic drawing of a stage 4 embryo. Dotted line indicates the presumptive neural plate. Red box indicates prospective FB explants used for in vitro studies.</p><p>(B–E) Explants were cultured alone or exposed to Wnt, RA, and/or FGF4 for 44 h, then washed and exposed to Shh-N (15 nM) for an additional 22 h. Bars represent mean ± s.e.m. number of <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup>, Hb9 <sup>+</sup>/Isl <sup>+</sup>, and Hb9 <sup>+</sup>/Hoxc9 <sup>+</sup> cells, respectively, as percentage of total cell number. </p><p>(B–E) Each row represents consecutive sections from a single explant.</p><p>(B) Stage 4 FB explants cultured alone, before exposure of Shh-N, generated Isl <sup>+</sup> cells but no <italic>Tbx20 <sup>+</sup>, </italic> Hb9 <sup>+</sup>, or Hoxc9 <sup>+</sup> cells ( <italic>n</italic> = 6 explants). </p><p>(C) Stage 4 FB explants cultured in the presence of Wnt3A (˜150 ng/ml) and RA (10 nM), before exposure of Shh-N, generated <italic>Tbx20 <sup>+</sup></italic>/Isl <sup>+</sup> cells but no, or very few, Hb9 <sup>+</sup>/Isl <sup>+</sup> cells and no Hoxc9 <sup>+</sup> cells ( <italic>n</italic> = 12 explants). </p><p>(D) Cultivation in the presence of Wnt3A (˜150 ng/ml), RA (10 nM), and FGF4 (30 ng/ml), before exposure of Shh-N, generated Hb9 <sup>+</sup>/Isl <sup>+</sup> cells and only a few <italic>Tbx20</italic> <sup>+</sup>/Isl <sup>+</sup> and Hb9 <sup>+</sup>/Hoxc9 <sup>+</sup> cells ( <italic>n</italic> = 14 explants). </p><p>(E) Cultivation in the presence of Wnt3A (˜150 ng/ml) and FGF4 (60 ng/ml), before exposure of Shh-N, generated Hb9 <sup>+</sup>/Isl <sup>+</sup> and Hb9 <sup>+</sup>/Hoxc9 <sup>+</sup> cells but no <italic>Tbx20 <sup>+</sup></italic> cells ( <italic>n</italic> = 9 explants). Scale bars represent 100 μm (Isl) and 50 μm (double labels), respectively. </p></caption><graphic xlink:href="pbio.0040252.g010"/></fig><table-wrap id="pbio-0040252-t001" position="float"><label>Table 1</label><caption><p>Quantification of Transcription Factor Expression in Neural Plate Explants</p></caption><graphic xlink:href="pbio.0040252.t001"/></table-wrap><p>Collectively, these findings provide evidence that Wnt signals in combination with RA and/or FGF exposure, induce <italic>Hox</italic> profiles in neural cells that predict the later position-specific emergence of dMNs and vMNs characteristic of the hindbrain and spinal cord. These observations support the idea that early exposure to Wnt signals, together with later RA and FGF signals imposes <italic>Hox</italic> profiles that anticipate the patterned generation of dMN and vMN subclasses in the developing hindbrain and spinal cord. </p></sec></sec><sec id="s3"><title>Discussion</title><p>This study has examined the link between extrinsic patterning signals, regionally restricted profiles of transcription factor expression in neural progenitor cells, and the specification of MN subtype along the rostrocaudal axis of the hindbrain and spinal cord. Our results support four main conclusions: (i) Wnt signaling is required to specify cells of spinal cord character, (ii) the initial specification of spinal cord progenitor cells appears to require prolonged, or higher level Wnt signaling than does the specification of cells of hindbrain character, (iii) early Wnt signaling provides a positional context for the later actions of RA and FGF signals in specifying the rostrocaudal regional identity of hindbrain and spinal cord cells, and (iv) the interplay of Wnt, retinoid, and FGF signals establish distinction in progenitor cell <italic>Cdx</italic> and <italic>Hox</italic> profiles that anticipate the rostrocaudal position of generation of dMNs and vMNs in the hindbrain and spinal cord. Below, we discuss the evidence that supports each of these conclusions. </p><sec id="s3a"><title>Interplay between Early Wnt and Later FGF and RA Signals in the Assignment of Caudal and Neural Fates</title><p>The generation of different subclasses of MNs along the rostrocaudal axis of the hindbrain and spinal cord depends on two crucial early steps of caudal neural development: first, the early specification of cells of hindbrain and spinal cord character, and second, the subsequent refinement of rostrocaudal regional character of hindbrain and spinal cord progenitor cells. Wnt signaling has been implicated in the generation of caudal neural cells [ <xref rid="pbio-0040252-b048" ref-type="bibr">48</xref>, <xref rid="pbio-0040252-b060" ref-type="bibr">60</xref>– <xref rid="pbio-0040252-b069" ref-type="bibr">69</xref>], and results in chick have provided evidence that FGF and graded Wnt signaling in neural cells specify cells of caudal forebrain, midbrain, and rostral hindbrain character [ <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>, <xref rid="pbio-0040252-b066" ref-type="bibr">66</xref>]. The present study provides evidence that early Wnt signaling is also essential to impose caudal hindbrain and spinal cord character on neural progenitor cells. The results also support the view that the specification of spinal cord progenitor cells requires prolonged, or higher level Wnt signaling than is required for the specification of cells of hindbrain character, and they are consistent with previous findings that <italic>Cdx</italic> genes respond to Wnt signals and act upstream of 5′ <italic>Hox</italic> genes in neural progenitor cells [ <xref rid="pbio-0040252-b023" ref-type="bibr">23</xref>, <xref rid="pbio-0040252-b070" ref-type="bibr">70</xref>– <xref rid="pbio-0040252-b075" ref-type="bibr">75</xref>]. Thus, taken together, our results support the idea that in the presence of FGF signals, graded Wnt signaling imposes midbrain, hindbrain, and spinal cord character on prospective caudal neural plate cells. </p><p>The signals and mechanisms that act in the subsequent step to impose rostrocaudal regional identity on hindbrain and spinal cord progenitor cells have been examined previously. From early somite stages, RA supplied by the paraxial mesoderm and newly formed somites, promotes the expression of <italic>Hox</italic> genes characteristic of the cHB and rostral levels of the spinal cord [ <xref rid="pbio-0040252-b006" ref-type="bibr">6</xref>, <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>, <xref rid="pbio-0040252-b029" ref-type="bibr">29</xref>, <xref rid="pbio-0040252-b031" ref-type="bibr">31</xref>, <xref rid="pbio-0040252-b067" ref-type="bibr">67</xref>, <xref rid="pbio-0040252-b076" ref-type="bibr">76</xref>, <xref rid="pbio-0040252-b077" ref-type="bibr">77</xref>]. FGF signals derived from the regressing primitive streak promote the expression of progressively more caudal Hox-c proteins in a concentration-dependent manner [ <xref rid="pbio-0040252-b007" ref-type="bibr">7</xref>, <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>]. Thus, RA and FGF signals act in an opponent manner to impose rostrocaudal regional identity on hindbrain and spinal cord progenitor cells [ <xref rid="pbio-0040252-b006" ref-type="bibr">6</xref>, <xref rid="pbio-0040252-b078" ref-type="bibr">78</xref>]. Our findings extend these results by showing that early Wnt signaling establishes a positional context for the later actions of RA and FGF signals in specifying hindbrain and spinal cord cells of rostral and caudal regional identity. </p><p>Collectively, our results suggest a model of how hindbrain and spinal cord cells of early rostrocaudal regional identity are generated ( <xref ref-type="fig" rid="pbio-0040252-g011">Figure 11</xref>). At gastrula stages, prospective caudal neural plate cells are exposed to Wnt signals derived from the emerging caudal paraxial mesoderm and from epiblast cells; and to FGF signals derived from the primitive streak [ <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>, <xref rid="pbio-0040252-b046" ref-type="bibr">46</xref>, <xref rid="pbio-0040252-b047" ref-type="bibr">47</xref>, <xref rid="pbio-0040252-b079" ref-type="bibr">79</xref>, <xref rid="pbio-0040252-b080" ref-type="bibr">80</xref>]. In response to convergent Wnt and FGF signaling, prospective caudal neural plate cells are initially specified either as cells of rHB character or as <bold><italic>Hoxb4 <sup>+</sup>/b8 <sup>+</sup></italic>/ <italic>c9 <sup>+</sup></italic></bold> cells characteristic of caudal/thoracic spinal cord. At early somite stages, caudal paraxial mesoderm and newly formed somites located adjacent to the prospective cHB and rSC, express high levels of <italic>Raldh2,</italic> providing a local source of RA [ <xref rid="pbio-0040252-b059" ref-type="bibr">59</xref>, <xref rid="pbio-0040252-b081" ref-type="bibr">81</xref>]. Our results suggest that RA specifies cells of r7/r8 cHB character by inducing <italic>Hoxb4</italic> expression in prospective rHB cells and cells of rSC identity by preventing the expression of <italic>Hoxc9</italic> in prospective cSC cells. At these stages, several <italic>Fgfs</italic> are expressed in the regressing Hensen's node and primitive streak adjacent to the developing spinal cord [ <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>, <xref rid="pbio-0040252-b046" ref-type="bibr">46</xref>, <xref rid="pbio-0040252-b079" ref-type="bibr">79</xref>]. We find that FGF signals maintain the specification of cSC cells, and in the presence of RA, promote the generation of rSC cells. This model is strengthened by our data providing evidence that distinct combinations of Wnt, RA, and/or FGF signals can reconstitute rostral and caudal hindbrain and spinal cord character in naive prospective FB cells in a predictable manner. </p><fig id="pbio-0040252-g011" position="float"><label>Figure 11</label><caption><title>Combinatorial Wnt, RA, and FGF Signals Specify Progenitor Cell Identity That Prefigure MN Subtype in the Developing Hindbrain and Spinal Cord</title><p>Combinatorial actions of Wnt, FGF, and RA signals specify neural progenitor cells expressing <italic>Hox</italic> gene profiles characteristic of the cHB, rSC, and cSC that generate patterns of differentiated MNs, with dMN or vMN exit points, characteristic of hindbrain and spinal cord, in response to Shh signaling. </p></caption><graphic xlink:href="pbio.0040252.g011"/></fig><p>Genetic analyses have provided evidence that inactivation of <italic>Wnt</italic> genes, expressed in the caudal regions of gastrula stage mouse and zebrafish embryos, perturbs the development of the caudal neural plate [ <xref rid="pbio-0040252-b060" ref-type="bibr">60</xref>, <xref rid="pbio-0040252-b062" ref-type="bibr">62</xref>, <xref rid="pbio-0040252-b063" ref-type="bibr">63</xref>, <xref rid="pbio-0040252-b069" ref-type="bibr">69</xref>, <xref rid="pbio-0040252-b082" ref-type="bibr">82</xref>]. However, the formation of paraxial mesoderm, which serves as a local source of neural caudalizing signals [ <xref rid="pbio-0040252-b063" ref-type="bibr">63</xref>, <xref rid="pbio-0040252-b083" ref-type="bibr">83</xref>– <xref rid="pbio-0040252-b085" ref-type="bibr">85</xref>], is also impaired in these mutant embryos [ <xref rid="pbio-0040252-b045" ref-type="bibr">45</xref>, <xref rid="pbio-0040252-b069" ref-type="bibr">69</xref>, <xref rid="pbio-0040252-b086" ref-type="bibr">86</xref>]. Thus, these genetic studies left unresolved the issue whether the effect of perturbed Wnt signaling on caudal neural development reflects the impaired formation of paraxial mesoderm or reflects direct Wnt signaling in neural cells. Our in vitro studies establish direct effects of Wnt signals on neural tissue in the absence of other tissues and clarify the integrative mechanisms that control the early development of the hindbrain and spinal cord. </p><p>FGF and retinoid signals also regulate the temporal pattern of differentiation of caudal neural progenitor cells. FGF has been shown to keep cells in a stem zone–like state [ <xref rid="pbio-0040252-b014" ref-type="bibr">14</xref>, <xref rid="pbio-0040252-b078" ref-type="bibr">78</xref>, <xref rid="pbio-0040252-b087" ref-type="bibr">87</xref>], whereas RA promotes the differentiation of neural cells [ <xref rid="pbio-0040252-b014" ref-type="bibr">14</xref>, <xref rid="pbio-0040252-b078" ref-type="bibr">78</xref>, <xref rid="pbio-0040252-b088" ref-type="bibr">88</xref>]. Consistent with these roles of FGF and RA, exposure of naive neural explants, cultured with Wnt and FGF to low levels of RA (2 nM), does not change the rostrocaudal identity of neural progenitor cells but results in an increased number of differentiated MNs (unpublished data), whereas under these conditions, increased levels of FGFs greatly reduce MN differentiation (unpublished data). These findings fit well with the suggested opponent activities for FGF and RA in deciding the balance between neural cell proliferation and differentiation [ <xref rid="pbio-0040252-b087" ref-type="bibr">87</xref>]. </p></sec><sec id="s3b"><title> <italic>Hox</italic> Gene Profiles of Hindbrain and Spinal Cord Progenitor Cells Predict the Pattern of dMNs and vMNs </title><p>The patterned expression of <italic>Hox</italic> genes in neural progenitor cells appears to be a major determinant of the identity of different MN populations [ <xref rid="pbio-0040252-b001" ref-type="bibr">1</xref>, <xref rid="pbio-0040252-b007" ref-type="bibr">7</xref>, <xref rid="pbio-0040252-b011" ref-type="bibr">11</xref>, <xref rid="pbio-0040252-b016" ref-type="bibr">16</xref>, <xref rid="pbio-0040252-b018" ref-type="bibr">18</xref>]. Earlier studies have provided evidence that RA and FGF act in an opponent manner on caudal neural cells to establish the rostrocaudal pattern of distinct subclasses of differentiated MNs [ <xref rid="pbio-0040252-b007" ref-type="bibr">7</xref>– <xref rid="pbio-0040252-b009" ref-type="bibr">9</xref>]. dMNs and vMNs represent two major MN subclasses that are generated in distinct rostrocaudal patterns in the hindbrain and spinal cord [ <xref rid="pbio-0040252-b010" ref-type="bibr">10</xref>, <xref rid="pbio-0040252-b089" ref-type="bibr">89</xref>]. Our findings provide evidence that Wnt signaling in neural progenitor cells is required for the generation of both vMNs and dMNs. We also show that Wnt, RA, and/or FGF signals can induce cells with <italic>Hox</italic> gene profiles characteristic of cHB, rSC, and cSC progenitor cells that differentiate into corresponding dMNs and vMNs when exposed to Shh-N. These findings therefore reveal a tight link between Wnt, RA, and FGF signals, profiles of progenitor cell <italic>Hox</italic> expression, and the rostrocaudal pattern of dMN and vMN generation in the hindbrain and spinal cord. </p><p>Other recent studies have revealed a determinative role of <italic>Hox</italic> genes in MN subtype specification. In the hindbrain, Hoxb1 is expressed throughout r4, and has been shown to be required for the specification of facial branchiomotor neurons [ <xref rid="pbio-0040252-b017" ref-type="bibr">17</xref>– <xref rid="pbio-0040252-b019" ref-type="bibr">19</xref>]. Similarly, targeted expression of Hoxa3 in the rHB leads to the generation of ectopic somatic MNs [ <xref rid="pbio-0040252-b011" ref-type="bibr">11</xref>]. In the spinal cord, the rostrocaudal profile of genes of the Hox6 to Hox9 paralog group have been shown to establish distinctions in MN columnar and pool subtype [ <xref rid="pbio-0040252-b007" ref-type="bibr">7</xref>, <xref rid="pbio-0040252-b016" ref-type="bibr">16</xref>]. It seems likely, therefore, that the initial profiles of <italic>Hox</italic> expression, shown here to depend on early Wnt signaling, are involved in establishing domains of dMN and vMN formation. Furthermore, <italic>Hox</italic> genes appear to represent a common regulatory target for the three classes of signaling factors—Wnts, FGFs, and RA [ <xref rid="pbio-0040252-b090" ref-type="bibr">90</xref>]—that conspire to regulate the position of generation of dMNs and vMNs. </p><p>Thus, our results reveal that an early Wnt-based program is required to interact with a later RA- and FGF-mediated mechanism to generate a pattern of neural progenitor cells with <italic>Cdx</italic> and <italic>Hox</italic> profiles that prefigures the generation of two major subclasses of MNs in the developing hindbrain and spinal cord. Further studies will reveal how these three signals are integrated at the molecular level to regulate <italic>Cdx</italic> and <italic>Hox</italic> gene profiles, leading to the subsequent differentiation of dMN and vMN classes. </p></sec></sec><sec id="s4"><title>Materials and Methods</title><sec id="s4a"><title>Embryos.</title><p>Fertilized white leghorn chicken eggs were obtained from Agrisera AB, Umeå, Sweden. Chick embryos were staged according to Hamburger and Hamilton (HH) [ <xref rid="pbio-0040252-b091" ref-type="bibr">91</xref>]. </p></sec><sec id="s4b"><title>Isolation of tissue explants.</title><p>Prospective neural plate explants were isolated from HH stage 4, 5, 6, 7 (1-somite), and 8 (3–4-somites) chick embryos. For stage 5–8 embryos, Dispase I (Roche Diagnostics) was used to facilitate removal of the underlying mesoderm.</p></sec><sec id="s4c"><title>Culture of tissue explants.</title><p>Explants were cultured in vitro as previously described [ <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>]. To enable tracing the rostrocaudal orientation of the dissected explants, rostrocaudally asymmetrical explants were isolated and then placed in a defined orientation in collagen matrix, where their orientation was maintained during cultivation, fixation, and cryo-sectioning (see <xref ref-type="fig" rid="pbio-0040252-g005">Figure 5</xref>A); or the caudal margin of the explants was labeled with DiI crystals (Molecular Probes) during the dissection process ( <xref ref-type="fig" rid="pbio-0040252-g005">Figures 5</xref>D, <xref ref-type="fig" rid="pbio-0040252-g007">7</xref>B, <xref ref-type="fig" rid="pbio-0040252-g007">7</xref>C, and 7D). Recombinant human FGF4 (R & D Systems, Minneapolis, Minnesota, United States) was used at 30 and 60 ng/ml. The FGF receptor inhibitor SU5402 (Calbiochem, EMD Biosciences, San Diego, California, United States) was used at 3 and 5 μM. All- <italic>trans</italic> retinoic acid (RA) (Sigma-Aldrich, St. Louis, Missouri, United States) was used at 10–40 nM. Purified recombinant mouse Wnt3A (R & D Systems) was used at 150 ng/ml. Soluble Wnt3A and control-conditioned media [ <xref rid="pbio-0040252-b092" ref-type="bibr">92</xref>] were obtained as described [ <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>] and used at 50–100 μl/ml, which for Wnt3A conditioned medium, mimicked the activity 75–150 ng/ml of Wnt3A protein (R & D Systems). Soluble mouse Frizzled 8 (mFrz8CRD-IgG) [ <xref rid="pbio-0040252-b050" ref-type="bibr">50</xref>] and control-conditioned medium were generated as described [ <xref rid="pbio-0040252-b032" ref-type="bibr">32</xref>] and used at 300–500 μl/ml of culture medium. Explants cultured with control-conditioned medium behaved like explants cultured alone. In MN differentiation studies, Shh-N (R & D Systems) was used at 15 nM, and cultivated explants were washed to remove Wnt, RA, and/or FGF4 from the medium before addition of Shh-N. </p></sec><sec id="s4d"><title>In situ hybridization, immunohistochemistry, and quantification of cells.</title><p>Chick <italic>Hoxb4, Hoxb8, Hoxc9, CdxC, CdxB, Tbx20,</italic> and <italic>Mox1</italic> (see <xref ref-type="supplementary-material" rid="sg003">Figure S3</xref>, [ <xref rid="pbio-0040252-b055" ref-type="bibr">55</xref>]) and <italic>Brachyur</italic>y ( <italic>Bra)</italic> (see <xref ref-type="supplementary-material" rid="sg003">Figure S3</xref>, [ <xref rid="pbio-0040252-b056" ref-type="bibr">56</xref>]) expression was detected using in situ RNA hybridization on consecutive cryo-sections using digoxigenin-labeled probes and was carried out essentially as described [ <xref rid="pbio-0040252-b093" ref-type="bibr">93</xref>], before being labeled with DAPI for quantification. Hoxc9 protein and all further markers were detected using fluorescent immunohistochemistry. Neural tissue was detected with rabbit anti-Sox1 antiserum, kindly provided by S. Wilson. Rabbit anti-Otx2 was kindly provided by G. Corte. Rabbit anti-Krox20 was obtained from Babco. Rabbit anti-Isl1/2, mouse anti-Hb9, and rabbit anti-Hoxc9 were used as described [ <xref rid="pbio-0040252-b008" ref-type="bibr">8</xref>, <xref rid="pbio-0040252-b094" ref-type="bibr">94</xref>]. Co-localization of <italic>Tbx20</italic>/Isl was determined by double labeling with RNA probe and antibody, and co-localization of Hb9/Isl and Hoxc9/Hb9 was determined by double labeling with antibodies. Images of consecutive sections were collected using a Nikon (Tokyo, Japan) E800 light/epi-flourescent microscope. Images presented are representative of the number of explants indicated, for each experiment, in the figure legends. For each experiment, all explants were sectioned, and cells from consecutive sections were counted from 3–5 representative explants using nuclear Dapi staining. Each explant usually generated ˜10–18 8-μm sections, and the cell counts in <xref ref-type="fig" rid="pbio-0040252-g003">Figures 3</xref>, <xref ref-type="fig" rid="pbio-0040252-g007">7</xref>, and 9 indicate the mean percentage of total cell number ± s.e.m. per section in <italic>Hox</italic>-gene <sup>+</sup> domains that were compared by overlay of consecutive sections. The number of Isl1/2 <sup>+</sup> cells, <italic>Tbx20 <sup>+</sup></italic>/ Isl1/2 <sup>+</sup> cells, Hb9 <sup>+</sup>/ Isl1/2 <sup>+</sup> cells, and Hb9 <sup>+</sup>/ Hoxc9 <sup>+</sup> cells in each experiment was acquired from two to four explants (two sections from each explant) as described above ( <xref ref-type="fig" rid="pbio-0040252-g004">Figures 4</xref>, <xref ref-type="fig" rid="pbio-0040252-g006">6</xref>, and 10). </p></sec></sec><sec sec-type="supplementary-material" id="s5"><title>Supporting Information</title><supplementary-material content-type="local-data" id="sg001"><label>Figure S1</label><caption><title>Wnt3A or FGF4 Alone Does Not Induce More Caudal Character in Prospective rHB Cells</title><p>(A) Schematic drawing of a stage 4 embryo. Dotted line indicates the presumptive neural plate. Boxed region indicates caudal (C) neural plate explants used for in vitro studies.</p><p>(B–D) Stage 4 C explants cultured alone (B), in the presence of FGF4 (120 ng/ml) ( <italic>n</italic> = 18 explants) (C), or Wnt3A (150 ng/ml) ( <italic>n</italic> = 9 explants) (D) for 44 h generated a domain of Krox20 <sup>+</sup> cells (30%–60% of total cell number) and a domain of <italic>Hoxb4</italic> <sup>+</sup> <italic>/Hoxb8</italic> <sup>+</sup> <italic>/Hoxc9</italic> <sup>+</sup> cells (40%–70% of the total cell number). Each row represents consecutive sections from a single explant. Scale bar represents 100 μm. </p><p>(3.2 MB TIF)</p></caption><media xlink:href="pbio.0040252.sg001.tif"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sg002"><label>Figure S2</label><caption><title>Wnt, FGF or RA Signals Alone Do Not Induce More Caudal Character in Prospective FB Cells</title><p>(A) Schematic drawing of a stage 4 embryo. Dotted line indicates the presumptive neural plate. Red box indicates prospective FB explants that were used for in vitro studies. (B–F) Sox1 was used as a general neural marker. Each row represents consecutive sections from a single explant.</p><p>(B–F) Stage 4 FB explants cultured alone ( <italic>n</italic> = 24 explants) (B), or in the presence of Wnt3A (75 ng/ml) ( <italic>n</italic> = 12 explants) (C), FGF4 (60 ng/ml) ( <italic>n</italic> = 9 explants) (D), RA (10 nM) ( <italic>n</italic> = 17 explants) (E), or FGF (60 ng/ml) and RA (10 nM) in combination ( <italic>n</italic> = 12 explants) (F) for 44 h generated Sox1 <sup>+</sup>/ Otx2 <sup>+</sup> cells (75%–98% of total cell number), but no caudal neural cells. Scale bar represents 100 μm. </p><p>(3.1 MB TIF)</p></caption><media xlink:href="pbio.0040252.sg002.tif"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sg003"><label>Figure S3</label><caption><title>Wnt3A, FGF4, and RA in Combination Do Not Block Neural, or Induce Mesodermal Character, in Prospective FB or Caudal Neural Cells</title><p>(A) Schematic drawing of a stage 4 chick embryo. Dotted line indicates the presumptive neural plate. Red box indicates FB explants and black box indicates caudal neural plate tissue explants (4C) that were used for in vitro studies.</p><p>(B–D) All explants were cultured for 24 h (˜ corresponding to a stage 10 embryo)</p><p>(B) Stage 4 FB explants exposed to Wnt3A (150 ng/ml), FGF4 (20ng/ml), and RA (10 nM) generated Sox1 <sup>+</sup> and Sox2 <sup>+</sup> cells, but no <italic>Mox1-</italic> or <italic>Bra-</italic>expressing cells. </p><p>(C) Stage 4 FB explants exposed to Wnt3A (150 ng/ml) and FGF4 (60 ng/ml) generated Sox1 <sup>+</sup> and Sox2 <sup>+</sup> cells and few <italic>Bra</italic> <sup>+</sup> cells, but no <italic>Mox1-</italic>expressing cells. </p><p>(D) Stage 4C explants grown in the presence of Wnt3A (˜75 ng/ml) and FGF4 (60 ng/ml), in combination, generated Sox1 <sup>+</sup> and Sox2 <sup>+</sup> cells and few <italic>Bra</italic> <sup>+</sup> cells, but no <italic>Mox1-</italic>expressing cells. </p><p>(E–G) Transversal sections of a stage 10 embryo.</p><p>(E) Schematic drawing of a stage 10 embryo. The lines indicate the level of the transverse sections shown in the corresponding panels (F and G).</p><p>(F) Sox1 and Sox2 are expressed in the neural tube. <italic>Mox1</italic> is expressed in the adjacent somites (s), whereas <italic>Bra</italic> is expressed in the notochord (NC). </p><p>(G) Sox1 and Sox2 are expressed in the neural plate (NP). <italic>Mox1</italic> is expressed in the presomitic mesoderm (PSM) at the neural plate level and <italic>Brachyury</italic> is expressed in the neural plate and in the mesendoderm. </p><p>(3.1 MB TIF)</p></caption><media xlink:href="pbio.0040252.sg003.tif"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material></sec>
Accelerated Growth following Poor Early Nutrition Impairs Later Learning	Could not extract abstract	<contrib contrib-type="author"><name><surname>Gross</surname><given-names>Liza</given-names></name></contrib>	PLoS Biology	<p>The conditions an organism experiences early in life can have critical impacts on its subsequent health and well being, both over the short and long term. Aside from facing a greater risk of death, low birth weight human babies (5.5 pounds and under) have increased risk of developmental disabilities throughout life. Recent evidence indicates that many organisms can offset some of the changes associated with early poor nutrition by modifying their physical development. For example, poorly nourished children can undergo a period of accelerated growth once their diet improves, ultimately appearing normal as an adult.</p><p>But such compensatory measures often come at a price, with cognitive or other developmental disabilities emerging later in life, suggesting that growth rates are optimized to avoid such costs. Poor nutrition early in life can impair neural development, leading to lower IQ in humans and flawed song learning in birds. A recent study found that full-term, low birth weight babies who grew quickly when fed an enriched diet had lower cognitive skills when tested at nine months than did babies given a normal diet. But questions remain about the relative consequences of compensatory growth versus impaired growth and poor nutrition on the observed cognitive defects.</p><p>In a new study, Michael Fisher, Rudolph Nager, and Pat Monaghan explored the connection between early poor nutrition, compensatory growth, and learning ability in adulthood. To circumvent the confounding variables inherent in human studies and to control for genetic effects, the researchers compared the learning performance of zebra finch siblings reared on different quality diets after hatching. Only food quality, not quantity, was changed. The rate at which adult birds could learn a simple task, they found, depended on the rate of compensatory growth the birds showed following a period on lower-quality food early in life—not on the diet itself or on the degree of stunted growth.</p><p>Because zebra finch hatchlings are totally dependent on their parents for food, the researchers could vary food quality for the chicks by manipulating the food available. The birds, facile learners that are often used in cognitive studies, were reared on ad libitum diets of different nutritional quality and then tested for learning performance as adults. (Handily, zebra finches show little sex differences in size as adults.) After hatching, siblings were raised on either a normal or low-quality diet for 20 days, then switched to the higher-quality standard diet. While on the low-quality diet, birds grew slower and were lighter than their control siblings by the end of the 20 days. Once they were switched to the standard diet, birds reared on the poor diet then grew significantly more than their normally fed siblings and reached the same adult size.</p><p>The extent to which birds’ growth was depressed during the poor nutrition phase of the experiment varied considerably, as did the degree of accelerated growth after the switch to a normal diet. As it happened, birds with the most stunted growth (relative to their control siblings) and those with the most accelerated growth (after switching diets) fell into different groups, allowing the researchers to distinguish cognitive effects associated with stunted growth from those associated with compensatory growth.</p><p>To test the adult birds’ learning performance, the researchers tested them on an associative learning task. Birds were placed in a circular foraging area with corridors leading to a screen with cups of seed behind it, and were trained to associate a yellow screen with food. Though all the birds eventually learned the task, their learning rate depended on the rate of compensatory growth they had undergone as chicks. Undernourished birds that had grown fastest after switching to the normal diet performed poorest on the learning task compared to their control siblings. Since the undernourished birds were the only group that showed this relationship between growth rate and learning speed, the researchers concluded that it is the compensatory growth following reduced nutrition that accounts for poor learning performance in adulthood.</p><p>These results suggest that poor early nutrition can have long-lasting negative consequences for cognitive ability—for finches as well as humans, given similar findings in human infants. While it’s unclear whether the learning defects stem from behavioral, hormonal, or neural changes, it’s likely that resources normally dedicated to these pathways are diverted to support accelerated growth, shortchanging the co-opted pathway. Future study is needed to identify the underlying causes of impaired learning speed, an essential step in determining how to manage growth and nutrition for low birth weight babies and avoid the costs associated with compensatory growth.</p><sec><title/><fig id="pbio-0040270-g001" position="float"><caption><p>Adult zebra finches who grew fastest as chicks after a nutrition-poor diet had trouble learning an association task.</p></caption><graphic xlink:href="pbio.0040270.g001"/></fig></sec>
Compensatory Growth Impairs Adult Cognitive Performance	<p>Several studies have demonstrated that poor early nutrition, followed by growth compensation, can have negative consequences later in life. However, it remains unclear whether this is attributable to the nutritional deficit itself or a cost of compensatory growth. This distinction is important to our understanding both of the proximate and ultimate factors that shape growth trajectories and of how best to manage growth in our own and other species following low birth weight. We reared sibling pairs of zebra finches on different quality nutrition for the first 20 d of life only and examined their learning performance in adulthood. Final body size was not affected. However, the speed of learning a simple task in adulthood, which involved associating a screen colour with the presence of a food reward, was negatively related to the amount of growth compensation that had occurred. Learning speed was not related to the early diet itself or the amount of early growth depression. These results show that the level of compensatory growth that occurs following a period of poor nutrition is associated with long-term negative consequences for cognitive function and suggest that a growth-performance trade-off may determine optimal growth trajectories.</p>	<contrib contrib-type="author"><name><surname>Fisher</surname><given-names>Michael O</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Nager</surname><given-names>Ruedi G</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author" corresp="yes"><name><surname>Monaghan</surname><given-names>Pat</given-names></name><email>[email protected]</email></contrib>	PLoS Biology	<sec id="s1"><title>Introduction</title><p>It has recently been recognised that many organisms are capable of adjusting their phenotypic development to mitigate at least some of the potentially negative effects of poor early nutrition [ <xref rid="pbio-0040251-b001" ref-type="bibr">1</xref>, <xref rid="pbio-0040251-b002" ref-type="bibr">2</xref>]. While it has also been recognised that these adjustments may themselves carry costs, possibly not evident until later in life, the processes responsible are poorly understood [ <xref rid="pbio-0040251-b003" ref-type="bibr">3</xref>– <xref rid="pbio-0040251-b006" ref-type="bibr">6</xref>]. </p><p>A widespread example of such a mitigating response is “catch-up” or “compensatory” growth. Following an episode of poor nutrition during which growth is depressed, individuals can then accelerate or prolong their growth such that their external morphology at adulthood is indistinguishable from that of nonaffected individuals [ <xref rid="pbio-0040251-b004" ref-type="bibr">4</xref>, <xref rid="pbio-0040251-b007" ref-type="bibr">7</xref>, <xref rid="pbio-0040251-b008" ref-type="bibr">8</xref>]. That many organisms have the capacity to grow faster than they generally do suggests that growth rate is optimised rather than maximised, varying in response to the balance of associated fitness costs and benefits [ <xref rid="pbio-0040251-b004" ref-type="bibr">4</xref>, <xref rid="pbio-0040251-b005" ref-type="bibr">5</xref>, <xref rid="pbio-0040251-b007" ref-type="bibr">7</xref>, <xref rid="pbio-0040251-b009" ref-type="bibr">9</xref>]. While offsetting disadvantages resulting from small adult size, compensatory growth might carry costs operational over varying time scales [ <xref rid="pbio-0040251-b010" ref-type="bibr">10</xref>– <xref rid="pbio-0040251-b013" ref-type="bibr">13</xref>] but nonetheless be favoured given the net fitness benefits relative to no compensation [ <xref rid="pbio-0040251-b004" ref-type="bibr">4</xref>]. However, we still know very little about how such costs may occur, and hence about the selection pressures that shape optimal growth rates. </p><p>One important route through which early nutrition could profoundly influence subsequent performance is by effects on neural development and hence on cognitive ability [ <xref rid="pbio-0040251-b014" ref-type="bibr">14</xref>]. Impairment of cognitive performance can obviously have pervasive effects on many important individual attributes, such as foraging ability, antipredator behaviour, mate acquisition, parenting, and the maintenance of physiological homeostasis through appropriate behavioural responses to changing environmental circumstances. Several studies in birds and mammals have provided evidence that poor nutrition in early life can adversely influence neural development, with, in the case of birds, adverse consequences for song learning [ <xref rid="pbio-0040251-b015" ref-type="bibr">15</xref>– <xref rid="pbio-0040251-b019" ref-type="bibr">19</xref>] and, in humans, subsequent intelligence quotients [ <xref rid="pbio-0040251-b020" ref-type="bibr">20</xref>– <xref rid="pbio-0040251-b024" ref-type="bibr">24</xref>]. However, the role played by the level of compensatory growth that occurs following the episode of poor nutrition has been little studied. In humans, it has recently been found that full-term babies born at low birth weight, and in whom subsequent compensatory growth was induced through provision of higher-quality feed (protein, mineral, and vitamin enriched), showed reduced cognitive performance at the end of the enrichment period (9 mo of age) relative to a control group not receiving such enrichment [ <xref rid="pbio-0040251-b025" ref-type="bibr">25</xref>]. Moreover, the group most negatively affected was that which showed the greatest compensatory growth, in this case, girls [ <xref rid="pbio-0040251-b025" ref-type="bibr">25</xref>, <xref rid="pbio-0040251-b026" ref-type="bibr">26</xref>]. This suggests that the level of compensatory growth that occurs following nutritional deficit may be very important in determining aspects of early cognitive performance. </p><p>In examining the links between compensatory growth and cognitive abilities, it can be difficult to control confounding variables both during and after the manipulation period. Even within an experimental treatment, the changes in growth trajectories that occur may vary between subjects, and it is thus important to examine growth responses at the individual level and relate these to individual cognitive performance. Furthermore, it can be difficult to follow effects of growth on performance in the long term, especially in humans. Altricial birds offer a useful system in this context since they hatch in a relatively embryonic state; food is then provided by the parents without further processing, making it possible to directly manipulate the quality of nutrition available. The growth period is clearly defined and relatively short, and birds generally have good learning abilities and are widely used in learning studies.</p><p>Using same-sex sibling zebra finches <named-content content-type="genus-species">Taeniopygia guttata</named-content> exposed to different nutritional environments during the early posthatching period, we measured the degree of growth depression that occurred during an episode of nutritional deficit in early life and the degree of compensatory growth when normal nutrition was restored. We then tested whether subsequent learning performance in adulthood was linked to variation in these early growth trajectories. </p></sec><sec id="s2"><title>Results</title><sec id="s2a"><title>Early Nutrition and Growth</title><p>The zebra finch is a size-monomorphic species [ <xref rid="pbio-0040251-b027" ref-type="bibr">27</xref>], and there were no sex differences in growth rates during either the deficit (day 0 to 20, <italic>t</italic>-test: <italic>t</italic> <sub>32</sub> = 0.81, <italic>p</italic> = 0.42) or postdeficit periods (day 20 to 70, <italic>t</italic> <sub>32</sub> = 1.29, <italic>p</italic> = 0.21). We therefore pooled the pairs of brothers ( <italic>n</italic> = 7) and pairs of sisters ( <italic>n</italic> = 10) for analysis. <xref ref-type="fig" rid="pbio-0040251-g001">Figure 1</xref> shows the average difference in the mass of the sib receiving a normal diet (N siblings) and that of its same-sex sib experiencing an early nutritional deficit (D siblings) before, during, and after the diet manipulation. Siblings allocated to the two early nutrition treatments did not differ in hatching mass (D birds = 1.0 ± 0.03 g; N birds = 1.0 ± 0.04 g; paired <italic>t</italic>-test: <italic>t</italic> <sub>16</sub> = 1.16, <italic>p</italic> = 0.26). Nutritional deficit slowed the nestling growth of D birds relative to their N siblings so that, by the end of the diet manipulation period (20 d posthatching), offspring in the D group were on average significantly lighter (11.6 ± 0.40 g) than those in the N group (13.4 ± 0.26 g; paired <italic>t</italic>-test comparing siblings: <italic>t</italic> <sub>16</sub> = 5.61, <italic>p</italic> < 0.001; <xref ref-type="fig" rid="pbio-0040251-g001">Figure 1</xref>). In the postdeficit period (day 20 to 70 posthatching, all birds now on standard nutrition), most of the D birds then grew more than their N sibling; during this period D birds gained an additional 46.50 ± 4.78% of their day 20 body mass, compared to a 29.49 ± 2.91% average gain in the N birds (paired <italic>t</italic> <sub>16</sub> = 3.07, <italic>p</italic> = 0.01). By day 70 posthatching, the birds were effectively fully grown. D birds (16.8 ± 0.46 g) were no longer significantly lighter than their N siblings (17.4 ± 0.52 g; paired <italic>t</italic>-test: <italic>t</italic> <sub>16</sub> = 1.23, <italic>p</italic> = 0.24; <xref ref-type="fig" rid="pbio-0040251-g001">Figure 1</xref>). Body mass changed very little thereafter; when the birds were weighed again following attainment of sexual maturity (around 100 d), body mass had changed on average 3% in the D birds and 5% in the N birds, and this change did not differ between them ( <italic>p</italic> = 0.28), and probably represents fat deposition rather than tissue growth. </p><fig id="pbio-0040251-g001" position="float"><label>Figure 1</label><caption><title>The Mean Difference in Body Mass between Pairs of Same-Sex Siblings at Different Stages of Development</title><p>The differences were calculated as the body mass of the D sib, which experienced the early nutritional deficit minus that of the N sib reared under normal conditions, and are expressed as a percentage of the N sib's body mass. Thus negative values indicate that the D sib is lighter than its N sib.</p></caption><graphic xlink:href="pbio.0040251.g001"/></fig><p>However, there was substantial variation in the extent to which growth slowed in response to the dietary deficit regimen and in the degree of catch-up growth that occurred following resumption of normal feeding (coefficient of variation of within sibling pair differences in instantaneous growth rate between 0 and 20 d = 75.5% and between 20 and 70 d = 90.6%). Furthermore, the D birds that showed the strongest reductions in growth during the deficit period, relative to their N siblings, were not the same birds that produced the most extreme compensatory responses once the deficit period ended (correlation between sib differences in day 0 to 20 instantaneous growth rate and day 20 to 70 instantaneous growth rate: <italic>r</italic> <sub>15</sub> = −0.31, <italic>p</italic> = 0.23). This allows us to separate the effects on cognitive abilities of the degree of slowing of growth during the period of nutritional deficit from effects associated with the degree of later compensatory growth. </p></sec><sec id="s2b"><title>Early Nutrition and Adult Learning Ability</title><p>All birds reduced the number of errors they made over the course of the 16 associative learning trials, and thus all birds seemed capable of eventually learning the task; this was not influenced by biological or foster parent or early diet ( <xref ref-type="fig" rid="pbio-0040251-g002">Figure 2</xref>). To examine differences in the speed of learning of siblings in relation to growth pattern, we used the sibling difference in the number of trials it took them to achieve two trials without error and sibling differences in growth. The difference in learning speed between siblings was strongly related to the extent to which growth compensation occurred and not related to the magnitude of growth depression during the deficit period (multiple regression: sibling differences in growth from day 20 to 70: <italic>F</italic> <sub>1,14</sub> = 8.03, <italic>p</italic> = 0.01; sibling differences in growth from day 0 to 20: <italic>F</italic> <sub>1,14</sub> = 0.02, <italic>p</italic> = 0.90, <xref ref-type="fig" rid="pbio-0040251-g003">Figure 3</xref>A and <xref ref-type="fig" rid="pbio-0040251-g003">3</xref>B). The D birds that showed a strong compensatory growth response relative to their N sibling performed relatively poorly on the associative learning task ( <xref ref-type="fig" rid="pbio-0040251-g003">Figure 3</xref>B). Learning speed was not influenced by early diet itself (paired <italic>t</italic> <sub>16</sub> = 1.03, <italic>p</italic> = 0.32) or by body mass at testing ( <italic>r</italic> <sub>32</sub> = 0.26, <italic>p =</italic> 0.14). If the sib groups are considered separately, then in neither group is there any significant relationship between learning speed and growth rate during the deficit period (N sibs <italic>p</italic> = 0.17, D sibs <italic>p</italic> = 0.68); only in the D group was there a relationship between learning speed and growth rate during the compensatory period (N sibs r <sub>15</sub> = 0.02, <italic>p</italic> = 0.95, D sibs r <sub>15</sub> = 0.62, <italic>p</italic> = 0.008). Thus, the data clearly demonstrate that those birds that grew relatively fast when normal nutrition was restored following a period of early deficit had poor learning performance in adulthood. </p><fig id="pbio-0040251-g002" position="float"><label>Figure 2</label><caption><title>The Mean (± SE) Number of Errors per Trial Made on the Associative Learning Task by Experimental Birds (see <xref ref-type="sec" rid="s4">Materials and Methods</xref>) </title><p>All birds (i.e., both sexes and birds from both early nutrition treatment groups) are included ( <italic>n</italic> = 34). The dotted line indicates the number of errors expected by chance. The dashed line denotes a 24-h retention interval (i.e., trial 9 was the first trial of day 2). Sex, early diet, and family of origin did not influence the pattern of decline in errors with trial number (mixed model using a binomial error distribution where trial number was a repeated measure within bird identity [random effect]: trial number: <italic>z</italic> = 14.07, <italic>p</italic> < 0.0001; bird identity: <italic>z</italic> = 0.13, <italic>p</italic> = 0.45; family of origin [random factor]: <italic>z</italic> = 1.12, <italic>p</italic> = 0.13; sex: <italic>F</italic> <sub>1,13.4</sub> = 0.05, <italic>p</italic> = 0.82; early diet: <italic>F</italic> <sub>1,13.7</sub> = 0.74, <italic>p</italic> = 0.40), nor did the foster family in which the experimental birds were reared affect learning performance ( <italic>F</italic> <sub>18, 15</sub> = 0.56, <italic>p</italic> = 0.88). </p></caption><graphic xlink:href="pbio.0040251.g002"/></fig><fig id="pbio-0040251-g003" position="float"><label>Figure 3</label><caption><title>Relationship between Compensatory Growth following a Period of Low-Quality Early Nutrition and Associative Learning Performance in Adulthood</title><p>Within-sibling differences (D sib – N sib) in the growth increment from (A) day 0 to 20 and (B) day 20 to 70 and learning performance are plotted so that each data point represents a sibling pair. Positive values for growth indicate that the D sib grew faster during the postdeficit period than its same-sex N sib that had experienced normal nutrition throughout growth. The learning performance was scored on the basis of the number of trials it took them to achieve two trials without error. Birds that did not reach this criterion obtained a score of zero and the fastest learners obtained the highest score. The figure shows the difference in score between the sibling pairs, and negative values mean that the D sib had a poorer learning performance than its N sib. There was no correlation between the intersib difference in growth during the deficit period (0 to 20 d) and learning performance in adulthood (A: <italic>r</italic> <sub>15</sub> = 0.10, <italic>p</italic> = 0.70); however, the faster the D sib grew relative to its N sib when normal nutrition was restored, the poorer was its learning performance in adulthood (B: <italic>r</italic> <sub>15</sub> = 0.63, <italic>p</italic> = 0.007). In this relationship, no point had a large influence (Cook's distance for all points <0.4). </p></caption><graphic xlink:href="pbio.0040251.g003"/></fig></sec></sec><sec id="s3"><title>Discussion</title><p>Episodes of poor early nutrition have been linked to impairment of performance in adulthood in a number of taxa [ <xref rid="pbio-0040251-b006" ref-type="bibr">6</xref>, <xref rid="pbio-0040251-b028" ref-type="bibr">28</xref>– <xref rid="pbio-0040251-b031" ref-type="bibr">31</xref>]. The extent to which such impairments are associated with a shortage of nutrients and consequent slowing of growth during crucial stages of development, or a negative consequence of the compensatory growth that often follows the period of poor nutrition, is unclear. It is generally very difficult to separate the two, since the nutritional deficit induces compensatory growth if normal nutrition is restored and thus the two are closely linked. However, distinguishing between the two is important if we are to understand what determines optimal growth rate. Individuals do vary in their responses to particular developmental conditions, such that, even within a particular experimental treatment, both the magnitude of growth depression and growth compensation can vary amongst individuals, possibly due to genetic or other developmental differences [ <xref rid="pbio-0040251-b004" ref-type="bibr">4</xref>, <xref rid="pbio-0040251-b005" ref-type="bibr">5</xref>]. In our experiment, we manipulated early posthatching nutrition in zebra finches, controlling for genetic effects by using siblings, and monitored the growth responses of individuals both during the deficit and after the restoration of normal levels of nutrition. Those individuals in the deficit group that showed the greatest growth depression relative to their genetic sibling were not those that showed the most growth compensation. This allowed us to separate to some extent the magnitude of the early deficit from the degree of compensatory growth that then occurred. Furthermore, we also standardised the environment and level of nutrition the birds experienced until testing at adulthood. Our data on zebra finches show that it is the magnitude of growth compensation in response to the nutritional deficit, rather than the early diet itself or the degree of growth depression during the deficit, that is related to subsequent learning performance in adulthood. </p><p>Another possibility is that our results are a consequence of some underlying relationship between activity levels and cognitive performance, in that if the slower learners are generally less active than brighter birds, the former may put on more mass than the latter. In our experimental conditions, there is in fact rather little scope for substantial differences in activity levels amongst the young birds, since they are bred in nest boxes in cages and fed by their parents. However, more important, our data suggest that this explanation is unlikely. If there was a general relationship between activity levels (and thereby mass gain) and cognitive ability, we would then expect to see this relationship throughout the growth period, i.e., also during days 0 to 20. However, during this period there was no relationship between growth rate and adult learning performance when examined in terms of sib differences, or in the D or in the N birds when these are considered separately. We would also expect to see this general relationship within N as well as the D birds in the later growth period, days 20 to 70. This is not so—the relationship between growth and learning performance only occurs in those birds that experienced an early nutritional deficit and were then undergoing compensatory growth to differing degrees.</p><p>What we have found is that the more compensatory growth following an early deficit, the poorer is the learning performance in adulthood. This is in accordance with effects reported over a relatively short-term study of human infants involving similar relatively small differences in the quality of the food provided (1.45% protein formula milk feed compared with feed containing 1.85% protein plus enriched vitamin and mineral content) [ <xref rid="pbio-0040251-b025" ref-type="bibr">25</xref>]. In Morley et al.'s human study, the difference in cognitive performance recorded in babies at 9 mo (the end of the early nutrition treatment period) was not detected at 18 mo. In our study, effects of compensatory growth on cognitive performance, measured as the speed of learning of a relatively simple task, were detectable in adulthood. The effects in human offspring may be more transient, be more difficult to measure at later stages, or have been offset by environmental differences. It is possible, for example, that mothers with children whose cognitive development seems slow may provide more stimulation that has a beneficial effect. It is obviously difficult to control this experimentally. </p><p>While it has been suggested that development of the central nervous system is prioritised during development [ <xref rid="pbio-0040251-b032" ref-type="bibr">32</xref>], it is not entirely buffered from nutritional effects. Mammalian experimental studies have shown that nutritional conditions during growth and development can have long-term effects on brain structure and function [ <xref rid="pbio-0040251-b015" ref-type="bibr">15</xref>, <xref rid="pbio-0040251-b021" ref-type="bibr">21</xref>]. In birds, early nutrition has been found to affect the development of song centres in the male brain [ <xref rid="pbio-0040251-b018" ref-type="bibr">18</xref>, <xref rid="pbio-0040251-b019" ref-type="bibr">19</xref>]. However, even though considerable neurogenesis is known to occur in non-song brain structures during juvenile and adult life [ <xref rid="pbio-0040251-b033" ref-type="bibr">33</xref>], no previous studies have experimentally examined the impact of early growth on general cognitive ability in birds. This study suggests that compensatory growth can have long-lasting negative effects on cognitive performance. There are a number of ways in which such effects could come about. In this study, it was the speed of learning in particular that was affected, and all birds did eventually learn the task. This slow learning could involve behavioural, endocrinological, and neural changes. A diversion of resources from neural development to fuel growth of other body parts may occur and/or there may be a shift in investment in tissue growth at the expense of maturation as a consequence of a growth/maturation trade-off [ <xref rid="pbio-0040251-b034" ref-type="bibr">34</xref>, <xref rid="pbio-0040251-b035" ref-type="bibr">35</xref>]. Prolonged stress during a period of compensatory growth may produce chronically elevated levels of corticosterone, which could influence behaviour; it is known to result in irreversible damage to the hippocampus, a brain region closely associated with learning and memory [ <xref rid="pbio-0040251-b036" ref-type="bibr">36</xref>]. Furthermore, chronically high levels of corticosterone could potentially have more widespread effects in the avian than the mammalian brain, given the substantial neurogenesis that can occur in adulthood in birds [ <xref rid="pbio-0040251-b037" ref-type="bibr">37</xref>]. The vertebrate brain also contains a large number of polyunsaturated fatty acids that are especially susceptible to free radical damage [ <xref rid="pbio-0040251-b038" ref-type="bibr">38</xref>], which may increase during rapid growth. Any such negative effects of growth pattern on cognitive performance are likely to have far reaching consequences for individual fitness. Further work is required in order to identify the cause of the slower learning speed, both at the behavioural level, since there are many different ways in which the slower association between screen colour and food reward could come about, and at the level of brain growth and function. Understanding how growth rate may influence organism function is an important area of research in both the context of understanding the evolution of growth rates and developing appropriate procedures for managing growth following low birth weight. </p></sec><sec id="s4"><title>Materials and Methods</title><sec id="s4a"><title>Early nutritional deficit.</title><p>In order to evaluate the effects of early growth, while allowing for genetic effects, we compared the cognitive performance of genetic full siblings of the same sex reared under different qualities of posthatching nutrition. Zebra finches with previous breeding experience were randomly paired and allowed to nest in individual breeding cages (60 cm × 45 cm × 40 cm). All birds were provided with the standard diet recommended for captive finches of ad libitum seed (J.E. Haith Ltd, Cleethorpes, United Kkingdom; foreign finch mixture, containing 11.6% protein); a vitamin and mineral supplement (Daily Essentials 2; The Birdcare Company, Nailsworth, United Kingdom), plus Haith's conditioning food (13.6% protein), was provided three times a week. Zebra finch parents carry food provided in the cage to the nestlings in the nest box for about the first 20 d posthatching, following which the chicks gradually begin to fledge, leaving the box and taking food in the cage themselves. It is thus not possible to control or measure directly the food fed to each chick. We therefore manipulated the quality of the food available to the chick-rearing parents. When the first chick in each nest hatched, nests were allocated randomly to the two early nutrition treatment groups, one of which involved provision of the high-quality rearing diet normally provided during chick rearing and the other, a lower-quality diet. Nests in the normal diet (N) group were provided with the ad libitum mixed seed, and the vitamin and conditioning food was provided daily, together with a homogenised egg mixture containing around 12% protein. Nests in the low-quality early diet treatment were provided only with ad libitum mixed seed and water (Early Deficit Group-D), and none of the higher-protein conditioning food or the vitamin and mineral supplement. These diets were maintained for the first 20 d posthatching. From then on, all nests were provided with the standard diet as described above for parents prior to chicks hatching. All birds were then maintained on this standard diet for the remainder of the experiment.</p><p>Since we did not know the sex of chicks at hatching, we ensured that pairs of same-sex siblings experienced different dietary regimens by using the following experimental protocol. On the day of hatching, each genetic sibling was fostered to a different nest, with sibs being fostered alternately to D nests and N nests. Fostering was arranged so that, after chicks had been moved, each foster nest contained three unrelated chicks. If it was later found that more than one same-sex sibling had been allocated to the same treatment ( <italic>n</italic> = 4 cases), one of these birds was chosen at random for use in the learning study. This rearing procedure therefore also controlled for any effects of fostering, since no chicks were reared by their genetic parents. Each chick was weighed on the day it hatched and again at 20 and 70 d posthatching; instantaneous growth rates (log mass <sub>2</sub> – log mass <sub>1</sub>/t <sub>2</sub> – t <sub>1</sub>) for the period of nutritional deficit (day 0 to 20) and the posttreatment period (day 20 to 70) were calculated. From 35 d posthatching onward, offspring can be sexed from plumage [ <xref rid="pbio-0040251-b027" ref-type="bibr">27</xref>], and they were then removed from their foster nest and placed in single-sex groups of four to six young birds, with same-sex siblings being in the same group. Seventeen pairs of same-sex siblings that had experienced different early nutrition were available for study in adulthood (ten sister-sister pairs, seven brother-brother pairs). </p></sec><sec id="s4b"><title>Learning task.</title><p>All 34 birds were given an associative learning task in adulthood (approximately 150 d old at testing). These trials were conducted in a circular foraging arena (height 40 cm, diameter 150 cm) under full spectrum lighting (bird lamp' Arcadia, Croydon, United Kingdom). The arena contained seven corridors leading from a central area. At the end of each corridor was a white screen behind which cups containing seed were placed. Initially, birds were given shaping sessions to ensure that they were familiar with the arena and the food cups. Once all birds were readily feeding from the cups, birds were then entered into trials in which they were required to learn an association between the colour of a screen and the presence of food. In this situation, only one of the screens, coloured yellow, now had food behind it; the other six screens were black. The location of the yellow screen was randomly altered between trials. Each bird was given eight consecutive trials per day, for two consecutive days, before the first feeding in the morning. Each trial consisted of a 10-min acclimation period (inside a pulley-operated mesh holding chamber in the centre of the arena) to enable the bird to recover from handling; this was followed by a search phase in the arena. During the search phase, the number of errors (i.e., the number of unrewarded screens visited) that a bird made prior to finding the rewarded tray was recorded. Once the bird located the rewarded tray, it was allowed to feed for 30 s and then removed from the arena. Following each feeding from behind the yellow screen, the bird was given a 10-min rest interval prior to the start of the next trial. The amount of feeding was such that birds remained hungry throughout the consecutive trials on each day, following which they were returned to their home cage and allowed to feed normally.</p></sec></sec>
The “Ultimate Prize” for Big Tobacco: Opening the Chinese Cigarette Market by Cigarette Smuggling	Could not extract abstract	<contrib contrib-type="author"><name><surname>Novotny</surname><given-names>Thomas E</given-names></name></contrib>	PLoS Medicine	<sec id="s1"><title>The Prize</title><p>China has long been identified as the most important international cigarette market, now and in the future—O'Sullivan and Chapman called China the “ultimate prize” among the world's emerging tobacco markets [ <xref rid="pmed-0030279-b1" ref-type="bibr">1</xref>]. Almost two-thirds of Chinese men (63%), and 3.8% of Chinese women, are smokers [ <xref rid="pmed-0030279-b2" ref-type="bibr">2</xref>], giving a total of 350 million smokers in China, with more on the way. Young Chinese women are likely to be an important target group for growth for Big Tobacco. </p><p>According to a 1998 survey, among Chinese adolescents aged 11–20 years, almost half (47.8%) of the boys and 12.8% of the girls surveyed were experimenting with tobacco [ <xref rid="pmed-0030279-b3" ref-type="bibr">3</xref>]. This survey would surely warm the cold heart of any multinational tobacco company executive. Robert Fletcher, regional public affairs manager for the tobacco company Rothman, said that “thinking about Chinese smoking statistics is like trying to think about the limits of space” [ <xref rid="pmed-0030279-b4" ref-type="bibr">4</xref>]. </p><p>China has been a member of the World Trade Organization since 2001, and this membership may reduce barriers to market entry through lowered tariffs on imports by transnational tobacco companies (TTCs) [ <xref rid="pmed-0030279-b5" ref-type="bibr">5</xref>]. Thus, it is important to understand how these markets will be entered by the TTCs, as market entry also means globalizing the imagery, advertising, and health effects of smoking [ <xref rid="pmed-0030279-b6" ref-type="bibr">6</xref>]. Market entry into China also means that with greatly increased numbers of smokers, the burden of non-communicable diseases attributable to smoking will explode, affecting China's transitional health-care system and its economic stability. </p></sec><sec id="s2"><title>Smuggling as a Marketing Tool</title><p>TTCs have used cigarette smuggling as a market-opening tool throughout recent history (such as in Latin America [ <xref rid="pmed-0030279-b7" ref-type="bibr">7</xref>] and the former Soviet Union [ <xref rid="pmed-0030279-b8" ref-type="bibr">8</xref>]). Now, in an important new study published in <italic>PLoS Medicine</italic> [ <xref rid="pmed-0030279-b9" ref-type="bibr">9</xref>], Lee and Collin document evidence that British American Tobacco (BAT) has exploited China's large cigarette smuggling problem. </p><p>Smuggling of cigarettes into China is an especially critical issue in light of China's desire both to open trade through membership in the World Trade Organization and to participate more fully as a global partner in health. With the 2002 SARS epidemic and the potential avian flu pandemic serving as grim reminders of the need for international collaboration in health, China now participates more openly in other health agreements. Specifically, China ratified the World Health Organization's first ever global health treaty, the Framework Convention on Tobacco Control (FCTC) [ <xref rid="pmed-0030279-b10" ref-type="bibr">10</xref>], on October 11, 2005. This treaty will have as one of its major areas of focus the control of global cigarette smuggling [ <xref rid="pmed-0030279-b11" ref-type="bibr">11</xref>]. Although the United States has not ratified the FCTC, it may join in a protocol that addresses smuggling. At present, one-quarter of total global cigarette exports are illegally traded. This illegal trade is a big, multinational problem, and multinational efforts are needed to control it; in addition, there is still insufficient scholarly research on this problem [ <xref rid="pmed-0030279-b12" ref-type="bibr">12</xref>]. </p><disp-quote><p>“Thinking about Chinese smoking statistics is like trying to think about the limits of space.”</p></disp-quote></sec><sec id="s3"><title>Getting Inside Information</title><p>Lee and Collin used BAT documents from the Guildford Depository in the United Kingdom. These documents were made public as a result of legal actions against the TTCs. The terms of a legal settlement in Minnesota, United States, stipulated that the public should be allowed, for ten years, access to documents produced during litigation against the tobacco industry, via the creation of two depositories, one in Minnesota and one in Guildford [ <xref rid="pmed-0030279-b13" ref-type="bibr">13</xref>]. </p><p>Document research is of specific value in examining the smuggling issue, since data on smuggling are not easily found in other public sources. The authors found suggestions of large-scale smuggling of foreign cigarettes into China to circumvent barriers to market access. By triangulating official documents on legal imports subject to tariffs, reported imports by all TTCs, and BAT sales figures in China, they found significant discrepancies between what was reported and what was sold. The sales reported in internal company documents made China one of the larger profit centers for BAT Industries, with 25% of profits coming from “transit” trade to China, a code word for smuggled cigarettes. Moreover, the Guildford documents show evidence of BAT's efforts to build market presence in competition with other popular brands, using whatever it takes to succeed in the new market, including smuggling.</p></sec><sec id="s4"><title>What Must Be Done</title><p>The FCTC came into force on February 27, 2005. The signatory member states now have the best-ever opportunity to shut down the pipeline of smuggled tobacco products and the associated illicit profits for BAT and other TTCs through multinational efforts to better track exports and enforce customs regulations. There are three major reasons why it is crucial to stop tobacco smuggling: (1) the profits from smuggling fuel TTCs' efforts to buy political influence and maintain economic strength across non-tobacco sectors; (2) smuggling leads to countries losing tariff revenue from the legal trade of cigarettes; and, most importantly, (3) the health of target groups such as young Chinese women will suffer immeasurably unless the FCTC is effective in reducing TTC market success.</p><p>Much can be done to combat smuggling as a tool of the TTCs, including implementing more aggressive litigation to assure corporate liability, pursuing anti-money-laundering actions to remove the influence of organized crime [ <xref rid="pmed-0030279-b14" ref-type="bibr">14</xref>], improving cigarette tracking systems, and strictly enforcing record-keeping of shipments and receipts by TTCs. There must be binding obligations to carry out such anti-smuggling activities, enforced through a multinational protocol of the FCTC, not voluntary compliance as proposed by the TTCs. The world is a global marketplace, and global problems such as cigarette smuggling need global solutions as put forth in the FCTC. </p></sec>
Insensitivity of Paediatric HIV-1 Subtype C Viruses to Broadly Neutralising Monoclonal Antibodies Raised against Subtype B	<sec id="st1"><title>Background</title><p>A Phase I clinical trial has been proposed that uses neutralising monoclonal antibodies (MAbs) as passive immunoprophylaxis to prevent mother-to-child transmission of HIV-1 in South Africa. To assess the suitability of such an approach, we determined the sensitivity of paediatric HIV-1 subtype C viruses to the broadly neutralising MAbs IgG1b12, 2G12, 2F5, and 4E10.</p></sec><sec id="st2"><title>Methods and Findings</title><p>The gp160 envelope genes from seven children with HIV-1 subtype C infection were cloned and used to construct Env-pseudotyped viruses that were tested in a single-cycle neutralisation assay. The epitopes defining three of these MAbs were determined from sequence analysis of the envelope genes. None of the seven HIV-1 subtype C pseudovirions was sensitive to 2G12 or 2F5, which correlated with the absence of crucial N-linked glycans that define the 2G12 epitope and substitutions of residues integral to the 2F5 epitope. Four viruses were sensitive to IgG1b12, and all seven viruses were sensitive to 4E10.</p></sec><sec id="st3"><title>Conclusions</title><p>Only 4E10 showed significant activity against HIV-1 subtype C isolates, while 2G12 and 2F5 MAbs were ineffective and IgG1b12 was partly effective. It is therefore recommended that 2G12 and 2F5 MAbs not be used for passive immunization experiments in southern Africa and other regions where HIV-1 subtype C viruses predominate.</p></sec>	<contrib contrib-type="author"><name><surname>Gray</surname><given-names>Elin Solomonovna</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Meyers</surname><given-names>Tammy</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Gray</surname><given-names>Glenda</given-names></name><xref ref-type="aff" rid="aff3"> <sup>3</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Montefiori</surname><given-names>David Charles</given-names></name><xref ref-type="aff" rid="aff4"> <sup>4</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Morris</surname><given-names>Lynn</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref><xref ref-type="corresp" rid="cor1">*</xref></contrib>	PLoS Medicine	<sec id="s1"><title>Introduction</title><p>Only four broadly neutralising monoclonal antibodies (MAbs) against HIV-1 have been generated to date, all of which were derived from patients with HIV-1 subtype B infection. IgG1b12 recognizes an epitope overlapping the CD4 binding site in the envelope glycoprotein complex [ <xref rid="pmed-0030255-b001" ref-type="bibr">1</xref>– <xref rid="pmed-0030255-b005" ref-type="bibr">5</xref>], and 2G12 recognizes a mannose-rich epitope on the silent face of gp120 [ <xref rid="pmed-0030255-b006" ref-type="bibr">6</xref>– <xref rid="pmed-0030255-b010" ref-type="bibr">10</xref>]. The 2F5 and 4E10 linear MAbs are located in the membrane-proximal external region of gp41 [ <xref rid="pmed-0030255-b011" ref-type="bibr">11</xref>– <xref rid="pmed-0030255-b013" ref-type="bibr">13</xref>]. Passive transfer studies in primates using combinations of these MAbs have provided strong evidence that MAbs are able to control viral replication [ <xref rid="pmed-0030255-b014" ref-type="bibr">14</xref>– <xref rid="pmed-0030255-b017" ref-type="bibr">17</xref>] and prevent HIV-1 infection parenterally and through mucosal tissues [ <xref rid="pmed-0030255-b018" ref-type="bibr">18</xref>, <xref rid="pmed-0030255-b019" ref-type="bibr">19</xref>]. More recent data have shown that in some individuals with HIV infection, these MAbs can reduce the rate of viral rebound following a structured treatment interruption [ <xref rid="pmed-0030255-b020" ref-type="bibr">20</xref>]. Furthermore, oral challenge studies in neonatal macaque monkeys support the use of neutralising MAbs for prevention of virus transmission to human infants [ <xref rid="pmed-0030255-b021" ref-type="bibr">21</xref>, <xref rid="pmed-0030255-b022" ref-type="bibr">22</xref>]. </p><p>Mother-to-child transmission (MTCT) of HIV-1 infection remains a significant problem in developing countries. While the use of single-dose nevirapine, acting to prevent intrapartum transmission, has reduced the number of infections, more potent interventions are needed, particularly to prevent postpartum transmissions. It is estimated that in South Africa alone, approximately 96,000 children with HIV-1 infection were born in 2003 [ <xref rid="pmed-0030255-b023" ref-type="bibr">23</xref>]. Passive immunization using neutralising MAbs has been suggested as a strategy to prevent breast milk–borne infections [ <xref rid="pmed-0030255-b024" ref-type="bibr">24</xref>, <xref rid="pmed-0030255-b025" ref-type="bibr">25</xref>]. Whether this approach is valid is likely to depend on the efficacy of these MAbs against the targeted viruses. </p><p>The most common subtype of HIV-1 infection in southern Africa as well as globally is subtype C ( <ext-link ext-link-type="uri" xlink:href="http://www.unaids.org">http://www.unaids.org</ext-link>). Results from a previous study indicated that a combination of the MAbs 2F5, 2G12, IgG1b12, and 4E10 successfully neutralised 100% of HIV-1 subtype C isolates tested [ <xref rid="pmed-0030255-b026" ref-type="bibr">26</xref>]. However, other studies have shown that 2F5 and 2G12 MAbs are usually ineffective against HIV-1 subtype C viruses, while 4E10 is able to neutralise isolates from all subtypes [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030255-b028" ref-type="bibr">28</xref>]. To further address whether 2G12, 2F5, IgG1b12, and 4E10 are active against HIV-1 subtype C viruses, we tested them in an Env-pseudotyped virus infectivity assay. We chose to use specifically those viruses derived from infants and children who had perinatally acquired HIV-1 infection to determine whether or not these MAbs are effective in vitro as an indication of their potential use for prevention of MTCT. </p></sec><sec id="s2"><title>Methods</title><sec id="s2a"><title>HIV-1 Subtype C Viral Isolates</title><p>Viruses were isolated from the blood of children with HIV-1 infection by standard co-culture techniques using peripheral blood mononuclear cells (PBMCs) [ <xref rid="pmed-0030255-b029" ref-type="bibr">29</xref>, <xref rid="pmed-0030255-b030" ref-type="bibr">30</xref>]. Blood samples were collected from children residing in an orphanage or receiving medical care at the Chris Hani Baragwanath Hospital in Johannesburg between 1999 and 2002 ( <xref ref-type="table" rid="pmed-0030255-t001">Table 1</xref>) [ <xref rid="pmed-0030255-b029" ref-type="bibr">29</xref>]. Informed consent was obtained from either a parent or a guardian of each child at the time of blood collection. This study received ethical approval from the University of the Witwatersrand Committee for Research on Human Subjects (Medical) (Johannesburg, South Africa). </p><table-wrap id="pmed-0030255-t001" position="float"><label>Table 1</label><caption><p>Patient Information and Viral Isolate Characteristics for HIV-1 Subtype C Cloned Envelope Genes</p></caption><graphic xlink:href="pmed.0030255.t001"/></table-wrap></sec><sec id="s2b"><title>MAbs, sCD4, and Plasma Samples</title><p>MAbs were obtained from the National Institutes of Health Reference and Reagent Program (Germantown, Maryland, United States) and the International AIDS Vaccine Initiative Neutralizing Antibody Consortium (New York, New York, United States), and used at a starting concentration of 50 μg/ml. Recombinant soluble CD4 (sCD4) comprising the extracellular domain of human CD4 produced in Chinese hamster ovary cells was obtained from Progenics Pharmaceuticals (Tarrytown, New York, United States), and tested at 50 μg/ml. Two plasma samples (BB12 and IBU21) from blood donors with HIV-1 subtype C infection were tested at a starting dilution of 1:50.</p></sec><sec id="s2c"><title>Cell Lines</title><p>JC53-bl cells were obtained from the National Institutes of Health Reference and Reagent Program (catalog number 8129). These cells were derived from a HeLa cell clone that expresses CD4, CCR5, and CXCR4 constitutively [ <xref rid="pmed-0030255-b031" ref-type="bibr">31</xref>] and contains two reporter genes: firefly luciferase and <named-content content-type="genus-species">Escherichia coli</named-content> β-galactosidase under the control of the HIV-1 LTR promoter [ <xref rid="pmed-0030255-b032" ref-type="bibr">32</xref>]. The 293T cells used for transfection were obtained from the American Type Culture Collection (Manassas, Virginia, United States). Both cell lines were cultured in D-MEM containing 10% heat-inactivated fetal bovine serum. Cell monolayers were disrupted at confluency by treatment with 0.25% trypsin in 1mM EDTA. </p></sec><sec id="s2d"><title>Cloning of Envelope Genes and Production of Pseudovirions</title><p>Proviral DNA extracted from in vitro infected PBMCs was used to amplify full-length envelope genes. The 3-kilobase PCR fragments, generated using envA and envM primers [ <xref rid="pmed-0030255-b033" ref-type="bibr">33</xref>], were cloned into the pCDNA 3.1-TOPO vector (Invitrogen, Carlsbad, California, United States) and bacterial colonies screened by PCR for insertion and correct orientation using T7 and envM primers. The Env-pseudotyped virus stocks were generated by co-transfecting 2 μg of the <italic>env</italic> encoding plasmid DNA with 3.3 μg of the HIV genomic vector SG3delta <italic>env</italic> (a gift from Beatrice Hahn) into an 80% confluent monolayer of 293T cells in a T-25 culture flask in the presence of 40 μl of PolyFect Transfection Reagent (Qiagen, Heidelberg, Germany). The media was replaced 6–8 h after transfection; 48 h later, culture supernatant containing the pseudoviruses was harvested, filtered (0.45 μm), and stored at −70 °C. The tissue culture infectious doses (TCID <sub>50</sub>) were quantified by infecting JC53-bl cells with serial 5-fold dilutions of the supernatant in quadruplicate in the presence of DEAE dextran (30 μg/ml) (Sigma, St. Louis, Missouri, United States). The infection was monitored 48 h later by evaluating the luciferase activity using the Bright Glo Reagent (Promega) following manufacturer instructions. Luminescence was measured in a Wallac 1420 Victor Multilabel Counter (Perkin Elmer, Wellesley, California, United States). TCID <sub>50</sub> was calculated as described [ <xref rid="pmed-0030255-b034" ref-type="bibr">34</xref>]. Wells with relative light units greater than 2.5 times the negative control (mock infection) were considered positive for infection. </p></sec><sec id="s2e"><title>Single-Cycle Neutralisation Assay</title><p>Neutralisation was measured as a reduction in luciferase gene expression after a single-round infection of JC53-bl cells with Env-pseudotyped viruses [ <xref rid="pmed-0030255-b035" ref-type="bibr">35</xref>]. Briefly, 200 TCID <sub>50</sub> of pseudoviruses in 50 μl culture media was incubated with 100 μl of serially diluted MAbs, plasma, or sCD4 using D-MEM with 10% fetal bovine serum in a 96-well plate in triplicate for 1 h at 37 °C. MAbs were either tested singly starting at 50 μg/ml (before addition of cells) or in combination also at 50 μg/ml for each MAb. Thus, TriMab contained 2G12, IgG1b12, and 2F5 (50:50:50 μg/ml) and TriMab plus 4E10 contained 2G12, IgG1b12, 2F5, and 4E10 (50:50:50:50 μg/ml). A 100-μl solution of JC53-bl cells (1 × 10 <sup>4</sup> cells/well) containing 75 μg/ml DEAE dextran was added; the cultures were then incubated at 37 °C in 5% CO <sub>2</sub>/95% air for 48 h. Infection was monitored by evaluating the luciferase activity. Titres were calculated as inhibitor concentration (IC <sub>50</sub>) or reciprocal plasma dilution (ID <sub>50</sub>) values causing 50% reduction of relative light units compared to the virus control (wells with no inhibitor) after subtracting the background (wells without virus infection). IC <sub>50</sub> values obtained for MAb combinations were compared to MAbs tested singly. The HIV-1 subtype B pseudovirus QH692.42 was included as a positive control, because this virus has been known to be sensitive to all four of the test MAbs [ <xref rid="pmed-0030255-b036" ref-type="bibr">36</xref>, <xref rid="pmed-0030255-b037" ref-type="bibr">37</xref>]. </p></sec><sec id="s2f"><title>gp160 Sequencing</title><p>Cloned <italic>env</italic> genes were sequenced using the ABI PRISM Big Dye Terminator Cycle Sequencing Ready Reaction kit (Perkin-Elmer Applied Biosystems, Foster City, California, United States) and resolved on an ABI 3100 automated genetic analyzer. The full-length gp160 sequences were assembled and edited using Sequencher (version 4.0) software (Gene Codes, Ann Arbor, Michigan, United States). </p></sec></sec><sec id="s3"><title>Results</title><sec id="s3a"><title>HIV-1 Subtype C Cloned Envelopes from Paediatric Patients</title><p>We cloned complete (gp160) envelope genes from seven HIV-1 subtype C isolates cultured from the blood of children with perinatally acquired HIV-1 infection. Five of these isolates were from rapidly progressing infants (RP and COT) who developed severe clinical symptoms within the first year of life, most of whom died shortly after blood collection ( <xref ref-type="table" rid="pmed-0030255-t001">Table 1</xref>). Two isolates were from children who had survived for between 6 and 9 y and were moderately symptomatic with illnesses, such as lymphocytic interstitial pneumonitis. All isolates used the CCR5 co-receptor, while viruses from one rapidly progressing infant (RP1) also used the CXCR4 co-receptor and was therefore dualtropic [ <xref rid="pmed-0030255-b029" ref-type="bibr">29</xref>]. The Env-pseudotyped virus derived from the latter isolate was able to use only CXCR4 as co-receptor, while the other six pseudoviruses used CCR5 ( <xref ref-type="table" rid="pmed-0030255-t001">Table 1</xref>). All cloned envelopes were sequenced and compared to the original viral isolate. Phylogenetic analysis indicated that all samples were HIV-1 subtype C isolates; sequences from the same individual clustered together with high bootstrap values (unpublished data). </p></sec><sec id="s3b"><title>Neutralisation Sensitivity of HIV-1 Subtype C Env-Pseudotyped Viruses to MAbs</title><p>The HIV-1 subtype C envelope clones were used to generate Env-pseudotyped viruses by co-transfection with a subgenomic plasmid. These pseudoviruses were tested for their sensitivity to neutralisation by the MAbs IgG1b12, 2G12, 2F5, and 4E10. The MAbs 2G12 and 2F5 failed to neutralise any of the seven HIV-1 subtype C pseudoviruses at 50 μg/ml, whereas the HIV-1 subtype B virus QH692.42 had IC <sub>50</sub> values of 0.8 and 7.1, respectively ( <xref ref-type="table" rid="pmed-0030255-t002">Table 2</xref>). The IgG1b12 neutralised four of the seven HIV-1 subtype C viruses as well as the HIV-1 subtype B control. The IC <sub>50</sub> values of the sensitive pseudoviruses ranged from 0.2 μg/ml to 12 μg/ml, indicating high potency of this MAb. The MAb 4E10 neutralised all the viruses. The IC <sub>50</sub> values were generally high, supporting the notion that this MAb has broad specificity but lower potency than other MAbs [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>]. </p><table-wrap id="pmed-0030255-t002" position="float"><label>Table 2</label><caption><p>Sensitivity of HIV-1 Subtype C Pseudovirions to Anti-HIV MAbs, sCD4, and Plasma</p></caption><graphic xlink:href="pmed.0030255.t002"/></table-wrap></sec><sec id="s3c"><title>Neutralisation Using Combinations of MAbs</title><p>Synergistic neutralisation among MAbs that recognize different specificities in the envelope glycoprotein has been suggested [ <xref rid="pmed-0030255-b038" ref-type="bibr">38</xref>, <xref rid="pmed-0030255-b039" ref-type="bibr">39</xref>], although it has been a controversial topic. We decided, therefore, to test combinations of these MAbs using equimolar concentrations of 2G12, IgG1b12, and 2F5 (TriMAb), and TriMAb plus 4E10. The IC <sub>50</sub> values in the presence of TriMAb were similar to those for IgG1b12 alone ( <xref ref-type="table" rid="pmed-0030255-t002">Table 2</xref>), indicating that the activity in TriMAb was probably due to the activity of IgG1b12. When 4E10 was added to TriMAb, it was not surprising that neutralisation of all isolates was achieved, because 4E10 was active against all isolates when used alone at this concentration range. </p><p>Analysis of the dose-response curves confirmed the lack of significant synergy among MAbs. Those viruses sensitive to IgG1b12 (RP4.3, RP6.6, TM7.9, and COT9.6) had similar neutralisation curves in the presence of IgG1b12 alone or when tested as part of TriMAb with or without 4E10 ( <xref ref-type="fig" rid="pmed-0030255-g001">Figure 1</xref>A). However, among isolates insensitive to IgG1b12 (COT6.15, TM3.8, and RP1.12), slightly greater potency was observed with TriMAb plus 4E10, compared to 4E10 alone ( <xref ref-type="fig" rid="pmed-0030255-g001">Figure 1</xref>B). </p><fig id="pmed-0030255-g001" position="float"><label>Figure 1</label><caption><title>Neutralisation Dose-Response Curves of the MAbs 2G12, 2F5, IgG1b12, and 4E10, Alone and in Combination</title><p>The MAb concentrations in the triple and quadruple combination are represented as the concentration of each MAb in the equimolar mix starting at 50 μg/ml. Results are shown as the reduction of virus infectivity relative to the virus control (without MAbs) with 50% inhibition indicated by a dotted line. Note those viruses sensitive to IgG1b12 and 4E10 (A) and those viruses sensitive to 4E10 alone (B).</p></caption><graphic xlink:href="pmed.0030255.g001"/></fig></sec><sec id="s3d"><title>Sensitivity to sCD4 and Polyclonal Anti-HIV Antibodies</title><p>Given the relative resistance of the HIV-1 subtype C pseudovirions to neutralisation by MAbs, we chose to test their responses to sCD4 and polyclonal antibodies from individuals with HIV-1 infection. sCD4, which blocks gp120 binding to the CD4 receptor, neutralised all of the pseudovirions ( <xref ref-type="table" rid="pmed-0030255-t002">Table 2</xref>), indicating that the CD4 binding site is accessible on the pseudotyped envelope glycoproteins. The IgG1b12 binding site overlaps with the CD4 binding site; however, there was no correlation between the ID <sub>50</sub> values for sCD4 and IgG1b12 in this assay, similar to what others have reported [ <xref rid="pmed-0030255-b028" ref-type="bibr">28</xref>, <xref rid="pmed-0030255-b040" ref-type="bibr">40</xref>]. </p><p>All pseudovirions except TM7.9 were neutralised by one or both of the plasma samples with a wide variation in IC <sub>50</sub> titres, as is often seen when using polyclonal antibodies, suggesting that these envelopes were not atypical in their ability to be neutralised ( <xref ref-type="table" rid="pmed-0030255-t002">Table 2</xref>). </p></sec><sec id="s3e"><title>Analysis of Amino Acid Sequences Comprising the Neutralisation Epitopes</title><p>Sequence analysis of the predicted N-linked glycosylation (PNG) sites at positions 295, 332, and 392, which are critical for the 2G12 epitope, indicated that all HIV-1 subtype C isolates lacked the glycan 295. TM7.9 also lacked the glycan 392 ( <xref ref-type="table" rid="pmed-0030255-t003">Table 3</xref>). Another site (position 386), reported to play an indirect role in the formation of the 2G12 epitope, was also absent from one of the HIV-1 subtype C envelopes. The HIV-1 subtype B pseudovirus QH692.42 was the only virus possessing all five PNG sites and was the only virus sensitive to 2G12. These data suggest that the lack of the glycan 295 renders isolates resistant to 2G12, as previously suggested [ <xref rid="pmed-0030255-b009" ref-type="bibr">9</xref>, <xref rid="pmed-0030255-b010" ref-type="bibr">10</xref>]. </p><table-wrap id="pmed-0030255-t003" position="float"><label>Table 3</label><caption><p>Amino Acid Sequences of MAb Epitopes in Cloned Subtype C Envelope Genes</p></caption><graphic xlink:href="pmed.0030255.t003"/></table-wrap><p>The 2F5 epitope is centred on the sequence ELDKWA [ <xref rid="pmed-0030255-b011" ref-type="bibr">11</xref>]. Mutagenesis studies have revealed that the amino acid residues DKW are indispensable for the recognition by this MAb [ <xref rid="pmed-0030255-b013" ref-type="bibr">13</xref>, <xref rid="pmed-0030255-b041" ref-type="bibr">41</xref>]. In particular, substitutions at residue K665 appear to be the major determinant of resistance [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>]. In this study, all HIV-1 subtype C isolates had substitutions at position 665 with the lysine (K) residue replaced by serine (S) or other amino acids (R or N), while the HIV-1 subtype B pseudovirus QH692.42 had no such substitution. These data support the finding that the residue K665 is crucial for neutralisation by 2F5. </p><p>4E10 recognizes an epitope containing the sequence NWF(D/N)IT [ <xref rid="pmed-0030255-b012" ref-type="bibr">12</xref>, <xref rid="pmed-0030255-b042" ref-type="bibr">42</xref>] at the C-terminal of the 2F5 epitope. Mutagenesis experiments have shown that the residues W672, F673, and W680 are indispensable for recognition by 4E10 [ <xref rid="pmed-0030255-b013" ref-type="bibr">13</xref>], while the crystal structure of the Fab 4E10-epitope complex indicates that W672, F673, I675, and T676 are the key residues in this interaction [ <xref rid="pmed-0030255-b043" ref-type="bibr">43</xref>]. All the viruses analyzed in this study had a conserved 4E10 epitope (W672, F673, W680), consistent with their phenotypic sensitivity to this MAb. </p></sec></sec><sec id="s4"><title>Discussion</title><p>The neutralisation sensitivity of HIV-1 subtype C isolates derived from children appears similar to previously reported sensitivity of isolates from adults with HIV-1 subtype C infection [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030255-b028" ref-type="bibr">28</xref>]. Thus the broadly cross-reactive neutralising MAbs 2G12 and 2F5 are ineffective against both paediatric and adult HIV-1 subtype C viruses, while IgG1b12 potently neutralised approximately 50% of the tested viruses. Only 4E10 showed broad activity against HIV-1 subtype C viruses, although its potency was low. Collectively, these data caution against the use of 2G12 and 2F5 MAbs for passive immunization in areas where HIV-1 subtype C viruses are highly prevalent. </p><p>In this study, we have used cloned envelope genes in a single-cycle neutralisation assay, which is a high-throughput assay that, to our knowledge, is rapidly becoming the method of choice for measuring antibody neutralisation [ <xref rid="pmed-0030255-b037" ref-type="bibr">37</xref>, <xref rid="pmed-0030255-b044" ref-type="bibr">44</xref>]. Comparative studies have shown a positive correlation between results derived from this assay and the more traditional PBMC-based neutralisation assay (Taylor et al., unpublished data) [ <xref rid="pmed-0030255-b037" ref-type="bibr">37</xref>]. However, the 293T-derived pseudovirions were found to be more sensitive to neutralisation by MAbs and serum samples when compared to the uncloned PBMC-derived viruses [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030255-b037" ref-type="bibr">37</xref>]. It has been suggested that this effect is due to the cells used to generate the pseudoviruses [ <xref rid="pmed-0030255-b044" ref-type="bibr">44</xref>] and not the nature of the target cells or the clonal nature of the envelope [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030255-b045" ref-type="bibr">45</xref>]. Overall, we can be confident that the observed resistance of HIV-1 subtype C isolates to 2G12 and 2F5 is not due to the use of an Env-pseudotyped virus-based neutralisation assay. Instead, the extra sensitivity of the latter assay might be expected to generate false-positive and not false-negative outcomes. </p><p>It has been shown in multiple studies that 2G12 is generally ineffective against HIV-1 subtype C isolates [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030255-b028" ref-type="bibr">28</xref>]. The 2G12 epitope binds a cluster of mannose residues; the absence of an N-linked glycan at position 295 appears to correlate with resistance to this MAb [ <xref rid="pmed-0030255-b009" ref-type="bibr">9</xref>, <xref rid="pmed-0030255-b010" ref-type="bibr">10</xref>]. The absence of N295 may prevent the correct processing and presentation of glycans at position 332, affecting antibody binding and therefore neutralisation [ <xref rid="pmed-0030255-b007" ref-type="bibr">7</xref>]. A recent study has shown that reintroduction of this PNG site into a subtype C isolate restored binding of 2G12, although sensitivity to neutralisation was not tested [ <xref rid="pmed-0030255-b046" ref-type="bibr">46</xref>]. An analysis of 339 HIV-1 subtype C envelope sequences obtained from Los Alamos Database showed that 83% of sequences lacked a glycosylation site at position 295. If the lack of the PNG at position 295 is indeed a cause of resistance to 2G12, then a majority of HIV-1 subtype C viruses would be insensitive to this MAb. </p><p>The 2F5 MAb has been shown to have broadly neutralising activity but has minimal efficacy against HIV-1 subtype C viruses [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030255-b028" ref-type="bibr">28</xref>]. An alanine scan over the ELDKWAS epitope defined the motif DKW in positions 664–666 as a determinant for 2F5 recognition [ <xref rid="pmed-0030255-b013" ref-type="bibr">13</xref>], although some viruses with this epitope are insensitive to this MAb [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>]. However, all viruses with a substitution at residue K665 are resistant to 2F5 [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>]. Similarly, in this study, we found that all resistant viruses had a substitution at K665 while the subtype B virus did not. Analysis of 324 sequences in Los Alamos Database showed that the subtype C consensus for the 2F5 epitope is ALDSWA, with only approximately 12% bearing a K at position 665. This suggests that the majority of HIV-1 subtype C viruses will also be resistant to 2F5. However, a geographical clustering of some HIV-1 subtype C variants that may be sensitive to 2F5 due to the presence of the DKW epitope has been suggested [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>]. </p><p>Our data with IgG1b12 agree with other studies in that this MAb is more effective than 2F5 or 2G12 at neutralising HIV-1 subtype C viruses, although IgG1b12 inhibited only approximately 50% of the isolates tested [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030255-b028" ref-type="bibr">28</xref>]. Among sensitive isolates, this MAb is particularly potent and requires very low antibody concentrations for 50% inhibition. Due to the conformational nature of the IgG1b12 epitope, it is difficult to predict resistance to this MAb by simple sequence analysis. Some studies have described neutralisation escape mutations for this MAb, such as a proline-to-alanine substitution in position 369 in the C3 region of gp120 [ <xref rid="pmed-0030255-b047" ref-type="bibr">47</xref>, <xref rid="pmed-0030255-b048" ref-type="bibr">48</xref>]. In this study, we found no correlation between the presence of a proline at this position and sensitivity to IgG1b12, which suggests that this escape mutation was specific to the isolate used in the referred study. </p><p>The 4E10 epitope appears to be the most broadly cross-reactive MAb described to date, neutralising all viruses so far tested. In previous studies, 4E10 has been shown to neutralise 100% of viruses in a comprehensive panel that included all genetic subtypes of HIV-1 group M and some recombinant forms [ <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030255-b049" ref-type="bibr">49</xref>]. However, 4E10 is a low-potency antibody generally requiring high concentrations to reduce infectivity by 50%, as seen in this and other studies [ <xref rid="pmed-0030255-b020" ref-type="bibr">20</xref>, <xref rid="pmed-0030255-b027" ref-type="bibr">27</xref>]. Whether this is a property of the antibody or inaccessibility of the epitope remains to be determined. The motif WF on the 4E10 epitope was 100% conserved in 324 sequences of this portion of gp41 from HIV-1 subtype C viruses in Los Alamos Database. This suggests that HIV-1 subtype C viruses will be universally sensitive to 4E10. </p><p>Some studies have suggested that MAbs can act synergistically to increase neutralisation potency against HIV-1 [ <xref rid="pmed-0030255-b026" ref-type="bibr">26</xref>, <xref rid="pmed-0030255-b038" ref-type="bibr">38</xref>, <xref rid="pmed-0030255-b039" ref-type="bibr">39</xref>, <xref rid="pmed-0030255-b050" ref-type="bibr">50</xref>, <xref rid="pmed-0030255-b051" ref-type="bibr">51</xref>]. However, this has been a controversial topic, and isolate dependency has been observed [ <xref rid="pmed-0030255-b051" ref-type="bibr">51</xref>, <xref rid="pmed-0030255-b052" ref-type="bibr">52</xref>] with different results obtained with T-cell line-adapted virus and primary isolates [ <xref rid="pmed-0030255-b039" ref-type="bibr">39</xref>, <xref rid="pmed-0030255-b050" ref-type="bibr">50</xref>, <xref rid="pmed-0030255-b052" ref-type="bibr">52</xref>]. In this study, we did not observe strong synergy among these MAbs. The combination of the four MAbs neutralised all the tested viruses in agreement with other study results for HIV-1 subtype C isolates [ <xref rid="pmed-0030255-b026" ref-type="bibr">26</xref>]. This is likely due to the neutralisation activity of individual MAbs rather than the combined effect of them, because a significant increase in potency was not observed with the mixtures. There may have been a slight synergistic effect for RP1.12, TM3.8, and COT6.15 as demonstrated by increased neutralisation when 4E10 was combined with IgG1b12, 2F5, and 2G12. Such an effect is probably due to IgG1b12, given the absence of the 2G12 and 2F5 epitopes in these viruses. A more thorough analysis of synergism would require titrating 4E10 against IgG1b12 and evaluating the data based on the Chou-Talalay method [ <xref rid="pmed-0030255-b039" ref-type="bibr">39</xref>, <xref rid="pmed-0030255-b053" ref-type="bibr">53</xref>]. It is also possible that the clonal nature of the envelope glycoproteins used in this study precluded the detection of synergism. Some researchers have suggested that the heterogeneity of the virus is the cause of the synergistic effects of some neutralising antibody combinations [ <xref rid="pmed-0030255-b054" ref-type="bibr">54</xref>]. However, others have observed no differences between virus isolates passaged in PBMCs and cloned envelope pseudotype viruses [ <xref rid="pmed-0030255-b039" ref-type="bibr">39</xref>]. </p><p>The MTCT of HIV-1 infection is usually associated with transmission of single variants [ <xref rid="pmed-0030255-b026" ref-type="bibr">26</xref>]. In this study, four of the cloned envelopes were from children infected for fewer than 12 mo, two of which were infected for 4 mo and therefore represent relatively early variants. Although these clones may not have been the earliest transmitted variants, it is unlikely that earlier variants would differ in their neutralisation sensitivity to these MAbs. We base this assumption on the fact that the MAb sensitivities of viruses from adults with HIV-1 subtype C infection, who would be the source of infection in perinatally infected children, are similar. In addition, we did not observe variation in the susceptibility to neutralisation or in the epitope sequences that can be related to the age of the child: infants and children in this study had identical phenotypic and genotypic profiles. Overall, we feel confident that the MAb neutralisation profiles of the viruses analyzed in this study would be representative of the earliest transmitted variants. </p><p>Based on our results, we question the use of MAb combinations that include 2F5 and 2G12 as a prophylactic treatment in regions where HIV-1 subtype C viruses predominate, even if such combinations were to include 4E10 and IgG1b12. In passive immunoprophylaxis studies using a single MAb, protection was not observed even when the challenge strain was successfully neutralised in vitro. Only a combination of at least three MAbs with bona fide neutralisation activity against the challenge strain offered complete protection [ <xref rid="pmed-0030255-b017" ref-type="bibr">17</xref>]. Such a combination is not likely to be achievable against HIV-1 subtype C isolates. Furthermore, a recent study using a combination of 2G12, 2F5, and 4E10 for the treatment of individuals with HIV-1 infection has denoted that the ability of 2F5 and 4E10 to affect the virus in vivo is unclear and may require very high serum concentrations of these MAbs [ <xref rid="pmed-0030255-b020" ref-type="bibr">20</xref>]. This further questions the use of MAb combinations in which only 4E10 has the potential to be 100% effective. </p><p>Overall, we believe that the use of these MAbs to prevent MTCT of HIV-1 subtype C infection is unlikely to be efficacious; therefore, a clinical trial should not be conducted. A recent study has confirmed our viewpoint that these MAbs would have limited benefit when used to prevent MTCT in populations with HIV-1 non-B subtype infection [ <xref rid="pmed-0030255-b055" ref-type="bibr">55</xref>]. In addition, recent data have suggested that the MAbs 2F5 and 4E10 react against self-antigens, such as cardiolipin, and the MAb IgG1b12 reacts with double-stranded DNA [ <xref rid="pmed-0030255-b056" ref-type="bibr">56</xref>]. Although safety concerns exist surrounding the use of these MAbs for treatment [ <xref rid="pmed-0030255-b057" ref-type="bibr">57</xref>], no adverse effects have yet been reported in treated adults [ <xref rid="pmed-0030255-b020" ref-type="bibr">20</xref>]. It should be noted that this work remains to be corroborated by others. Nevertheless, if these findings on autoreactivity prove to be true, then the utility of these MAbs for in vivo use is in further doubt, particularly if they are to be used in infants. </p><p>The study of the epitopes recognized by these broadly neutralising MAbs contributes to the knowledge necessary for the rational design of an immunogen capable of inducing a broad and potent neutralisation response against HIV-1 infection. Considerable efforts have been invested in designing immunogens based on these epitopes [ <xref rid="pmed-0030255-b041" ref-type="bibr">41</xref>]. However, given the subtype constraints of some of these epitopes, new, more broadly occurring epitopes need to be found for the design of vaccines that will be able to elicit an efficient neutralising response against a broad spectrum of HIV subtypes. </p></sec><sec sec-type="supplementary-material" id="s5"><title>Supporting Information</title><sec id="s5a"><title>Accession Numbers</title><p>The GenBank ( <ext-link ext-link-type="uri" xlink:href="http://www.ncbi.nlm.nih.gov/Genbank">http://www.ncbi.nlm.nih.gov/Genbank</ext-link>) accession numbers for nucleotide sequences of the cloned envelope genes discussed in this paper are DQ447266–DQ447272 ( <xref ref-type="table" rid="pmed-0030255-t001">Table 1</xref>). </p></sec></sec>
Routine HIV Testing in Botswana: A Population-Based Study on Attitudes, Practices, and Human Rights Concerns	<sec id="st1"><title>Background</title><p>The Botswana government recently implemented a policy of routine or “opt-out” HIV testing in response to the high prevalence of HIV infection, estimated at 37% of adults.</p></sec><sec id="st2"><title>Methods and Findings</title><p>We conducted a cross-sectional, population-based study of 1,268 adults from five districts in Botswana to assess knowledge of and attitudes toward routine testing, correlates of HIV testing, and barriers and facilitators to testing, 11 months after the introduction of this policy. Most participants (81%) reported being extremely or very much in favor of routine testing. The majority believed that this policy would decrease barriers to testing (89%), HIV-related stigma (60%), and violence toward women (55%), and would increase access to antiretroviral treatment (93%). At the same time, 43% of participants believed that routine testing would lead people to avoid going to the doctor for fear of testing, and 14% believed that this policy could increase gender-based violence related to testing. The prevalence of self-reported HIV testing was 48%. Adjusted correlates of testing included female gender (AOR = 1.5, 95% CI = 1.1–1.9), higher education (AOR = 2.0, 95% CI = 1.5–2.7), more frequent healthcare visits (AOR = 1.9, 95% CI = 1.3–2.7), perceived access to HIV testing (AOR = 1.6, 95% CI = 1.1–2.5), and inconsistent condom use (AOR = 1.6, 95% CI = 1.2–2.1). Individuals with stigmatizing attitudes toward people living with HIV and AIDS were less likely to have been tested for HIV/AIDS (AOR = 0.7, 95% CI = 0.5–0.9) or to have heard of routine testing (AOR = 0.59, 95% CI = 0.45–0.76). While experiences with voluntary and routine testing overall were positive, 68% felt that they could not refuse the HIV test. Key barriers to testing included fear of learning one's status (49%), lack of perceived HIV risk (43%), and fear of having to change sexual practices with a positive HIV test (33%).</p></sec><sec id="st3"><title>Conclusions</title><p>Routine testing appears to be widely supported and may reduce barriers to testing in Botswana. As routine testing is adopted elsewhere, measures should be implemented to assure true informed consent and human rights safeguards, including protection from HIV-related discrimination and protection of women against partner violence related to testing.</p></sec>	<contrib contrib-type="author"><name><surname>Weiser</surname><given-names>Sheri D</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref><xref ref-type="aff" rid="aff3"> <sup>3</sup> </xref><xref ref-type="corresp" rid="cor1">*</xref></contrib><contrib contrib-type="author"><name><surname>Heisler</surname><given-names>Michele</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref><xref ref-type="aff" rid="aff4"> <sup>4</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Leiter</surname><given-names>Karen</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Percy-de Korte</surname><given-names>Fiona</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Tlou</surname><given-names>Sheila</given-names></name><xref ref-type="aff" rid="aff5"> <sup>5</sup> </xref></contrib><contrib contrib-type="author"><name><surname>DeMonner</surname><given-names>Sonya</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Phaladze</surname><given-names>Nthabiseng</given-names></name><xref ref-type="aff" rid="aff5"> <sup>5</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Bangsberg</surname><given-names>David R</given-names></name><xref ref-type="aff" rid="aff3"> <sup>3</sup> </xref><xref ref-type="aff" rid="aff6"> <sup>6</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Iacopino</surname><given-names>Vincent</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref><xref ref-type="aff" rid="aff7"> <sup>7</sup> </xref></contrib>	PLoS Medicine	<sec id="s1"><title>Introduction</title><p>There has been widespread concern about the slow uptake of voluntary counseling and testing (VCT) in many parts of sub-Saharan Africa [ <xref rid="pmed-0030261-b001" ref-type="bibr">1</xref>, <xref rid="pmed-0030261-b002" ref-type="bibr">2</xref>]. VCT is a cornerstone of cost-effective HIV prevention and linkage to expanding HIV treatment in low-resource settings [ <xref rid="pmed-0030261-b003" ref-type="bibr">3</xref>, <xref rid="pmed-0030261-b004" ref-type="bibr">4</xref>]. Some of the most significant barriers to HIV testing identified in sub-Saharan Africa include lack of access to VCT and to high quality clinical services, lack of access to antiretroviral (ARV) therapy, and HIV-related stigma [ <xref rid="pmed-0030261-b001" ref-type="bibr">1</xref>, <xref rid="pmed-0030261-b005" ref-type="bibr">5</xref>, <xref rid="pmed-0030261-b006" ref-type="bibr">6</xref>]. </p><p>With a seroprevalence of 37% of adults ages 15–49 [ <xref rid="pmed-0030261-b007" ref-type="bibr">7</xref>, <xref rid="pmed-0030261-b008" ref-type="bibr">8</xref>], Botswana established universal access to antiretroviral treatment (ART) beginning in 2002 for all patients with CD4 counts less than 200 or with an AIDS-defining illness [ <xref rid="pmed-0030261-b009" ref-type="bibr">9</xref>– <xref rid="pmed-0030261-b011" ref-type="bibr">11</xref>]. By January 2004, however, only 17,500 patients were enrolled in the Botswana National Treatment Program out of an estimated 110,000 eligible individuals [ <xref rid="pmed-0030261-b009" ref-type="bibr">9</xref>]. Slow enrollment in HIV treatment was thought to be due in part to underutilization of HIV testing [ <xref rid="pmed-0030261-b009" ref-type="bibr">9</xref>, <xref rid="pmed-0030261-b011" ref-type="bibr">11</xref>, <xref rid="pmed-0030261-b012" ref-type="bibr">12</xref>]; by mid-2003, only 70,000 tests in total had been performed in Botswana out of a population of 1.7 million [ <xref rid="pmed-0030261-b013" ref-type="bibr">13</xref>]. HIV stigma was identified by government and press sources as one possible impediment to HIV testing and hence to the success of the new ART program, in that individuals may avoid testing and treatment facilities to avoid potential stigma and discrimination [ <xref rid="pmed-0030261-b008" ref-type="bibr">8</xref>, <xref rid="pmed-0030261-b011" ref-type="bibr">11</xref>, <xref rid="pmed-0030261-b013" ref-type="bibr">13</xref>]. We previously reported that social stigma and fear of positive test results significantly delayed testing among a group of patients treated in the private sector in 2000 [ <xref rid="pmed-0030261-b014" ref-type="bibr">14</xref>]. </p><p>In an attempt to increase the uptake of HIV testing and ART, the Botswana government introduced the policy of routine HIV testing in early 2004, whereby nearly all patients would be tested as a routine part of medical visits unless they explicitly refused [ <xref rid="pmed-0030261-b013" ref-type="bibr">13</xref>, <xref rid="pmed-0030261-b015" ref-type="bibr">15</xref>]. While this approach to testing is provider-initiated, all patients should receive essential information about HIV testing and be informed of their right to refuse. In addition, there is typically greater emphasis on post-test compared with pre-test counseling [ <xref rid="pmed-0030261-b016" ref-type="bibr">16</xref>]. Studies in resource-rich settings have shown that routine HIV testing can be cost-effective and life-saving, both by increasing the life expectancy of individuals with HIV and by reducing the annual HIV transmission rate [ <xref rid="pmed-0030261-b017" ref-type="bibr">17</xref>– <xref rid="pmed-0030261-b021" ref-type="bibr">21</xref>]. In June 2004, as part of a change in testing policy recommendations, UNAIDS and the World Health Organization recommended the routine offer of HIV testing by healthcare providers in a wide range of clinical encounters based in part on the Botswana experience [ <xref rid="pmed-0030261-b022" ref-type="bibr">22</xref>, <xref rid="pmed-0030261-b023" ref-type="bibr">23</xref>]. The goal of routine testing is to increase the proportion of individuals aware of their status, and thereby reduce “HIV exceptionalism,” lessen HIV-related stigma, and provide more people access to life-saving therapy [ <xref rid="pmed-0030261-b016" ref-type="bibr">16</xref>, <xref rid="pmed-0030261-b024" ref-type="bibr">24</xref>]. While provider-initiated approaches to testing are gaining popularity, there have been concerns that routine testing policies are potentially coercive, that counseling will no longer be practiced, that people may be dissuaded from visiting their doctors for fear of being tested, and that this policy may increase testing-related partner violence [ <xref rid="pmed-0030261-b015" ref-type="bibr">15</xref>, <xref rid="pmed-0030261-b025" ref-type="bibr">25</xref>– <xref rid="pmed-0030261-b027" ref-type="bibr">27</xref>]. </p><p>As routine testing is increasingly being recommended as an option in other countries [ <xref rid="pmed-0030261-b017" ref-type="bibr">17</xref>, <xref rid="pmed-0030261-b018" ref-type="bibr">18</xref>, <xref rid="pmed-0030261-b028" ref-type="bibr">28</xref>– <xref rid="pmed-0030261-b030" ref-type="bibr">30</xref>], it is important to improve our understanding of the consequences and specific human rights concerns associated with implementation of this policy in Botswana. We therefore assessed: 1) knowledge of and attitudes toward routine testing in Botswana with a focus on human rights concerns associated with this policy; 2) factors associated with whether respondents had heard of routine testing, and with positive attitudes toward this policy; and 3) the prevalence and correlates of HIV testing, barriers and facilitators to testing, and reported experiences with testing 11 months after the introduction of routine testing in Botswana. </p></sec><sec id="s2"><title>Methods</title><p>In November and December 2004, we conducted a cross-sectional study using structured survey instruments among a probability sample of 1,268 adults selected from the five districts of Botswana with the highest number of HIV-infected individuals. These districts included Gaborone, Kweneng East, Francistown, Serowe/Palapye, and Tutume, and cover a population of 725,000 out of a total population of 1.7 million individuals in Botswana. We used a stratified two-stage probability sample design for the selection of the population-based sample with the assistance of the Central Statistics Office at the Ministry of Finance and Development Planning in Botswana. In the first stage of sampling, 89 enumeration areas were selected with probability proportional to measures of size, where measures of size are the number of households in the enumeration area as defined by the 2001 Population and Housing Census. At the second stage of sampling, households were systematically selected in each enumeration area by trained field researchers under the guidance of field supervisors. With a target sample of 1,200 households, and 15% over-sampling for an anticipated 85% response rate, 1,433 households were selected. Within each household, random number tables were used to select one adult member who met our inclusion criteria, and up to two repeat visits were made.</p><p>Participants were excluded if they were older than 49 or younger than 18 years old, if they had cognitive disabilities, or if there was inadequate privacy to conduct the interviews. The 45- to 60-minute survey was conducted in either English or Setswana in a private setting, and written consent was obtained from all study participants. Our structured survey instrument and consent forms were pilot-tested among 20 individuals from Gaborone, and then translated into Setswana and back-translated into English. All study procedures were approved by the Human Subjects Committee at the University of California San Francisco (San Francisco, California, United States), as well as by the Botswana Ministry of Health Research and Development Committee.</p><sec id="s2a"><title>Measures</title><p>Domains of inquiry for our 234-item survey ( <xref ref-type="supplementary-material" rid="sd001">Protocol S1</xref>) included demographics, HIV knowledge, experiences with HIV testing, barriers and facilitators to HIV testing, attitudes toward routine testing, HIV risk behaviors, HIV-related stigma, depression, beliefs about gender roles and gender discrimination, and measures of healthcare access and utilization. Based on an extensive literature review [ <xref rid="pmed-0030261-b002" ref-type="bibr">2</xref>, <xref rid="pmed-0030261-b006" ref-type="bibr">6</xref>, <xref rid="pmed-0030261-b031" ref-type="bibr">31</xref>– <xref rid="pmed-0030261-b037" ref-type="bibr">37</xref>] and discussions with key informants, we developed a conceptual model that guided the selection of variables for our multivariate model for correlates of testing, as shown in <xref ref-type="fig" rid="pmed-0030261-g001">Figure 1</xref>. Relevant variables are explained below. </p><fig id="pmed-0030261-g001" position="float"><label>Figure 1</label><caption><title>Hypothesized Model for Predictors of Testing</title></caption><graphic xlink:href="pmed.0030261.g001"/></fig><sec id="s2a1"><title>Knowledge of and attitudes toward routine testing.</title><p>Participants were asked whether they had heard of routine testing and were given a detailed explanation of this policy (see <xref ref-type="table" rid="pmed-0030261-t001">Table 1</xref>). Participants then indicated the extent to which they are in favor of routine testing and whether they think this policy affects HIV-related stigma, barriers to testing, violence against women related to testing, and uptake of ARVs. From questions assessing attitudes toward routine testing ( <xref ref-type="table" rid="pmed-0030261-t001">Table 1</xref>), we constructed an ordinal outcome of positive attitudes toward this policy. Participants were categorized as having zero to one, two, three, or four positive views toward routine testing. (See <xref ref-type="table" rid="pmed-0030261-t001">Tables 1</xref> and <xref ref-type="table" rid="pmed-0030261-t002">2</xref> for specific items.) </p><table-wrap id="pmed-0030261-t001" position="float"><label>Table 1</label><caption><p>Attitudes toward Routine Testing</p></caption><graphic xlink:href="pmed.0030261.t001"/></table-wrap><table-wrap id="pmed-0030261-t002" position="float"><label>Table 2</label><caption><p>Respondent Characteristics Associated with Having Heard of Routine Testing and with Positive Attitudes Toward Routine Testing in Multivariate Analyses ( <italic>n</italic> = 1,168) </p></caption><graphic xlink:href="pmed.0030261.t002"/></table-wrap></sec><sec id="s2a2"><title>HIV testing.</title><p>Participants were asked whether they had ever been tested for HIV (by either VCT or routine testing). If so, they were asked detailed questions about their experiences with pre-test and post-test counseling, confidentiality, facilitators to testing, and personal repercussions of testing. If not, they were asked a series of questions related to barriers to testing adapted from the CDC HIV Testing Instrument, version 9.00, and about their intention to be tested within the next six months. HIV status was not asked in order to maximize response rate and hence the generalizability of the population-based sample.</p></sec><sec id="s2a3"><title>HIV-related stigma.</title><p>Respondents were asked seven questions representing potential stigmatizing attitudes adapted from the UNAIDS general population survey and the Department of Health Services AIDS module, which have been used successfully in previous studies in Botswana [ <xref rid="pmed-0030261-b038" ref-type="bibr">38</xref>]. Anyone who reported a discriminatory attitude on any of four principal questions was registered as having stigmatizing attitudes per the UNAIDS scoring system. Since participants may not always openly endorse stigmatizing views toward people living with HIV and AIDS (PLWA) due to social desirability bias, as an additional measure of stigma, we also asked individuals to project the type of responses they would anticipate from others if they were to test positive for HIV and divulge their status to others. We converted this information to a nine-item index on “projected HIV stigma” with higher scores associated with a greater number of reported adverse social consequences associated with testing positive. This index had high internal reliability with a Cronbach alpha of 0.77. </p></sec><sec id="s2a4"><title>HIV knowledge.</title><p>Participants were asked 15 questions about their knowledge of HIV transmission and prevention, based on questions modified from the UNAIDS General Population Survey and the Department of Health Services AIDS module. Using the UNAIDS knowledge indicator scoring system, individuals were scored as having HIV knowledge if they correctly identified the two most common modes of HIV prevention in Botswana.</p></sec><sec id="s2a5"><title>Depression.</title><p>As depression is known to impede access to care and to worsen HIV outcomes in Western settings, we included depression in our analysis [ <xref rid="pmed-0030261-b039" ref-type="bibr">39</xref>– <xref rid="pmed-0030261-b041" ref-type="bibr">41</xref>]. Symptoms of depression were measured using the 15-item Hopkins Symptom Checklist for Depression [ <xref rid="pmed-0030261-b042" ref-type="bibr">42</xref>] which has been validated previously in locations in Africa and elsewhere [ <xref rid="pmed-0030261-b043" ref-type="bibr">43</xref>]. </p></sec></sec><sec id="s2b"><title>Analysis</title><p>We used standard procedures for data entry and quality control. All data were analyzed with S TATA statistical software. Outcomes of interest included: a) having heard of routine testing; b) number of positive attitudes toward routine testing (categorized as an ordinal variable consisting of the following categories: zero to one, two, three, and four positive statements about routine testing); c) self-reported HIV testing (by either VCT or routine testing); d) having been tested under routine testing; and e) planning to test within the next six months (among people who had not tested). </p><p>The following covariates were included in our analyses: 1) age (continuous); 2) sex; 3) income (≥population mean, <population mean); 4) education (≥high school, <high school); 5) residence type (rural, urban, urban village); 6) marital status (married, living with partner, other); 7) knowledge surrounding HIV/AIDS; 8) HIV-related stigma; 9) symptoms of depression (dichotomous variable); 10) frequency of visits to a medical provider (<once/year, 1–2 times/year, ≥3 times/year); 11) perceived access to good quality medical clinics; 12) access to ARV therapy in the respondent's community; 13) access to HIV testing sites; 14) projected HIV stigma (continuous index); 15) history of not being consistently treated with respect by health providers; 16) health status (very good or good versus fair or poor); and 17) history of inconsistent condom use over the past year.</p><p>We used descriptive statistics to examine sample characteristics and experiences with and attitudes toward testing. We then conducted univariate and multivariate logistic regression analyses to explore the independent association of each covariate with having heard of routine testing, having had a prior HIV test, having been tested for HIV by routine testing, and planning to test for HIV among those not tested. We used ordinal logistic regression to assess factors associated with number of positive attitudes toward routine testing. All variables from our conceptual model were included in our final multivariate models. Regression diagnostic procedures yielded no evidence of multi-collinearity or overly influential outliers in any of the models. No variable had more than 3% missing data.</p></sec></sec><sec id="s3"><title>Results</title><sec id="s3a"><title>Description of Study Population</title><p>1,268 (89%) respondents completed the survey. Among 165 non-respondents, 46 (28%) were unavailable after two repeat visits, 78 (47%) refused or did not meet criteria, and 41 (25%) were unable to complete the interview. Demographic and behavioral characteristics of the study population are shown in <xref ref-type="table" rid="pmed-0030261-t003">Table 3</xref>. </p><table-wrap id="pmed-0030261-t003" position="float"><label>Table 3</label><caption><p>Baseline Characteristics of Respondents ( <italic>n</italic> = 1,268) </p></caption><graphic xlink:href="pmed.0030261.t003"/></table-wrap></sec><sec id="s3b"><title>Knowledge of and Attitudes toward Routine Testing</title><p>Fifty-four percent of respondents had heard of routine testing before the survey interview ( <xref ref-type="table" rid="pmed-0030261-t001">Table 1</xref>). After adjusting for all independent variables (see <xref ref-type="fig" rid="pmed-0030261-g001">Figure 1</xref>), higher education, higher income, being married, having better health, and having more frequent medical visits were each associated with higher odds of having heard of routine testing ( <xref ref-type="table" rid="pmed-0030261-t002">Table 2</xref>, column 2). Respondents who reported stigmatizing attitudes toward PLWA had lower odds of having heard of routine testing (AOR = 0.59, 95% CI = 0.45–0.76), as did respondents with more fears of being stigmatized if they tested positive, people in rural areas, and people who reported inconsistent condom use. </p></sec><sec id="s3c"><title>Attitudes toward Routine Testing</title><p>Although approximately half of respondents had heard of routine testing before the interview, a majority endorsed positive views toward routine testing after the policy was explained ( <xref ref-type="table" rid="pmed-0030261-t001">Table 1</xref>). Eighty-one percent were “very much” or “extremely” in favor of routine testing. A majority agreed that routine testing results in decreased discrimination of HIV-positive people (60%), leads to decreased violence against women (55%), and makes it easier for people to get tested (89%) and to gain access to ART (93%). On the other hand, 43% believed that routine testing would cause people to avoid seeing their health provider for fear of being tested, and 14% thought that routine testing would lead to more violence against women. There were fewer than 2% mutually incompatible response pairs in each of our questions on routine testing. </p><p>In ordinal logistic regression analyses, with number of positive views toward routine testing as the outcome variable ( <xref ref-type="table" rid="pmed-0030261-t002">Table 2</xref>, column 3), those with more fears of being stigmatized themselves if they tested positive for HIV had significantly fewer positive views than those without such fears. Similarly, those who reported unsafe sexual practices had fewer positive attitudes. </p></sec><sec id="s3d"><title>Prevalence and Correlates of Having Been Tested for HIV</title><p>Overall, 605 respondents (48%) reported that they had been tested for HIV. The proportion tested according to demographic and behavioral attributes are shown in <xref ref-type="table" rid="pmed-0030261-t004">Table 4</xref>. Factors associated with having been tested in unadjusted analyses included: older age, female gender, higher education, higher income, self-reported excellent or good health status, more frequent medical visits, access to good healthcare, access to HIV testing, being treated with respect consistently by health providers, lack of stigmatizing attitudes toward PLWA, and consistent condom use ( <xref ref-type="table" rid="pmed-0030261-t004">Table 4</xref>). In adjusted analyses, associations remained among all these covariates and HIV testing except for older age, higher income, and being treated with respect consistently by health providers. </p><table-wrap id="pmed-0030261-t004" position="float"><label>Table 4</label><caption><p>Unadjusted and Adjusted Odds of Having Been Tested for HIV ( <italic>n</italic> = 1,164) </p></caption><graphic xlink:href="pmed.0030261.t004"/></table-wrap><p>Fifteen percent of tested respondents reported having been tested by routine testing. In a multivariate logistic regression model with being tested by routine testing as the dependent outcome, and including all covariates listed in <xref ref-type="fig" rid="pmed-0030261-g001">Figure 1</xref>, the only independent correlates of getting routine testing were being married (AOR = 2.67, CI = 1.29–5.53) and seeing the medical provider more than three times per year (AOR = 2.95, CI = 1.41–6.20). In addition, people who held stigmatizing attitudes toward PLWA were significantly less likely to get routine testing (AOR = 0.52, CI = 0.32–0.84). </p></sec><sec id="s3e"><title>Experiences with Testing</title><p>Among those tested, 54% were tested at VCT centers, 26% at public hospitals, and the rest were tested in outpatient clinics or private hospitals. Sixty-two percent of participants who had undergone testing reported that they last tested in 2004 (after the introduction of routine testing). Almost all respondents who had been tested reported that they made the decision on their own to get tested (93%); however, 68% of participants believed that they could not refuse the HIV test whether or not they had made the initial decision to test. Ninety-eight percent reported no ill treatment related to testing, and an equal proportion claimed that they did not regret getting tested. Most participants had obtained the results of their tests (94%) and reported that confidentiality had been strictly maintained at the testing centers (95%). Nearly all participants (99%) denied that their partner had hit or threatened them for being tested. Ninety-six percent reported receiving pre-testing counseling, 90% reported receiving post-testing counseling, and 92% reported that their experience with testing led them to encourage others to undergo testing. Individuals who tested by VCT reported pre-test counseling more frequently than those who tested by routine testing (97% versus 93%, <italic>p</italic> = 0.04) and less poor treatment from others related to testing (2% versus 6%, <italic>p</italic> = 0.03). </p></sec><sec id="s3f"><title>Barriers and Facilitators to Testing</title><p> <xref ref-type="table" rid="pmed-0030261-t005">Table 5</xref> presents reported impediments to HIV testing among respondents who had not been tested ( <italic>n</italic> = 658). Participants indicated whether any of the listed factors served as a barrier for them; they could agree with multiple possible responses. Almost half agreed that a key barrier to testing was that they were “afraid to know” if they were HIV-positive. Forty-three percent reported that they had no reason to believe that they were infected, and 33% did not test because testing positive would force them to stop some of their sexual practices. There were several significant gender differences in the cited barriers to testing. Women were significantly more likely than men to report lack of permission from their spouse or partner (10% versus 3%, <italic>p</italic> < 0.01). Men were more likely to cite frequent migration (25% versus 15%, <italic>p</italic> = 0.01), not wanting to change sexual practices (39% versus 27%, <italic>p</italic> < 0.01), and concerns about lack of social supports if they tested positive (20% versus 12%, <italic>p</italic> < 0.01). </p><table-wrap id="pmed-0030261-t005" position="float"><label>Table 5</label><caption><p>Principal Barriers to Testing among Respondents Who Had Not Been Tested ( <italic>n</italic> = 664) </p></caption><graphic xlink:href="pmed.0030261.t005"/></table-wrap><p>Among those who had not been tested, 71% reported that they intended to be tested in the next six months. The most commonly cited factors that would facilitate testing included knowing that they could get treatment for HIV/AIDS (67%), and being tested with their spouse or main partner (64%). In a multivariate logistic model assessing planning to test as the dependent outcome and including all of our hypothesized correlates of getting tested ( <xref ref-type="fig" rid="pmed-0030261-g001">Figure 1</xref>), respondents with stigmatizing attitudes had significantly lower odds of planning to get tested than those without stigmatizing attitudes (AOR = 0.44, 95% CI = 0.28–0.70). Respondents who reported unprotected sex had significantly higher odds of planning to test (AOR = 2.21, 95% CI = 1.42–3.44). The only other respondent characteristics associated with planning to test were urban location and self-reported very good or good health. </p><p>The most common facilitating factors among those tested were TV or radio messages (69%), knowing that treatment was available (65%), and knowing that the test results would be confidential (64%) ( <xref ref-type="table" rid="pmed-0030261-t006">Table 6</xref>). Men were significantly more likely than women to list treatment availability (74% versus 58%, <italic>p</italic> < 0.01), advice from family or friends (44% versus 34%, <italic>p</italic> = 0.02), messages from the media (77% versus 63%, <italic>p</italic> < 0.01), encouragement or support from someone who had been tested (55% versus 33%, <italic>p</italic> < 0.01), and confidentiality of testing (74% versus 56%, <italic>p</italic> < 0.01) as factors that influenced them to get tested. Women were significantly more likely to report encouragement from prenatal programs (31% versus 13%, <italic>p</italic> < 0.01) as a facilitator to testing. </p><table-wrap id="pmed-0030261-t006" position="float"><label>Table 6</label><caption><p>Principal Facilitators to Testing among Respondents Who Had Been Tested ( <italic>n</italic> = 590) </p></caption><graphic xlink:href="pmed.0030261.t006"/></table-wrap></sec></sec><sec id="s4"><title>Discussion</title><p>This is the first study to our knowledge to assess knowledge, attitudes, and experiences surrounding the new policy of routine testing in Botswana. We found that 11 months after the introduction of routine testing in Botswana, there was widespread support for this policy in a population-based survey, with 81% of participants reporting that they were either extremely or very much in favor of routine testing and an additional 8% reporting that they were somewhat in favor of this policy. A majority of respondents felt that routine testing would decrease barriers to testing, HIV-related stigma, and violence toward women, and would increase uptake of ARVs through the Botswana National Treatment Program. These results, in conjunction with data showing a significant increase in testing and treatment uptake since the introduction of this policy [ <xref rid="pmed-0030261-b009" ref-type="bibr">9</xref>, <xref rid="pmed-0030261-b012" ref-type="bibr">12</xref>, <xref rid="pmed-0030261-b044" ref-type="bibr">44</xref>], suggest that this policy is beneficial in improving access to testing and life-saving treatment. Consistent with this, a study of several prenatal clinics in Botswana found that routine prenatal HIV testing was associated with a 15% increase in the proportion of pregnant women undergoing HIV testing between February and April 2004 (after routine testing was introduced) when compared with the last few months of 2003 [ <xref rid="pmed-0030261-b044" ref-type="bibr">44</xref>]. Figures also indicate a more than 2-fold increase in treatment enrollment since the introduction of this policy, with 42,000 individuals enrolled in treatment as of March 2005 [ <xref rid="pmed-0030261-b045" ref-type="bibr">45</xref>]. </p><p>Evaluating our findings in the context of potential human rights burdens, we found that few individuals reported violence (1%), discrimination (2%), or a breach of confidentiality by healthcare workers (5%) associated with VCT or routine testing. Highlighting some potential problems with routine testing, 43% believed that routine testing would lead to avoidance of doctor visits for fear of being tested, and 14% felt that this policy could lead to increased violence against women. In addition, approximately two-thirds of participants who were tested by either routine testing or VCT felt that they could not refuse the HIV test, suggesting that the voluntary nature of both routine testing and VCT is not fully understood. These findings underscore the importance of implementing HIV testing policies with measures in place to ensure informed consent, protection of confidentiality, and protection of women from gender-based violence related to testing. Careful monitoring and evaluation of Botswana's routine testing program will help to ensure that the significant benefits of this program in terms of linkage to care and prevention of HIV transmission are not associated with potential adverse impacts.</p><p>Detailed guidelines for the implementation of routine testing were not introduced until February 2004, and the training of healthcare practitioners and the development of training materials were still ongoing in early 2005 [ <xref rid="pmed-0030261-b015" ref-type="bibr">15</xref>]. Consequently, at the time of our study, there was still some confusion surrounding the details of implementation of this policy, including the extent to which routine testing should be provided as opt-out (all patients are automatically tested unless they refuse) or as routine-offer (all patients are offered a test, and they must provide explicit informed consent). The current policy has moved toward routine-offer HIV testing in accordance with the recommendations of WHO and UNAIDS; both organizations emphasize that the central principles of testing should include confidentiality, counseling, and informed consent [ <xref rid="pmed-0030261-b022" ref-type="bibr">22</xref>, <xref rid="pmed-0030261-b024" ref-type="bibr">24</xref>, <xref rid="pmed-0030261-b047" ref-type="bibr">47</xref>]. As counseling has been found to account for some of the benefits of VCT in terms of reduced HIV transmission risk behavior [ <xref rid="pmed-0030261-b046" ref-type="bibr">46</xref>] and linkage to subsequent medical care, reinforcing the importance of counseling in routine testing programs may help ensure that these programs help to maximize sexual-risk reduction and access to care. Additional conditions should be considered when implementing routine testing policies elsewhere, including the need to increase human resources and to expand the use of rapid testing. </p><p>Consistent with the documented role of HIV-related stigma as an impediment to testing in studies in Africa and elsewhere [ <xref rid="pmed-0030261-b006" ref-type="bibr">6</xref>, <xref rid="pmed-0030261-b031" ref-type="bibr">31</xref>, <xref rid="pmed-0030261-b037" ref-type="bibr">37</xref>, <xref rid="pmed-0030261-b048" ref-type="bibr">48</xref>– <xref rid="pmed-0030261-b050" ref-type="bibr">50</xref>], we found that HIV-related stigma was associated with decreased odds of having been tested for HIV, of getting routine testing, and of planning to test among people not previously tested. In addition, respondents with more stigmatizing views about HIV and a greater number of fears related to HIV stigma were significantly less likely to have heard of routine testing after adjusting for possible confounders, attesting to the association between poor information and HIV-related stigma. Addressing HIV-related stigma should comprise an integral part of ongoing HIV testing programs and policies in Botswana, including measures to protect people with HIV/AIDS from discrimination in healthcare, work, and other settings. Policies that target HIV-related stigma may also prevent a reduction of clinical visits related to people's fears of being tested. Increasing testing and decreasing stigma will likely work together to reinforce one another, with more testing leading to a reduction in HIV-related stigma, which in turn will work to further increase testing. Botswana already has several innovative programs in place aimed to address stigma directly, including media campaigns, the public testing of President Festus Mogae and other national leaders, and the annual “Miss HIV Stigma Free” competition [ <xref rid="pmed-0030261-b011" ref-type="bibr">11</xref>, <xref rid="pmed-0030261-b014" ref-type="bibr">14</xref>]. Additional progress toward stigma reduction will require a deeper understanding of the structural dimensions of HIV-related stigma, and the mechanisms by which stigma reinforces and generates social inequalities related to gender, ethnicity, and class [ <xref rid="pmed-0030261-b051" ref-type="bibr">51</xref>]. </p><p>We found a relatively high prevalence of self-reported HIV testing in Botswana in the era of routine testing, compared with its neighboring countries. While 48% of our sample reported having been tested for HIV, results from Zimbabwe suggest that only 10%–12% of people are aware of their HIV status [ <xref rid="pmed-0030261-b052" ref-type="bibr">52</xref>], and a nationwide community based–survey in South Africa in 2002 found that only 20% of people aware of VCT services had been tested for HIV [ <xref rid="pmed-0030261-b053" ref-type="bibr">53</xref>]. In addition to the policy of routine testing, universal access to ARVs and to HIV testing likely contributes to the relatively high prevalence of testing in Botswana. Consistent with this, perceived access to testing was associated with 60% higher odds of having received an HIV test among respondents in our study, and the availability of ART was cited as a leading facilitator to testing. In addition, a national survey from Botswana in 2001 showed that fewer than 20% of individuals ages 15–49 had ever received an HIV test [ <xref rid="pmed-0030261-b054" ref-type="bibr">54</xref>], suggesting a more than 2-fold increase in testing prevalence since the introduction of both universal ART access and routine testing. On the other hand, because over 50% of our sample had not yet been tested, our results reinforce the fact that availability of testing facilities and ART, while essential, may not be sufficient to guarantee HIV testing for many [ <xref rid="pmed-0030261-b031" ref-type="bibr">31</xref>]. </p><p>Study results should be interpreted in the context of a number of limitations. First, as this study was cross-sectional, causality cannot be determined from our findings. Second, while we interviewed individuals from both rural and urban areas, and covered the five most populated districts in Botswana, because we did not interview individuals in all districts of Botswana, our results may not be generalizable to the entire Botswana population. In addition, Botswana has a number of unique features that may limit generalizability to neighboring African countries, such as its relatively high per capita income, comparatively extensive healthcare infrastructure, strong donor involvement, and strong government commitment to combating HIV. Third, as the policy of routine testing was not yet implemented in a uniform way across all medical facilities in Botswana, and different facilities were at different stages of implementation, it was impossible to conduct a more systematic evaluation of the impacts of this policy. Moreover, since routine testing is a relatively new policy in Botswana, only a small proportion of those tested (15%) had been tested by routine testing at the time of our study. Finally, self-report can introduce misclassification and bias. To maximize validity we did not ask about HIV status, assured confidentiality and privacy, and asked survey questions in a culturally sensitive, nonjudgmental manner. To reduce social desirability bias, interviewers were not informed of key research hypotheses, and study aims were presented to participants in general terms.</p><sec id="s4a"><title>Concluding Remarks</title><p>In the face of a devastating epidemic that has already infected nearly half of its adult population, the government of Botswana has taken strong steps to improve access to testing and to ensure the right to life-sustaining treatment for all of its citizens. Early evidence of widespread support for the policy of routine testing in this study holds significant promise for the prevention and treatment of HIV/AIDS in Botswana and elsewhere. Concerted efforts to scale up HIV testing, however, must also be accompanied by appropriate monitoring of testing practices to ensure that they are implemented in accordance with international guidelines on human rights and HIV/AIDS [ <xref rid="pmed-0030261-b055" ref-type="bibr">55</xref>, <xref rid="pmed-0030261-b056" ref-type="bibr">56</xref>]. </p></sec></sec><sec sec-type="supplementary-material" id="s5"><title>Supporting Information</title><supplementary-material content-type="local-data" id="sd001"><label>Protocol S1</label><caption><title>Botswana Community Survey</title><p>(338 KB DOC)</p></caption><media xlink:href="pmed.0030261.sd001.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material></sec>
<italic>MPLW515L</italic> Is a Novel Somatic Activating Mutation in Myelofibrosis with Myeloid Metaplasia	<sec id="st1"><title>Background</title><p>The <italic>JAK2V617F</italic> allele has recently been identified in patients with polycythemia vera (PV), essential thrombocytosis (ET), and myelofibrosis with myeloid metaplasia (MF). Subsequent analysis has shown that constitutive activation of the JAK-STAT signal transduction pathway is an important pathogenetic event in these patients, and that enzymatic inhibition of JAK2V617F may be of therapeutic benefit in this context. However, a significant proportion of patients with ET or MF are <italic>JAK2V617F-</italic>negative. We hypothesized that activation of the JAK-STAT pathway might also occur as a consequence of activating mutations in certain hematopoietic-specific cytokine receptors, including the erythropoietin receptor (EPOR), the thrombopoietin receptor (MPL), or the granulocyte-colony stimulating factor receptor (GCSFR). </p></sec><sec id="st2"><title>Methods and Findings</title><p>DNA sequence analysis of the exons encoding the transmembrane and juxtamembrane domains of EPOR, MPL, and GCSFR, and comparison with germline DNA derived from buccal swabs, identified a somatic activating mutation in the transmembrane domain of MPL (W515L) in 9% (4/45) of <italic>JAKV617F-</italic>negative MF. Expression of MPLW515L in 32D, UT7, or Ba/F3 cells conferred cytokine-independent growth and thrombopoietin hypersensitivity, and resulted in constitutive phosphorylation of JAK2, STAT3, STAT5, AKT, and ERK. Furthermore, a small molecule JAK kinase inhibitor inhibited MPLW515L-mediated proliferation and JAK-STAT signaling in vitro. In a murine bone marrow transplant assay, expression of MPLW515L, but not wild-type MPL, resulted in a fully penetrant myeloproliferative disorder characterized by marked thrombocytosis (Plt count 1.9–4.0 × 10 <sup>12</sup>/L), marked splenomegaly due to extramedullary hematopoiesis, and increased reticulin fibrosis. </p></sec><sec id="st3"><title>Conclusions</title><p>Activation of JAK-STAT signaling via MPLW515L is an important pathogenetic event in patients with <italic>JAK2V617F-</italic>negative MF. The bone marrow transplant model of MPLW515L-mediated myeloproliferative disorders (MPD) exhibits certain features of human MF, including extramedullary hematopoiesis, splenomegaly, and megakaryocytic proliferation. Further analysis of positive and negative regulators of the JAK-STAT pathway is warranted in <italic>JAK2V617F-</italic>negative MPD. </p></sec>	<contrib contrib-type="author"><name><surname>Pikman</surname><given-names>Yana</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Lee</surname><given-names>Benjamin H</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Mercher</surname><given-names>Thomas</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>McDowell</surname><given-names>Elizabeth</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Ebert</surname><given-names>Benjamin L</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref><xref ref-type="aff" rid="aff3"> <sup>3</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Gozo</surname><given-names>Maricel</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Cuker</surname><given-names>Adam</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Wernig</surname><given-names>Gerlinde</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Moore</surname><given-names>Sandra</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Galinsky</surname><given-names>Ilene</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>DeAngelo</surname><given-names>Daniel J</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Clark</surname><given-names>Jennifer J</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Lee</surname><given-names>Stephanie J</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Golub</surname><given-names>Todd R</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref><xref ref-type="aff" rid="aff3"> <sup>3</sup> </xref><xref ref-type="aff" rid="aff4"> <sup>4</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Wadleigh</surname><given-names>Martha</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Gilliland</surname><given-names>D. Gary</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref><xref ref-type="aff" rid="aff4"> <sup>4</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Levine</surname><given-names>Ross L</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref><xref ref-type="corresp" rid="cor1">*</xref></contrib>	PLoS Medicine	<sec id="s1"><title>Introduction</title><p>The BCR-ABL negative chronic myeloproliferative disorders (MPD) include polycythemia vera (PV), essential thrombocytosis (ET), and myelofibrosis with myeloid metaplasia (MF) [ <xref rid="pmed-0030270-b001" ref-type="bibr">1</xref>]. Although clonal hematopoiesis was observed in these disorders more than three decades ago, the molecular etiology of these disorders was not known until recently when several groups reported a somatic mutation in the JAK2 tyrosine kinase ( <italic>JAK2V617F</italic>) in most patients with PV and in a subset of patients with ET and MF [ <xref rid="pmed-0030270-b002" ref-type="bibr">2</xref>– <xref rid="pmed-0030270-b006" ref-type="bibr">6</xref>]. Recent estimates using high sensitivity mutation detection techniques indicate that <italic>JAK2V617F</italic> is present in ˜95–100% of PV, 60%–70% of ET, and 50% of MF [ <xref rid="pmed-0030270-b007" ref-type="bibr">7</xref>, <xref rid="pmed-0030270-b008" ref-type="bibr">8</xref>]. JAK2V617F is a constitutively active tyrosine kinase [ <xref rid="pmed-0030270-b009" ref-type="bibr">9</xref>] that activates downstream signal transduction pathways and transforms hematopoietic cells to cytokine-independent growth [ <xref rid="pmed-0030270-b004" ref-type="bibr">4</xref>, <xref rid="pmed-0030270-b010" ref-type="bibr">10</xref>], and these cells are sensitive to a small molecule JAK Inhibitor [ <xref rid="pmed-0030270-b002" ref-type="bibr">2</xref>]. In addition, expression of JAK2V617F in a murine bone marrow transplant assay results in a MPD most similar to PV [ <xref rid="pmed-0030270-b004" ref-type="bibr">4</xref>, <xref rid="pmed-0030270-b011" ref-type="bibr">11</xref>]. These data indicate that constitutive activation of JAK-STAT signaling by the mutant JAK2V617F kinase plays a central role in the pathogenesis of <italic>JAK2V617F-</italic>positive PV, ET, and MF. </p><p>Despite the recent discovery of the <italic>JAK2V617F</italic> allele, questions remain regarding the molecular pathogenesis of PV, ET, and MF. In particular, the mutation(s) responsible for <italic>JAK2V617F-</italic>negative ET and MF remain to be identified. Although more sensitive techniques can identify <italic>JAK2V617F</italic> mutations in a small proportion of clonal cells, we recently demonstrated that the majority of <italic>JAK2V617F-</italic>negative ET and MF patients have clonal granulocytes [ <xref rid="pmed-0030270-b007" ref-type="bibr">7</xref>], suggesting that alternative mutation(s) result in clonal granulopoiesis in this subset of patients. </p><p>We hypothesized that <italic>JAK2V617F-</italic>negative patients with MPD might have mutations involving other components of the JAK-STAT signal transduction pathway that include cytokine receptors, other JAK family members, or STAT family members. Screens for additional mutations in other JAK-STAT signaling intermediates in this clinical context including <italic>JAK1</italic>, <italic>JAK3</italic>, <italic>TYK2</italic>, <italic>STAT3</italic>, or <italic>STAT5</italic> have been negative [ <xref rid="pmed-0030270-b012" ref-type="bibr">12</xref>]. </p><p>We have recently shown that expression of a homodimeric type I cytokine receptor, such as the erythropoietin receptor (EPOR), the thrombopoietin receptor (MPL), or the granulocyte-colony stimulating factor receptor (GCSFR), is required for JAK2V617F-mediated transformation of hematopoietic cells and for activation of downstream signaling [ <xref rid="pmed-0030270-b010" ref-type="bibr">10</xref>]. These data suggested the possibility that mutations in the regions of these cytokine receptors that are critical for receptor dimerization (transmembrane domain) and for JAK2 binding (juxtamembrane domain) might lead to activation of JAK-STAT signaling in <italic>JAK2V617F-</italic>negative MPD. Indeed, such mutations in the context of <italic>EPOR</italic> have been identified in rare familial cases of polycythemia, though these have not been reported in acquired MPD. Heretofore, <italic>MPL</italic> has been sequenced in a small cohort of patients with MF and ET, but no mutations were identified [ <xref rid="pmed-0030270-b013" ref-type="bibr">13</xref>], and multiple groups have reported the absence of <italic>EPOR</italic> mutations in small numbers of patients with PV [ <xref rid="pmed-0030270-b014" ref-type="bibr">14</xref>, <xref rid="pmed-0030270-b015" ref-type="bibr">15</xref>]. High throughput DNA sequence analysis and the collection of a large number of MPD patient samples [ <xref rid="pmed-0030270-b002" ref-type="bibr">2</xref>] has enabled evaluation of a larger series of patients for mutations in candidate genes, including cytokine receptors. We therefore investigated patients with <italic>JAK2V617F-</italic>negative MPD for somatic activating mutations in <italic>EPOR, MPL,</italic> or <italic>GCSFR</italic>. </p></sec><sec id="s2"><title>Methods</title><sec id="s2a"><title>Sample Collection</title><p>Granulocyte DNA samples and matched normal DNA samples were collected from patients with PV, ET, and MF who were enrolled in the Harvard Myeloproliferative Disorders Study as previously described [ <xref rid="pmed-0030270-b002" ref-type="bibr">2</xref>]. Additional samples from patients with MF were collected on a separate Dana-Farber Cancer Institute protocol. All subjects provided informed consent on protocols approved by the Dana-Farber Cancer Institute Institutional Review Board. </p></sec><sec id="s2b"><title>DNA Sequence Analysis and Genotyping for <italic>MPLW515L</italic> </title><p>PCR amplification and DNA sequencing of select exons of <italic>MPL, EPOR,</italic> and <italic>GCSFR</italic> was performed using M13-tailed primers as previously described [ <xref rid="pmed-0030270-b002" ref-type="bibr">2</xref>], and specific primer sequences are listed in <xref ref-type="supplementary-material" rid="st001">Table S1</xref>. Sequence analysis of bidirectional sequence traces was performed using Mutation Surveyor version 2.28 (SoftGenetics, State College, Pennsylvania, United States). Candidate mutations were reamplified and sequenced from original DNA for independent verification, and sequence analysis of buccal DNA was performed to ascertain whether non-synonymous mutations were constitutional or somatic in origin. Identity between granulocyte and buccal DNA for individual patients was confirmed using eight informative synonymous single nucleotide polymorphisms. Genotypic analysis of the HapMap panel of normal patients was performed using a mass spectrometric assay as previously described [ <xref rid="pmed-0030270-b002" ref-type="bibr">2</xref>]. </p></sec><sec id="s2c"><title>Expression Vectors and Cell Culture</title><p>The MSCV-MPL-Neo and MSCV-MPL-IRES-EGFP retroviral vectors were generously provided by W. Tong and H. Lodish. The <italic>MPLW515L</italic> mutation was generated using site-directed mutagenesis (Quickchange-XL, Stratagene, La Jolla, California, United States) and confirmed by full-length DNA sequencing. 293T cells were grown in DMEM with 10% fetal bovine serum (FBS). Transient co-transfection of 293T cells and generation of retroviral supernatant were performed using Fugene (Roche, Nutley, New Jersey, United States) according to manufacturer's guidelines. 32D and Ba/F3 cells were grown in RPMI medium 1640 containing 10% FBS and 10% WEHI-3B cell supernatant as a source of IL3. UT7 cells were grown in IMDM medium supplemented with 10% FBS and GMCSF (1 ng/mL). 32D cells, Ba/F3 cells, and murine bone marrow cells were transduced with viral supernatant, and UT7 cells were transduced by electroporation (Amaxa, Gaithersburg, Maryland, United States). Cells were then selected in G418 (1mg/mL). To assess for factor-independent growth and for thrombopoietin hypersensitivity, cells were first washed three times in PBS, and then 1 × 10 <sup>5</sup> viable cells/mL were resuspended in appropriate cytokine-free media or in media containing different concentrations of thrombopoietin (Sigma, St. Louis, Missouri, United States) as noted. The number of viable cells was determined by trypan blue exclusion. </p></sec><sec id="s2d"><title>Western Blot Analysis</title><p>Cells were collected and lysed in lysis buffer and separated by electrophoresis as described previously [ <xref rid="pmed-0030270-b016" ref-type="bibr">16</xref>]. Nitrocellulose membrane was blocked in TBST/5% milk and incubated with one of the following antibodies: anti-pSTAT5 (Cell Signaling, Beverly, Massachusetts, United States), anti-pSTAT3 (Cell Signaling), anti-pERK (Cell Signaling), anti-pAKT (Cell Signaling), anti-pTYK2 (Cell Signaling), or anti-pJAK2 (Cell Signaling). For loading controls, blots were stripped and reprobed using anti-JAK2 (Santa Cruz Biotechnology, Santa Cruz, California, United States), anti-TYK2 (Santa Cruz Biotechnology), anti-STAT5 (Santa Cruz Biotechnology), anti-STAT3 (Cell Signaling), anti-ERK (Cell Signaling), anti-AKT (Cell Signaling), and anti-TYK2 (Santa Cruz Biotechnology) antibodies as appropriate. </p></sec><sec id="s2e"><title>JAK Inhibitor Assays</title><p>For cell-proliferation assays, 32D cells stably expressing FIP1L1-PDGFRA [ <xref rid="pmed-0030270-b016" ref-type="bibr">16</xref>] and MPLW515L were grown in RPMI/10% FBS and then incubated with appropriate concentrations of JAK Inhibitor I (Calbiochem, San Diego, California, United States) for 72 hours. 32D cells expressing MPLWT were grown in RPMI/10%FBS/mTPO in the presence of JAK Inhibitor I. The number of viable cells was determined using CellTiter 96 Aqueous One Cell Proliferation Assay (Promega, San Luis Obispo, California, United States). For Western blotting, cell lines were incubated in the presence of JAK Inhibitor I for four hours, and phosphorylation of relevant proteins was assessed as described above. </p></sec><sec id="s2f"><title>Murine Bone Marrow Transplant Assay</title><p>The murine bone marrow transplant assay was performed as previously described [ <xref rid="pmed-0030270-b017" ref-type="bibr">17</xref>]. MSCV retroviral supernatants were titered by determining the percentage of GFP positive cells 48 hours after infection of Ba/F3 and 32D cells (1 mL supernatant used to infect 1 × 10 <sup>6</sup> cells). MPLWT and MPLW515L retroviral supernatants were able to reproducibly infect 40%–60% of Ba/F3 and 32D cells as assessed by flow cytometry. Balb/C donor mice were treated with 5-Florouracil (150 mg/kg) seven days prior to bone marrow harvest. Bone marrow cells were harvested from donor mice, treated with red blood cell lysis buffer and cultured for 24 hours in transplantation medium (RPMI + 10% FBS + 6 ng/mL IL-3, 10 ng/mL IL-6, and 10 ng/mL stem-cell factor (SCF)). Cells were treated by spin infection with retroviral supernatants (1 mL supernatant per 4 × 10 <sup>6</sup> cells, plus polybrene) and centrifuged at 1,800 <italic>g</italic> for 90 minutes. The spin infection was repeated 24 hours later. The cells were then washed and resuspended in Hank's balanced salt solution, and injected into lateral tail veins of lethally irradiated (2 × 4.5 Gy [450 rad]) Balb/C recipient mice (Taconic, Germantown, New York, United States) at 0.5 to 1.0 × 10 <sup>6</sup> cells/mouse. Animals were humanely killed when they had palpable splenomegaly or were moribund. Animals transplanted with bone marrow that was transduced with MPLWT viral supernatant were killed at the time of manifestation of MPLW515L disease for endpoint analysis. </p></sec><sec id="s2g"><title>Mouse Analysis</title><p>Peripheral blood was collected from the retro-orbital cavity using EDTA-treated glass capillary tubes and analyzed by automated complete and differential blood cell counts and blood smears (Wright-Giemsa stained). Single-cell suspensions of spleen and bone marrow were prepared by pressing tissue through a cell strainer, followed by red blood cell lysis. For histopathology, tissues were fixed in formalin and then embedded in paraffin for histology analysis as previously described [ <xref rid="pmed-0030270-b018" ref-type="bibr">18</xref>]. </p><p>For flow cytometry, cells were washed in PBS + 1% bovine serum albumin and stained with monoclonal antibodies in PBS + 1% bovine serum albumin for 30 minutes on ice. Antibodies used were allophycocyanin–conjugated Gr-1, B220, Ter119, CD8, cKit; and phycoerythrin–conjugated Mac-1, CD19, CD71, CD4, CD41 (BD Pharmingen, San Diego, California, United States). Flow cytometry analysis was performed on a FACSCalibur instrument (BD Biosciences, San Jose, California, United States) and analyzed with CellQuest software (BD Biosciences). Viability was assessed by incubating cells with 7-AAD (7-amino-actinomycin D) (BD Pharmingen) for five minutes prior to flow cytometry. Cells were gated for viability (using forward/side scatter and 7-AAD) and GFP positivity, and 10,000 events were analyzed from this subset for marker expression.</p><p>The study protocol was reviewed and approved by the Boston (Massachusetts, United States) Children's Hospital Animal Care and Use Committee.</p></sec><sec id="s2h"><title>In Vitro Colony-Forming Assays</title><p>Myeloid colony–forming assays were performed in methylcellulose-based medium (M3434) containing 3U/mL erythropoietin, 10 ng/mL IL-3, 10 ng/mL IL-6, and 50 ng/mL SCF, and in M3231 medium containing no additional cytokines, as per manufacturer's protocols (StemCell Technologies, Vancouver, British Columbia, Canada). Cells were plated at 1.5 × 10 <sup>4</sup> cells per dish for bone marrow cells and 5 × 10 <sup>4</sup> cells per dish for spleen cells, in duplicate. Plates were incubated at 37 °C for seven days. Colonies were scored by morphology, and representative colonies were confirmed by Wright-Giemsa stained slides of colony cytospins. Graphs represent total number of colonies counted for three representative mice in each group, with two plates counted for each mouse. Megakaryocyte colony–forming assays were performed in collagen-containing medium (MegaCult-C; StemCell Technologies) containing TPO, IL-11, IL-3, and IL-6 for seven days. Slides were fixed and stained for six hours with acetylcholine iodide (Sigma) as per MegaCult protocol, and were counterstained with Harris hematoxylin solution (Sigma) as per manufacturer's protocol. </p><p>To assess megakaryocyte ploidy, bone marrow cells were cultured for four days in RPMI/10% FBS containing 50 ng/mL mTPO and 10 mg/mL SCF. Cells were collected and stained with FITC-rat anti-mouse CD41 antibody (BD Biosciences), followed by 50 ug/mL propidium iodide in 0.1% sodium citrate buffer with 50 ug/mL RNAase, as previously described [ <xref rid="pmed-0030270-b019" ref-type="bibr">19</xref>]. Data was acquired on a FACSCalibur (BD Biosciences) and analyzed with CellQuest software. </p></sec></sec><sec id="s3"><title>Results</title><sec id="s3a"><title> <italic>MPLW515L</italic> Mutation in JAK2 <italic>V617F</italic>-Negative MPD </title><p>We sequenced the transmembrane and juxtamembrane domains of <italic>EPOR, GCSFR,</italic> and <italic>MPL</italic> in 15 <italic>JAK2V617F</italic>-negative MF patients, 16 <italic>JAK2V617F</italic>-negative patients with ET, and 4 <italic>JAK2V617F</italic>-negative PV patients. Although we found no mutations in <italic>EPOR</italic> or <italic>GCSFR,</italic> we identified a guanine to thymidine substitution in <italic>MPL</italic> at nucleotide 1544 which results in a tryptophan to leucine substitution at codon 515 ( <italic>MPLW515L</italic>) in two patients with MF ( <xref ref-type="fig" rid="pmed-0030270-g001">Figure 1</xref>A and <xref ref-type="fig" rid="pmed-0030270-g001">1</xref>B). Sequencing of 30 additional <italic>JAK2V617F-</italic>negative MF samples identified two additional <italic>MPLW515L</italic>-positive samples, for a total frequency of 4/45 (9%) patients. <italic>MPLW515L</italic> was not detected in any JAK2 <italic>V617F</italic>-negative ET ( <italic>n</italic> = 50) or PV ( <italic>n</italic> = 10) patient samples. Sequence analysis of matched normal DNA derived from buccal swabs did not identify the <italic>MPLW515L</italic> substitution ( <xref ref-type="fig" rid="pmed-0030270-g001">Figure 1</xref>A, upper trace), demonstrating that <italic>MPLW515L</italic> is a somatic mutation in hematopoietic cells. Sequence analysis of the entire open reading frame of <italic>MPL</italic> in all <italic>JAK2V617F-</italic>negative ET and MF patients did not reveal additional somatic mutations. To assess the prevalence of the <italic>MPLW515L</italic> allele in the general population, we genotyped a standard panel of 270 samples collected by the International HapMap Consortium [ <xref rid="pmed-0030270-b020" ref-type="bibr">20</xref>]. All 270 HapMap samples were homozygous for the wild-type <italic>MPLW515</italic> allele. </p><fig id="pmed-0030270-g001" position="float"><label>Figure 1</label><caption><title>MPLW515L Mutation Is Found in JAK2V617F-Negative MF and Causes Cytokine-Independent Growth in 32D and UT7 Cells, and Constitutively Activates the JAK-STAT Signaling Pathway</title><p>(A) Forward (middle trace) and reverse (lower trace) sequence traces demonstrating a heterozygous guanine to thymine substitution (arrows) present in granulocyte DNA from a patient with MF. The mutation is not present in buccal DNA from the same patient (upper trace).</p><p>(B) DNA sequence and protein translation for both the wild-type and mutant MPL alleles. The mutation results in a tryptophan-to-leucine substitution at codon 515.</p><p>(C) <italic>Upper:</italic> 32D cells transduced with MPLW515L exhibit cytokine-independent growth compared with MPLWT (left). Cell lines grown in the presence of IL3 show equal rates of growth (right). Error bars denote the standard deviation for each sample measured in triplicate. <italic>Lower:</italic> UT7 cells transformed with MPLW515L exhibit cytokine-independent growth compared with MPLWT (left). Cell lines grown in the presence of TPO (5 ng/mL) show equal rates of growth (right). Error bars denote the standard deviation for each sample measured in triplicate. </p><p>(D) 32D cells, 32D MPLWT cells, and 32D MPLW515L cells were deprived of cytokines and then analyzed by Western blots, demonstrating phosphorylation of JAK2, STAT5, STAT3, AKT, and ERK in MPLW515L compared with MPLWT.</p></caption><graphic xlink:href="pmed.0030270.g001"/></fig><p>No significant clinical differences were observed between <italic>MPLW515L</italic>-positive versus <italic>MPLW515L</italic>-negative patients with MF with regard to median age at presentation and to disease duration, though the statistical power of this analysis was limited by the small number of patients with MF in our cohort. For all <italic>MPLW515L-</italic>positive MF patients for which bone marrow histopathology was available (3/4), marked megakaryocyte hyperplasia was noted. In addition, two of the four <italic>MPLW515L-</italic>positive patients exhibited leukocytosis and thrombocytosis at the time of disease presentation. One of the four <italic>MPLW515L-</italic>positive patients had a rapidly progressive clinical course, with death two years after initial presentation from complications related to MF. </p></sec><sec id="s3b"><title>Functional Analysis of MPLW515L and Wild-Type MPL</title><p>In cell culture assays, expression of MPLW515L, but not MPLWT, conferred cytokine-independent growth to 32D, UT7, or Ba/F3 cell lines ( <xref ref-type="fig" rid="pmed-0030270-g001">Figure 1</xref>C and unpublished data). Western blotting demonstrated constitutive phosphorylation of the JAK/STAT signaling proteins, including JAK2, STAT3, and STAT5 ( <xref ref-type="fig" rid="pmed-0030270-g001">Figure 1</xref>D). Downstream targets of this pathway were also activated, with constitutive phosphorylation of p42/44 ERK and AKT in 32D MPLW515L cells consistent with activation of the RAS/MAPK and PI3K/AKT pathways ( <xref ref-type="fig" rid="pmed-0030270-g001">Figure 1</xref>D). In addition, expression of MPLW515L in 32D and UT7 cells resulted in hypersensitivity to TPO compared with MPLWT as assessed by cell growth ( <xref ref-type="fig" rid="pmed-0030270-g002">Figure 2</xref>A and <xref ref-type="fig" rid="pmed-0030270-g002">2</xref>B) and by Western blotting after TPO stimulation ( <xref ref-type="fig" rid="pmed-0030270-g002">Figure 2</xref>C and <xref ref-type="fig" rid="pmed-0030270-g002">2</xref>D). </p><fig id="pmed-0030270-g002" position="float"><label>Figure 2</label><caption><title>MPLW515L-Expressing Cells Show Hyper-Responsiveness to TPO</title><p>(A) 32D MPLWT cells and 32D MPLW515L cells were cultured in triplicate at an initial concentration of 1 × 10 <sup>5</sup> cells/mL, in RPMI/10% FBS containing TPO 2 ng/mL, 1 ng/mL, 0.1 ng/mL, 0.01 ng/mL, 0.001 ng/mL, or in the absence of TPO for four days, and then cell numbers were assessed. Error bars denote the standard deviation for each sample measured in triplicate. </p><p>(B) UT7 MPLWT cells and UT7 MPLW515L cells were cultured in triplicate, at an initial concentration of 1 × 10 <sup>5</sup> cells/mL, in IMDM/10% FBS containing TPO 5 ng/mL, 1 ng/mL, 0.1 ng/mL, 0.01 ng/mL, 0.001ng/mL, or in the absence of TPO for six days, and then cell numbers were assessed. Error bars denote the standard deviation for each sample measured in triplicate. </p><p>(C) 32D MPLWT or 32D MPLW515L cells were deprived of cytokine overnight and then stimulated with TPO 5 ng/mL, 1 ng/mL, 0.1 ng/mL, 0.01 ng/mL, or without TPO for seven minutes and then analyzed by Western blot for phosphorylation of JAK2 and STAT3, demonstrating increased JAK2/STAT3 phosphorylation in response to TPO in 32D MPLW515L cells as compared with 32D-MPLWT cells.</p><p>(D) Phosphorylation of JAK2, TYK2, STAT3, STAT5, AKT, and ERK in response to seven-minute stimulation with TPO (50ng/mL). 32D MPLWT or 32 MPLW515L cells were deprived of cytokine overnight and then stimulated with TPO 50 ng/mL for seven minutes and then analyzed by Western blot for phosphorylation of JAK2, TYK2, STAT3, STAT5, AKT, and ERK, demonstrating increased phosphorylation of these proteins in response to TPO in 32D MPLW515L cells as compared with 32D MPLWT cells.</p></caption><graphic xlink:href="pmed.0030270.g002"/></fig></sec><sec id="s3c"><title>A Murine Model for MPLW515L-Induced Myeloproliferative Disease</title><p>To assess the in vivo effects of the MPLW515L mutant receptor, we developed a murine model of MPLW515L-induced myeloproliferative disease. Bone marrow cells were transduced with MSCV-IRES-EGFP vectors containing MPLWT or MPLW515L and transplanted into lethally irradiated Balb/C mice. MPLW515L-transduced animals, but not MPLWT-transduced animals, developed a short latency (median latency = 18 days post BMT), fully penetrant, lethal MPD ( <xref ref-type="fig" rid="pmed-0030270-g003">Figure 3</xref>A) notable for marked thrombocytosis (mean platelet count = 3.414 × 10 <sup>12</sup>/L) as well as leukocytosis (mean WBC = 201 × 10 <sup>9</sup>/L) ( <xref ref-type="fig" rid="pmed-0030270-g003">Figure 3</xref>B). There was no significant difference in the hematocrit of MPLW515L mice compared with MPLWT mice. At the time of sacrifice, MPLW515L-expressing mice showed splenomegaly and hepatomegaly ( <xref ref-type="fig" rid="pmed-0030270-g003">Figure 3</xref>C and <xref ref-type="fig" rid="pmed-0030270-g003">3</xref>D), and staining of bone marrow demonstrated increased reticulin fibrosis in MPLW515L, but not MPLWT-expressing mice ( <xref ref-type="fig" rid="pmed-0030270-g003">Figure 3</xref>E). </p><fig id="pmed-0030270-g003" position="float"><label>Figure 3</label><caption><title>Bone Marrow Transplant with MPLW515L-Transduced Bone Marrow Causes a Rapid Myeloproliferative Disease</title><p>(A) Kaplan-Meier survival plot of Balb/C mice transduced with MPLW515L ( <italic>n</italic> = 9) and MPLWT ( <italic>n</italic> = 6) showing death of all MPLW515L mice between 17 and 32 days post-BMT compared with MPLWT ( <italic>n</italic> = 6), which were sacrificed for endpoint analysis without evidence of disease. </p><p>(B) Complete blood counts show leukocytosis and thrombocytosis in MPLW515L BMT model compared with MPLWT. There is no difference in hematocrits. Standard deviation is indicated.</p><p>(C) Spleen weights of MPLW515L and MPLWT mice shows splenomegaly in MPLW515L mice but not in MPLWT mice with average spleen weight equal to 1,171 mg.</p><p>(D) Liver weights of MPLW515L and MPLWT mice shows hepatomegaly in MPLW515L mice but not in MPLWT mice, with average liver weight equal to 2,390 mg (compared with 1,222 mg in MPLWT mice).</p><p>(E) Bone marrow shows significantly increased bone marrow fibrosis by reticulin staining at 17 days post-BMT in MPLW515L mice, but not in MPLWT-expressing mice.</p></caption><graphic xlink:href="pmed.0030270.g003"/></fig><p>Histopathology analysis was consistent with the presence of a MPD in MPLW515L-expressing mice, but not in MPLWT-expressing mice ( <xref ref-type="fig" rid="pmed-0030270-g004">Figure 4</xref>). Examination of peripheral blood smears from MPLW515L mice revealed marked thrombocytosis and leukocytosis composed of predominantly mature myeloid elements admixed with occasional to frequent numbers of nucleated erythroid cells ( <xref ref-type="fig" rid="pmed-0030270-g004">Figure 4</xref>A and <xref ref-type="fig" rid="pmed-0030270-g004">4</xref>B). Compared with MPLWT animals, MPLW515L mice demonstrated markedly hypercellular bone marrow with a predominance of maturing myeloid forms and with increased numbers of atypical and dysplastic megakaryocytes which could be found in frequent clusters and exhibited emperipolesis of neutrophils in megakaryocyte cytoplasm ( <xref ref-type="fig" rid="pmed-0030270-g004">Figure 4</xref>C and <xref ref-type="fig" rid="pmed-0030270-g004">4</xref>D). Spleens from MPLW515L mice revealed complete effacement of normal splenic architecture ( <xref ref-type="fig" rid="pmed-0030270-g004">Figure 4</xref>E), with a prominent expansion of red pulp composed of an admixture of maturing myeloid and erythroid elements, and numerous abnormal megakaryocytes. By comparison, spleens from MPLWT animals displayed a relative preservation of normal splenic architecture with only a mild expansion of red pulp by maturing erythroid elements ( <xref ref-type="fig" rid="pmed-0030270-g004">Figure 4</xref>F and <xref ref-type="fig" rid="pmed-0030270-g004">4</xref>H) that were also noted in small, focal clusters in the livers of these animals ( <xref ref-type="fig" rid="pmed-0030270-g004">Figure 4</xref>J and <xref ref-type="fig" rid="pmed-0030270-g004">4</xref>L). In contrast, extensive extramedullary hematopoiesis was observed in the livers of MPLW515L animals, which was composed primarily of maturing erythroid elements with frequent atypical megakaryocytes and relatively fewer numbers of mature myeloid cells ( <xref ref-type="fig" rid="pmed-0030270-g004">Figure 4</xref>I and <xref ref-type="fig" rid="pmed-0030270-g004">4</xref>K). </p><fig id="pmed-0030270-g004" position="float"><label>Figure 4</label><caption><title>MPLW515L and MPLWT Bone Marrow Transplant Model Histopathology</title><p>Histology of MPLW515L-transduced and MPLWT-transduced Balb/C mice showing images of peripheral blood (A and B) and histopathology in representative sections of bone marrow (C and D), spleen (E–H), and liver (I–L). Peripheral blood smear (B) (600×, Wright-Giemsa) of a representative MPLWT animal displays an unremarkable white blood cell and platelet count. In contrast, peripheral blood smear (A) (600×, Wright-Giemsa) of a representative MPLW515L mutant animal reveals marked thrombocytosis and leukocytosis comprising a predominant population of maturing myeloid cells as well as frequent nucleated erythroid forms. Bone marrow images from MPLWT animals display preserved marrow architecture with maturing trilineage hematopoiesis (D) (600×, hematoxylin and eosin [H&E]). Comparatively, bone marrow sections from MPLW515L mutant animals demonstrate marrow elements comprising a prominent population of maturing myeloid cells with increased numbers of megakaryocytes including atypical and dysplastic forms occurring in frequent clusters (C) (600×, H&E) and showing emperipolesis of neutrophils in megakaryocyte cytoplasm. Spleen sections from MPLW515L mice display complete effacement of normal splenic architecture (E) (40×, H&E) with a marked expansion of red pulp that is composed of an admixture of maturing myeloid and erythroid elements and numerous numbers of atypical megakaryocytes (G) (600×, H&E) compared with MPLWT spleens (F and H) (40× and 600×, H&E), which display a relative preservation of normal spleen architecture and the presence of only maturing erythroid forms in the red pulp. Liver images from MPLW515L mice illustrate evidence of extensive extramedullary hematopoiesis in a perivascular and sinusoidal distribution (I) (100×, H&E) composed predominantly of a population of maturing erythroid elements with frequent large atypical megakaryocytes and occasional admixed myeloid forms (K) (600×, H&E). In comparison, only small, focal areas of nucleated erythroid cells were observed in livers from MPLWT animals (J and L) (100× and 600×, H&E).</p></caption><graphic xlink:href="pmed.0030270.g004"/></fig><p>Consistent with the histopathologic findings, flow cytometry analysis of MPLW515L bone marrow cells demonstrated an ˜4-fold increase in Mac1+/Gr1+ mature myeloid cells as compared with bone marrow cells from MPLWT animals ( <xref ref-type="fig" rid="pmed-0030270-g005">Figure 5</xref>A, upper panels). These findings were also reflected in the splenocyte populations examined from MPLW515L animals that revealed a nearly 10-fold increase in Mac1+/Gr1+ cells versus MPLWT spleens ( <xref ref-type="fig" rid="pmed-0030270-g005">Figure 5</xref>B, upper panels). Further analysis of MPLWT and MPLW515L spleen and bone marrow cells demonstrated a significant erythroid population ( <xref ref-type="fig" rid="pmed-0030270-g005">Figure 5</xref>B, middle panel) in both sets of animals, although a shift to a more immature erythroid population was observed in MPLW515L compared with MPLWT mice as reflected by an ˜2-fold increase in CD71+/Ter119- cells ( <xref ref-type="fig" rid="pmed-0030270-g005">Figure 5</xref>A and <xref ref-type="fig" rid="pmed-0030270-g005">5</xref>B, middle panels). The marked increase in megakaryocytes seen in the histopathology of MPLW515L mice was corroborated by an ˜15-fold increase in the percentage of CD41+ cells in MPLW515L bone marrow compared with MPLWT ( <xref ref-type="fig" rid="pmed-0030270-g005">Figure 5</xref>A, bottom panels). Similar increases were observed in MPLW515L spleen cells compared with MPLWT with an ˜30-fold increase in the proportion of CD41+ cells observed in MPLW515L spleen cells compared with MPLWT ( <xref ref-type="fig" rid="pmed-0030270-g005">Figure 5</xref>B, bottom panels). Analysis of B cells demonstrated a proportionate decrease in B220 positive cells in MPLW515L bone marrow and spleen (unpublished data). There was no difference in CD4/CD8+ or c-Kit positive cell populations between MPLW515L and MPLWT bone marrow and spleen cells (unpublished data). </p><fig id="pmed-0030270-g005" position="float"><label>Figure 5</label><caption><title>Flow Cytometry Analysis of BM and Spleen in Mice Transduced with MPLW515L and MPLWT</title><p>(A) Flow-cytometry analysis of bone marrow cells shows a 4-fold increase in Mac1+/Gr1+ cells and a shift to a more immature erythroid population. There is a 15-fold increase in CD41+ cells in bone marrow expressing MPLW515L compared with MPLWT.</p><p>(B) Flow-cytometry analysis of spleen cells shows a 10-fold increase in Mac1+/Gr1+ cells in MPLW515L. There is also a shift to a more immature erythroid population in MPLW515L with a greater percentage of CD71+/Ter119- cells. CD41+ cells are increased 30-fold in MPLW515L spleen cells compared with MPLWT spleen cells.</p></caption><graphic xlink:href="pmed.0030270.g005"/></fig></sec><sec id="s3d"><title>MPLW515L Confers a Proliferative Advantage to Bone Marrow and Spleen Cells</title><p>Megakaryocyte colony–forming assays of bone marrow cells derived from MPLW515L- and MPLWT-expressing mice demonstrated a similar number of megakaryocyte colony-forming units, although there was a marked increase in megakaryocyte colony–forming units formed from spleen cells from MPLW515L-expressing mice compared with MPLWT ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>A). In addition, acetylcholinesterase staining demonstrated MPLW515L megakaryocyte colonies ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>B, left picture) that were significantly larger than MPLWT megakaryocyte colonies ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>B, right picture). The difference in megakaryocyte colony growth and morphology was not explained by a difference in ploidy of megakaryocyte cells in MPLW515L compared with MPLWT bone marrow cells ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>C). Perhaps the most striking finding was the ability of bone marrow or spleen cells derived from MPLW515L animals, but not MPLWT, to grow in methylcellulose culture in the absence of any cytokines ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>D and <xref ref-type="fig" rid="pmed-0030270-g006">6</xref>E). There was also an increase in the number of myeloid colonies in methylcellulose supplemented with cytokines from MPLW515L-expressing spleen cells ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>E), but not bone marrow cells ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>D), consistent with a shift in hematopoiesis from the bone marrow to the spleen in MPLW515L-expressing mice. There was no difference in the distribution of myeloid colonies grown from MPLWT- and MPLW515L-expressing mice ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>D and <xref ref-type="fig" rid="pmed-0030270-g006">6</xref>E). </p><fig id="pmed-0030270-g006" position="float"><label>Figure 6</label><caption><title>MPLW515L Increases the Number of Megakaryocyte and Myeloid Colonies in Spleen, without Affecting Megakaryocyte Ploidy, and Causes Cytokine-Independent Myeloid Colony Growth</title><p>(A) Megakaryocyte colony–forming assay in the presence of TPO, IL3, IL11, and IL6 demonstrates similar numbers of megakaryocyte colonies obtained from MPLW515L-expressing bone marrow and an increase in the number of megakaryocyte colonies from spleen cells compared with MPLWT.</p><p>(B) Acetylcholinesterase staining of megakaryocyte colonies derived from bone marrow demonstrates much larger colony size in MPLW515L as compared with MPLWT megakaryocyte colonies.</p><p>(C) Megakaryocyte ploidy analysis shows the same distribution for MPLW515L-expressing cells and MPLWT-expressing cells.</p><p>(D and E) Total myeloid colony formation from bone marrow cells (D) and spleen cells (E) demonstrates cytokine-independent colony formation in MPLW515L bone marrow and spleen. There is no difference in colony distribution between MPLWT- and MPLW515L-expressing cells. Colony counts reflect only positively identifiable colonies, with thorough megakaryocyte colony analysis done in MegaCult assay ( <xref ref-type="fig" rid="pmed-0030270-g006">Figure 6</xref>A), and thus are excluded from methylcellulose colony analysis. Colony numbers represent a total of three representative mice per group, in duplicate. </p></caption><graphic xlink:href="pmed.0030270.g006"/></fig></sec><sec id="s3e"><title>JAK Kinase Inhibition Results in Inhibition of MPLW515L-Induced Proliferation and Signal Transduction</title><p>Treatment of 32D MPLW515L or 32D MPLWT cells but not 32D FIP1L1-PDGFRA cells with a small molecule JAK kinase inhibitor [ <xref rid="pmed-0030270-b021" ref-type="bibr">21</xref>] resulted in a dose-dependent inhibition of cell growth ( <xref ref-type="fig" rid="pmed-0030270-g007">Figure 7</xref>A). Dose-dependent inhibition of phosphorylation of STAT3, ERK, and AKT was observed in 32D MPLW515L and MPLWT cells but not in 32D FIP1L1-PDGFRA cells ( <xref ref-type="fig" rid="pmed-0030270-g007">Figure 7</xref>B; unpublished data). These data indicate that MPLW515L-mediated proliferation and activation of JAK-STAT signaling is sensitive to inhibition with a small molecule JAK kinase inhibitor. </p><fig id="pmed-0030270-g007" position="float"><label>Figure 7</label><caption><title>32D MPLW515L Cells Are Sensitive to JAK Inhibitor I</title><p>(A) Dose-dependent inhibition of growth of 32D MPLW515L and 32D MPLWT cells but not 32D FIP1L1-PDGFRA cells with increasing doses of JAK Inhibitor I.</p><p>(B) Reduction in STAT3 phosphorylation in 32D MPLW515L cells but not 32D FIP1L1-PDGFRA cells with increasing doses of JAK Inhibitor I. Cells were incubated with varying drug concentrations for four hours and then collected for Western blot analysis.</p></caption><graphic xlink:href="pmed.0030270.g007"/></fig></sec></sec><sec id="s4"><title>Discussion</title><p>In this report we describe a somatic mutation in the transmembrane region of <italic>MPL</italic> ( <italic>MPLW515L</italic>) in a subset of <italic>JAK2V617F-</italic>negative MF. Expression of MPLW515L transforms hematopoietic cells to cytokine-independent proliferation, and results in constitutive activation of JAK-STAT signaling. The relevance of this mutation is underscored by its effect in the murine bone marrow transplant assay, as expression of MPLW515L induces myeloproliferation characterized by splenomegaly, leukocytosis, marked thrombocytosis, extramedullary hematopoeisis, and myelofibrosis. In addition, a small-molecule JAK kinase inhibitor inhibits proliferation of MPLW515L-transformed cells, suggesting that targeted inhibition of JAK-STAT signaling may be an effective therapy for both <italic>JAK2V617F-</italic>positive and <italic>MPLW515L-</italic>positive MPD. </p><p>Although the constitutively active JAK2V617F mutant kinase is present in many patients with MPD, there are a significant proportion of MF and ET patients who are <italic>JAK2V617F-</italic>negative. We hypothesized that activation of JAK–STAT signaling in these patients might result from activating mutations in cytokine receptors, which led to the discovery of a novel mutation in <italic>MPL</italic> ( <italic>MPLW515L</italic>) in patients with <italic>JAK2V617F-</italic>negative MF. The mutation is present in granulocytes but not in non-hematopoietic germline DNA, demonstrating that <italic>MPLW515L</italic> is a somatic mutation present in clonally derived hematopoietic cells. Moreover, the <italic>MPLW515L</italic> allele was not observed in a panel of normal individuals, demonstrating that it is not a common single nucleotide polymorphism. These data provide genetic evidence that <italic>MPLW515L</italic> is a pathogenetic mutation in this subset of <italic>JAK2V617F-</italic>negative MF. </p><p>In contrast to previous reports [ <xref rid="pmed-0030270-b013" ref-type="bibr">13</xref>– <xref rid="pmed-0030270-b015" ref-type="bibr">15</xref>], we have identified a mutant MPL receptor that results in constitutive activation of downstream signal transduction pathways. MPL signals via its association with JAK2 and TYK2, resulting in phosphorylation of both MPL and JAK2/TYK2 and consequent activation of the STAT3, STAT5, ERK/MAPK, and PI3K/AKT signaling pathways. MPLW515L activates JAK-STAT signal transduction independently of ligand binding, and, in contrast with MPLWT, confers cytokine-independent growth to 32D, Ba/F3, and UT7 cells. Moreover, Western blot analysis shows ligand-independent phosphorylation of downstream signaling proteins, which may effect both proliferation and differentiation. Finally, hematopoietic cells that express MPLW515L are hypersensitive to stimulation with TPO as assessed by cell proliferation and Western blot analysis, suggesting MPLW515L-positive hematopoietic progenitors may have a selective proliferative advantage as compared with wild-type progenitors in MF patients. </p><p>The discovery of this mutation provides a novel mechanism for activation of signal transduction in hematopoietic malignancies. Somatic mutations in MPL have not previously been described, though previous genetic and biochemical data suggest that the cytoplasmic-transmembrane junction (which includes W515) is important in its activation. These data include the germline S505N allele in <italic>MPL</italic> associated with familial ET [ <xref rid="pmed-0030270-b022" ref-type="bibr">22</xref>, <xref rid="pmed-0030270-b023" ref-type="bibr">23</xref>], the W515S mutation that has been reported as a spontaneous mutation in Ba/F3 cells stably transduced with MPL[ <xref rid="pmed-0030270-b024" ref-type="bibr">24</xref>], and the observation that in vitro deletion of the RWQFP domain that includes W515 results in MPL activation [ <xref rid="pmed-0030270-b025" ref-type="bibr">25</xref>]. The mechanism of activation of MPL by these alleles is not fully understood, and it will be of value to obtain structural insight into the role of the cytoplasmic transmembrane–junction region in the regulation of MPL signal transduction. </p><p>Expression of MPLW515L in murine bone marrow recapitulated many of the phenotypic characteristics of MF in humans, including atypical megakaryocytic hyperplasia, splenomegaly due to extramedullary hematopoiesis, and thrombocytosis that may be attenuated in humans by splenic sequestration [ <xref rid="pmed-0030270-b026" ref-type="bibr">26</xref>, <xref rid="pmed-0030270-b027" ref-type="bibr">27</xref>]. This is in contrast to murine bone marrow transplant experiments with constitutively activated tyrosine kinase alleles evaluated in our laboratory and by others, including TEL-JAK2, FLT3-ITD, and FIP1L1-PDGFRA [ <xref rid="pmed-0030270-b017" ref-type="bibr">17</xref>, <xref rid="pmed-0030270-b028" ref-type="bibr">28</xref>, <xref rid="pmed-0030270-b029" ref-type="bibr">29</xref>], that each result in a neutrophilic MPD without involvement of the megakaryocyte lineage. It is of interest that JAK2V617F, which is associated with ET and MF in humans, results in polycythemia and leukocytosis, but not thrombocytosis, in the same murine BMT model [ <xref rid="pmed-0030270-b011" ref-type="bibr">11</xref>]. Of note, absolute leukocytosis in both the JAK2V617F and the MPLW515L-mediated BMT models is higher than is typically observed in human PV and MF, respectively. Also, MPLW515L mice develop rapidly fatal disease. These differences between the murine model and the human phenotype are likely accounted for by the retroviral transduction model of disease, and is similar to what has been observed for the rapidly fatal BCR-ABL-mediated disease in the murine BMT model as compared with human CML [ <xref rid="pmed-0030270-b030" ref-type="bibr">30</xref>]. </p><p>MPLW515L-expressing bone marrow and spleen cells exhibit cytokine-independent colony formation, another feature of clinical ET and MF. In addition, the MPLW515L murine bone marrow transplant model shares some phenotypic characteristics with other murine models of MF, including that induced by overexpression of thrombopoietin, either in transgenic [ <xref rid="pmed-0030270-b031" ref-type="bibr">31</xref>] or in bone marrow transplant models [ <xref rid="pmed-0030270-b032" ref-type="bibr">32</xref>, <xref rid="pmed-0030270-b033" ref-type="bibr">33</xref>], and in the transgenic GATA-1 <sup>low</sup> mouse [ <xref rid="pmed-0030270-b034" ref-type="bibr">34</xref>], though the severity and chronicity of the MF phenotype varies in these different models, which, unlike the current model, are not based on somatic disease alleles observed in human MF. </p><p>MPLW515L is an attractive target for therapeutic intervention with small molecule inhibitors. Current treatments for MF include empiric therapies such as hydroxyurea or interferon, and splenectomy for symptomatic splenomegaly [ <xref rid="pmed-0030270-b026" ref-type="bibr">26</xref>, <xref rid="pmed-0030270-b027" ref-type="bibr">27</xref>]. Allogeneic stem cell transplantation is a viable option for some patients with MF, but most patients are diagnosed at an advanced age and are not candidates for stem cell transplantation. Recent clinical trials with thalidomide demonstrated modest efficacy in MF, largely by slowing the rate of fibrosis in these patients rather than by reversal of fibrosis [ <xref rid="pmed-0030270-b035" ref-type="bibr">35</xref>]. Allogeneic BMT data does indicate that eradication of the malignant hematopoietic clone in MF does result in complete reversal of fibrosis [ <xref rid="pmed-0030270-b036" ref-type="bibr">36</xref>], and suggests that drugs that target mutant alleles such as MPLW515L may be effective disease-remitting agents. JAK2 inhibitor development is well under way as a potential therapeutic strategy for treating MPD patients with the <italic>JAK2V617F</italic> allele. Data presented here indicate that the subset of <italic>JAK2V617F-</italic>negative MF patients with the <italic>MPLW515L</italic> mutation will also be sensitive to inhibition with small molecule JAK2 inhibitors. These findings also suggest that a more expansive genomic screen for involvement of other components of the JAK-STAT signal transduction pathway is warranted in MPD patients. </p></sec><sec sec-type="supplementary-material" id="s5"><title>Supporting Information</title><supplementary-material content-type="local-data" id="st001"><label>Table S1</label><caption><title>Primer Sequences Used in This Study</title><p>(25 KB DOC)</p></caption><media xlink:href="pmed.0030270.st001.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><sec id="s5a"><title>Accession Numbers</title><p>The UniGene ( <ext-link ext-link-type="uri" xlink:href="http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=unigene">http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=unigene</ext-link>) accession numbers used in this paper are <italic>MPL</italic> (Hs.82906 and Mm.4864), <italic>EPOR</italic> (Hs.127826), and <italic>GCSFR</italic> (Hs.524517). </p></sec></sec>
Core Verbal Autopsy Procedures with Comparative Validation Results from Two Countries	<sec id="st1"><title>Background</title><p>Cause-specific mortality statistics remain scarce for the majority of low-income countries, where the highest disease burdens are experienced. Neither facility-based information systems nor vital registration provide adequate or representative data. The expansion of sample vital registration with verbal autopsy procedures represents the most promising interim solution for this problem. The development and validation of core verbal autopsy forms and suitable coding and tabulation procedures are an essential first step to extending the benefits of this method.</p></sec><sec id="st2"><title>Methods and Findings</title><p>Core forms for peri- and neonatal, child, and adult deaths were developed and revised over 12 y through a project of the Tanzanian Ministry of Health and were applied to over 50,000 deaths. The contents of the core forms draw upon and are generally comparable with previously proposed verbal autopsy procedures. The core forms and coding procedures based on the International Statistical Classification of Diseases (ICD) were further adapted for use in China. These forms, the ICD tabulation list, the summary validation protocol, and the summary validation results from Tanzania and China are presented here.</p></sec><sec id="st3"><title>Conclusions</title><p>The procedures are capable of providing reasonable mortality estimates as adjudged against stated performance criteria for several common causes of death in two countries with radically different cause structures of mortality. However, the specific causes for which the procedures perform well varied between the two settings because of differences in the underlying prevalence of the main causes of death. These differences serve to emphasize the need to undertake validation studies of verbal autopsy procedures when they are applied in new epidemiological settings.</p></sec>	<contrib contrib-type="author"><name><surname>Setel</surname><given-names>Philip W</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="corresp" rid="cor1">*</xref></contrib><contrib contrib-type="author"><name><surname>Rao</surname><given-names>Chalapati</given-names></name><xref ref-type="aff" rid="aff3">3</xref></contrib><contrib contrib-type="author"><name><surname>Hemed</surname><given-names>Yusuf</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff4">4</xref></contrib><contrib contrib-type="author"><name><surname>Whiting</surname><given-names>David R</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff5">5</xref></contrib><contrib contrib-type="author"><name><surname>Yang</surname><given-names>Gonghuan</given-names></name><xref ref-type="aff" rid="aff6">6</xref></contrib><contrib contrib-type="author"><name><surname>Chandramohan</surname><given-names>Daniel</given-names></name><xref ref-type="aff" rid="aff7">7</xref></contrib><contrib contrib-type="author"><name><surname>Alberti</surname><given-names>K. G. M. M</given-names></name><xref ref-type="aff" rid="aff8">8</xref></contrib><contrib contrib-type="author"><name><surname>Lopez</surname><given-names>Alan D</given-names></name><xref ref-type="aff" rid="aff3">3</xref></contrib>	PLoS Medicine	<sec id="s1"><title>Introduction</title><p>Globally, only about a third of all deaths are registered with age, sex, and cause [<xref rid="pmed-0030268-b001" ref-type="bibr">1</xref>]. The vast majority of these are in developed countries. In sub-Saharan Africa, where premature mortality accounts for about 80% of the total burden of disease [<xref rid="pmed-0030268-b002" ref-type="bibr">2</xref>], the need to remedy this situation is urgent. What is known about causes of death in these areas comes primarily from demographic surveillance sites and is largely limited to causes of death among children [<xref rid="pmed-0030268-b003" ref-type="bibr">3</xref>–<xref rid="pmed-0030268-b006" ref-type="bibr">6</xref>]. While more is known about child mortality than that of adults, knowledge remains patchy for neonatal and perinatal mortality [<xref rid="pmed-0030268-b007" ref-type="bibr">7</xref>].</p><p>In a 2003 address to the World Health Organization (WHO) staff, Director General Jong-Wook Lee succinctly highlighted the urgency of improving knowledge about vital events: “To make people count, we first need to be able to count people” [<xref rid="pmed-0030268-b008" ref-type="bibr">8</xref>]. The WHO has likened the continued lack of quality health information in lower-income countries, including data on vital events, to a “gathering storm” [<xref rid="pmed-0030268-b009" ref-type="bibr">9</xref>]. The crisis is being precipitated by the rapid escalation in national data demands and in reporting requirements for international initiatives, many of which require summary measures of survival and/or cause-specific mortality as indicators of program impact.</p><p>Improving the monitoring of vital events, and generating representative mortality statistics in lower-income countries in particular, will require new techniques, new technologies, and new thinking about sustainable, representative, and reliable systems for registering deaths and determining causes [<xref rid="pmed-0030268-b010" ref-type="bibr">10</xref>].</p><p>Sample or sentinel mortality surveillance using standardized “verbal autopsy” (VA) procedures represents a viable mid- or long-term strategy for improving mortality information [<xref rid="pmed-0030268-b010" ref-type="bibr">10</xref>]. A VA is an interview administered to caregivers or family members after a death occurs. A wide range of interview instruments and cause-of-death attribution procedures have been developed for this purpose [<xref rid="pmed-0030268-b011" ref-type="bibr">11</xref>–<xref rid="pmed-0030268-b018" ref-type="bibr">18</xref>]. Although VA is a limited tool [<xref rid="pmed-0030268-b019" ref-type="bibr">19</xref>], the procedure has demonstrated the ability to produce valid estimates of the mortality cause structure in many settings [<xref rid="pmed-0030268-b014" ref-type="bibr">14</xref>,<xref rid="pmed-0030268-b020" ref-type="bibr">20</xref>–<xref rid="pmed-0030268-b025" ref-type="bibr">25</xref>]. Some assessments of the validity and cross-comparability of VA-derived mortality estimates for child mortality have also been conducted [<xref rid="pmed-0030268-b004" ref-type="bibr">4</xref>–<xref rid="pmed-0030268-b006" ref-type="bibr">6</xref>]. VA has been applied in numerous countries, among children and adults, and for the purposes of both exploring specific causes of death in research projects and developing an overall description of the mortality structure at the community or population level. The WHO and the United Nations Children's Fund have called for the expanded use of the technique to monitor child mortality for at least a decade [<xref rid="pmed-0030268-b026" ref-type="bibr">26</xref>].</p><p>This article presents a proposed set of core VA procedures and the summary results of a two-country validation study conducted in Tanzania and China. These procedures are proposed for adaptation to a variety of settings, particularly in the context of sample or sentinel vital registration.</p><p>Experiences from India [<xref rid="pmed-0030268-b027" ref-type="bibr">27</xref>], China [<xref rid="pmed-0030268-b028" ref-type="bibr">28</xref>], and Tanzania [<xref rid="pmed-0030268-b029" ref-type="bibr">29</xref>,<xref rid="pmed-0030268-b030" ref-type="bibr">30</xref>] have shown how information generated through community-based mortality surveillance using VA can influence health policy, practice, monitoring, and evaluation. Generating data from VA procedures follows a simple, stepwise process. First, deaths are registered using some form of active, community-based reporting system. Second, VA interviews are obtained by trained interviewers who visit the households of the deceased within a specified period after the death. Third, physician certifiers use these completed VA interview forms to assign a specific cause of death, and write death certificates according to protocols based on the <italic>International Statistical Classification of Diseases and Related Health Problems,</italic> 10th revision (ICD-10) [<xref rid="pmed-0030268-b031" ref-type="bibr">31</xref>–<xref rid="pmed-0030268-b033" ref-type="bibr">33</xref>]. Lastly, mortality data are tabulated on a periodic basis and fed into routine reporting, planning, and monitoring processes and are used to analyze mortality structures, levels, and trends.</p><p>The ultimate impact of mortality surveillance will hinge upon the validity, comparability, and consistency of tools and methods used to obtain the “raw” data from representative sample or sentinel populations. In order to contribute to the expanded use of VA in sample and sentinel registration, as well as in research, this paper proposes a core set of VA procedures that have been validated in China and Tanzania. Where relevant, we have compared these procedures with those used in other settings.</p></sec><sec id="s2"><title>Methods</title><p>Between 1992 and 2004, the VA procedures presented here were developed as part of a long-term national system of sentinel demographic surveillance in Tanzania. The forms, coding methods, and mortality surveillance activities were integrated into the routine functions of local health authorities in Tanzania [<xref rid="pmed-0030268-b034" ref-type="bibr">34</xref>–<xref rid="pmed-0030268-b036" ref-type="bibr">36</xref>] and were applied in more than 50,000 deaths. In 2001, the procedures were further refined with reference to other existing and recommended tools. They were then translated with additional slight modifications for use in the Chinese Disease Surveillance Points System and vital registration system. The Chinese Disease Surveillance Points System is China's national sample vital registration system, covering 6% of the population in 160 urban and rural clusters [<xref rid="pmed-0030268-b037" ref-type="bibr">37</xref>].</p><p>The procedures discussed in this article were the subject of a 4-y validation study in Tanzania and China. The details of the study protocol and results for all age groups from both countries have been published elsewhere [<xref rid="pmed-0030268-b038" ref-type="bibr">38</xref>,<xref rid="pmed-0030268-b039" ref-type="bibr">39</xref>]. Briefly, the protocol entailed collection of VA and medical record information for the same individuals. In Tanzania, data were collected from urban and rural sentinel demographic surveillance areas operated by the Tanzanian Ministry of Health through the Adult Morbidity and Mortality Project, and from nearby health facilities [<xref rid="pmed-0030268-b040" ref-type="bibr">40</xref>]. Deaths from the sentinel areas were eligible for inclusion if the deceased visited a health facility during the period during which the “terminal” events leading to the death occurred. This did not necessarily mean that the death took place in the facility. For all eligible deaths an attempt was made to trace the medical records after informed consent was obtained from surviving family members. For deaths that occurred in participating health facilities during the study period, all were eligible, provided they met a geographic restriction criterion to ensure comparability with deaths from the sentinel surveillance sites [<xref rid="pmed-0030268-b038" ref-type="bibr">38</xref>]. In these cases, the medical records were obtained from the health facility, and the relatives, if they gave consent, were traced to their homes, usually within 1 mo after the death, and a VA interview was administered. In China, data were collected from urban areas through collaboration with the national Disease Surveillance Points System. Deaths were included from 100 tertiary hospitals in six cities. The number of deaths selected for each cause was based on the frequency in the routine system, with oversampling of some rarer causes and undersampling of some very common causes.</p><p>Physician panels assigned causes of death to all VAs and medical records using standard procedures, including blinding. Medical records and VA data were handled identically in this regard, and both sources were used to produce standard death certificates. No physician assigned causes of death using both the VA and medical record for the same individual. Entries were then coded to ICD-10 at the core code and four-digit levels and tabulated according to the list in <xref ref-type="table" rid="pmed-0030268-t001">Table 1</xref>. The table contains a 57-item VA tabulation list with ICD-10 core codes in the third column. The list is organized according to International Statistical Classification of Diseases (ICD) principles, and contains the causes that are amenable to detection by VA and are relevant for guiding policy and program development. It is important to use such a tabulation list as a minimum standard for reporting in order to maintain international comparability of mortality datasets.</p><table-wrap id="pmed-0030268-t001" content-type="1col" position="float"><label>Table 1</label><caption><p>ICD Mortality Tabulation List for VA Data</p></caption><graphic xlink:href="pmed.0030268.t001"/></table-wrap></sec><sec id="s3"><title>Results</title><sec id="s3a"><title>Content of Core VA Interview Forms</title><p>In order to function well as part of routine systems, the forms had to be easy to use by interviewers with varying degrees of clinical skills and knowledge. They were also used to record relevant contextual information (such as use of health facilities in the period before death and data on risk factors). Additionally, the forms made use of any documentary evidence available from the household of the deceased that might aid in determining the probable cause of death. Lastly, the forms had to provide physicians with enough data to produce internationally comparable mortality statistics based on ICD coding guidelines, and be amenable to developing data-derived algorithms to determine the probable cause of death [<xref rid="pmed-0030268-b041" ref-type="bibr">41</xref>].</p><p>All short core VA forms referred to in this article are available online. They include the forms for perinatal events and neonatal deaths (<xref ref-type="supplementary-material" rid="pmed-0030268-sg001">Figure S1</xref>), deaths in post-neonatal children under age 5 (<xref ref-type="supplementary-material" rid="pmed-0030268-sg002">Figure S2</xref>), and deaths among persons aged 5 y and above (<xref ref-type="supplementary-material" rid="pmed-0030268-sg003">Figure S3</xref>). Each form follows the same basic structure: identifying information about the deceased (including age, sex, and place of death), cause of death according to respondent, short narrative history, symptom duration checklist, health services used in the period before death, and any medical evidence available at the household, including whether a health worker informed the respondent of the cause of death.</p><p>A section on the condition of the mother during and after pregnancy and birth is included on the neonatal form. Deaths to women of reproductive age, and maternal deaths in particular, are addressed in a subsection of the form for deaths over age 5. The questions contained in the symptom duration checklist are generally arranged by anatomical system. They are intended to provide strong support for a positive diagnosis of probable cause of death, and the confident exclusion of differential diagnoses.</p><p> <xref ref-type="table" rid="pmed-0030268-t002">Table 2</xref> compares the VA form used at some sites that are members of the INDEPTH Network [<xref rid="pmed-0030268-b042" ref-type="bibr">42</xref>], the VA form used by the Indian Sample Registration System (Indian SRS) [<xref rid="pmed-0030268-b015" ref-type="bibr">15</xref>–<xref rid="pmed-0030268-b018" ref-type="bibr">18</xref>], and a VA form from the WHO [<xref rid="pmed-0030268-b012" ref-type="bibr">12</xref>], and notes key areas of difference in content. VA interview forms tailored to specific age groups generally use the same standard cut-off for neonatal mortality (i.e., death before 28 d), although there is some variation in the age range for the application of forms for post-neonatal child deaths [<xref rid="pmed-0030268-b017" ref-type="bibr">17</xref>,<xref rid="pmed-0030268-b042" ref-type="bibr">42</xref>–<xref rid="pmed-0030268-b044" ref-type="bibr">44</xref>]. For reasons of cost and ease of implementation, the layout and length of the proposed core forms were limited to no more than two A4-sized pages. In this they are similar to forms used in the Indian SRS, and much briefer than most other VA forms presented in the literature [<xref rid="pmed-0030268-b012" ref-type="bibr">12</xref>,<xref rid="pmed-0030268-b013" ref-type="bibr">13</xref>,<xref rid="pmed-0030268-b041" ref-type="bibr">41</xref>].</p><table-wrap id="pmed-0030268-t002" content-type="1col" position="float"><label>Table 2</label><caption><p>Content Comparison of Verbal Autopsy Forms</p></caption><graphic xlink:href="pmed.0030268.t002"/></table-wrap><p>Aside from the differences noted in <xref ref-type="table" rid="pmed-0030268-t002">Table 2</xref>, the use of long “open history” sections in other forms is another major difference between the proposed core forms and other published VA tools. The Indian SRS form for adult deaths, for example, relies almost exclusively on narrative histories of the events preceding death to provide evidence about the cause of death. If administered as a clinical history, these sections can provide relevant information to physicians who assign probable causes of death. The short core forms in <xref ref-type="supplementary-material" rid="pmed-0030268-sg001">Figures S1</xref>–<xref ref-type="supplementary-material" rid="pmed-0030268-sg003">S3</xref> allow for brief narrative histories, but emphasize a “symptom duration checklist” approach for use in cause-of-death attribution.</p><p>Experience in implementing VA procedures suggests that lengthy clinical history sections cannot be standardized and vary substantially depending on the clinical skills and medical training of the interviewer. In addition, interviewers may introduce bias into data collection by recording histories that neatly fit into known or familiar disease descriptions or are based on the interviewers' initial impressions of the likely cause of death. Therefore, a core symptom duration checklist may be more systematic and to produce a more complete inventory of the signs and symptoms before death than would a heavier reliance on open histories.</p></sec><sec id="s3b"><title>Interviewing Protocols and Cross-Cultural Applicability of Procedures</title><p>A VA interview is conducted similarly to any confidential health-related interview, with the added consideration that the subject matter concerns a topic that could hardly be more distressing—the recent death of a family or household member. This, in part, speaks to the need to enroll respected community members in areas where VA will be implemented to help build local awareness and acceptance of what is, generally speaking, a new and unfamiliar mechanism of collecting health information.</p><p>There is a range of opinion about whether medical training should be a preferred qualification for VA interviewers or whether educated but non-medically trained persons are more suitable. Local experience will determine the optimal solution. Training should include discussion of symptoms and their description in local languages. In addition, a clear understanding of how live births and stillbirths can be accurately differentiated using appropriate terminology is important. Ideally, the interview should happen as soon as possible after a death with due consideration to culturally appropriate mourning periods. All questions on the VA form (aside from the appropriate skips) must be asked of the respondent regardless of the opinion of the interviewer as to their relevance. Quality assurance should be performed routinely. If feasible, re-interview of a 10%–15% sample of VAs would offer a strict standard. However, given the sensitive nature of VA, it may be sufficient to verify for a similar proportion of deaths that (a) the death indeed occurred and (b) the VA interview in fact took place at the household of the deceased, with an appropriate respondent. Interviewer retraining and supportive supervision are probably the most important components of quality assurance for VA.</p><p>As with any survey instrument intended for cross-cultural and cross-linguistic application, care must be taken in translation into local languages and field testing so that all questions are understood by respondents in the way they are intended [<xref rid="pmed-0030268-b022" ref-type="bibr">22</xref>]. Part of the cultural validation of VA should include observations of VA by a medical anthropologist or sociologist, and interviews with community members to ensure accurate understanding of terms used in the VA form. VA interviewers should be informed about any areas of potential confusion due to local or colloquial expressions. This will help ensure that in addition to building community rapport for administering VA, culturally appropriate and sensitive terms, idioms, and expressions are used in interviews without sacrificing precision and cross-comparability of results. A balance must be struck between clinically precise terminology, which can be confusing or even offensive in the context of a VA interview, and colloquial expressions or local terms that might impede accurate cause of death attribution and ICD coding.</p><p>It is also important to consider linguistic and cultural issues in implementing proposed core VA forms in very different settings. The experience of transferring these procedures from Tanzania to China has been instructive in this regard. It is felt that both the brevity of the interviews and the efforts expended in establishing rapport contributed to attaining response rates in both settings of over 90%. Minimal, though important, modifications were required to translate specific questions and variables from the original Tanzanian forms into the Chinese context. For instance, the question “Was [the deceased] breathless on lying flat?” employed in Tanzania was not clearly interpreted in China, and a question on “breathlessness interfering with sleep” was substituted. Chinese interviewers readily adopted the protocols, and physician reviewers in both sites were able to certify causes of death using an international death certificate. Finally, statistics could be compiled from both countries according to the proposed tabulation list, yielding internationally comparable data.</p></sec><sec id="s3c"><title>ICD Coding, Cause-of-Death Attribution, and Tabulation List</title><p>Because VA may serve as the best or even sole evidence on cause of death in many settings, establishing international comparability is important [<xref rid="pmed-0030268-b004" ref-type="bibr">4</xref>–<xref rid="pmed-0030268-b006" ref-type="bibr">6</xref>,<xref rid="pmed-0030268-b026" ref-type="bibr">26</xref>,<xref rid="pmed-0030268-b045" ref-type="bibr">45</xref>]. The lack of standard interview forms, cause-of-death categories, and coding practices has hindered attempts to synthesize results from various applications of VA to assessing precise causes of child mortality [<xref rid="pmed-0030268-b004" ref-type="bibr">4</xref>,<xref rid="pmed-0030268-b006" ref-type="bibr">6</xref>,<xref rid="pmed-0030268-b045" ref-type="bibr">45</xref>]. Therefore, it is recommended that the causes of death as determined through physician review of VA be recorded using a four-line death certificate (i.e., underlying, immediate, associated, and contributory causes). Subsequently, a physician or medical recorder should select and code the underlying cause to the core 3 character code using standard ICD-10 rules.</p><p>In certain cases it may be possible to code to the fourth digit. On the other hand, ICD-10 rules may frequently preclude the use of certain three-digit codes in the VA context. For example, ICD codes starting with B50, B51, and B52 refer to malaria. Use of these codes requires both confirmation of parasite infection and identification of the malaria species. It is unlikely that evidence of such confirmation would be available at the household level. In such instances there are usually three-digit codes available for use (in this case either “B53 other parasitologically confirmed malaria” or “B54 unspecified malaria—clinically diagnosed malaria without parasitological confirmation”) that would not affect the outcome of tabulating and reporting VA data, or the main public-health interpretations and policy implications of the tabulated data.</p><p>Physicians who review the completed VA forms usually require training in cause-of-death certification using ICD rules and international death certificates. An explanation of the structure and content of the ICD classification, and of the rules for selection and coding of the underlying cause of death, is also necessary to ensure uniformity of data across different coders. Details of coding guidelines and criteria, manuals, and options for organizing VA coding can be obtained from the authors.</p><p>The issue of reporting single versus multiple causes of death in VAs, particularly for children, has been addressed extensively. Most sources recommend or employ multiple cause-of-death attribution in children without providing a single underlying cause [<xref rid="pmed-0030268-b043" ref-type="bibr">43</xref>,<xref rid="pmed-0030268-b044" ref-type="bibr">44</xref>,<xref rid="pmed-0030268-b046" ref-type="bibr">46</xref>–<xref rid="pmed-0030268-b048" ref-type="bibr">48</xref>], and at least one source does so for adults [<xref rid="pmed-0030268-b013" ref-type="bibr">13</xref>]. We recommend the use of standard death certificates for all ages in accordance with ICD convention ([<xref rid="pmed-0030268-b032" ref-type="bibr">32</xref>], p. 31). This enables recording, coding, and analysis of multiple causes of death while retaining comparability of mortality data based on the tabulation of a single underlying cause, as prescribed by ICD. Although ICD does recommend a specially designed death certificate for perinatal deaths ([<xref rid="pmed-0030268-b032" ref-type="bibr">32</xref>], p. 90), few countries have implemented it. For the present, therefore, perinatal deaths (and stillbirths, if desired) may be recorded, together with neonatal deaths, on a conventional death certificate.</p><p>The ICD recommends two “condensed” tabulation lists for mortality reporting [<xref rid="pmed-0030268-b033" ref-type="bibr">33</xref>]. These lists contain many causes that can be accurately identified only with specific diagnostic or clinical information. In the case of VA, the smaller list of causes presented in <xref ref-type="table" rid="pmed-0030268-t001">Table 1</xref> is more appropriate. The ICD specifically sanctions the development of such tailored tabulation lists [<xref rid="pmed-0030268-b032" ref-type="bibr">32</xref>].</p></sec><sec id="s3d"><title>Validation and Comparative Findings from China and Tanzania in Deaths over Age 5</title><p>There were 25 causes in China and 26 causes in Tanzania (for at least one age group) for which there were at least five deaths in both the VA and the medical records. <xref ref-type="table" rid="pmed-0030268-t003">Table 3</xref> shows summary results for causes of death for which at least five deaths were validated, sensitivity was greater than 50%, and the relative difference in the cause-specific mortality fraction (CSMF) in the VA (CSMF<sub>VA</sub>) and medical record (CMFS<sub>MR</sub>) was equal to or less than 20%. These criteria are based on threshold values for sensitivity and CSMF suggested for assessing accuracy of adult VA [<xref rid="pmed-0030268-b049" ref-type="bibr">49</xref>]. The comparison of CSMFs was based on the relative difference in the proportion of deaths due to cause <italic>X</italic> in the medical records dataset (CSMF<sub>MR</sub>) from the proportion of deaths due to the same cause in the verbal autopsy dataset (CSMF<sub>VA</sub>). We calculated sensitivities and specificities for all causes using conventional two-by-two table analysis, although results are displayed only for those causes reaching the threshold sensitivity. For the over-five age group, data were available for 1,912 deaths in 42 cause-of-death categories from Tanzania, and 2,029 deaths in 37 categories from rural areas in China. Of these, 140 deaths from Tanzania and 170 deaths from China were coded to either “all other specified diseases” or “undetermined”; these are excluded from the comparison.</p><table-wrap id="pmed-0030268-t003" content-type="1col" position="float"><label>Table 3</label><caption><p>Validation Results for Causes of Adult Mortality in Tanzania and China</p></caption><graphic xlink:href="pmed.0030268.t003"/></table-wrap><p>Of the 20 causes of death listed in <xref ref-type="table" rid="pmed-0030268-t003">Table 3</xref>, three (VA-02 tuberculosis, VA-27 cerebrovascular diseases, and VA-50 transport accidents) met the threshold criteria in both countries. Six causes reached the threshold in Tanzania only, and 11 causes met the threshold levels in China only. At the upper bound of the 95% confidence level for sensitivity, seven additional causes reach the threshold in Tanzania (VA-09 remainder of infectious and parasitic diseases, VA-15 malignant neoplasm of liver, VA-19 other neoplasms, VA-21 diabetes mellitus, VA-25 hypertensive diseases, VA-38 other maternal causes, and VA-57 all other external causes).</p><p>Significantly, for six causes (VA-09 remainder of infectious and parasitic diseases, VA-17 malignant neoplasm of breast, VA-21 diabetes mellitus, VA-25 hypertensive diseases, VA-29 asthma/chronic obstructive pulmonary disease, and VA-51 falls) the relative difference in sensitivity of VA was less than 25%. It should be born in mind that because of the small samples for certain causes of death it was not possible in the studies to validate all the causes contained in <xref ref-type="table" rid="pmed-0030268-t001">Table 1</xref>.</p></sec></sec><sec id="s4"><title>Discussion</title><p>The increasing importance of VA is reflected in the growing number of meta-analyses of VA-based datasets on child mortality from demographic surveillance sites and special studies [<xref rid="pmed-0030268-b004" ref-type="bibr">4</xref>–<xref rid="pmed-0030268-b006" ref-type="bibr">6</xref>,<xref rid="pmed-0030268-b045" ref-type="bibr">45</xref>], all of which make the case for standardized procedures. The procedures presented here are the product of over a decade of application, trial, assessment, and refinement, and have considerable commonalities with other forms in the public domain.</p><p>The main purpose of these tools is to supply countries that have no source of reliable mortality reporting and cause-of-death data with the means to confidently produce and use accurate, repeatable, and internationally comparable measurements of the cause structure of mortality for the most important diseases and conditions, and that are free from major systematic misclassification. To be sure, VA is a crude substitute for proper medical certification of cause of death—which can be a dubious “gold standard” even in developed countries [<xref rid="pmed-0030268-b050" ref-type="bibr">50</xref>]. The Tanzania–China experience has shown that the transfer of this technology from one setting to another is feasible and can produce results with acceptable sensitivity and CSMFs for important causes of death. In wider application, local validation studies should be considered an essential part of implementing VA procedures intended for national monitoring, evaluation, priority-setting, and policy-making.</p><p>These VA procedures performed quite differently for different causes in China and Tanzania. In Tanzania, where more data were available to analyze VA performance in younger age groups (analysis not presented), VA yielded good sensitivity, specificity, and CSMFs for several important causes including pneumonia, but did not perform as well for others, including childhood malaria. This re-emphasizes the need to bear in mind previous findings that both the number of different causes and their underlying prevalence vary by age and across settings where the use VA procedures is appropriate, and that this variation affects VA performance [<xref rid="pmed-0030268-b019" ref-type="bibr">19</xref>,<xref rid="pmed-0030268-b051" ref-type="bibr">51</xref>,<xref rid="pmed-0030268-b052" ref-type="bibr">52</xref>]. Thus, wherever feasible, VA procedures should be accompanied by a validation study, and revalidation should be undertaken periodically if there are indications of major shifts in causes of mortality—either as a result of successful large-scale intervention, or due to epidemics. Validation studies should also take into consideration that the “gold standard” of medical record diagnosis is often an imperfect one, at best. A carefully conducted VA may be superior to poorly maintained or scanty medical records, as seen, for example, in stillbirth [<xref rid="pmed-0030268-b053" ref-type="bibr">53</xref>]. It should also be acknowledged that for some conditions, such as malaria mortality among adults, for which no reliable statistics exist, neither VA nor medical records may form a suitable evidence base.</p><p>In addition to the use of standard procedures, the following are needed in order to make the best informed use of VA: further validation studies for less prevalent causes and whenever the procedures are applied in a new setting, further investigation into the effect of recall period [<xref rid="pmed-0030268-b054" ref-type="bibr">54</xref>] and respondent characteristics (e.g., relationship to the deceased and education), further development of guidelines and criteria for assigning cause of death, systematic handling of misclassification error, and candor with respect to any insuperable limitations of VA [<xref rid="pmed-0030268-b014" ref-type="bibr">14</xref>,<xref rid="pmed-0030268-b019" ref-type="bibr">19</xref>]. The WHO's leadership in the future development of VA procedures will be critical to establishing international standards.</p><p>Even with extensive validation, the weight given to VA-derived mortality data is likely to be an ongoing topic of debate. Ultimately, the interpretation of how well VA performs is entirely dependent upon the function the technique is meant to perform. VA will never meet the standards of proper medical certification of death at the time of its occurrence. The technique has inherent shortcomings including the prevalence dependency of its accuracy, and the serious effects that variations in sensitivity and specificity can have on comparative estimates of cause-specific mortality across populations or over time in the same population [<xref rid="pmed-0030268-b019" ref-type="bibr">19</xref>,<xref rid="pmed-0030268-b051" ref-type="bibr">51</xref>]. Nevertheless, for purposes of broad priority-setting, tracking trends in mortality due to major conditions of public-health importance, and providing broad burden-of-disease measures, it may be deemed preferable to the current state of near ignorance with regard to direct measures of cause-specific mortality, particularly among adults.</p><p>Standard and validated VA procedures are only part of the solution to maximal utility of VA. It is critical to do validation studies so that the degree of uncertainty, which will vary by cause of death, can be factored into mortality burden estimations. While the use of proportional mortality models based on VA data to estimate mortality burdens [<xref rid="pmed-0030268-b004" ref-type="bibr">4</xref>,<xref rid="pmed-0030268-b006" ref-type="bibr">6</xref>,<xref rid="pmed-0030268-b045" ref-type="bibr">45</xref>] remains controversial, these models have the virtue of attempting to make use of the only body of data available on cause-specific mortality for the populations concerned. More significant progress in producing mortality statistics that are valid, comparable, and representative, however, will depend on an expanded commitment to sample vital registration systems that use VA—not through reliance on disease-specific research studies, household surveys, or research demographic surveillance systems. Future research will allow a better understanding of the degree to which these core tools, with locally appropriate modifications, can achieve the ultimate aim of generating reliable and internationally comparable cause-specific mortality statistics.</p></sec><sec sec-type="supplementary-material" id="s5"><title>Supporting Information</title><supplementary-material content-type="local-data" id="pmed-0030268-sd001"><label>Alternative Language Abstract S1</label><caption><title>Translation of the Abstract into French</title><p>(28 KB DOC)</p></caption><media xlink:href="pmed.0030268.sd001.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="pmed-0030268-sd002"><label>Alternative Language Abstract S2</label><caption><title>Translation of the Abstract into Spanish</title><p>(29 KB DOC)</p></caption><media xlink:href="pmed.0030268.sd002.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="pmed-0030268-sg001"><label>Figure S1</label><caption><title>Core VA Form 1: Death of Child under 29 d</title><p>(130 KB DOC)</p></caption><media xlink:href="pmed.0030268.sg001.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="pmed-0030268-sg002"><label>Figure S2</label><caption><title>Core VA Form 2: Death of Child Aged 29 d to under 5 y</title><p>(129 KB DOC)</p></caption><media xlink:href="pmed.0030268.sg002.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="pmed-0030268-sg003"><label>Figure S3</label><caption><title>Core VA Form 3: Death of Person Aged 5 y and Above</title><p>(181 KB DOC)</p></caption><media xlink:href="pmed.0030268.sg003.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material></sec>
Aberrant Receptor-Mediated Endocytosis of <italic>Schistosoma mansoni</italic> Glycoproteins on Host Lipoproteins	<sec id="st1"><title>Background</title><p>Bilharzia is one of the major parasitic infections affecting the public health and socioeconomic circumstances in (sub) tropical areas. Its causative agents are schistosomes. Since these worms remain in their host for decades, they have developed mechanisms to evade or resist the immune system. Like several other parasites, their surface membranes are coated with a protective layer of glycoproteins that are anchored by a lipid modification.</p></sec><sec id="st2"><title>Methods and Findings</title><p>We studied the release of glycosyl-phosphatidylinositol (GPI)-anchored proteins of <named-content content-type="genus-species">S. mansoni</named-content> and found them in the circulation associated with host lipoprotein particles. Host cells endocytosed schistosomal GPI-anchored proteins via their lipoprotein receptor pathway, resulting in disturbed lysosome morphology. In patients suffering from chronic schistosomiasis, antibodies attacked the parasite GPI-anchored glycoproteins that were associated with the patients' own lipoprotein particles. These immunocomplexes were endocytosed by cells carrying an immunoglobulin-Fc receptor, leading to clearance of lipoproteins by the immune system. As a consequence, neutral lipids accumulated in neutrophils of infected hamsters and in human neutrophils incubated with patient serum, and this accumulation was associated with apoptosis and reduced neutrophil viability. Also, <italic>Trypanosoma brucei,</italic> the parasite that causes sleeping sickness, released its major GPI-anchored glycoprotein VSG221 on lipoprotein particles, demonstrating that this process is generalizable to other pathogens/parasites.</p></sec><sec id="st3"><title>Conclusions</title><p>Transfer of parasite antigens to host cells via host lipoproteins disrupts lipid homeostasis in immune cells, promotes neutrophil apoptosis, may result in aberrant antigen presentation in host cells, and thus cause an inefficient immune response against the pathogen.</p></sec>	<contrib contrib-type="author"><name><surname>Sprong</surname><given-names>Hein</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="corresp" rid="cor1"></xref></contrib><contrib contrib-type="author"><name><surname>Suchanek</surname><given-names>Monika</given-names></name><xref ref-type="aff" rid="aff2">2</xref></contrib><contrib contrib-type="author"><name><surname>van Dijk</surname><given-names>Suzanne M</given-names></name><xref ref-type="aff" rid="aff1">1</xref><xref ref-type="aff" rid="aff3">3</xref></contrib><contrib contrib-type="author"><name><surname>van Remoortere</surname><given-names>Alexandra</given-names></name><xref ref-type="aff" rid="aff4">4</xref></contrib><contrib contrib-type="author"><name><surname>Klumperman</surname><given-names>Judith</given-names></name><xref ref-type="aff" rid="aff3">3</xref></contrib><contrib contrib-type="author"><name><surname>Avram</surname><given-names>Diana</given-names></name><xref ref-type="aff" rid="aff5">5</xref></contrib><contrib contrib-type="author"><name><surname>van der Linden</surname><given-names>Joke</given-names></name><xref ref-type="aff" rid="aff6">6</xref></contrib><contrib contrib-type="author"><name><surname>Leusen</surname><given-names>Jeanette H. W</given-names></name><xref ref-type="aff" rid="aff6">6</xref></contrib><contrib contrib-type="author"><name><surname>van Hellemond</surname><given-names>Jaap J</given-names></name><xref ref-type="aff" rid="aff7">7</xref></contrib><contrib contrib-type="author"><name><surname>Thiele</surname><given-names>Christoph</given-names></name><xref ref-type="aff" rid="aff2">2</xref><xref ref-type="corresp" rid="cor1"></xref></contrib>	PLoS Medicine	<sec id="s1"><title>Introduction</title><p>Schistosomiasis, also known as bilharzia, is a chronic disease caused by parasitic <italic>Schistosoma</italic> flatworm species [<xref rid="pmed-0030253-b001" ref-type="bibr">1</xref>] that affects more than 200 million people worldwide. The schistosome is a prime example of a complex multicellular organism that can flourish in human hosts for decades despite the development of a pronounced immune response. The molecular basis for the prolonged survival of adult schistosomes is an outstanding issue [<xref rid="pmed-0030253-b002" ref-type="bibr">2</xref>].</p><p>The major surface antigens of adult schistosomes are tightly associated with the extracellular leaflet of the outer membranes by a glycosyl-phosphatidylinositol (GPI) anchor [<xref rid="pmed-0030253-b003" ref-type="bibr">3</xref>,<xref rid="pmed-0030253-b004" ref-type="bibr">4</xref>]. GPI-anchored proteins are abundant in parasites and have an important role in cell viability and defence against the host immune system [<xref rid="pmed-0030253-b005" ref-type="bibr">5</xref>,<xref rid="pmed-0030253-b006" ref-type="bibr">6</xref>]. They can be shed into the medium by the action of proteases or phospholipases [<xref rid="pmed-0030253-b007" ref-type="bibr">7</xref>] or released on membranous particles [<xref rid="pmed-0030253-b008" ref-type="bibr">8</xref>−<xref rid="pmed-0030253-b011" ref-type="bibr">11</xref>]. A third and largely unexplored mechanism is their release on lipoprotein particles. Lipid-modified proteins on the exoplasmic face of the plasma membrane might fit well into the phospholipid monolayer of such a particle, as demonstrated by the presence of GPI-anchored CD59 in HDL particles [<xref rid="pmed-0030253-b012" ref-type="bibr">12</xref>]. These lipoprotein particles might further transfer GPI-anchored proteins onto other cells, resulting in a net intercellular transfer. Interestingly, when transfected into cultured cells, GPI-anchored proteins can be transferred between cells [<xref rid="pmed-0030253-b013" ref-type="bibr">13</xref>−<xref rid="pmed-0030253-b016" ref-type="bibr">16</xref>]; this process was also observed in mice transgenic for human CD59 [<xref rid="pmed-0030253-b017" ref-type="bibr">17</xref>]. Trypanosome surface GPI protein has been observed on host erythrocytes [<xref rid="pmed-0030253-b018" ref-type="bibr">18</xref>], and retrograde transfer of host GPI protein onto the parasite was found in schistosomes [<xref rid="pmed-0030253-b019" ref-type="bibr">19</xref>]. It is not known how this transfer works mechanistically nor whether lipoprotein particles are involved.</p><p>We recently found that the lipid-linked morphogens Hedgehog and Wingless associate with <italic>Drosophila</italic> lipoprotein particles, and that their intercellular transport depends on the presence of lipoproteins [<xref rid="pmed-0030253-b020" ref-type="bibr">20</xref>]. Similarly, parasites such as schistosomes might utilize host lipoprotein particles as carriers for their GPI-linked surface antigens. Here, we demonstrate that parasite GPI-anchored glycoproteins are released onto host lipoprotein particles that are taken up by host cells via the low-density lipoprotein (LDL) receptor and the Fc receptors. This intercellular transfer of parasite GPI-anchored proteins to its host has important pathophysiological implications.</p></sec><sec id="s2"><title>Methods</title><sec id="s2a"><title>Materials</title><p>Chemicals, unless indicated otherwise, were from Sigma (St. Louis, Missouri, United States) and used in the highest purity available. Silica TLC plates were from Merck (Darmstadt, Germany), organic solvents were from Riedel de Haën (Darmstadt, Germany), and cell culture media and reagents were from Invitrogen (Breda, The Netherlands). Cell culture plasticware was from Costar (Cambridge, Massachusetts, United States). Phosphatidylinositol-specific phospholipase C (PI-PLC) from <named-content content-type="genus-species">Bacillus cereus</named-content> was from Molecular Probes (Eugene, Oregon, United States). Tran[<sup>35</sup>S]label (>36 TBq/mmol) was from ICN (Costa Mesa, California, United States). [U-<sup>14</sup>C]palmitic acid (18 TBq/mol) was from Amersham (Buckinghamshire, UK). Jürgen Gent generously provided us with a plasmid containing the human LDL receptor.</p></sec><sec id="s2b"><title>Antibodies</title><p>The synthetic peptide CIKAHDYKLLTKILAARQLQDLFDNDKN, which corresponds to amino acids 29–55 from the GPI-anchored surface protein SM200 of <named-content content-type="genus-species">S. mansoni</named-content> [<xref rid="pmed-0030253-b021" ref-type="bibr">21</xref>], was conjugated with its amino-terminal cysteine to keyhole limpet haemocyanin, and used for immunization of rabbits (Eurogentec, Seraing, Belgium). Antibodies to SM200 were affinity purified using the same peptide coupled to Affigel-15 columns (BioRad, Hercules, California, United States). The mouse monoclonal antibody (mab) to SM200, 307D5 [<xref rid="pmed-0030253-b004" ref-type="bibr">4</xref>], was kindly provided by Karl Hoffmann (University of Cambridge, UK). The mouse mab 120–1B10 to circulating anodic antigen (CAA) from <named-content content-type="genus-species">S. mansoni</named-content>—which specifically recognizes schistosomal glycoproteins (sGPs) carrying O-linked GlcAβ1–3GalNacβ1–6(GlcAβ1–3)GalNAcβ1 repeating units—was described previously [<xref rid="pmed-0030253-b022" ref-type="bibr">22</xref>,<xref rid="pmed-0030253-b023" ref-type="bibr">23</xref>]. The mouse mab 114–5B1, which recognizes the schistosome-specific oligosaccharide structure GalNacβ1–4[Fucα1–2Fucα1–3]GlcNacβ-R (or LDN-DF), was described previously [<xref rid="pmed-0030253-b024" ref-type="bibr">24</xref>,<xref rid="pmed-0030253-b025" ref-type="bibr">25</xref>].</p><p>Polyclonal rabbit antisera recognizing the human LDL receptor [<xref rid="pmed-0030253-b026" ref-type="bibr">26</xref>] and calreticulin [<xref rid="pmed-0030253-b027" ref-type="bibr">27</xref>] were a generous gift of Jürgen Gent and Ineke Braakman (Utrecht University). Blocking antibody against human FcγRIIa was mab IV.3 from Medarex (Annandale, New Jersey, United States). Rabbit anti-human ApoA1 and ApoB were from Calbiochem (La Jolla, California, United States), rabbit anti-mouse ApoA1 was from Biodesign (Saco, Maine, United States). Rat mab 1D4B to mouse LAMP-1 was from Santa Cruz Biotechnology (Santa Cruz, California, United States). Biotinylation of affinity-purified antibodies with biotin-X succinimidyl ester (Molecular Probes) was performed according to the vendor's instructions. Fluorescently labelled secondary goat antibodies were obtained from Jackson ImmunoResearch Laboratories (West Grove, Pennsylvania, United States). Horseradish peroxidase-conjugated secondary goat antibodies were from DAKO (Glostrup, Denmark).</p><p>Antibodies used in immuno-electron microscopy were anti-120–1B10-biotin conjugate, rabbit anti-SM200-biotin, rabbit anti-apolipoprotein B, goat anti-biotin conjugated to 10 nm gold (Aurion, The Netherlands) and 10 nm protein A-gold (Cell Microscopy Center, Utrecht, The Netherlands)</p></sec><sec id="s2c"><title>Sera</title><p>Sera of <named-content content-type="genus-species">S. mansoni</named-content>-infected individuals were obtained from the WHO/TDR Reference Serum Bank for African Schistosomiasis. Control sera were from donors with no known history of schistosomiasis.</p><p>Heat-inactivated foetal calf serum (FCS) was delipidated (DL-FCS) by solvent extraction. Serum was mixed with an equal volume of a 2:1 mixture of diisopropyl ether:<italic><sc>n</sc></italic>-butanol, stirred at room temperature for 30 min, and phases were separated by centrifugation at 5,000 <italic>g</italic> for 30 min. The aqueous phase was re-extracted as before with an equal volume of diisopropyl ether, dialyzed against PBS, and filter-sterilized. This treatment resulted in the total loss of intact lipoprotein particles, as judged by the loss of apolipoprotein A and B from the top fraction of a KBr gradient centrifugation analysed by SDS-PAGE and Coomassie brilliant blue staining. The cholesterol content was reduced to less than 5% of untreated serum, as determined by a colorimetric assay kit (Boehringer-Mannheim, Indianapolis, Indiana, United States).</p></sec><sec id="s2d"><title>Cell Culture, Transfection, and Uptake Experiments</title><p>Peripheral blood from normal human volunteers was collected in heparin tubes, and polymorphonuclear neutrophils (PMNs) were isolated by density gradient centrifugation using Ficoll-Hypaque [<xref rid="pmed-0030253-b028" ref-type="bibr">28</xref>]. The neutrophils were divided into 4-ml samples with 10<sup>6</sup> cells/ml in RPMI 1640 medium containing 25 mM Hepes (pH 7.2), 2 mM nonessential amino acids, 1 mM sodium pyruvate, 2 U/ml interferon γ, and 20% heat-inactivated human serum. After cells were incubated for 8 h with 20% patient or control sera at 37 °C with 5% CO<sub>2</sub>, they were washed extensively with ice-cold PBS, lysed in reducing Laemmli sample buffer, and subjected to Western blotting.</p><p>Mutant Chinese hamster ovary cells lacking the mature LDL receptor (LDLA cells) [<xref rid="pmed-0030253-b029" ref-type="bibr">29</xref>] were kindly provided by Ineke Braakman (Utrecht University, the Netherlands) and grown in Ham's F12 medium with 5% FCS, 2 mM glutamine at 37 °C with 5% CO<sub>2</sub>. LDLA cells were transfected with the human LDL receptor in pCDNA3 [<xref rid="pmed-0030253-b026" ref-type="bibr">26</xref>] or the empty vector using Lipofectamine 2000. Transfectants were cultured in normal culture medium containing 1 mg/ml geneticin. Stable cell lines were obtained by subcloning individual colonies. Positive clones were selected by immunofluorescence microscopy and tested for expression of hLDL-R by Western blotting. LDLA cells were transfected with the human FcγRIIa expression vector FcγRIIa-pRC-CMV [<xref rid="pmed-0030253-b030" ref-type="bibr">30</xref>] or the empty vector using Lipofectamine 2000. Expression of FcγRIIa was induced by 5 mM sodium butyrate 14–16 h prior to experiment. For uptake experiments, confluent cells on 3-cm dishes, or subconfluent cells on glass coverslips, were incubated with Ham's F12 medium containing 2 mM glutamine and 20% human serum for 6 h at 37 °C with 5% CO<sub>2</sub>. Cell were washed extensively with ice-cold PBS and subjected to either Western blotting or immunofluorescence microscopy as described [<xref rid="pmed-0030253-b031" ref-type="bibr">31</xref>], except that the cells were examined with a Nikon D-eclipse C1 confocal microscope using separate filters for each fluorochrome viewed (FITC: L<sub>ex</sub> = 488 nm and L<sub>em</sub> = 515 LP; Texas red: L<sub>ex</sub> = 568 nm and L<sub>em</sub> = 585 LP). Single-labelled cells with each primary/secondary antibody combination were examined and confirmed that no bleed-through occurred for the given conditions.</p></sec><sec id="s2e"><title>Metabolic Labelling of Adult <named-content content-type="genus-species">S. mansoni</named-content> Worms</title><p> <named-content content-type="genus-species">S. mansoni</named-content> adult worms were collected by perfusion of the hepatic portal system of golden hamsters at 7 wk after infection with ~ 500 cercariae per hamster [<xref rid="pmed-0030253-b032" ref-type="bibr">32</xref>]. About 100 adult worms pairs were incubated with 5 ml RPMI 1640 medium containing 100 mM Hepes (pH 7.2), 2 mM sodium pyruvate, 100 U/ml penicillin, and 100 μg/ml streptomycin. For radiolabelling of worm proteins and lipids, 18 MBq/ml Tran[<sup>35</sup>S]-label and 18 kBq/ml [U-<sup>14</sup>C]palmitic acid were added to the medium, respectively. After worms were incubated for 24 h at 37 °C with 5% CO<sub>2</sub>, the medium was subjected to isopycnic density centrifugation. The worms were washed with ice-cold PBS, and the worm proteins and lipids were analysed as described below.</p></sec><sec id="s2f"><title>Neutrophil Apoptosis and FACS Analysis</title><p>PMNs were isolated from healthy volunteers as described and cultured in RPMI 1640 medium containing 20% control or patient serum, or 10% serum preincubated with 10 mg of rabbit anti-ApoB antibodies or rabbit IgG (control) for 30 min at room temperature, and immunocomplexes were allowed to form. After 20 h of culture, cells were counted and washed in binding buffer consisting of 10 mM Hepes, 150 mM NaCl, 5 mM KCl, 1.8 mM CaCl<sub>2</sub>, and 1 mM MgCl<sub>2</sub>. Annexin V FITC (BD Biosciences Pharmingen) was added to the cells and incubated for 15 min at room temperature, and cells were measured on a FACSCalibur (Becton Dickinson, San Jose, California, United States) to determine the percentage of dead cells. Survival was calculated as the percentage of living cells relative to the input number of PMNs.</p></sec><sec id="s2g"><title>Isopycnic Density Centrifugation and PI-PLC Treatment</title><p>Serum or medium was spun for 3 h at 39,000 rpm at 4 °C in a SW41 rotor (Beckman). Pellet and supernatant were designated as P120 and S120, respectively. In some cases, the S120 was split in two equal fractions and either treated with 4 U/ml PI-PLC or mock-treated for 6 h at 37 °C. Lipoprotein particles were separated from soluble proteins by isopycnic density centrifugation. In short, KBr was added to S120 to a final concentration of 0.33 g/ml, and the sample was spun for 2 d at 39,000 rpm at 4 °C in a SW41 rotor. Six or twelve fractions were taken from the top. The top and bottom fractions had a density of 1.22 and 1.40 g/cm<sup>3</sup>, respectively. The proteins were precipitated with chloroform/methanol, and the lipids in the chloroform phase were analysed separately. Lipids were separated by thin-layer chromatography using silica gel 60 plates and chloroform/methanol/25% NH<sub>4</sub>OH (65:25:4 v/v) as running solvent. Radiolabelled spots were detected by phosphor-imaging on a Storm phosphor-imager. Spots were identified by comparison to standards and quantified using the Imagequant Software (Molecular Dynamics, Sunnyvale, California). Radiolabelled proteins were analysed by SDS-PAGE followed by autoradiography or by Western blotting.</p></sec><sec id="s2h"><title>Immunoprecipitation</title><p>S120 was precleared during a 2-h incubation with 0.25 volumes of streptavidin-agarose or empty Sepharose beads. A fraction of each supernatant was used to determine relative amounts of ApoA1 and ApoB by Western blotting. The remainder was incubated with biotinylated anti-rabbit SM200 or a mixture of biotinylated 120–1B10 and 114–5B1 antibodies preadsorbed to streptavidin-agarose. For immunoprecipitations of human antibodies from serum, the samples were incubated with protein A-Sepharose beads without addition of a primary antibody. Immunoprecipitates were washed at least ten times with five volumes of PBS containing 0.5% bovine serum albumin and once with five volumes of PBS. Proteins were eluted from beads using either a detergent-containing buffer (150 mM NaCl, 2 mM EDTA, 100 mM Tris-Cl [pH 8.3], 0.5% Nonidet P 40, 0.5% sodium deoxycholate, and 0.1% SDS) or Laemmli sample buffer containing 5% β-mercaptoethanol.</p></sec><sec id="s2i"><title>Immuno-Electron Microscopy</title><p>Lipoprotein particles were separated from the blood of either the positive or the control patient by a KBr gradient. Of this fraction 3-μl droplets were placed on carbon-coated copper grids with a formvar film and incubated for 30 min. The particles attach to the formvar film. After washing with PBS, the particles were labelled as previously described [<xref rid="pmed-0030253-b033" ref-type="bibr">33</xref>] with biotinylated primary antibodies followed by goat anti-biotin directly coupled to gold. To detect human antibodies on host lipoproteains, purified particles were labelled with protein A-gold only.</p></sec></sec><sec id="s3"><title>Results</title><sec id="s3a"><title>Schistosomes Release Proteins and Lipids on Lipoprotein Particles</title><p>Adult schistosomes release proteins and lipids from their surface into the circulation [<xref rid="pmed-0030253-b034" ref-type="bibr">34</xref>]. To investigate whether they are present on lipoprotein particles, sera from patients chronically infected with <named-content content-type="genus-species">S. mansoni</named-content> were cleared of parasite remnants and membrane fragments by high-speed centrifugation and analysed by isopycnic density centrifugation (<xref ref-type="fig" rid="pmed-0030253-g001">Figure 1</xref>A). The top fraction contained all lipoprotein particles, whereas soluble proteins were found in fractions of relatively higher density. Remarkably, several GPI-anchored sGPs as well as the GPI-anchored surface protein SM200 selectively accumulated in the lipoprotein fraction of patient but not control sera (<xref ref-type="fig" rid="pmed-0030253-g001">Figure 1</xref>A, top lanes).</p><fig id="pmed-0030253-g001" position="float"><label>Figure 1</label><caption><title>Lipoprotein Particles of Infected Humans Contain Schistosomal Surface Proteins</title><p>(A) Lipoprotein particles of control and patient sera were separated from soluble proteins by KBr density gradient centrifugation. Proteins from gradient fractions were analysed by SDS-PAGE and Western blotting against schistosome-specific glycoproteins (sGP), schistosomal GPI-anchored protein SM200, and human apolipoprotein B (ApoB). Note the presence of sGP and SM200 in the top fractions that contain the lipoprotein particles. Strong background staining is due to high protein concentrations in serum; data are representative of six different patient sera.</p><p>(B) Control (C) and patient (P1 and P2) sera were mock-treated (−) or treated with PI-PLC (+) and subjected to immunoprecipitation with biotinylated antibodies against either sGP (left) or SM200 (right). Immunocomplexes were isolated with streptavidin-agarose beads and analysed by SDS-PAGE and Western blotting for ApoB. Streptavidin beads alone did not pull down ApoB in patient sera (unpublished data). Antibodies were raised against a peptide sequence or a common glycoconjugate and recognize the protein irrespective of the presence of the diacylglycerol moiety.</p></caption><graphic xlink:href="pmed.0030253.g001"/></fig><p>Next, we tested whether ApoB, the major protein constituent of human LDL particles, could be coimmunoprecipitated with antisera against sGP or SM200. To avoid effects resulting from the presence of anti-<italic>Schistosoma</italic> antibodies in patient serum (see below), we used biotinylated antibodies and streptavidin-agarose for pull-downs throughout. ApoB was coimmunoprecipitated from patient sera, but not from control serum, by antibodies against both schistosomal antigens (<xref ref-type="fig" rid="pmed-0030253-g001">Figure 1</xref>B), indicating a physical association of schistosomal GPI-anchored proteins with human lipoprotein particles. The association of sGP was sensitive to treatment with PI-PLC (<xref ref-type="fig" rid="pmed-0030253-g001">Figure 1</xref>B, left), suggesting that they bound to lipoprotein particles via the GPI anchors. Immunoprecipitation of the GPI-anchored protein SM200 after PI-PLC treatment did not result in a significant reduction of recovered ApoB (<xref ref-type="fig" rid="pmed-0030253-g001">Figure 1</xref>B, right), consistent with a previous report that the GPI anchor of SM200 is resistant to PI-PLC [<xref rid="pmed-0030253-b035" ref-type="bibr">35</xref>].</p><p>The localization of schistosomal proteins on lipoprotein particles was further assessed by electron microscopy (<xref ref-type="fig" rid="pmed-0030253-g002">Figure 2</xref>). Lipoprotein particles of 10–60 nm diameter were detected in both control and patients sera by immunostaining against ApoB (<xref ref-type="fig" rid="pmed-0030253-g002">Figure 2</xref>E and <xref ref-type="fig" rid="pmed-0030253-g002">2</xref>F). No membranous fragments or parasite remnants were detected in any of the samples. Purified lipoprotein particles from patient serum contained both sGP and SM200 (<xref ref-type="fig" rid="pmed-0030253-g002">Figure 2</xref>B and <xref ref-type="fig" rid="pmed-0030253-g002">2</xref>D), which were lacking in the control sample (<xref ref-type="fig" rid="pmed-0030253-g002">Figure 2</xref>A and <xref ref-type="fig" rid="pmed-0030253-g002">2</xref>C). Note the close proximity of gold particles and lipoprotein particles; no gold dot is found that is not associated with a particle, ruling out background binding to nonparticulate material or to a possible soluble, free form of the antigen. We performed statistical analysis on the data in <xref ref-type="fig" rid="pmed-0030253-g002">Figure 2</xref>A and <xref ref-type="fig" rid="pmed-0030253-g002">2</xref>B. In three independent fields of different patients or control individuals we counted total lipoprotein particles (altogether 9,700 in control and 5,900 in patient fields) and gold particles (two controls, 48 patients) to give values of 0.5 ± 0.5 gold/1,000 lipoprotein particles for control and 8.2 ± 2 gold/1,000 lipoprotein particles for patients (<italic>p</italic> < 0.001).</p><fig id="pmed-0030253-g002" position="float"><label>Figure 2</label><caption><title>Schistosomal GPI-Linked Surface Proteins Directly Bind to Lipoprotein Particles</title><p>Lipoprotein particles isolated from control (A, C, and E) and patient (B, D, and F) sera were stained by immunogold labelling for schistosome-specific glycoconjugates using biotinylated antibody 120–1B10 (A and B), anti-SM200 (C and D), and anti-ApoB (E and F) and goat anti-biotin conjugated to 10 nm gold and visualized by electron microscopy. Bar, 100 nm. Data are representative of three infected individuals.</p></caption><graphic xlink:href="pmed.0030253.g002"/></fig></sec><sec id="s3b"><title>Transfer of Schistosomal Proteins onto Lipoprotein Particles In Vitro Depends on Intact Host Lipoprotein Particles</title><p>To reconstitute the transfer of GPI-anchored proteins onto lipoprotein particles outside the host organism, we kept adult worms in culture medium supplemented with [<sup>35</sup>S]Cys/Met and [<sup>14</sup>C]palmitate and analysed the medium for radiolabelled parasite proteins and lipids. After worms and membrane fragments were removed by high-speed centrifugation, about 2% of the radioactive proteins were found in the high speed supernatant S120 (unpublished data). The lipoprotein fraction was separated from the rest of the S120 by isopycnic density centrifugation on a KBr gradient. <xref ref-type="fig" rid="pmed-0030253-g003">Figure 3</xref>A shows the two top gradient fractions of the experiment done in normal or in delipidated medium; lipoproteins accumulate in the very top fraction 1. At least six distinct parasite proteins were found in this lipoprotein particle fraction (<xref ref-type="fig" rid="pmed-0030253-g003">Figure 3</xref>A, full, fraction 1), suggesting association with lipoprotein particles. Incubation of schistosomes in solvent-extracted delipidated medium, free of lipoprotein particles, resulted in a loss of parasite proteins from the top fraction of the gradient (<xref ref-type="fig" rid="pmed-0030253-g003">Figure 3</xref>A, DL, fraction 1). Schistosomal lipids, synthesized from the added [<sup>14</sup>C]palmitate, were also found in the top fraction of the full, but not the delipidated medium (<xref ref-type="fig" rid="pmed-0030253-g003">Figure 3</xref>A, [<sup>14</sup>C]-lipids). This indicates that <named-content content-type="genus-species">S. mansoni</named-content> transfers not only protein but also lipids onto host lipoprotein particles. Mass spectrometrical analysis confirmed the presence of schistosome-specific C20:1 fatty acid on phosphatidylcholine in the lipoprotein fraction of patient, but not of control serum (unpublished data).</p><fig id="pmed-0030253-g003" position="float"><label>Figure 3</label><caption><title>Schistosomes Transfer Glycoproteins and Lipids to Lipoprotein Particles In Vitro</title><p>Adult <named-content content-type="genus-species">S. mansoni</named-content> worms were incubated overnight in medium with [<sup>35</sup>S]-amino acids and [<sup>14</sup>C]-palmitic acid in the presence of normal (full) or delipidated (DL) FCS, and media were precleared by high-speed centrifugation.</p><p>(A) Media were subjected to KBr density gradient centrifugation, and fractions were analysed for schistosomal proteins by SDS-PAGE followed by autoradiography (top) and for schistosomal lipids by TLC followed by autoradiography (bottom). The top and second top fractions of the gradients are shown. Note the presence of prominent protein bands at 25, 70, and 250 kDa in the top fraction of full but not of delipidated medium.</p><p>(B) Top fractions from KBr density gradient centrifugations of mock-treated (− PI-PLC) or PI-PLC treated (+ PI-PLC) normal media (full) or delipidated media (DL) were analysed for schistosome-specific glycoproteins (top) and for SM200 (middle) by Western blotting and for schistosomal lipids by thin-layer chromatography followed by autoradiography (bottom).</p></caption><graphic xlink:href="pmed.0030253.g003"/></fig><p>Both SM200 and sGP were found in the lipoprotein fraction of the same culture medium (<xref ref-type="fig" rid="pmed-0030253-g003">Figure 3</xref>B), indicating that at least some of the radioactive schistosomal proteins in the lipoprotein fraction are GPI-anchored. This association was sensitive to PI-PLC treatment prior to density centrifugation, which caused sGP, but not SM200, to disappear from the top fraction (<xref ref-type="fig" rid="pmed-0030253-g003">Figure 3</xref>B, full +), consistent with the observations made with patient sera (<xref ref-type="fig" rid="pmed-0030253-g001">Figure 1</xref>). We conclude that the transfer of GPI-linked schistosomal surface proteins to lipoprotein particles can take place in vitro, requires the presence of host lipoprotein particles in the serum, and is independent of lipoprotein turnover by the host cells.</p></sec><sec id="s3c"><title>LDL Receptor-Dependent Uptake of Schistosomal Glycoproteins</title><p>To distribute lipids to peripheral tissues, LDL particles are constantly endocytosed by recipient cells. Thus, schistosomal antigens on lipoprotein particles should spread not only in the circulation, but also into host cells in the periphery of the body. To address this possibility, CHO-LDLA cells, which lack a functional LDL receptor [<xref rid="pmed-0030253-b029" ref-type="bibr">29</xref>], were stably retransfected with the human LDL receptor or were mock-transfected and incubated with medium containing control (C in <xref ref-type="fig" rid="pmed-0030253-g004">Figure 4</xref>) or patient (P1 and P2 in <xref ref-type="fig" rid="pmed-0030253-g004">Figure 4</xref>) serum. As shown by immunoblotting of cell lysates, only cells with the LDL receptor (+ in <xref ref-type="fig" rid="pmed-0030253-g004">Figure 4</xref>), but not mock-transfected controls (− in <xref ref-type="fig" rid="pmed-0030253-g004">Figure 4</xref>) were able to take up SM200 and other schistosomal glycoproteins from patient serum. To follow the intracellular localization of the internalized material, confocal immunofluorescence microscopy was performed (<xref ref-type="fig" rid="pmed-0030253-g005">Figure 5</xref>). Cells with (LDLR in <xref ref-type="fig" rid="pmed-0030253-g005">Figure 5</xref>) or without (LDLA in <xref ref-type="fig" rid="pmed-0030253-g005">Figure 5</xref>) the LDL receptor were incubated in patient serum, then stained for either sGP or SM200 (red) and counterstained for the lysosomal marker protein LAMP-1 (green). As a negative control, cells with LDL receptor were also incubated in control serum and stained for the same markers. Both sGP and SM200 became internalized into structures positive for the lysosomal marker protein LAMP-1 (LDLR + P in <xref ref-type="fig" rid="pmed-0030253-g005">Figure 5</xref>), as demonstrated by a virtually complete colocalization of the two markers (overlay, yellow). These lysosomes were characteristically increased in size (diameter mean ± standard deviation: control, 0.4 ± 0.2 μm; patient serum, 1.0 ± 0.3 μm; <italic>n</italic> = 38 each; <italic>p</italic> < 0.001) and decreased in number compared to cells that either expressed no receptor (<xref ref-type="fig" rid="pmed-0030253-g005">Figure 5</xref>, LDLA + P) or expressed receptor but received control serum (<xref ref-type="fig" rid="pmed-0030253-g005">Figure 5</xref>, LDLR + C). A small amount of schistosomal antigens was also internalized in LDLA cells (<xref ref-type="fig" rid="pmed-0030253-g005">Figure 5</xref>, LDLA + P), but did not colocalize with the lysosomal marker.</p><fig id="pmed-0030253-g004" position="float"><label>Figure 4</label><caption><title>The LDL Receptor Pathway Can Mediate Endocytosis of Schistosomal Surface Proteins</title><p>CHO-LDLA cells lacking (−) or retransfected with (+) LDL receptor were incubated with control (C) or patient (P1 and P2) sera. Cell lysates were subjected to SDS-PAGE and Western blotting for endocytosed SM200 or schistosome-specific glycoproteins. As loading controls, the medium was probed for ApoB, and the cells for calreticulin and for the presence of transfected LDL-receptor.</p></caption><graphic xlink:href="pmed.0030253.g004"/></fig><fig id="pmed-0030253-g005" position="float"><label>Figure 5</label><caption><title>Endocytosed Schistosomal Surface Proteins Accumulate in Enlarged Lysosomes</title><p>CHO-LDLA cells lacking (LDLA) or re-transfected with LDL receptor (LDLR) were incubated with control (C) or patient (P) sera, fixed, and processed for immunofluorescence microscopy. Localization of schistosomal glycoproteins (sGP, red) or SM200 (SM200, red) and of the lysosomal marker protein LAMP-1 (green) was imaged by confocal laser fluorescence microscopy. Note the enlarged lysosomes in cells (LDLR + P) that take up schistosomal proteins on lipoprotein particles. Bars, 10 μm.</p></caption><graphic xlink:href="pmed.0030253.g005"/></fig></sec><sec id="s3d"><title>Antibody-Antigen Complexes on Lipoprotein Particles in Patient Serum</title><p>The presence of schistosomal antigens on host lipoprotein particles enables host antibodies to bind indirectly to lipoproteins via the attached antigens. In patient serum, but not in control serum, ApoB was precipitated by IgG-binding protein A beads without the addition of antibodies against schistosomal antigens, demonstrating binding of host antibodies to lipoprotein particles (<xref ref-type="fig" rid="pmed-0030253-g006">Figure 6</xref>A). Ultrastructural analysis (<xref ref-type="fig" rid="pmed-0030253-g006">Figure 6</xref>B) also showed human IgG on lipoprotein particles from patient serum.</p><fig id="pmed-0030253-g006" position="float"><label>Figure 6</label><caption><title>Lipoproteins from Infected Humans Are Decorated with Antibodies and Can Be Endocytosed Via the FcγRIIa Receptor</title><p>(A) Control (C) and patient (P1 and P2) sera were incubated with protein A-Sepharose beads. Bound material was eluted and subjected to SDS-PAGE and Western blotting for ApoB.</p><p>(B) Lipoprotein particles were isolated from control and patient sera, stained for endogenous antibodies present on lipoprotein particles using protein A-gold, and subjected to electron microscopy.</p><p>(C) CHO-LDLA cells lacking (−) or transfected with (+) FCγRIIa receptor were incubated with control (C) or patient (P1 and P2) sera. Cell lysates were subjected to SDS-PAGE and Western blotting for endocytosed ApoB (ApoB [Cells]). A proteolytic fragment [<xref rid="pmed-0030253-b056" ref-type="bibr">56</xref>] of ApoB was detected at ~80 kDa. As load controls, the medium was probed for ApoB (ApoB [Medium]), and the cells for calreticulin.</p></caption><graphic xlink:href="pmed.0030253.g006"/></fig><p>This result suggests that cells expressing low-affinity IgG receptors such as FcγRIIa could specifically bind and endocytose liporotein particles from infected patients. To study this possibility, we transfected CHO-LDLA cells with FcγRIIa and analysed the amount of ApoB that was endocytosed from control and patient serum (<xref ref-type="fig" rid="pmed-0030253-g006">Figure 6</xref>C). Presence of the receptor caused an increase in endocytosed lipoprotein particles from patient but not from control serum.</p></sec><sec id="s3e"><title>Neutrophils Accumulate Lipids from Patient Serum by IgG Receptor-Mediated Lipoprotein Uptake</title><p>Human PMNs physiologically express the FcγIIRa receptor and might endocytose antibody-lipoprotein complexes from schistosomiasis patient serum. We found that freshly isolated human neutrophils took up lipoprotein particles from patient serum (P1 and P2 in <xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>A), but not from control serum (<xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>A and <xref ref-type="fig" rid="pmed-0030253-g007">7</xref>C). This uptake actually depended on the FcγIIRa receptor, since it could be blocked by addition of a blocking antibody, mab IV.3 [<xref rid="pmed-0030253-b036" ref-type="bibr">36</xref>], against this receptor (<xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>B). This result suggests that clearance by the immune system is one of the reasons for the low levels of lipoprotein particles observed in schistosomiasis patient circulation [<xref rid="pmed-0030253-b037" ref-type="bibr">37</xref>].</p><fig id="pmed-0030253-g007" position="float"><label>Figure 7</label><caption><title>Lipoproteins of Patient Serum Are Endocytosed by Neutrophils and Lead to Intracellular Accumulation of Neutral Lipids and Enhanced Neutrophil Apoptosis</title><p>(A, B, and C) Freshly isolated human PMNs were incubated for 16 h with control (C) and patient (P, P1, P2, and Pat) sera. For (A and B), cells were washed extensively and cell lysates probed for ApoB and LDL-receptor by SDS-PAGE and Western blotting. A proteolytic fragment [<xref rid="pmed-0030253-b056" ref-type="bibr">56</xref>] of ApoB was detected at ~80 kDa. LDL receptor served as a load control. For (B and C), the FcγRIIa blocking mab IV.3 was added during the incubation (P1+ [B]; Pat. + mab IV.3 [C]). For (C), cells were fixed, stained with Nile Red and subjected to confocal fluorescence microscopy.</p><p>(D) PMNs of control or schistosome-infected hamster were stained with Nile Red and subjected to confocal fluorescence microscopy. Data are representative of four infected and control animals.</p><p>(E) PMNs were incubated for 20 h with 20% (left) or 10% (right) control or patient serum as indicated. For the right bar graph, 10 μg/ml normal rabbit antibody (Control) or anti-ApoB antibody were preincubated with control serum, and subsequently added to PMNs. Cells were counted before and after incubation with serum, stained with Annexin V-FITC (which indicates apoptosis), and analysed by FACS. Results are expressed as percentage of Annexin V-negative cells, relative to the number of cells before incubation with serum. Statistical analysis was done by Student's t-test, ** <italic>p</italic> < 0.01.</p></caption><graphic xlink:href="pmed.0030253.g007"/></fig><p>Uptake of lipoprotein particles by neutrophils delivers not only apolipoproteins but also the associated lipids, which could affect the cells' lipid balance. To address this question, we stained neutrophils that had been incubated in control or patient serum with a dye, Nile red, that specifically detects accumulations of neutral lipid. Increased accumulations of neutral lipid in multiple round structures, likely lipid droplets, were found after incubation in patient serum (Pat, <xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>C), again inhibited by the presence of mab IV.3 during the incubation (Pat + mab IV.3, <xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>C). To test if this neutral lipid accumulation also occurs in the acute disease state, we isolated PMNs from control or <named-content content-type="genus-species">S. mansoni</named-content>-infected hamsters and examined these cells by Nile red staining. We found an accumulation of lipid in cells of infected but not of control animals (<xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>D).</p><p>If this uptake of antibody-antigen-lipoprotein complex were harmful to the neutrophils, it might contribute to the parasites' evasion of host immune response. We cultured freshly isolated PMNs with control or patient serum, which led to apoptosis of a major fraction of the cells under both conditions, with a significantly reduced survival in patient serum (left bar graph, <xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>E). To test whether reduced survival could be caused by antibody-lipoprotein complexes, we repeated the experiment in normal serum that was preincubated with control antibodies or anti-ApoB antibodies, to allow anti-ApoB antibodies to bind directly to lipoprotein particles. This treatment resulted in nearly complete apoptosis (right bar graph, <xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>E), suggesting that antibody-lipoprotein complexes indeed destroy the most abundant phagocytes in the blood and thereby inhibit the immune response.</p></sec><sec id="s3f"><title>Presence of GPI-Linked Surface Antigens on Lipoproteins is a Widespread Phenomenon</title><p>The potential to release GPI-linked surface antigens on lipoprotein particles may not be restricted to schistosomes but shared by other blood-dwelling parasites such as <italic>Trypanosoma brucei,</italic> whose VSG surface proteins have been found on host erythrocytes [<xref rid="pmed-0030253-b018" ref-type="bibr">18</xref>]. Conditioned medium of cultivated <named-content content-type="genus-species">T. brucei</named-content> was precleared by high-speed centrifugation and analysed by isopycnic density gradient centrifugation (<xref ref-type="fig" rid="pmed-0030253-g008">Figure 8</xref>). The major GPI-linked surface antigen VSG221 accumulated in the top fraction of the gradient (culture − PI-PLC, <xref ref-type="fig" rid="pmed-0030253-g008">Figure 8</xref>A) and was sensitive to treatment with PI-PLC (culture + PI-PLC, <xref ref-type="fig" rid="pmed-0030253-g008">Figure 8</xref>A). VSG221 was specifically coimmunoprecipitated from the medium by antibodies against mouse ApoA1 (<xref ref-type="fig" rid="pmed-0030253-g008">Figure 8</xref>B), and was also sensitive to pretreatment with PI-PLC. ApoA1 was chosen because, in contrast to human serum, mouse serum HDL levels are much higher than LDL levels. We concluded, also, that VSG221 associates with lipoprotein particles via its GPI-anchor.</p><fig id="pmed-0030253-g008" position="float"><label>Figure 8</label><caption><title>Variant Surface Glycoprotein VSG221 of <named-content content-type="genus-species">T. brucei</named-content> Associates with Lipoprotein Particles</title><p>(A) <named-content content-type="genus-species">T. brucei</named-content> was cultured in vitro in the presence of mouse serum. Control medium (control) and medium from <named-content content-type="genus-species">T. brucei</named-content> in vitro cultures that was treated with PI-PLC (VSG221 + PI-PLC) or left untreated (VSG221) was subjected to KBr density gradient centrifugation. Gradient fractions were analysed by SDS-PAGE and Western blotting for the presence of VSG221.</p><p>(B) Medium from <named-content content-type="genus-species">T. brucei</named-content> in vitro cultures was treated with PI-PLC (+) or left untreated (−) and subjected to immunoprecipitation using anti-human (h) or anti-mouse (m) antibodies against ApoA1. Immunoprecipitates were subjected to SDS-PAGE and probed for VSG221. The lower blots shows loading controls for VSG221 and mouse ApoA1 from nonimmunoprecipitated medium.</p></caption><graphic xlink:href="pmed.0030253.g008"/></fig></sec></sec><sec id="s4"><title>Discussion</title><sec id="s4a"><title>GPI-Anchored Proteins as Mobile Intercellular Elements</title><p>GPI is a complex glycolipid structure that acts as a membrane anchor for many cell surface proteins of eukaryotes [<xref rid="pmed-0030253-b005" ref-type="bibr">5</xref>]. Release of GPI-anchored proteins from the cell surface by specific phospholipases may play a key role in regulation of their surface expression and functional properties. It has been demonstrated that GPI-anchored proteins can undergo intercellular transfer in vitro [<xref rid="pmed-0030253-b013" ref-type="bibr">13</xref>–<xref rid="pmed-0030253-b016" ref-type="bibr">16</xref>]. The present study shows, to our knowledge for the first time, release of endogenous parasite GPI-anchored proteins to human lipoproteins. The lipoprotein complexes formed were then transferred to host cells. In vitro and in vivo data strongly suggest that this process is connected to the pathology of parasite infection. The mechanism of transfer may either be spontaneous diffusion, enhanced by tight binding of human lipoproteins to the schistosome surface [<xref rid="pmed-0030253-b038" ref-type="bibr">38</xref>], or be supported by an unknown mechanism at the plasma membrane or in the endocytic pathway. In <italic>T. brucei,</italic> both GPI-linked surface proteins and LDL share the same endocytic pathway [<xref rid="pmed-0030253-b039" ref-type="bibr">39</xref>], suggesting that transfer takes places in endocytic or recycling organelles. Identification of the mechanism of transfer (which would be a potential drug target) and study of its regulation, are important goals of future research. The principle demonstrated here—GPI-protein transfer between organisms—may also apply to prion diseases such as bovine spongiform encephalopathy and Creutzfeldt-Jakob disease, which are caused by infective GPI-anchored proteins. Intercellular transfer of prion protein has been observed in cell culture [<xref rid="pmed-0030253-b015" ref-type="bibr">15</xref>], but it was claimed that spreading occurs on exosomes [<xref rid="pmed-0030253-b040" ref-type="bibr">40</xref>]. In light of our findings, it seems worthwhile to investigate the role of lipoproteins in prion protein spreading in infected animals.</p></sec><sec id="s4b"><title>Immunologic and Pathophysiologic Aspects</title><p>To evade the immune response, parasites have developed various strategies [<xref rid="pmed-0030253-b041" ref-type="bibr">41</xref>], such as the rapid antigenic variation of trypanosomal surface VSG proteins [<xref rid="pmed-0030253-b042" ref-type="bibr">42</xref>]. It is unclear how schistosomes protect themselves from the immune response, in particular adult worms, which can live for years in the portal vein [<xref rid="pmed-0030253-b002" ref-type="bibr">2</xref>]. It appears that release of antigens is a major factor determining the immune response against schistosomes [<xref rid="pmed-0030253-b043" ref-type="bibr">43</xref>,<xref rid="pmed-0030253-b044" ref-type="bibr">44</xref>], but the effects of these released antigens are subject to dispute [<xref rid="pmed-0030253-b045" ref-type="bibr">45</xref>,<xref rid="pmed-0030253-b046" ref-type="bibr">46</xref>]. In general, released antigens have negative modulating effects on the immune system, such as blocking of antibodies and effector cells, induction of tolerance, or activation of suppressor cells. Also, it has been noted that the immune response against a lipopeptide construct that can bind to surface lipids is different from that against the corresponding free peptide for both a viral [<xref rid="pmed-0030253-b047" ref-type="bibr">47</xref>] and a schistosomal antigen [<xref rid="pmed-0030253-b048" ref-type="bibr">48</xref>]. We also speculate that travelling on a lipoprotein surface would change the antigenic properties of a protein, although so far we have found only GPI-linked proteins and other lipidated proteins (Hedgehog and Wingless [<xref rid="pmed-0030253-b020" ref-type="bibr">20</xref>]) on lipoprotein surfaces. Since these proteins are also normally found on cell surface lipids or membrane fragments, uptake, processing, and presentation by antigen-presenting cells should not be grossly altered.</p><p>What seems more important is that spreading of antigens via lipoproteins intermixes two pathways—antigen phagocytosis and lipid turnover—that normally are kept separate from each other. On the one hand, cells that are not specialized in antigen phagocytosis will endocytose schistosomal surface antigens via the LDL receptor pathway. As an example, endocytosis of lipoprotein particles from schistosomiasis patient serum changed the morphology of lysosomes in fibroblast cells, which probably reflects problems in the lysosomal degradation of schistosomal glycoconjugates. Similar lysosomal phenotypes are observed in glycolipid storage diseases such as Fabry's and Gaucher's disease, with accumulations in kidney tubule cells and cardiac muscle [<xref rid="pmed-0030253-b049" ref-type="bibr">49</xref>], or in macrophages [<xref rid="pmed-0030253-b050" ref-type="bibr">50</xref>], respectively. In schistosomiasis, lysosomal abnormalities or defects have not been described so far. On the other hand, immune cells that are specialized for endocytosis of antibody-antigen complexes will attack lipoprotein particles. Neutrophils, which are the most important phagocytes of the blood, take up lipoproteins of schistosomiasis patients via FcγRIIa receptors, as demonstrated in this study. This should contribute to the reduction of lipoprotein levels in patients [<xref rid="pmed-0030253-b037" ref-type="bibr">37</xref>], which has the interesting secondary effect of protecting the host from atherosclerosis [<xref rid="pmed-0030253-b051" ref-type="bibr">51</xref>,<xref rid="pmed-0030253-b052" ref-type="bibr">52</xref>]. With the endocytosed lipoproteins, these cells take up large amounts of lipids, giving rise to intracellular accumulation of lipids, shown in <xref ref-type="fig" rid="pmed-0030253-g007">Figure 7</xref>C and <xref ref-type="fig" rid="pmed-0030253-g007">7</xref>D both in vitro and in vivo. Damage induced by the lipid accumulation may contribute to neutrophil apoptosis and low neutrophil counts in schistosomiasis patients [<xref rid="pmed-0030253-b053" ref-type="bibr">53</xref>], supporting immune evasion by the parasite. In support of this, schistosomiasis patient serum promotes neutrophil apoptosis. We can mimic this effect by using normal serum to which anti-ApoB antibodies were added, which effectively is a treatment with a lipoprotein-antibody complex similar to the ones generated in schistosomiasis by antibody-GPI-protein complexes on the lipoprotein particles. Peripheral immunocomplexes in schistosomiasis also act on B cells, which also carry the Fc receptor, to reduce proliferation and MHC class II expression [<xref rid="pmed-0030253-b054" ref-type="bibr">54</xref>]. Other cells that endocytose antigen via Fc receptors are macrophages and dendritic cells, which also are exposed to antigen-loaded lipoproteins. Released schistosomal antigens strongly modulate their behavior to favour a Th-2 type immune response [<xref rid="pmed-0030253-b002" ref-type="bibr">2</xref>]. It is tempting to speculate that, similar to the alterations of lysosomal structure in fibroblast in response to antigen-loaded lipoproteins, antigen processing by these specialized cells might also be altered. Since lysosomal activity is an important parameter influencing efficient antigen presentation [<xref rid="pmed-0030253-b055" ref-type="bibr">55</xref>], this might ultimately contribute to a modified immune response.</p></sec></sec>
IL-32: An Emerging Player in the Immune Response Network against Tuberculosis?	Could not extract abstract	<contrib contrib-type="author"><name><surname>Kundu</surname><given-names>Manikuntala</given-names></name></contrib><contrib contrib-type="author"><name><surname>Basu</surname><given-names>Joyoti</given-names></name><xref ref-type="corresp" rid="cor1">*</xref></contrib>	PLoS Medicine	<p>Interleukin-32 (IL-32), which was previously called natural killer cell transcript 4, has recently been recognized as a proinflammatory cytokine (see Glossary) [<xref rid="pmed-0030274-b001" ref-type="bibr">1</xref>]. The main sources of IL-32 are natural killer cells, T cells, epithelial cells, and blood monocytes. Four transcripts of IL-32 are known at present. IL-32 has emerged as an important player in innate and adaptive immune responses, and information is emerging on synergism between IL-32 and other well-characterized players in innate immunity.</p><p>The innate immune response depends on the recognition of pathogen-associated molecular patterns by families of pathogen recognition receptors. The best characterized among these are the Toll-like receptor (TLR) [<xref rid="pmed-0030274-b002" ref-type="bibr">2</xref>] and the nucleotide-binding oligomerization domain (NOD) [<xref rid="pmed-0030274-b003" ref-type="bibr">3</xref>] families of proteins. Recent studies have shown that IL-32 synergizes with NOD1 and NOD2 ligands to stimulate IL-1ß and IL-6 release in a caspase-1-dependent manner [<xref rid="pmed-0030274-b004" ref-type="bibr">4</xref>]. These findings are of potential clinical importance in settings where NOD2 plays a protective role, such as in Crohn's disease, where NOD2-dependent production of defensins and cytokines contributes to antimicrobial defense in the gut. These findings are also likely to be of significance in tuberculosis. Individuals homozygous for the 3020ins C NOD2 mutation show a defective cytokine response to <italic>Mycobacterium tuberculosis</italic> [<xref rid="pmed-0030274-b005" ref-type="bibr">5</xref>]. It is in this context that the study reported by Netea et al. in <italic>PLoS Medicine</italic>, in which the authors explored the regulation of IL-32 production by primary cells of the immune system, is of potential importance [<xref rid="pmed-0030274-b006" ref-type="bibr">6</xref>].</p><sec id="s2"><title>Stimulation of IL-32 by Mycobacteria</title><p>Netea et al. explored the role of IL-32 in the context of <italic>M. tuberculosis</italic> infection. In their study, freshly obtained human peripheral blood mononuclear cells (PBMCs) were stimulated with different TLR agonists, and gene expression and synthesis of IL-32 was determined. The authors showed that <italic>M. tuberculosis</italic> could elicit IL-32 release from PBMCs as well as from purifed monocyte populations. <disp-quote><p>IL-32 has emerged as an important player in innate and adaptive immune responses.</p></disp-quote>This is the first documented instance of a probable role of IL-32 in the immune response elicited by an intracellular pathogen. The authors also found that other heat-killed organisms, such as <italic>Staphylococcus aureus</italic>, <italic>Candida albicans</italic>, or <italic>Aspergillus fumigatus</italic>, do not stimulate IL-32 production, although these organisms are potent inducers of the proinflammatory cytokines IL-6 and TNF-α. </p><p>This study showed that the expression of the genes of isoforms α and γ of IL-32 is stimulated by <italic>M. tuberculosis</italic>, whereas the gene for the ß isoform is constitutively expressed (i.e., always expressed, even without a stimulus). The authors provide evidence in favor of a role of IFN-γ in the induction of IL-32 synthesis. They went on to show that <italic>M. tuberculosis</italic> elicits caspase-1-dependent cleavage of the precursor of IL-18, its release, and IL-18-dependent production of IFN-γ, which could be blocked by IL-18 blocking peptide. The authors also claim that TNF-α release from PBMCs challenged with <italic>M. tuberculosis</italic> occurs in an IL-32-independent manner.</p></sec><sec id="s3"><title>Strengths and Weaknesses of the Study</title><p>The crucial role of IFN-γ in antimycobacterial immunity has been shown in patients with defects in the IFN-γ receptor [<xref rid="pmed-0030274-b007" ref-type="bibr">7</xref>]. Netea and colleagues' findings raise important questions regarding (a) how IL-32 levels vary in patients with defects in IFN-γ receptors who are susceptible to tuberculosis and (b) how these variations may impact the course of the infection. One of the deficiencies of their study stems from the fact that the role of IL-32 in controlling <italic>M. tuberculosis</italic>-dependent cytokine networks has not been investigated directly by neutralizing or blocking IL-32. Rather, the authors' inferences have been drawn indirectly through blocking IFN-γ and IL-18 signaling. PR3 has recently been identified as an IL-32 binding protein [<xref rid="pmed-0030274-b008" ref-type="bibr">8</xref>], and therefore researchers can now explore the effect of neutralizing IL-32 using enzymatically inactive PR3 or PR3-derived peptides.</p><p>In addition, the present report fails to address how the synthesis of the different isoforms of the IL-32 protein is regulated, which will require future exploration. IL-32 has been reported to induce the pro-inflammatory cytokines TNF-α and IL-1ß from murine peritoneal macrophages, as well as from phorbol ester-differentiated human THP-1 cells [<xref rid="pmed-0030274-b001" ref-type="bibr">1</xref>]. Netea et al. show that IL-32 production is not associated with TNF-α induction. This raises the question whether the differences in TNF-α induction are attributable to intrinsic differences in behavior of monocytic cells in comparison with differentiated macrophages.</p><p>The findings reported in this study raise interesting questions and open avenues for further exploration. One question is whether external stimuli such as IFN-γ regulate IL-32 at both the transcriptional as well as the translational level. The likely roles of the 5′- and 3′-untranslated regions (UTRs) of the IL-32 mRNA in regulating IL-32 production need to be deciphered. IL-32 production has been observed to be TLR independent but IFN-γ dependent [<xref rid="pmed-0030274-b001" ref-type="bibr">1</xref>]. MyD88, a classical TLR adapter protein [<xref rid="pmed-0030274-b002" ref-type="bibr">2</xref>] central to TLR signaling, has recently been shown to signal through the IFN-γ receptor [<xref rid="pmed-0030274-b009" ref-type="bibr">9</xref>] through a non-Toll-IL-1R domain. MyD88 stabilizes mRNA through classical AU-rich elements found in the 3′ UTRs of many mRNAs. Does MyD88 have any role in relation to the production of IL-32? This question is particularly important in view of the report showing fatal <italic>M. tuberculosis</italic> infection in mice in the absence of MyD88 [<xref rid="pmed-0030274-b010" ref-type="bibr">10</xref>]. In general, interactions between 5′ and 3′ UTRs resulting in the formation of an RNA loop increase translational efficiency [<xref rid="pmed-0030274-b011" ref-type="bibr">11</xref>], and RNA-binding proteins act at the level of enhancing or repressing these interactions to control translation. The control processes that regulate translation of IL-32 are additional areas of future exploration.</p></sec><sec id="s4"><title>Clinical Implications</title><p>For the clinician, IL-32 emerges as yet another cytokine whose role in the course of tuberculosis and related infections deserves evaluation. The cellular receptor (or interacting partner) for IL-32 needs to be identified in antigen-presenting cells, which are of particular relevance to tuberculosis. IL-32 contributes to the synovitis of rheumatoid arthritis, and the inflammation of rheumatoid arthritis correlates with IL-32 gene expression [<xref rid="pmed-0030274-b012" ref-type="bibr">12</xref>]. While considering intervention at the level of IL-32 for autoimmune disorders, its role in mounting an effective immune response against invading pathogens also needs to be considered. It would be pertinent to bear in mind that neutralization of IL-32 could render patients more susceptible to tuberculosis. The risk of tuberculosis in patients on anti-TNF-α therapy has already been documented [<xref rid="pmed-0030274-b013" ref-type="bibr">13</xref>].</p><p>The most important question raised by these studies is why <italic>M. tuberculosis</italic> induces IL-32 whereas other organisms that are known inducers of IFN-γ do not. The identification of likely additional IL-32 regulating pathways triggered uniquely by mycobacteria deserves immediate attention.</p><boxed-text position="float"><sec><title>Glossary</title><p> <bold>Cytokines:</bold> Broad group of signaling proteins that are in general produced by immune cells after cell activation and act as autocrine or paracrine regulators of the immune response.</p><p> <bold>Natural killer cells:</bold> Large, granular, bone-marrow-derived lymphocytes of the innate immune system that release cytolytic molecules to kill infected cells and tumor cells.</p><p> <bold>PAMPS:</bold> Specific conserved structures of pathogens recognized by pattern recognition receptors on macrophages and dendritic cells.</p><p> <bold>Toll-like receptor:</bold> Germline-encoded pattern recognition receptors that recognize conserved molecular patterns shared by microorganisms.</p><p> <bold>NOD families of proteins:</bold> A family of cytoplasmic proteins that contain a nucleotide-binding site and a leucine-rich repeat and function as cytosolic sensors involved in innate recognition of microorganisms and regulation of inflammatory responses.</p><p> <bold>Interleukins:</bold> Secreted regulatory proteins produced by immune cells such as monocytes and lymphocytes in response to stimuli that help the immune system fight infection and diseases such as cancer.</p><p> <bold>Caspase-1 dependent:</bold> Refers to the thiol protease caspase-1-mediated proteolytic processing of the precursor form of IL-18 to the mature, active form of the cytokine.</p><p> <bold>Defensins:</bold> Cationic, cysteine-rich peptides found in the cytoplasmic granules of neutrophils and macrophages possessing broad antimicrobial activity against bacteria, fungi, and enveloped viruses.</p><p> <bold>PR3:</bold> A granule serine protease present in neutrophils and monocytes capable of processing a variety of biological substrates.</p><p> <bold>MyD88:</bold> An adapter molecule that binds to the intracellular domains of TLRs and recruits a number of molecules to the TLR complexes to trigger signaling.</p></sec></boxed-text></sec>
Inflammatory Manifestations of Experimental Lymphatic Insufficiency	<sec id="st1"><title>Background</title><p>Sustained lymph stagnation engenders a pathological response that is complex and not well characterized. Tissue inflammation in lymphedema may reflect either an active or passive consequence of impaired immune traffic.</p></sec><sec id="st2"><title>Methods and Findings</title><p>We studied an experimental model of acute post-surgical lymphedema in the tails of female hairless, immunocompetent SKH-1 mice. We performed in vivo imaging of impaired immune traffic in experimental, murine acquired lymphatic insufficiency. We demonstrated impaired mobilization of immunocompetent cells from the lymphedematous region. These findings correlated with histopathological alterations and large-scale transcriptional profiling results. We found intense inflammatory changes in the dermis and the subdermis. The molecular pattern in the RNA extracted from the whole tissue was dominated by the upregulation of genes related to acute inflammation, immune response, complement activation, wound healing, fibrosis, and oxidative stress response.</p></sec><sec id="st3"><title>Conclusions</title><p>We have characterized a mouse model of acute, acquired lymphedema using in vivo functional imaging and histopathological correlation. The model closely simulates the volume response, histopathology, and lymphoscintigraphic characteristics of human acquired lymphedema, and the response is accompanied by an increase in the number and size of microlymphatic structures in the lymphedematous cutaneous tissues. Molecular characterization through clustering of genes with known functions provides insights into processes and signaling pathways that compose the acute tissue response to lymph stagnation. Further study of genes identified through this effort will continue to elucidate the molecular mechanisms and lead to potential therapeutic strategies for lymphatic vascular insufficiency.</p></sec>	<contrib contrib-type="author"><name><surname>Tabibiazar</surname><given-names>Raymond</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Cheung</surname><given-names>Lauren</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Han</surname><given-names>Jennifer</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Swanson</surname><given-names>Jeffrey</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Beilhack</surname><given-names>Andreas</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>An</surname><given-names>Andrew</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Dadras</surname><given-names>Soheil S</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Rockson</surname><given-names>Ned</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Joshi</surname><given-names>Smita</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Wagner</surname><given-names>Roger</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref></contrib><contrib contrib-type="author"><name><surname>Rockson</surname><given-names>Stanley G</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref><xref ref-type="corresp" rid="cor1">*</xref></contrib>	PLoS Medicine	<sec id="s1"><title>Introduction</title><p>Acquired lymphedema is a common, important, and often devastating consequence of successful surgical and adjuvant therapy of breast cancer and other malignancies [ <xref rid="pmed-0030254-b001" ref-type="bibr">1</xref>, <xref rid="pmed-0030254-b002" ref-type="bibr">2</xref>]. The biology of regional lymphatic vascular insufficiency (lymphedema) is complex and, as yet, poorly understood. Consequently, there is a paucity of effective treatment strategies in patients with lymphedema [ <xref rid="pmed-0030254-b002" ref-type="bibr">2</xref>]. There is an obvious need for better molecular characterization of this disease process to elucidate the pathobiology of lymphatic vascular insufficiency. </p><p>The tissue response to lymph stagnation is rather complex. The profound structural and functional abnormalities in the lymphedematous tissues reflect a multicellular response to impaired extracellular fluid mobilization [ <xref rid="pmed-0030254-b003" ref-type="bibr">3</xref>]. It has been suggested that lymphedema provokes an inflammatory tissue response in the skin [ <xref rid="pmed-0030254-b004" ref-type="bibr">4</xref>]. While it is conceivable that the inflammatory nature of the tissue response to lymph stagnation reflects either the active or passive consequences of impaired immune traffic [ <xref rid="pmed-0030254-b005" ref-type="bibr">5</xref>], direct experimental confirmation is lacking. </p><p>Transcriptional profiling has been utilized in the molecular characterization of isolated lymphatic endothelia [ <xref rid="pmed-0030254-b006" ref-type="bibr">6</xref>, <xref rid="pmed-0030254-b007" ref-type="bibr">7</xref>], but the molecular end-organ response to lymph stagnation remains unaddressed and poorly understood. While lymphatic and blood vessel responses to injury might be predictable, an elucidation of whole tissue response to disease is likely to provide more relevant insights into the important interactions between the tissue matrix and the resident, heterogeneous cellular populations that likely compose the target-organ response to persistent lymph stagnation. </p><p>To investigate tissue responses to lymphatic vascular insufficiency, we have undertaken dynamic, in vivo imaging of the impaired immune traffic in a murine model of acquired lymphatic insufficiency that is intended to simulate, in part, the lymphatic dysfunction of post-surgical lymphedema [ <xref rid="pmed-0030254-b008" ref-type="bibr">8</xref>]. These observations were correlated with an assessment of the cutaneous histopathology in the lymphedema tissue. Furthermore, to investigate the molecular mechanisms of tissue response to lymphatic vascular insufficiency, we have undertaken a large-scale transcriptional profiling of the lymphedema tissue utilizing a comprehensive mouse cDNA microarray containing 42,300 features, representing over 25,000 unique genes and expressed sequence tags (ESTs) [ <xref rid="pmed-0030254-b009" ref-type="bibr">9</xref>]. The patterns of gene expression in lymphedema were contrasted with those observed in normal and surgical sham controls. </p></sec><sec id="s2"><title>Methods</title><p>This study was approved by the Administrative Panels on Laboratory Animal Care of Stanford University.</p><sec id="s2a"><title>Creation of Experimental Lymphedema</title><p>Post-surgical lymphedema was experimentally created in the tails of female hairless, immunocompetent SKH-1 mice (Charles River Laboratories, Boston, Massachusetts, United States). Prior to surgery, the mice were anesthetized with intraperitoneal injection of 0.07 cc of a solution containing ketamine, xylazine, and saline. For each intervention, the skin of the tail was circumferentially incised proximally, at a point 16 mm distal to its base. The major lymphatic trunks were identified through subcutaneous injection of methylene blue distal to the surgical incision, followed by controlled, limited cautery ablation of these structures. In surgical controls (sham animals), skin incision alone was performed, with methylene blue injection but without lymphatic cautery. The normal control animals did not undergo any surgical manipulation. All animal subjects were sacrificed on day 14 of observation. After sacrifice, 0.5-gm sections of the tail were harvested for paraffin embedding and RNA extraction.</p></sec><sec id="s2b"><title>Tail Volume Quantitation</title><p>Tail volume was quantitated in each animal subject immediately prior to sacrifice. Volumetric assessment was performed with a manually adjusted caliper, with serial measurement of the tail circumference at 5-mm intervals along its axis. The tail volume was quantitated with the truncated cone formula [ <xref rid="pmed-0030254-b010" ref-type="bibr">10</xref>]. </p></sec><sec id="s2c"><title>Histology</title><p>Immediately following sacrifice, 0.5-gm sections of the tail were harvested for histological analysis and RNA extraction. Sections extended from a point 4 mm proximal to the surgical incision to 8 mm beyond it. For examination of the responses remote from the point of injury, sections were harvested 4 cm distal to the surgical site. The specimens were fixed overnight in 4% paraformaldehyde. After paraffin embedding, 5-μm sections were stained with hematoxylin and eosin (Richard-Allan Scientific, Kalamazoo, Michigan, United States). For visualization of histiocytes/mast cells, the sections were stained with a 1% toluidine blue solution (LabChem, Pittsburgh, Pennsylvania, United States) diluted in 1% NaCl. After deparaffinization in xylene, sections were rehydrated though a series of graded alcohol steps starting with 100% EtOH and ending in 50% EtOH. Slides remained in toluidine blue for 2 min and were then dehydrated through graded alcohol washes and covered with Cytoseal (Richard-Allan Scientific).</p></sec><sec id="s2d"><title>LYVE-1 Immunohistochemical Staining</title><p>Paraffin sections 5 μm thick were deparaffinized in xylene, rehydrated in a graded series of ethanol, pretreated with target retrieval solution (Dako, Carpinteria, California, United States) in a pressure cooker, and incubated in a peroxidase block for 10 min. Sections were then incubated with rabbit polyclonal anti-LYVE-1 antibody (1:200, Upstate Cell Signaling Solutions, Lake Placid, New York, United States) for 1 h at room temperature, followed by horseradish-peroxidase-conjugated secondary antibody for 30 min at room temperature and detection with DAB for 4 min (Envision System Kit, Dako). Tissue sections were counterstained with Gill 1 hematoxylin (Richard-Allan Scientific) for 15 s, then dehydrated in graded ethanol and coverslipped with CoverSafe (American MasterTech Scientific, Lodi, California, United States).</p></sec><sec id="s2e"><title>Functional Imaging of Immune Traffic in the Lymphedema Model</title><p>Experimental lymphedema was created surgically in the tails of FVB/N female wild-type mice (Jackson Laboratory, Bar Harbor, Maine, United States; <italic>n =</italic> 3), using the technique described above. Surgical sham controls ( <italic>n =</italic> 5) were also created and compared with normal mice ( <italic>n =</italic> 5). For in vivo bioluminescence imaging, spleens from transgenic luciferase ( <italic>luc</italic> <sup>+</sup>) heterozygous animals were put into single-cell suspension, expressing firefly <italic>luc</italic> under the control of a chicken beta-actin promoter as previously described [ <xref rid="pmed-0030254-b011" ref-type="bibr">11</xref>, <xref rid="pmed-0030254-b012" ref-type="bibr">12</xref>]. The single-cell suspensions from mouse spleens consisted of different hematopoietic lineages: ˜40% were CD19+ B cells, ˜20% were CD4+ T cells, ˜10%–15% were CD8+ T cells, 3% were NK1.1+ NK cells, and the rest were GR.1+ granulocytes, Mac-1+ macrophages, CD11c+ dendritic cells, and rarer cell populations. A total of 4 × 10 <sup>6</sup> splenocytes (>97% CD45+) in PBS were injected in a volume of 20 ml into the tail interstitium, 1 cm caudal to the site of surgery, in both lymphedema mice and surgical shams. Normal mice were injected at the corresponding level of the tail. Injections were performed on post-surgical day 7. Thereafter, <italic>luc <sup>+</sup></italic> cells were repetitively imaged in vivo <italic>,</italic> at predetermined intervals following the cell injections. In brief, mice were anesthetized by intraperitoneal co-injection of a mixture of ketamine (1 mg/mouse), xylazine (μg/mouse) in PBS, and the substrate luciferin (150 mg/kg). </p><p>Ten minutes thereafter, dorsal images were obtained with an IVIS100 CCD imaging system (Xenogen, Alameda, California, United States). The efficiency of cellular lymphatic drainage was determined by direct imaging of light emission at each of the measured time points, with quantitation of the change in light emission relative to that observed 20 h after cell injection, which was defined as 100%.</p></sec><sec id="s2f"><title>Microsphere Quantitation of Arterial Perfusion of the Mouse Tail</title><p>The arterial perfusion of the tails of experimental and control mice was quantitated through intracardiac microsphere injection. After induction of general anesthesia, stable-labeled 15-μm microspheres (STERIspheres Gold, BioPAL, Worcester, Massachusetts, United States) were injected into the left ventricle. Each animal subject received 0.5 × 10 <sup>6</sup> microspheres (0.2 ml) injected directly into the left ventricle. The animals were sacrificed after 12 min. The tails were harvested and dried overnight at 70 °C. The assay to quantitate disintegrations per minute (dpm) was performed by BioPhysics Assay Laboratory (Worcester, Massachusetts, United States) as previously described [ <xref rid="pmed-0030254-b013" ref-type="bibr">13</xref>]. </p></sec><sec id="s2g"><title>Lymphoscintigraphy in Experimental Lymphedema</title><p>Whole body lymphoscintigraphy was performed after the intradermal injection of 100 μCi/0.02 ml of filtered <sup>99m</sup>Tc-sulfur colloid (100 nm size) into the tip of the tail. Dynamic and static images (255 × 255) were acquired using a parallel hole collimator in a microSPECT gamma camera (Lumigem, Gamma Medica, Northridge, California, United States). The dynamic images (1,000 frames; 0.5 s/frame) were started 60 s prior to the injection of the tracer. The injection lasted for 20 s. The static images (10 min) were acquired immediately after the dynamic acquisition. </p></sec><sec id="s2h"><title>Microarray Experimental Design, RNA Preparation, and Hybridization</title><p>Tissues were derived from nine mice for each of the three biological states under study (cutaneous specimens from normal, lymphedematous, and surgical sham animals), for a total of 27 mice. All microarray hybridizations were performed with three biological replicates, using pooled samples independently derived from three mice each, for a total of nine hybridizations. After tissue was harvested for histological examination, the remaining, distal portion of the tail was retrieved for RNA isolation. After completely separating the tail skin from the cartilage by blunt dissection, the tissue was separated into segments of 0.5 mm for further processing. Total RNA was isolated using a modified two-step purification protocol as described previously [ <xref rid="pmed-0030254-b014" ref-type="bibr">14</xref>]. RNA integrity was assessed using the Agilent 2100 Bioanalyzer System with RNA 6000 Pico LabChip Kit (Agilent, Palo Alto, California, United States). First-strand cDNA was synthesized from 15 μg of total RNA derived from each pool and from whole embryonic-day-17.5 embryo for reference RNA, in the presence of Cy3 and Cy5 dUTP, respectively, and hybridized to the Mouse Transcriptome Microarray [ <xref rid="pmed-0030254-b014" ref-type="bibr">14</xref>– <xref rid="pmed-0030254-b016" ref-type="bibr">16</xref>]. A continuously updated and annotated list of the cDNAs included on this array is available at the Stanford Microarray Database [ <xref rid="pmed-0030254-b017" ref-type="bibr">17</xref>] ( <xref ref-type="supplementary-material" rid="st001">Table S1</xref>). </p></sec><sec id="s2i"><title>Data Acquisition, Analysis, and Statistical Analysis</title><p>Image acquisition of the mouse cDNA microarrays was performed on an Agilent G2565AA Microarray Scanner System. Feature extraction was performed with GenePix 4.0 software (Bucher Biotec, Basel, Switzerland). Numerical raw data were migrated from GenePix, without processing, into an Oracle relational database (CoBi) that was designed specifically for microarray data analysis (GeneData, Basel, Switzerland). The data were then analyzed using Expressionist software (GeneData). After background subtraction and dye normalization, features with low signal intensity in the reference channel were filtered if signal was less than 2.5× background value, retaining a total of 8,353 features for further analysis. <italic>K</italic>-nearest-neighbor algorithm was applied to impute for missing values (<7% of remaining data) [ <xref rid="pmed-0030254-b018" ref-type="bibr">18</xref>]. For two-group comparisons, we used the significance analysis of microarrays (SAM) algorithm [ <xref rid="pmed-0030254-b019" ref-type="bibr">19</xref>, <xref rid="pmed-0030254-b020" ref-type="bibr">20</xref>]. Heat maps were generated using HeatMap Builder [ <xref rid="pmed-0030254-b021" ref-type="bibr">21</xref>, <xref rid="pmed-0030254-b022" ref-type="bibr">22</xref>]. For enrichment analysis we used the EASE analysis software, which uses Gene Ontology (GO) annotation and Fisher's exact test to derive biological themes within particular gene sets [ <xref rid="pmed-0030254-b023" ref-type="bibr">23</xref>]. </p></sec><sec id="s2j"><title>Quantitative Real-Time RT-PCR</title><p>Quantitative real-time RT-PCR (qRT-PCR) was performed as described [ <xref rid="pmed-0030254-b014" ref-type="bibr">14</xref>]. Primers and probes for ten representative differentially expressed genes were obtained from Applied Biosystems Assays-on-Demand (Applied Biosystems, Foster City, California, United States. cDNA was synthesized from 5 μg of total RNA using Taqman Reverse Transcription Reagents (Applied Biosystems), a set which includes MultiScribe reverse transcriptase, RNase inhibitor, dNTP mixture, oligo d(T) <sub>16</sub>, random hexamers, 10× RT buffer, and MgCl <sub>2</sub> solution. Amplification was performed in triplicate at 50 °C for 2 min and 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Reactions without template and/or enzyme were used as negative controls. 18S ribosomal RNA was used as an internal control. A standard curve derived from embryonic-day-17.5 mouse RNA was plotted for each target gene by linear regression using SPSS version 11.0 software (Applied Biosystems). RNA quantity was expressed relative to the corresponding 18S control. Fold differences were calculated by dividing the experimental results by the pooled normal results and were plotted on a log10 scale. The primers and probes utilized in this study are listed in <xref ref-type="table" rid="pmed-0030254-t001">Tables 1</xref> and <xref ref-type="table" rid="pmed-0030254-t002">2</xref>. </p><table-wrap id="pmed-0030254-t001" position="float"><label>Table 1</label><caption><p>Primer/Probe Sequences for the Taqman-Based qRT-PCR</p></caption><graphic xlink:href="pmed.0030254.t001"/></table-wrap><table-wrap id="pmed-0030254-t002" position="float"><label>Table 2</label><caption><p>Names of Taqman-Based qRT-PCR Probes</p></caption><graphic xlink:href="pmed.0030254.t002"/></table-wrap></sec></sec><sec id="s3"><title>Results</title><sec id="s3a"><title>Murine Model of Acute Experimental Lymphedema: Tail Volume Quantification</title><p>Forty-five 3-wk-old SKH-1 hairless mice were studied in this investigation. Of these, 18 underwent post-surgical lymphatic ablation, nine served as surgical sham controls, and the remaining 18 served as normal controls. Tail volume for each group of animals is depicted in <xref ref-type="fig" rid="pmed-0030254-g001">Figure 1</xref>. At post-surgical day 7, the lymphedema tail volumes were 200% ± 50% of baseline ( <italic>p <</italic> 0.008 when compared to surgical sham controls). In the animals subjected to lymphatic ablation, the edematous enlargement of the tails persisted until the day of sacrifice (day 14). Of note is the fact that cutaneous healing of the wound, both in the lymphedematous and surgical sham subjects, was complete by day 14. There was no statistically significant change in tail volume in either surgical sham or normal controls. </p><fig id="pmed-0030254-g001" position="float"><label>Figure 1</label><caption><title>Tail Volume Changes at Post-Surgical Day 7 and at Day 14</title></caption><graphic xlink:href="pmed.0030254.g001"/></fig></sec><sec id="s3b"><title>Histological Assessment of the Cutaneous Response to Lymphatic Interruption</title><p>Hematoxylin and eosin specimens derived from the lymphedematous tails were characterized by the presence of marked acute inflammatory changes ( <xref ref-type="fig" rid="pmed-0030254-g002">Figure 2</xref>B), when compared to the tissue derived from the normal tails ( <xref ref-type="fig" rid="pmed-0030254-g002">Figure 2</xref>A). There was a notable increase in cellularity, with an increase in the number of observed fibroblasts and histiocytes, as well as a large infiltration of neutrophils Granulation tissue was observed closer to the center of the section, with bystander destruction of muscle tissue. In addition, there was hyperkeratosis and spongiosis and edema of the epidermis, with irregularity of the epidermal/dermal junction, elongation of the dermal papillae, and a 2- to 3-fold expansion of tissue between the bone and the epidermis. Lymphedema specimens were characterized by the presence of numerous dilated lymphatics in the dermis and subdermis, as seen in <xref ref-type="fig" rid="pmed-0030254-g002">Figure 2</xref>B. In contrast, normal tail sections were devoid of these dilated structures. The normal tissues were characterized by the presence of a thin dermis and epidermis, with a normal epidermal/dermal junction ( <xref ref-type="fig" rid="pmed-0030254-g002">Figure 2</xref>A). The surgical sham controls were indistinguishable from normals, with no increased cellularity in dermis or epidermis, and no enlarged nuclei or hyperkeratosis. </p><fig id="pmed-0030254-g002" position="float"><label>Figure 2</label><caption><title>Histopathology of Experimental Lymphedema in the Murine Tail</title><p>Lymphedema was characterized by the presence of marked acute inflammatory changes, both adjacent to the surgical site and within distal regions of the tail, remote from the site of surgical ablation.</p><p>(A) Normal tail skin harvested 16 mm from the base of the tail is characterized by the presence of a thin dermis and epidermis, with a normal epidermal/dermal junction. Surgical sham controls were indistinguishable from normals, with no increased cellularity in dermis or epidermis, and no enlarged nuclei or hyperkeratosis.</p><p>(B) Lymphedematous skin harvested immediately distal to the site of prior surgical lymphatic ablation is characterized by the presence of marked acute inflammatory changes, absent in the tissue derived from the normal tails. There is a notable increase in cellularity, with an increase in the number of observed fibroblasts and histiocytes, as well as a large infiltration of neutrophils. There is hyperkeratosis and spongiosis and edema of the epidermis, with irregularity of the epidermal/dermal junction, elongation of the dermal papillae, and a 2- to 3-fold expansion of tissue between the bone and the epidermis. There are numerous dilated lymphatic microvessels in the dermis and subdermis (black arrows). In contrast, normal tail sections were devoid of these dilated structures.</p><p>(C) Normal skin derived from the distal tail. No inflammation, hypercellularity, or lymphatic dilatation is observed.</p><p>(D) Distal skin in lymphedema. Spongiosis and lymphatic microvascular dilatation (black arrows) are once again detectable.</p></caption><graphic xlink:href="pmed.0030254.g002"/></fig><p>In order to assess whether the lymphedematous changes created a uniform pathological response distal to the point of lymphatic ablation, the tissues were also sampled distally (4 cm distal to the point of surgical incision) in normal ( <xref ref-type="fig" rid="pmed-0030254-g002">Figure 2</xref>C) and lymphedematous ( <xref ref-type="fig" rid="pmed-0030254-g002">Figure 2</xref>D) tails. The observed changes were comparable to those observed adjacent to the surgical site: lymphedematous tissues were characterized by hypercellularity, inflammatory infiltration, and microlymphatic dilatation that were not present in the normal tissues. </p></sec><sec id="s3c"><title>Quantitative Assessment of Arterial Perfusion in the Murine Tail Lymphedema Model</title><p>While we took great care to avoid concurrent injury to adjacent vascular structures during surgical lymphatic ablation, we have undertaken an evaluation to exclude inadvertent arterial injury during surgery. The mouse tails remained grossly stable throughout the post-surgical observation phase, with no evidence of frank necrosis distal to the surgical site. In order to further substantiate the absence of an arterial ischemic contribution to the histological pathology observed in lymphedema, quantitative assessment of arterial perfusion was performed through intracardiac injection of stable 15-μm microspheres into the left ventricles of normal ( <italic>n =</italic> 3) and lymphedema ( <italic>n =</italic> 3) mice. Perfusion of the tail, measured in disintegrations per minute (dpm), did not differ statistically between the two categories (normal, 151,186 ± 69,213 dpm; sham, 95,581 ± 48,003 dpm), confirming preservation of arterial supply in the lymphedema animals. </p></sec><sec id="s3d"><title>LYVE-1 Immunohistochemical Staining</title><p>The nature of the lymphatic vascular response distal to the anatomic surgical ablation was assessed with quantitative assessment of lymphatic vessel number and size by immunohistochemical staining for LYVE-1( <xref ref-type="fig" rid="pmed-0030254-g003">Figure 3</xref>) [ <xref rid="pmed-0030254-b024" ref-type="bibr">24</xref>, <xref rid="pmed-0030254-b025" ref-type="bibr">25</xref>]. As observed in the hematoxylin and eosin sections, lymphedema was characterized by the presence of numerous dilated microlymphatic structures in the dermis and subdermis. Mean lymphatic vessel number was determined by averaging the number of total lymphatic vessels in all the fields of each slide at 10× magnification. Single brown-stained endothelial cells with a lumen were counted as individual lymphatic vessels. Quantitation was performed for normals ( <italic>n =</italic> 3), surgical shams ( <italic>n =</italic> 3), and lymphedema tails ( <italic>n =</italic> 3). Lymphedema was characterized by an increase in LYVE-1-positive vessel number per field that was not observed in shams: lymphedema, 7.0 ± 4.8; sham, 0.6 ± 0.5; and normal, 1.2 ± 0.8. </p><fig id="pmed-0030254-g003" position="float"><label>Figure 3</label><caption><title>LYVE-1 Immunohistochemical Staining</title><p>Immunohistochemical staining for LYVE-1 is depicted in surgical sham controls (A) and in lymphedema (B) (black arrows). The lymphedema response is characterized by the presence of numerous dilated microlymphatic structures in the dermis and subdermis. Lymphedema produces a statistically significant increase in average cross-sectional vessel area.</p></caption><graphic xlink:href="pmed.0030254.g003"/></fig><p>Vessel area was quantitated according to the formula π· <italic>r</italic> <sub>1</sub>· <italic>r</italic> <sub>2</sub>. The average lymphatic luminal area per field was 503 ± 158 μm <sup>2</sup> in normals, 436 ± 345 μm <sup>2</sup> in shams, and 51,344 ± 18,688 μm <sup>2</sup> in lymphedema. Normals and shams did not differ statistically, but the lymphedema group displayed a statistically significant increase in average vessel area when compared either to normals or to sham surgical animals ( <italic>p</italic> = 0.009 for each comparison). Thus, in summary, the experimental lymphedema is accompanied by an increase in vessel number and, even more notably, by an increase in lymphatic vascular cross-sectional area. </p></sec><sec id="s3e"><title>Lymphoscintigraphy of Experimental Lymphedema</title><p>Whole body lymphoscintigraphy was performed in normal ( <italic>n =</italic> 4) and lymphedema ( <italic>n =</italic> 4) mice. All non-operated mice showed lymphatic drainage from the tip of the tail through two lumbar lymph nodes, asymmetric para-aortic nodes, and mediastinal nodes with final visualization of the liver. The lymphatic flow speed in basal conditions was estimated to be 0.9 ± 0.66 mm/s. In the lymphedema animals, significant dermal backflow was present, but no flow was observed beyond the base of the tail. These lymphoscintigraphic findings closely simulate the qualitative changes observed in the analogous imaging of acquired human lymphedema. </p></sec><sec id="s3f"><title>Functional In Vivo Imaging of Immune Traffic</title><p>The lymphatic vasculature participates in the immune response through the continuous transportation of white blood cells and antigen-presenting cells. The constellation of histological observations in this model, otherwise unexplained by impaired interstitial fluid mobilization, suggests that derangements in lymphatic immune traffic might contribute—actively, passively, or both ways—to the biology of lymph stagnation. Accordingly, we chose to corroborate histopathology with observed, quantifiable changes in immune traffic.</p><p>Bioluminescence imaging was performed on days 3, 5, and 7.5 following the introduction of <italic>luc <sup>+</sup></italic> cells into the distal tail (corresponding to post-operative days 10, 12, and 17.5, respectively). In general, when compared to normals, the clearance of bioluminescent immunocytes was delayed in lymphedema, but remained unimpaired in the surgical sham controls. <xref ref-type="fig" rid="pmed-0030254-g004">Figure 4</xref> depicts a series of imaging experiments for a representative pair of lymphedema and normal control mice. Relative photon density, expressed as the percent of the observed value on day 1, was significantly greater in lymphedema than in the normals, both at day 3 and at day 7 post-injection ( <xref ref-type="fig" rid="pmed-0030254-g004">Figure 4</xref>). </p><fig id="pmed-0030254-g004" position="float"><label>Figure 4</label><caption><title>Dynamic Imaging of Immune Traffic in Experimental Lymphedema</title><p>(A) In vivo bioluminescence imaging of immune traffic. Bioluminescence imaging was performed at defined time points following the introduction of <italic>luc <sup>+</sup></italic> cells. This figure contains a representative series of imaging experiments for paired normal control (A) and lymphedema (B) mice. Photon densities range from red (high) to blue (low). In general, clearance of bioluminescent immunocytes from the lymphedematous tails was delayed, but remained unimpaired in the surgical sham controls. The left panel shows a perceptible increase in photon densities in lymphedema on day 3 post-injection (post-operative day 10). Within several days, the disparity in cellular clearance is even more evident (middle panel); as late as day 17 post-injection, there is still visible bioluminescence in the lymphedematous tail, while all activity has cleared from the normal tail (right panel). The original surgical site is depicted by the white arrows. The black marks on the tail denote 8-mm vertical distances; splenocyte injection was performed 24 mm below the surgical site. </p><p>(B) Quantitative assessment of in vivo bioluminescence imaging of immune traffic. Relative photon density, expressed as a percent of the observed value on day 1, was significantly greater in lymphedema than in normal controls, both at day 3 and at day 7 post-injection (, <italic>p <</italic> 0.05; §, <italic>p <</italic> 0.02). </p></caption><graphic xlink:href="pmed.0030254.g004"/></fig></sec><sec id="s3g"><title>Large-Scale Analysis of Cutaneous Gene Expression in Response to Lymphatic Vascular Insufficiency (Lymph Stasis)</title><p>cDNA microarrays containing a large portion of the mouse transcriptome were used to study the repertoire of genes expressed in the murine skin structures. Triplicate microarray experiments were performed using pooled RNA from the tail skin of female SKH-1 hairless mice representing three biological states: normal, lymphedematous, and surgical sham. Our analyses demonstrated significantly different patterns of gene expression in normal skin and the skin derived from lymphedematous mice. SAM, at a false detection rate (FDR) of 5%, identified 429 upregulated genes in the lymphedema state versus 183 downregulated genes ( <xref ref-type="fig" rid="pmed-0030254-g005">Figure 5</xref>). There were no statistically significant differences between normal mice and surgical control animals (SAM, FDR < 25%). A complete list of differentially regulated genes is provided in <xref ref-type="table" rid="pmed-0030254-t003">Tables 3</xref> and <xref ref-type="table" rid="pmed-0030254-t004">4</xref>. </p><fig id="pmed-0030254-g005" position="float"><label>Figure 5</label><caption><title>SAM Analysis of Microarray Data</title><p>At an FDR of 5%, SAM analysis identified 429 upregulated genes in the lymphedema state versus 183 downregulated genes. There were no statistically significant differences between normal mice and surgical control animals (SAM, FDR < 25%). Enrichment analysis with the Fisher's exact test (EASE software) demonstrated several statistically significant ontologies.</p></caption><graphic xlink:href="pmed.0030254.g005"/></fig><table-wrap id="pmed-0030254-t003" position="float"><label>Table 3</label><caption><p>Upregulated Genes in Lymphedema versus Normal Control (SAM, FDR < 0.05)</p></caption><graphic xlink:href="pmed.0030254.t003"/></table-wrap><table-wrap id="pmed-0030254-t004" position="float"><label>Table 4</label><caption><p>Downregulated Genes in Lymphedema versus Normal Control (SAM, FDR < 0.05)</p></caption><graphic xlink:href="pmed.0030254.t004"/></table-wrap><p>To identify important biological themes represented by genes differentially expressed in the atherosclerotic lesions, we functionally annotated the genes using GO terms. Enrichment analysis with the Fisher's exact test (EASE software) demonstrated several statistically significant ontologies ( <xref ref-type="fig" rid="pmed-0030254-g005">Figure 5</xref>; <xref ref-type="table" rid="pmed-0030254-t005">Tables 5</xref> and <xref ref-type="table" rid="pmed-0030254-t006">6</xref>), including several pathways associated with inflammation. The inflammatory processes, such as defense response, immune response, response to stress, response to pest/pathogen/parasite, and complement activation, represent both humoral immune response and innate immunity. </p><table-wrap id="pmed-0030254-t005" position="float"><label>Table 5</label><caption><p>Pathway Analysis for Upregulated Genes in Lymphedema versus Normal Control (SAM, FDR < 0.05)</p></caption><graphic xlink:href="pmed.0030254.t005"/></table-wrap><table-wrap id="pmed-0030254-t006" position="float"><label>Table 6</label><caption><p>Pathway Analysis for Downregulated Genes in Lymphedema versus Normal Control (SAM, FDR < 0.05)</p></caption><graphic xlink:href="pmed.0030254.t006"/></table-wrap><p>Further scrutiny of the list of genes whose expression is significantly altered in lymphedematous skin suggests that the disease process can be characterized by alterations within a relatively small set of functional attributes, as summarized in <xref ref-type="table" rid="pmed-0030254-t007">Table 7</xref>. These processes include acute inflammatory response, wound healing and fibrosis, angiogenesis, cytoskeletal organization, <italic>Wnt</italic> pathway activation, and adipogenesis. </p><table-wrap id="pmed-0030254-t007" position="float"><label>Table 7</label><caption><p>Functional Gene Expression Analysis in Experimental Lymphedema</p></caption><graphic xlink:href="pmed.0030254.t007"/></table-wrap></sec><sec id="s3h"><title>qRT-PCR Confirms the Accuracy of Microarray Hybridization Results</title><p>Differential expression of eight representative genes from various pathways was confirmed by qRT-PCR. The genes were selected to represent the spectrum of magnitude and direction of change of lymphedematous gene expression relative to normal. The genes assayed included <italic>calgranulin A, calgranulin B, matrix metalloproteinase 3 (MMP3), matrix metalloproteinase 14 (MMP14), myeloid differentiation primary response gene 88 (MYD88), hydroxysteroid (17-beta) dehydrogenase 2 (HADH2), cadherin 11,</italic> and <italic>clusterin</italic>. Overall, the results of the two methods correlated well ( <xref ref-type="fig" rid="pmed-0030254-g006">Figure 6</xref>). </p><fig id="pmed-0030254-g006" position="float"><label>Figure 6</label><caption><title>qRT-PCR Confirmation of the Results of Microarray Hybridization</title><p>The graph represents fold-changes of expression in lymphedema, relative to normal controls, for each of eight representative genes, by microarray hybridization and qRT-PCR. For <italic>MYD88</italic> by microarray and <italic>HADH2</italic> by qRT-PCR, the log (gene expression) equaled zero. <italic>HADH2, hydroxysteroid (17-beta) dehydrogenase; 2MMP, matrix metalloproteinase; MYD88, myeloid differentiation primary response gene 88.</italic> </p></caption><graphic xlink:href="pmed.0030254.g006"/></fig></sec></sec><sec id="s4"><title>Discussion</title><p>In this study, we have characterized a mouse model of lymphedema using in vivo functional imaging and histopathological correlation. This model of acute, acquired lymph stagnation closely simulates the volume response, histopathology, and lymphoscintigraphic characteristics of human acquired lymphedema. LYVE-1 immunohistochemistry demonstrates that this acute impairment of lymph transport is accompanied by an increase in the number and size of microlymphatic structures in the lymphedematous cutaneous tissues.</p><p>We have also undertaken molecular characterization of the disease process through comprehensive transcriptional profiling of the murine lymphedematous tail skin. We have identified a set of genes and molecular pathways that play a role in the unique biology of this cutaneous response to lymph stasis (lymphedema). Recognition of this molecular response pattern is likely to enhance our comprehension of the pathogenesis and biology of lymphedema.</p><p>The model has been elaborated to simulate the regional, acquired lymph stagnation that can arise after trauma, surgery, and cancer therapeutics [ <xref rid="pmed-0030254-b008" ref-type="bibr">8</xref>]. Despite apparent rapid healing of the external cutaneous wound, the model features a stable, persistent edematous increase in the volume of the tail, accompanied by a profound inflammatory response; neither edema nor inflammation is seen in surgical controls. </p><p>The cutaneous inflammatory response observed in this model replicates clinical descriptions of human acquired lymphedema, where there is frequently evidence of concomitant chronic inflammation, and regional immune responses are distorted [ <xref rid="pmed-0030254-b002" ref-type="bibr">2</xref>]. Architectural changes in the skin and subcutaneous tissues are often profound [ <xref rid="pmed-0030254-b003" ref-type="bibr">3</xref>]. Chronic lymph stasis typically stimulates an increase in the number of fibroblasts, adipocytes, and keratinocytes in the skin. Mononuclear cells (chiefly macrophages) often demarcate the chronic inflammatory response [ <xref rid="pmed-0030254-b003" ref-type="bibr">3</xref>]. In affected tissues, there is an increase in collagen deposition, accompanied by adipose and connective tissue overgrowth in the edematous regions [ <xref rid="pmed-0030254-b026" ref-type="bibr">26</xref>]. </p><p>In the current study, the molecular expression profile of lymphedema, observed in parallel with the histopathology and the dynamic immune traffic imaging, suggests that the deranged immune traffic plays at least a passive, if not an active, role in the pathogenesis of the disorder. In normal immune traffic, mononuclear phagocytes and lymphocytes from the tissues enter the afferent lymph vessels and the lymph nodes to elicit primary immune responses before reentering the vasculature [ <xref rid="pmed-0030254-b005" ref-type="bibr">5</xref>]. It is conceivable that, in chronic lymphedema, the impairment of lymphocyte and Langerhans cell trafficking from skin to regional lymph nodes leads to inefficient clearance of foreign antigens, and provides the substrate for chronic inflammatory changes [ <xref rid="pmed-0030254-b004" ref-type="bibr">4</xref>]. The complex biology of lymphedema is still quite poorly understood. Although it has been conjectured that inflammation may indeed trigger various forms of lymphangiogenesis [ <xref rid="pmed-0030254-b027" ref-type="bibr">27</xref>, <xref rid="pmed-0030254-b028" ref-type="bibr">28</xref>], the physiological sensors, signaling mechanisms, and cause-and-effect relationships that initiate post-natal lymphangiogenesis remain to be elucidated. </p><p>Transcriptional profiling has been utilized to identify genes activated in disease states and to refine targets for molecular therapy. Microarray technology has been applied to the elucidation of endothelial biology in health and disease [ <xref rid="pmed-0030254-b009" ref-type="bibr">9</xref>, <xref rid="pmed-0030254-b029" ref-type="bibr">29</xref>], as well as to the investigation of gene expression patterns in cutaneous diseases [ <xref rid="pmed-0030254-b030" ref-type="bibr">30</xref>] and in a wide array of non-neoplastic diseases that entail inflammatory or immune responses [ <xref rid="pmed-0030254-b031" ref-type="bibr">31</xref>]. This approach is particularly attractive for the problem of acquired lymphedema, where the heterogeneous cellular composition of the tissues exposed to lymph stagnation presupposes a very complex, interdependent pattern of gene expression. While the characteristic expression profiles of isolated lymphatic endothelia have previously been studied [ <xref rid="pmed-0030254-b006" ref-type="bibr">6</xref>, <xref rid="pmed-0030254-b007" ref-type="bibr">7</xref>], the current investigation represents the first in-depth molecular examination of the end-organ response to lymph stagnation. Despite the heterogeneous nature of the cellular material under investigation, the approach of high-throughput transcriptional profiling and statistical gene ontology analysis has disclosed discernable patterns of gene expression that appear to be representative of the disorder under scrutiny. </p><p>Transcriptional profiling can provide not only a gene-by-gene view of physiological alterations in a diseased state, but also a statistically rigorous identification of the biological processes that are induced or repressed in disease. This provides a much broader and more comprehensive view of the disease process as a whole than does a simple gene list, making generation of hypotheses about mechanisms more informed. Based on GO functional annotations for each gene on the array [ <xref rid="pmed-0030254-b032" ref-type="bibr">32</xref>], we used Fisher's exact test statistical analysis to identify functional processes that are significantly induced and repressed in this disease model ( <xref ref-type="table" rid="pmed-0030254-t007">Table 7</xref>). The results of this analysis were quite interesting, illustrating that whole panels of genes involved in the immune response, stress response, and complement activation are induced in lymphedema when compared to controls. Among the most interesting of the upregulated genes involved in these processes are many encoding proteins that reflect the inflammatory process. Calgranulin B, highly upregulated in this experimental model, belongs to a family of small calcium-binding proteins that are highly expressed in neutrophil and monocyte cytosol. These molecules are found at high levels in the extracellular milieu during inflammatory conditions [ <xref rid="pmed-0030254-b033" ref-type="bibr">33</xref>]. Calgranulins are potent stimulators of neutrophils and likely are involved in neutrophil migration to inflammatory sites. The levels of several of these proteins are markedly elevated in psoriasis, among other conditions [ <xref rid="pmed-0030254-b034" ref-type="bibr">34</xref>]. Tenascin C is strongly induced by various pro- and anti-inflammatory cytokines, and its de novo expression is a reliable molecular marker for acute inflammation [ <xref rid="pmed-0030254-b035" ref-type="bibr">35</xref>]. Peptidylprolyl isomerase B, also known as cyclophilin B, induces chemotaxis and integrin-mediated adhesion of T cells to the extracellular matrix in vitro [ <xref rid="pmed-0030254-b036" ref-type="bibr">36</xref>]. Basigin is also involved in inflammatory processes and is proposed to be a receptor of cyclophilin A. Stromal cell-derived factor 1, also known as CXCL12, is a highly efficacious lymphocyte chemoattractant [ <xref rid="pmed-0030254-b037" ref-type="bibr">37</xref>]. Platelet factor 4, also known as CXCL4, is a strong chemoattractant for neutrophils and fibroblasts. In addition to its putative role in inflammation, it has been implicated in the pathogenesis of atopic dermatitis. Upregulation of arachidonate 5-lipoxygenase activating protein suggests a role for leukotrienes in this acute inflammatory response; glutathione peroxidase may play an ancillary role. CD63 antigen can be interpreted as a marker of basophil activation and of degranulated neutrophils and monocytes. Legumain, an asparaginyl endopeptidase central to Class II major histocompatiblity complex presentation of microbial antigens, is a potential molecular marker of macrophage differentiation and function [ <xref rid="pmed-0030254-b038" ref-type="bibr">38</xref>]. Follistatin is an activin antagonist implicated in wound repair; activin is an important participant in inflammation, repair, and cytoprotection in various organs, but its induction is restricted to certain types of inflammation and its release is dependent upon the inflammatory setting [ <xref rid="pmed-0030254-b039" ref-type="bibr">39</xref>]. Nuclear factor kappa B is a transcription factor critical to the expression of a variety of chronic inflammatory disease states. The downregulation of gelsolin in this model is notable, inasmuch as hemostatic, inflammatory, and fibroblast responses are blunted in mice lacking gelsolin. Expression of nascent polypeptide-associated complex regulates formation of Fas-associated death domain (FADD) protein oligomers and modulates FADD-mediated signaling; FADD protein is a critical mediator of signal transduction pathways activated by several members of the tumor necrosis factor (TNF) receptor gene superfamily. Cathepsins are distinct intracellular acidic proteases that actively participate in the mechanism of antigen processing; conversely, the stefins are inhibitors of these cathepsins. </p><p>The immune response process is also statistically significantly induced in the lymphedema group versus controls. Proteins such as cytotoxic T lymphocyte-associated proteins are associated with activated T cell function and enhance TGF-β release by T cells [ <xref rid="pmed-0030254-b040" ref-type="bibr">40</xref>]. Leukocyte (or lymphocyte) specific protein 1 (LSP1) is a multifunctional protein involved in the regulation of neutrophil motility, chemotaxis, adhesion, and B lymphocyte apoptosis mediated by membrane immunoglobulin M (mIgM) [ <xref rid="pmed-0030254-b041" ref-type="bibr">41</xref>]. Beta-2 microglobulin is a major histocompatibility complex protein that presents peptide antigens on cell surfaces for recognition by T cell receptors. Lipocalin has recently been shown to participate in the response to bacterial growth [ <xref rid="pmed-0030254-b042" ref-type="bibr">42</xref>]. Galactose binding lectin is a participant in the acute phase response [ <xref rid="pmed-0030254-b043" ref-type="bibr">43</xref>]. Granulin is a high-molecular-weight secreted mitogen that is abundantly expressed in rapidly cycling epithelial cells and in the immune system [ <xref rid="pmed-0030254-b044" ref-type="bibr">44</xref>]. High affinity Fc receptor for IgE is a key molecule in triggering the allergic reaction; it might be considered to be a mast-cell-specific protein. Interferon gamma has an important role in activating macrophages in host defenses. </p><p>One of the most interesting findings of our study is that lymphedema mirrors many of the mechanisms of the inflammatory state, in the absence of a potent inflammatory stimulus. One would not necessarily hypothesize, a priori, that this would be the case, since the only difference between lymphedematous animals and the controls (both normals and surgical shams) was the presence of acute, acquired lymphatic stasis, with no additional inflammatory stimulus; thus, lymphatic stasis induces many of the same mechanisms as inflammatory stimuli. This important observation is novel and has led us to pursue follow-up investigation intended to determine the efficacy of inhibitors of the inflammatory pathways in reversing the pathologic responses of the tissues that are seen in lymphedema.</p><p>The apparent absence of lymphatic markers in this transcriptional profile suggests that, in principle, the findings might not be specific for lymphedema. However, in parallel work performed in the same animal model, we have identified several lymphatic markers including, most importantly, upregulation of VEGF-C (but not VEGF-A) and VEGFR-3. Therapeutic induction of a lymphangiogenic response through administration of exogenous, recombinant human VEGF-C produces both an amelioration of edema and a downregulation of both of these markers of the lymphatic vascular response to acquired vascular insufficiency (unpublished data). Finally, as seen in <xref ref-type="fig" rid="pmed-0030254-g003">Figure 3</xref>, LYVE-1 staining demonstrates specific lymphatic vascular changes in this model, including an increase in lymphatic microvascular size and density, in response to induced lymphatic vascular disruption. None of the morphological, histochemical, or gene expression changes referred to here are observed in animals that are subjected to surgery in the absence of induction of specific lymphatic injury (surgical controls), as described in this paper. Thus, it reasonable to state that the histological and molecular responses reflect lymph stagnation and not the more nonspecific responses of wound healing after surgical injury. </p><p>The cellular response to stress is another process that undergoes statistically significant induction during lymph stagnation. Among the stresses that can trigger this response are the elaboration of pathophysiological signals such as cytokines and eicosanoids [ <xref rid="pmed-0030254-b045" ref-type="bibr">45</xref>]. The expression of a variety of heat shock proteins is upregulated in our model [ <xref rid="pmed-0030254-b045" ref-type="bibr">45</xref>]. Additional evidence for the oxidative stress in lymphatic dysfunction is provided by the upregulation of heme oxygenase 1 (HO-1) [ <xref rid="pmed-0030254-b046" ref-type="bibr">46</xref>]; it is a downstream effector of the potent anti-inflammatory interleukin IL-10 [ <xref rid="pmed-0030254-b047" ref-type="bibr">47</xref>]. </p><p>Upregulation of gene expression related to wound repair, and importantly, to fibrosis is also prominently seen. During wound repair granulin promotes granulation and neovascularization, and regulates inflammation [ <xref rid="pmed-0030254-b048" ref-type="bibr">48</xref>]. The expression of fibulins is induced in the setting of injury, in response to various stimuli [ <xref rid="pmed-0030254-b049" ref-type="bibr">49</xref>]. Biglycan (BGN) has been implicated in the regulation of matrix assembly, cellular adhesion, migration, and TGF-β activity [ <xref rid="pmed-0030254-b050" ref-type="bibr">50</xref>]. Endoglin (CD 105) is a type III TGF-β1 receptor. It modulates the function of TGF-β1 by binding to and modulating signal transduction by the major type I and II TGF-β1 receptors. Lysyl oxidase plays a critical role in the biogenesis of connective tissue matrices. Alpha 2 actin has been identified as a marker of myofibroblast differentiation; all fibrocontractive diseases characterized by fibrosis entail the presence of myofibroblasts [ <xref rid="pmed-0030254-b051" ref-type="bibr">51</xref>]. </p><p>In addition to inflammatory/immune and stress responses, we have observed a gene expression profile that reflects alterations in the angiogenic response <bold>.</bold> Specifically, hypoxia inducible factor 1α has a key role in the cellular response to hypoxia, including the regulation of genes involved in energy metabolism, angiogenesis, and apoptosis. Alterations in the complement and <italic>Wnt</italic> pathways may also contribute significantly to the pathogenesis of the skin response to lymph stasis. </p><p>The observed differences between the lymphedematous animals and the surgical controls are noteworthy. In the absence of any observed delay in wound healing, overt infection, or inflammation, the gene expression profile in lymphedema, but not in surgical controls, is characterized by a remarkable induction of whole biological processes via coordinate upregulation of their component genes. These observations underscore the interpretation that lymphedema is a pathological process that is much more complex than a simple disorder of fluid homeostasis. Indeed, these gene expression profiles superficially resemble those of other recently elucidated inflammatory conditions, such as multiple sclerosis [ <xref rid="pmed-0030254-b031" ref-type="bibr">31</xref>], psoriasis [ <xref rid="pmed-0030254-b052" ref-type="bibr">52</xref>], and even atherosclerosis [ <xref rid="pmed-0030254-b053" ref-type="bibr">53</xref>, <xref rid="pmed-0030254-b054" ref-type="bibr">54</xref>]. </p><p>The differentially expressed genes in our study likely arise from a number of different cell types, including not only the cellular components of the inflammatory response, but also other involved cell types such as keratinocytes, vascular endothelial and smooth muscle cells, and fibroblasts. We feel that it is important to study the gene expression of cells within the complex cellular milieu of the damaged end organ in lymphedema, because precisely the genes we are interested in are those whose transcription results from cell–cell or cell–interstitial fluid interactions. When one removes cells from the milieu, there are immediate changes in transcription that are not reflective of the disease state. We are, of course, interested in defining the cellular compartments that are expressing specific genes, and this topic is the focus of ongoing research in our laboratory that will be published in the future.</p><p>In summary, we have used an animal model of lymphedema that shares many clinical and histopathological features with human lymphedema to identify the biological processes and genes that underlie these features. The fact that inflammatory and immune processes are significantly induced suggests that these observations will provide a useful avenue for the investigation of novel pharmacologic strategies for lymphatic dysfunction [ <xref rid="pmed-0030254-b002" ref-type="bibr">2</xref>]. This approach is particularly attractive in light of the observed parallels with other systemic inflammatory disease states for which effective therapies already exist. Ultimately, such therapies must successfully diminish the impact of the soft tissue fibrosis and adipose deposition that characterize the late disease [ <xref rid="pmed-0030254-b002" ref-type="bibr">2</xref>]; in this regard, it is interesting to contemplate that expression of several such genes is detectably altered in this model, long before architectural evidence of the tissue abnormality is present. This identification of such genes provides an avenue for future investigation and, specifically, provides early insights into the elaboration of molecular therapeutics for this disease. Our purpose here has been to establish an animal model that can be used in two ways: first, to generate new hypotheses about unsuspected genes and pathways that are involved in lymphedema and, second, to eventually test potential pharmacologic and therapeutic interventions for their efficacy in vivo. We recognize that not all interventions that show promise in animal models translate effectively into treatments for human disease, but such in vivo testing is an essential prerequisite to human application and testing, and the establishment of a good, relatively inexpensive animal model is a very important step in allowing more high-throughput analysis of putative therapeutic interventions. These studies contribute to our understanding of the pathogenesis of lymphedema. Further characterization of the genes and molecular pathways identified through this effort will likely provide insights into potential novel strategies for molecular therapies. </p></sec><sec sec-type="supplementary-material" id="s5"><title>Supporting Information</title><supplementary-material content-type="local-data" id="st001"><label>Table S1</label><caption><title>Annotated List of the cDNAs Included on the Mouse Transcriptome Microarray Utilized in These Investigations</title><p>(75,633 KB MPG)</p></caption><media xlink:href="pmed.0030254.st001.mpg"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><sec id="s5a"><title>Accession Numbers</title><p>The microarray data from these investigations have been deposited as dataset GSE4333 in the Gene Expression Omnibus database ( <ext-link ext-link-type="uri" xlink:href="http://www.ncbi.nlm.nih.gov/geo">http://www.ncbi.nlm.nih.gov/geo</ext-link>). Accession numbers for the genes and proteins referenced in tables refer to the LocusLink database ( <ext-link ext-link-type="uri" xlink:href="http://www.ncbi.nlm.nih.gov/projects/LocusLink">http://www.ncbi.nlm.nih.gov/projects/LocusLink</ext-link>). </p></sec></sec>
Evaluating Health Research Capacity Building: An Evidence-Based Tool	Could not extract abstract	<contrib contrib-type="author"><name><surname>Bates</surname><given-names>Imelda</given-names></name><xref ref-type="corresp" rid="cor1">*</xref></contrib><contrib contrib-type="author"><name><surname>Akoto</surname><given-names>Alex Yaw Osei</given-names></name></contrib><contrib contrib-type="author"><name><surname>Ansong</surname><given-names>Daniel</given-names></name></contrib><contrib contrib-type="author"><name><surname>Karikari</surname><given-names>Patrick</given-names></name></contrib><contrib contrib-type="author"><name><surname>Bedu-Addo</surname><given-names>George</given-names></name></contrib><contrib contrib-type="author"><name><surname>Critchley</surname><given-names>Julia</given-names></name></contrib><contrib contrib-type="author"><name><surname>Agbenyega</surname><given-names>Tsiri</given-names></name></contrib><contrib contrib-type="author"><name><surname>Nsiah-Asare</surname><given-names>Anthony</given-names></name></contrib>	PLoS Medicine	<p>An increasingly important goal of governments and external agencies in developing countries is the need for “capacity building” in health research. Although a poorly defined and understood concept, capacity building would essentially enable de novo health research programmes to be facilitated and existing programmes to be strengthened (see [<xref rid="pmed-0030299-b001" ref-type="bibr">1</xref>] and page 14 in [<xref rid="pmed-0030299-b002" ref-type="bibr">2</xref>]). For health research, the goal of building capacity is thus to improve the ability to conduct research, to use results effectively, and to promote demand for research (see page 14 in [<xref rid="pmed-0030299-b002" ref-type="bibr">2</xref>]). Prioritising the need for the international community to make a “quantum leap in capacity building”, as suggested in 1998 by the Director General of the World Health Organization (WHO), would improve health and reduce poverty in developing countries [<xref rid="pmed-0030299-b003" ref-type="bibr">3</xref>].</p><p>To achieve this goal, there is an urgent need for an evidence-based tool for determining whether the required infrastructure is present in a given setting, as well as for underpinning the design and evaluation of capacity-building programmes in health research. Here, we describe the development and use of such a tool through analysis of published models and effective capacity-building principles, together with structured reflection and action (see page 9 in [<xref rid="pmed-0030299-b002" ref-type="bibr">2</xref>]) by stakeholders at the Komfo Anokye Teaching Hospital (KATH) in Kumasi, Ghana.</p><sec id="s2"><title>Challenges Faced in Building and Supporting Research Capacity at KATH</title><p>KATH benefits from a new management team that is committed to developing the hospital and medical school into a regional centre of excellence for research, teaching, and clinical care. Although local clinicians had previously been involved in multinational research projects, these projects had largely been generated by external agencies. Local staff lacked experience in the conception and design of projects, and the hospital lacked local role models and tutors for generating de novo research. Consultant posts at KATH remained vacant because senior clinical trainees had difficulty in completing the prerequisite research component of their exit examinations for the West African Colleges. Tellingly, when asked why they had not completed their specialist exams, the most common reason given by health professionals in KATH was apprehension of beginning their own research programmes.</p><p>KATH management needed a tool that they could use to ensure that all necessary resources were in place to support local research. Unfortunately, the literature specifically describing the building of health research capacity is scarce and tends to emphasise microlevel activities, such as the choice of research trainees (e.g., Nchinda TC [<xref rid="pmed-0030299-b003" ref-type="bibr">3</xref>]), without considering how these activities can be integrated into the wider research system. Moreover, much of the available information on building research capacity is based on retrospective reports of external consultants, and the perspective of implementing capacity building in a developing country is almost never represented [<xref rid="pmed-0030299-b004" ref-type="bibr">4</xref>]. Our aim therefore was to develop an evidence-based tool that could be used to guide the design, implementation, and evaluation of capacity building in health research programmes.</p></sec><sec id="s3"><title>Developing an Appropriate Evaluation Tool</title><p>We used a three-stage approach: (1) searching the literature for existing tools and models; (2) analysing best-practice examples to guide the overall framework; and (3) adapting the framework into an operational tool that met the specific needs of KATH. By using translational research principles to analyse our findings, we systematically extracted and extrapolated stakeholders' evidential and experiential stories (see page 9 [<xref rid="pmed-0030299-b002" ref-type="bibr">2</xref>], and used this information to inform the overall design of our programme. <xref ref-type="fig" rid="pmed-0030299-g001">Figure 1</xref> outlines the stages of development and testing of the tool.</p><fig id="pmed-0030299-g001" position="float"><label>Figure 1</label><caption><title>Outline of Programme to Develop and Test Tool for Evaluating Research Capacity at KATH</title></caption><graphic xlink:href="pmed.0030299.g001"/></fig><sec id="s3a"><title>Literature search</title><p>We searched the following electronic databases: MEDLINE, Ingenta, and Science Direct using keywords, such as “capacity building”, “capacity development”, “developing countries,” and “Africa”. We retrieved the full text of any relevant papers, including articles cited in the reference lists of these papers. Because there is limited information about health research capacity building in peer-reviewed literature, we also consulted books, Web sites of organisations working on health and research capacity building (e.g., Web sites of WHO, United Nations agencies, the European Community, and the International Development Research Centre), and references provided by colleagues. This evidence was used to derive a definition of health research capacity building, to identify existing capacity-building models, and to synthesise best-practice examples to derive key principles. <xref ref-type="supplementary-material" rid="pmed-0030299-sd001">Dataset S1</xref> gives a detailed overview of our literature review on capacity building.</p><p>We found that many different definitions have been applied to capacity building according to the particular level—“micro”, “meso”, or “macro” (focused on in [<xref rid="pmed-0030299-b005" ref-type="bibr">5</xref>])—but that one of the most widely used definitions is “an ability of individuals, organisations or systems to perform appropriate functions effectively, efficiently and sustainably” [<xref rid="pmed-0030299-b006" ref-type="bibr">6</xref>].</p><p>By combining the definition for generic “capacity building” with published evidence and our practical experiences of developing a planning and evaluation tool, we have defined building capacity for health research as “an ability of individuals, organisations, or systems to perform and utilise health research effectively, efficiently, and sustainably.”</p></sec><sec id="s3b"><title>Using published best-practice examples to design the evaluation programme</title><p>No tools exist that are specific for evaluating health research capacity-building programmes. However, the literature review was helpful for identifying ineffective capacity-building strategies, such as “bolting-on” capacity building to research projects initiated by a specific donor in developing countries [<xref rid="pmed-0030299-b007" ref-type="bibr">7</xref>]. It was also useful for highlighting generic principles underlying successful capacity building.</p><p>We grouped the generic principles that consistently emerged from the literature as best practices into themes that emphasised the importance of three concepts. The first theme was a “phased approach”; this requires the sequential involvement of all stakeholders in assessing capacity gaps, developing strategies to fill these gaps, and evaluating outcomes [<xref rid="pmed-0030299-b006" ref-type="bibr">6</xref>]. The second theme was “strengthening of existing processes”; this is an iterative and flexible process that focuses on enhancing local ability to solve problems, define and achieve development needs, and then incrementally incorporate expanding circles of individuals, institutions, and systems [<xref rid="pmed-0030299-b008" ref-type="bibr">8</xref>]. The third theme was “partnerships”; for effective or sustained capacity building, the various partners involved must have similar concepts [<xref rid="pmed-0030299-b005" ref-type="bibr">5</xref>] and share responsibilities and obligations, with local partners taking ownership and leadership [<xref rid="pmed-0030299-b006" ref-type="bibr">6</xref>,<xref rid="pmed-0030299-b009" ref-type="bibr">9</xref>]. Thus, the role of external expertise is to facilitate the development of local skills through learning by experience, rather than acting as a “donor” who retains control of funds and expertise over a poorer “beneficiary” partner.</p></sec><sec id="s3c"><title>Developing and adapting the evaluation tool</title><p>An illuminating finding of the literature search was that there was no model that had been specifically designed with health research capacity building in mind. Indeed, the most useful model was one that had been developed for institutionalising quality assurance (QA) [<xref rid="pmed-0030299-b010" ref-type="bibr">10</xref>] because it focused on defining, measuring, and improving quality. This mirrors the processes required for capacity building in health research: defining the institutional systems needed to support research, enumerating existing and missing resources, and improving research support by addressing identified gaps. The QA institutionalisation framework represented a synthesis of over ten years' experience in developing countries, and was derived from a combination of organisational development and QA literature. The framework described organisations as passing through four phases when they implement innovation: awareness, experiential, expansion, and consolidation (<xref ref-type="table" rid="pmed-0030299-t001">Table 1</xref>).</p><table-wrap id="pmed-0030299-t001" position="float"><label>Table 1</label><caption><title>Framework for Designing and Evaluating a Health Research Capacity-Building Programme</title></caption><graphic xlink:href="pmed.0030299.t001"/></table-wrap><p>In the course of adapting our framework into a tool that was relevant to KATH, we were also influenced by a published framework for dissemination and implementation of evidence-based medicine (EBM) [<xref rid="pmed-0030299-b011" ref-type="bibr">11</xref>]. This prompted us to change the name of the “experiential” phase to “implementation”, as this was more appropriate to a research programme.</p><p>To meet the specific needs of KATH, local research stakeholders participated in adapting the tool. These stakeholders comprised ten KATH health professionals (nine clinicians and one physiotherapist), and senior hospital managers—including the chief executive, medical director, and heads of departments. Individual and group discussions took place during a workshop for the health professionals. Stakeholders considered each phase in the framework (<xref ref-type="table" rid="pmed-0030299-t002">Table 2</xref>), and suggested characteristics, activities, and indicators of progress in building research capacity that met the needs of KATH and that could be feasibly measured or shown. The stakeholders' suggestions were incorporated into the framework to create an operational tool that could be used to identify gaps in the research infrastructural support at KATH (<xref ref-type="table" rid="pmed-0030299-t002">Table 2</xref>). This ensured that a holistic approach was taken to developing the research capacity in the hospital, rather than a fragmented, unfocused approach.</p><table-wrap id="pmed-0030299-t002" position="float"><label>Table 2</label><caption><title>Adaptation of the Framework into a Tool That Was Relevant to the Needs of KATH</title></caption><graphic xlink:href="pmed.0030299.t002"/></table-wrap></sec></sec><sec id="s4"><title>Using the Evaluation Tool at KATH</title><sec id="s4a"><title>Identifying strengths and weaknesses in the research capacity</title><p>In the year following development of the tool, the ten health professionals undertook a research project as part of a work-based course to prepare them for the research component of their professional exams. By comparing their actual research experiences at KATH with the components itemised in the evaluation tool, they were able to identify aspects that were well supported by the institution and aspects where support was lacking or could only be provided by external facilitators. The comparison was achieved through group discussions and analysis of individual reflective statements about their research experiences, using a standard qualitative research approach known as “grounded theory”. Individual statements were scrutinised, and themes relating to research infrastructural support were extracted. Cycles of scrutinising, extracting data, and allocating it to themes were repeated until no new themes emerged [<xref rid="pmed-0030299-b012" ref-type="bibr">12</xref>].</p><p>A comparison between the themes that emerged from this process with the capacity-building evaluation tool identified strengths and weaknesses in the research infrastructural support. Strengths included the peer-support mechanisms within KATH, which occurred predominantly in three different contexts (peer group committees to review research proposals, small group work within course workshops, and cross-departmental research meetings). Peer support to promote work-based learning is an evidence-based educational approach [<xref rid="pmed-0030299-b013" ref-type="bibr">13</xref>], so the peer-support mechanisms in KATH corresponded to components of the evaluation tool. Weaknesses that emerged included gaps in knowledge concerning research resources available on the Internet, particularly systematic searching of the published literature.</p></sec><sec id="s4b"><title>Prioritising and implementing actions for addressing gaps in the research capacity</title><p>A nominal group technique [<xref rid="pmed-0030299-b014" ref-type="bibr">14</xref>] was used to achieve consensus among researchers about aspects of research support that were lacking in KATH and to agree on which of these should be prioritised. For this technique, researchers used their experiences of doing research and the evaluation tool to write their own observations on areas of research infrastructure that were lacking at KATH. These were categorised into themes by the whole group and ranked according to their importance for supporting research. Gaps that were identified as priorities included provision of local statistical expertise, lack of researcher skills in critical literature reviews, and inadequate Internet access. These gaps were presented by the researchers to senior managers in KATH as a list of recommendations, and the managers incorporated activities to address these recommendations in their annual plans and budgets in 2004/2005 and 2005/2006. Progress was reviewed with the managers and the researchers during the six-month course workshops (<xref ref-type="table" rid="pmed-0030299-t003">Table 3</xref>).</p><table-wrap id="pmed-0030299-t003" position="float"><label>Table 3</label><caption><title>Use of the Evaluation Tool to Identify and Address Gaps in Research Capacity within 18 Months</title></caption><graphic xlink:href="pmed.0030299.t003"/></table-wrap></sec></sec><sec id="s5"><title>What Was Achieved by Using the Tool?</title><p>Progress in strengthening the research infrastructure in KATH has been achieved both for individuals and for the institution. For individuals, a course to teach research skills has been established in partnership with the Liverpool School of Tropical Medicine (LSTM). Local facilitators have been trained to run the course and funding has been secured so that within three years the course will be wholly the responsibility of KATH staff, with LSTM providing external quality reviews for the course. At an institutional level, an Internet suite has been refurbished and equipped for use by researchers, research support meetings are now a regular monthly event, and KATH has trained its own clinical biostatistician to support its researchers. Within 18 months of the original recommendations, KATH management and researchers have achieved many of the indicators of progress listed in the evaluation tool, and have developed plans to achieve the remaining indicators within the next two years. Naturally, progress some indicators, particularly those relating to using research results to improve the quality of clinical care and encouraging whole departments to be more proactive about research, will be slow and could take several years to achieve.</p></sec><sec id="s6"><title>Discussion</title><sec id="s6a"><title>What have we learned?</title><p>The evaluation tool has enabled researchers and hospital managers to work together to achieve a common goal of improving the research capacity in KATH. They have monitored their progress against predetermined standards and have identified and filled gaps in research infrastructure.</p><p>The evaluation tool should be flexible enough to incorporate changes in the local environment and the needs of KATH, and consequently we plan to re-evaluate and amend it within five years. Because changing the research culture of an institution is a complex process, some important components that should have been included in the tool might have been overlooked. For example, dialogue between scientists and nonscientists, as well as non-health-sector workers, is important for developing and sustaining health research capacity [<xref rid="pmed-0030299-b003" ref-type="bibr">3</xref>]. Such interactions are not represented in our tool, which has focused instead on building institutional capacity.</p><p>The success of the process by which this tool was developed and tested confirms the importance of the generic principles underlying effective capacity building that we extracted from the literature. We used a phased approach to engage stakeholders in identifying strengths and weaknesses, and then to develop, implement, and monitor action plans to address these gaps [<xref rid="pmed-0030299-b006" ref-type="bibr">6</xref>]. Part of this process involved identifying and strengthening existing processes and building up local resources, rather than developing new parallel systems [<xref rid="pmed-0030299-b015" ref-type="bibr">15</xref>]. This strengthening process included formalising the peer-support meetings that researchers had found so helpful, and expanding the existing Internet facilities. The process is a good example of a genuine partnership for problem solving that is built on trust, common interest, long-term commitment, and shared responsibilities and obligations [<xref rid="pmed-0030299-b016" ref-type="bibr">16</xref>,<xref rid="pmed-0030299-b017" ref-type="bibr">17</xref>]. Although funding for the process was initially shared between KATH and LSTM, KATH has maintained ownership and leadership, and is now totally funding the capacity-building process. Each partner had clearly delineated roles, and mechanisms and timescales for transfer of skills from LSTM to KATH staff were agreed on early in the process.</p><p>Two important criteria for this project's success were the motivation of the researchers and the strong leadership and commitment of KATH managers. Participation of all stakeholders in the design of evaluation indicators is recognised to promote motivation and commitment (see Chapter 7 [<xref rid="pmed-0030299-b018" ref-type="bibr">18</xref>]). The rate of progress is likely to slow down over the next few years, as the institutional shift towards research begins to involve individuals who might not have the high motivation of the managers and researchers, but the tool nevertheless provides a means for maintaining focus on achieving some of the more difficult indicators.</p></sec><sec id="s6b"><title>How transferable are these lessons and the tool?</title><p>The generic principles of effective capacity building—phased approach, strengthening existing systems and partnerships for problem solving—were derived from contexts that were not health sector–specific, and yet they have been applied successfully here. However, the evaluation tool was developed for the context of health research at KATH, and its value and transferability in other contexts would need to be assessed.</p><p>Although the framework from which the tool was derived incorporated all the elements of a research process, such as problem identification, priority setting, and research use (see page 16 in [<xref rid="pmed-0030299-b002" ref-type="bibr">2</xref>]), the specific components used to produce the operational tool would need to be adapted to suit the specific needs of other institutions. Monitoring and evaluation is the most difficult and neglected component of capacity-building programmes because they can take over 20 years to achieve their objectives [<xref rid="pmed-0030299-b008" ref-type="bibr">8</xref>], and some outcomes, such as organisational culture, are difficult to measure [<xref rid="pmed-0030299-b019" ref-type="bibr">19</xref>]. Different users of evaluations will have different priorities, and the use of an evaluation tool helps to promote agreement on the purpose of the evaluation and the indicators [<xref rid="pmed-0030299-b020" ref-type="bibr">20</xref>]. The major advantages of our tool are that it enables an institution in a developing country to set its own priorities, to have control over local capacity building [<xref rid="pmed-0030299-b021" ref-type="bibr">21</xref>], and to evaluate progress in building capacity from its own perspective rather than from that of an external agency.</p></sec></sec><sec sec-type="supplementary-material" id="s7"><title>Supporting Information</title><supplementary-material content-type="local-data" id="pmed-0030299-sd001"><label>Dataset S1</label><caption><title>Further Details of the Literature Review on Capacity Building</title><p>(52 KB DOC).</p></caption><media xlink:href="pmed.0030299.sd001.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material></sec>
“Key to the Future”: British American Tobacco and Cigarette Smuggling in China	<sec id="st1"><title>Background</title><p>Cigarette smuggling is a major public health issue, stimulating increased tobacco consumption and undermining tobacco control measures. China is the ultimate prize among tobacco's emerging markets, and is also believed to have the world's largest cigarette smuggling problem. Previous work has demonstrated the complicity of British American Tobacco (BAT) in this illicit trade within Asia and the former Soviet Union.</p></sec><sec id="st2"><title>Methods and Findings</title><p>This paper analyses internal documents of BAT available on site from the Guildford Depository and online from the BAT Document Archive. Documents dating from the early 1900s to 2003 were searched and indexed on a specially designed project database to enable the construction of an historical narrative. Document analysis incorporated several validation techniques within a hermeneutic process. This paper describes the huge scale of this illicit trade in China, amounting to billions of (United States) dollars in sales, and the key supply routes by which it has been conducted. It examines BAT's efforts to optimise earnings by restructuring operations, and controlling the supply chain and pricing of smuggled cigarettes.</p></sec><sec id="st3"><title>Conclusions</title><p>Our research shows that smuggling has been strategically critical to BAT's ongoing efforts to penetrate the Chinese market, and to its overall goal to become the leading company within an increasingly global industry. These findings support the need for concerted efforts to strengthen global collaboration to combat cigarette smuggling.</p></sec>	<contrib contrib-type="author"><name><surname>Lee</surname><given-names>Kelley</given-names></name><xref ref-type="aff" rid="aff1"> <sup>1</sup> </xref><xref ref-type="corresp" rid="cor1">*</xref></contrib><contrib contrib-type="author"><name><surname>Collin</surname><given-names>Jeff</given-names></name><xref ref-type="aff" rid="aff2"> <sup>2</sup> </xref></contrib>	PLoS Medicine	<sec id="s1"><title>Introduction</title><p>Cigarette smuggling has emerged as a critical public health issue. As well as being the source of huge losses in government revenues, smuggling makes cigarettes more affordable, thus stimulating consumption and undermining control measures, most notably among youth and low-income consumers [ <xref rid="pmed-0030228-b001" ref-type="bibr">1</xref>, <xref rid="pmed-0030228-b002" ref-type="bibr">2</xref>]. The illicit trade in tobacco is served by three broad categories of supply, namely, counterfeiting, bootlegging, and large-scale smuggling. This paper focuses on large-scale smuggling of cigarettes and other tobacco products that circumvent tax regimes, which accounts for the vast majority of cigarettes smuggled globally [ <xref rid="pmed-0030228-b001" ref-type="bibr">1</xref>]. Recent estimates suggest that the trade in smuggled cigarettes accounts for around 6%–8% of global consumption [ <xref rid="pmed-0030228-b003" ref-type="bibr">3</xref>] and around one-quarter of total exports [ <xref rid="pmed-0030228-b004" ref-type="bibr">4</xref>]. The centrality of provisions to combat smuggling within the World Health Organization's Framework Convention on Tobacco Control (FCTC) [ <xref rid="pmed-0030228-b005" ref-type="bibr">5</xref>] indicate recognition of this worldwide problem. </p><p>Tobacco industry documents, released following litigation in the United States, have demonstrated the complicity of British American Tobacco (BAT) in large-scale smuggling [ <xref rid="pmed-0030228-b006" ref-type="bibr">6</xref>]. Whereas previous work has illustrated the strategic importance of smuggling to BAT's expansion in Asia [ <xref rid="pmed-0030228-b007" ref-type="bibr">7</xref>] and the former Soviet Union [ <xref rid="pmed-0030228-b008" ref-type="bibr">8</xref>], this paper offers the first detailed analysis of corporate involvement in contraband to one country. The significance of this work is enhanced by China's status as the world's largest cigarette market, comprising 350 million smokers and one-third of cigarettes smoked [ <xref rid="pmed-0030228-b009" ref-type="bibr">9</xref>]. So far dominated by domestic producers, China is seen as the “ultimate prize” [ <xref rid="pmed-0030228-b010" ref-type="bibr">10</xref>] among the world's emerging tobacco markets because of a potentially huge demand for foreign brands. </p><p>China has long been recognised as also having “possibly the largest cigarette smuggling problem in the world” [ <xref rid="pmed-0030228-b011" ref-type="bibr">11</xref>]. Legal imports of foreign cigarettes comprise about 3% of the 1.7 trillion (1.7 × 10 <sup>12</sup>) stick market. However, up to 50 billion (5 × 10 <sup>10</sup>) foreign cigarettes were believed to be entering the country illegally each year by the late 1990s [ <xref rid="pmed-0030228-b011" ref-type="bibr">11</xref>– <xref rid="pmed-0030228-b013" ref-type="bibr">13</xref>]. In recent years, Chinese government efforts to crack down on smuggling activities, and recover some of the US$1.8 billion in lost tax revenue annually [ <xref rid="pmed-0030228-b012" ref-type="bibr">12</xref>], have led to the strengthening of the General Administration of Customs, including administrative and criminal penalties for smuggling [ <xref rid="pmed-0030228-b014" ref-type="bibr">14</xref>]. While there have been criminal prosecutions of some individuals and widely reported media accounts of this illicit trade [ <xref rid="pmed-0030228-b006" ref-type="bibr">6</xref>, <xref rid="pmed-0030228-b015" ref-type="bibr">15</xref>], scholarly research on the role of tobacco companies remains limited. Using evidence from internal documents of the tobacco industry, this paper begins by describing the huge scale of this illicit trade and the key supply routes by which it has been conducted. It examines BAT's efforts to optimise earnings by restructuring operations and controlling the supply chain (how cigarettes are manufactured and legally exported, are handled by various “agents” or middlemen, and then sold illicitly as contraband) and the pricing of smuggled cigarettes. Indeed, documents produced in litigation demonstrate that smuggling has been strategically critical to BAT's ongoing efforts to penetrate the Chinese market, the “key to the future” [ <xref rid="pmed-0030228-b016" ref-type="bibr">16</xref>] in BAT's overall goal to become the leading company within an increasingly global industry. </p></sec><sec id="s2"><title>Methods</title><p>This paper is based on analysis of documents dating between the early 1900s and 2003 from the BAT Depository in Guildford, United Kingdom. The broad strengths and limitations of tobacco industry documents as a data source [ <xref rid="pmed-0030228-b017" ref-type="bibr">17</xref>, <xref rid="pmed-0030228-b018" ref-type="bibr">18</xref>], their application to the specific issue of smuggling [ <xref rid="pmed-0030228-b007" ref-type="bibr">7</xref>], and the distinctive value of, and problems posed by, the Guildford Depository [ <xref rid="pmed-0030228-b019" ref-type="bibr">19</xref>, <xref rid="pmed-0030228-b020" ref-type="bibr">20</xref>] have been previously described. Research for this paper was based on an iterative search strategy, commencing with searching the depository's crude file index in 2003 using the keywords China, Chinese, PRC (People's Republic of China), Sino*, Peking/Beijing, Shanghai, Canton/Guangdong, CNTC (China National Tobacco Corporation), and Hong Kong. A total of 210 files were identified and reviewed. This search was then supplemented by similar searches between October 2004 and September 2005 of the BAT Documents Archive (BATDA) Web site containing Guildford and additional BAT-related documents dating to 2003, exploiting the increasing online availability of BAT documents [ <xref rid="pmed-0030228-b020" ref-type="bibr">20</xref>]. Documents used in this analysis date from 1983–1999. </p><p>BATDA also enabled advanced Boolean searches via combined keywords, including well-recognised euphemisms for smuggling defined within BAT documents ([ <xref rid="pmed-0030228-b021" ref-type="bibr">21</xref>, <xref rid="pmed-0030228-b022" ref-type="bibr">22</xref>]; available as <xref ref-type="supplementary-material" rid="sd001">Text S1</xref> and <xref ref-type="supplementary-material" rid="sd002">S2</xref>, respectively), such as “transit”, “general trade” ( “GT”), “duty not paid” ( “DNP”), and “free market” [ <xref rid="pmed-0030228-b007" ref-type="bibr">7</xref>]. A total of 2,594 online documents were reviewed. From both on site and online sources, 882 relevant documents on China and 2,502 on Hong Kong were indexed on a specially designed project database to enable the construction of an historical narrative. Document analysis incorporated several validation techniques within a hermeneutic process [ <xref rid="pmed-0030228-b023" ref-type="bibr">23</xref>], including corroboration of interpretation between authors and attempted triangulation of findings via trade publications, newspaper articles, academic journals, and other secondary sources. </p></sec><sec id="s3"><title>Results</title><sec id="s3a"><title>Circumventing Barriers to Market Entry</title><p>The Chinese tobacco market remained closed to foreign companies for more than three decades, but in the late 1970s the “open door” policy reintroduced potential access to this market. Confronted with tight import quotas and high tariffs, however, transnational tobacco companies (TTCs) made limited progress in the world's largest market via legal trade [ <xref rid="pmed-0030228-b024" ref-type="bibr">24</xref>]. Under Chinese law, the State Tobacco Monopoly Administration is the sole agency allowed to trade in tobacco. A special agency under the State Tobacco Monopoly Administration runs all imports and exports of tobacco leaf and cigarettes. No individuals, state, or private enterprises are permitted to engage in this trade. State permits or licenses are needed for all imports, which are subject to tariffs. By the late 1980s, quantities of smuggled cigarettes confiscated rose markedly, although they represented only a fraction of the total contraband, estimated at 5%–10% of China's annual production. Between 1988 and 1992, 26,100 million sticks were seized by customs officials [ <xref rid="pmed-0030228-b025" ref-type="bibr">25</xref>]. In 1998, Premier Zhu Rongji initiated a national anti-smuggling campaign, resulting in regular seizures of millions of sticks [ <xref rid="pmed-0030228-b011" ref-type="bibr">11</xref>, <xref rid="pmed-0030228-b026" ref-type="bibr">26</xref>, <xref rid="pmed-0030228-b027" ref-type="bibr">27</xref>]. </p><p>Within this context, industry documents suggest large-scale smuggling of foreign cigarettes into China to circumvent barriers to market access. Official trade figures and other documents cite levels of legal imports, subject to State Tobacco Monopoly Administration quotas and tariffs (180% in 1985) and sold by licensed retailers, duty-free shops, and so-called friendship stores largely frequented by foreign visitors. Comparing official data for selected years, which represent imports by all TTCs, to BAT's sales figures to China reveals dramatic discrepancies ( <xref ref-type="table" rid="pmed-0030228-t001">Table 1</xref>). For example, during 1990, official imports from all TTCs totalled 10.5 billion sticks, whereas BAT documents describe company exports to China of over 20.3 billion sticks. The company recognised this trade as “one of the larger profit centres in BAT Industries” [ <xref rid="pmed-0030228-b028" ref-type="bibr">28</xref>], with 25% of BAT's profits coming “from transit trade to China” [ <xref rid="pmed-0030228-b029" ref-type="bibr">29</xref>]. </p><table-wrap id="pmed-0030228-t001" position="float"><label>Table 1</label><caption><p>Official Chinese Imports and Cited BAT Exports to China for Selected Years between 1982 and 2000</p></caption><graphic xlink:href="pmed.0030228.t001"/></table-wrap><p>The difficulty in accounting for such discrepancies is evident in a numbered list of questions and answers prepared by BAT's corporate communications department to prepare staff for answering potential questions about the proposed restructuring of BAT operations in China (discussed below). While direct questions about smuggling were answered with claims that high taxation caused such activity—a commonly recited argument unsupported by independent empirical evidence [ <xref rid="pmed-0030228-b001" ref-type="bibr">1</xref>, <xref rid="pmed-0030228-b002" ref-type="bibr">2</xref>, <xref rid="pmed-0030228-b030" ref-type="bibr">30</xref>]—no response was offered regarding the divergence between Chinese and BAT import figures: </p><disp-quote><p>11. Are you aware that cigarettes are smuggled into some Far East markets, particularly China</p></disp-quote><disp-quote><p>A. Yes! We are aware cigarettes are smuggled in some parts of the world. This is usually a result of tax differentials, providing an incentive for some to trade illegally in our products. We are opposed to such activities because they have an adverse impact on our legitimate sales…</p></disp-quote><disp-quote><p>15. How do you explain the differences in figures quoted by your company (M Broughton presentation) regarding exports to China and officially published figures by the government of PRC?</p></disp-quote><disp-quote><p>A. [ <xref rid="pmed-0030228-b031" ref-type="bibr">31</xref>] </p></disp-quote></sec><sec id="s3b"><title>Restructuring to “exploit opportunities for the sale of BAT Group products to private traders selling into China” [ <xref rid="pmed-0030228-b032" ref-type="bibr">32</xref>] </title><p>Control of the contraband supply chain was essential to optimise BAT sales within different markets, and to avoid unwanted attention from customs and excise officials [ <xref rid="pmed-0030228-b033" ref-type="bibr">33</xref>]. BAT initiated a restructuring of its China and Hong Kong operations in 1990 “to improve the effectiveness of the Group's penetration of the China market” [ <xref rid="pmed-0030228-b034" ref-type="bibr">34</xref>]. Problems over supply routes (see <xref ref-type="supplementary-material" rid="sd001">Protocol S1</xref> for detailed discussion of supply routes for smuggled cigarettes) were seen to require stronger central management of the export trade: </p><disp-quote><p>The reasons for transferring the business from Hong Kong to UK include the increasing need to control centrally the distribution and sourcing of the Group's international brands [ <xref rid="pmed-0030228-b035" ref-type="bibr">35</xref>]. </p></disp-quote><p>The impending handover of Hong Kong to Chinese rule in 1997 was a particular concern. BAT Hong Kong (BATHK) was responsible for the Hong Kong and Macau domestic markets, as well as what is described as “combined exports” to China. Sales to China in 1991 accounted for “the largest part of the [BATHK] business…representing approximately 92% of the total” [ <xref rid="pmed-0030228-b036" ref-type="bibr">36</xref>]. The existing structure of BAT operations in the region was seen as unsatisfactory by David Leach (Project Co-Ordinator, Far East Sales Unit [FESU], BATHK): </p><disp-quote><p>There is inadequate segregation of the management of the various markets currently served by BAT HK, marketing activities are inadequately co-ordinated and the structure is insufficiently flexible to respond to changes which may occur in the lead up to 1997 [ <xref rid="pmed-0030228-b036" ref-type="bibr">36</xref>]. </p></disp-quote><p>Reorganisation was seen as necessary to facilitate further expansion of the contraband trade across Asia. As Roy Salter (Group Planning Coordinator, BAT Industries) wrote, “Opportunities will emerge for sales to traders dealing in other markets, e.g. Thailand, possibly in conflict with established BATUKE [BAT United Kingdom and Export] and B&W [Brown and Williamson] agents, such as SUTL [Singapura United Tobacco Limited]” [ <xref rid="pmed-0030228-b037" ref-type="bibr">37</xref>]. </p><p>In 1990, a report was commissioned from consultancy firm Griffiths Management Limited to develop management and corporate structures that would provide BAT “with a competitive advantage over other international manufacturers” [ <xref rid="pmed-0030228-b038" ref-type="bibr">38</xref>] in China. The subsequent report envisaged a 3-fold division of responsibility. First, BATHK would be restricted “to manufacturing and selling in the [Hong Kong] domestic market and to manufacturing under contract for other markets as required” [ <xref rid="pmed-0030228-b039" ref-type="bibr">39</xref>]. Second, BAT China would be established, with responsibility for “direct [legal] imports to China, for managing the relationship with CNTC and for establishing a strong presence for BAT brands within the China market” [ <xref rid="pmed-0030228-b039" ref-type="bibr">39</xref>]. </p><p>The proposed restructuring would be completed with the creation of “an offshore company…domiciled in the UK” to handle “selling to trading organisations supplying China and other adjacent markets”. The offshore company was provisionally named BAT Distribution, and later referred to as International Tobacco (Overseas) Limited, and the report was explicit in the description of its proposed remit:</p><disp-quote><p>The transit trade for China, Taiwan and elsewhere in Asia should be undertaken by BAT Distribution Limited, incorporated in a tax haven of choice, probably UK. It could be owned by BAT Industries or whatever other company BAT Group prefers.</p></disp-quote><disp-quote><p>BAT Distribution should be little more than a brassplate company with very low overhead and the flexibility to establish branch offices wherever the transit traders move [ <xref rid="pmed-0030228-b038" ref-type="bibr">38</xref>]. </p></disp-quote><p>The primary objective of this complex model was to provide an operational divide between legal and illegal trade while, critically, ensuring clear oversight and control:</p><disp-quote><p>BAT Distribution is set up to enable BATHK to effect a formal separation of its transit from its official trade by transferring the transit trade to this company. BAT Distribution activities remain fully controlled by the BAT China Group (and not any of its subsidiaries) but structurally and formally, BAT Distribution is owned elsewhere in the BAT Group in a tax haven [ <xref rid="pmed-0030228-b040" ref-type="bibr">40</xref>]. </p></disp-quote><p>The strategic need for this separation is reaffirmed by BAT executives in discussing procedures for handling orders, invoicing, payment, and supply [ <xref rid="pmed-0030228-b041" ref-type="bibr">41</xref>]. The need for the contraband trade to be carried out “on an arms-length basis” [ <xref rid="pmed-0030228-b042" ref-type="bibr">42</xref>], which “removes BAT HK as an intermediary” [ <xref rid="pmed-0030228-b042" ref-type="bibr">42</xref>], was particularly noted. The company would “seek to identify trading organisations which might have the potential to function in a way similar to that of SUTL [Singapura United Tobacco Limited]” [ <xref rid="pmed-0030228-b039" ref-type="bibr">39</xref>]. The Griffiths report distinguished this approach from what it termed “the Philip Morris model”, seen as reducing both profitability and control over the transit trade, while noting that the proposed approach would not reduce BAT's ability to “pretend ignorance” of complicity in transit: </p><disp-quote><p>It has also been suggested that the Philip Morris model be adopted which places an intermediary between Philip Morris Asia and the transit traders. We have two objections. First the transit trade still originates with Philip Morris Asia and is controlled by that company. That is similar to the present situation with BATHK except for the intermediary. The second objection is that the intermediary adds another level of profit absorption and is harder to control than a dedicated group company. The apparent distancing has little effect since the business is transit trade. All the customers BATHK deals with are legitimate, properly incorporated businesses. Provided this remains the case the BAT Group can pretend ignorance that its cigarettes are being distributed through the transit trade as much and as justifiably as Philip Morris can [ <xref rid="pmed-0030228-b038" ref-type="bibr">38</xref>]. </p></disp-quote><p> <xref ref-type="fig" rid="pmed-0030228-g001">Figure 1</xref> illustrates the planned restructuring of BAT's China operations, with the contraband trade to be carried out via “exporters” under the responsibility of an “independent distributor”. </p><fig id="pmed-0030228-g001" position="float"><label>Figure 1</label><caption><title>Planned Restructuring of BAT Operations to Separate Supply of “Independent Distributors” to China</title><p>Source: British American Tobacco Documents Archive. Available: <ext-link ext-link-type="uri" xlink:href="http://bat.library.ucsf.edu/pageview?a=img%7B%7Bamp%7D%7Dtid=fm41a99%7B%7Bamp%7D%7Dtotal=2">http://bat.library.ucsf.edu/pageview?a=img&tid=fm41a99&total=2</ext-link>. Accessed 19 June 2006. </p></caption><graphic xlink:href="pmed.0030228.g001"/></fig><p>This proposed restructuring was submitted to BAT Chairman Sir Patrick Sheehy in March 1991 by former Chairman Barry Bramley, with the reported support of future Chairman Martin Broughton. The paper recommended that BAT Distribution be established with the following responsibilities:</p><disp-quote><p>(a) To manage and to exploit opportunities for the sale of BAT Group products to private traders selling into the China, and adjacent general trade markets, in a regulated manner, (effectively replacing the current role of BAT(HK)).</p></disp-quote><disp-quote><p>(b) Identification of new trading organisations with the potential to supply China and adjacent markets.</p></disp-quote><disp-quote><p>(c) To ensure the efficient distribution within the China markets of duty free BAT products [ <xref rid="pmed-0030228-b032" ref-type="bibr">32</xref>]. </p></disp-quote><p>Sheehy and fellow directors approved the reorganisation in April 1991, with BAT Distribution more discretely referred to as “responsible for sales to trading organisations supplying China and other adjacent territories” [ <xref rid="pmed-0030228-b034" ref-type="bibr">34</xref>]. </p><p>The planned reorganisation was then reviewed by in-house staff and Price Waterhouse in terms of the complex tax implications for BAT. Senior management sought to identify the optimal location of the three proposed companies to minimise tax liability, noting the “risk that the [UK] Inland Revenue may examine the operation carefully as a result of the overall reduced taxable profits” [ <xref rid="pmed-0030228-b037" ref-type="bibr">37</xref>]. </p><p>Given such considerations, a revised restructuring proposal, written by project co-ordinator David Leach and supported by Bramley and BAT Far East Director (now Chief Executive) Paul Adams, was submitted in November 1991 [ <xref rid="pmed-0030228-b043" ref-type="bibr">43</xref>]. This proposal advanced a simplified model under which “the role of servicing the China market and Hong Kong exports will be taken over by two new organisations, BAT China and a Far East Sales Unit (FESU) attached to BATUKE”. BAT China was to be wholly owned by BAT and have “responsibility for sales of cigarettes to China National Tobacco Corporation, Sino Russian Border Trade (SRBT) and China Duty Free Shops as well as relations between the BAT Group and the Government of PRC”. The FESU, headed by Adams, was to assume the role previously envisaged for BAT Distribution: </p><disp-quote><p>FESU, a Division of BATUKE, set up in the UK to co-ordinate the export business currently serviced by BAT HK, and to administer product invoicing and supply from the manufacturers to BAT China, BAT HK and independent distributors. This unit, which will be managed as a profit centre, will be responsible for co-ordinating sales forecasts, arranging contract manufacture and ensuring shipments are made in accordance with schedules [ <xref rid="pmed-0030228-b043" ref-type="bibr">43</xref>]. </p></disp-quote><p>This arrangement would enable the group to take advantage of the UK's lower corporate tax rates [ <xref rid="pmed-0030228-b043" ref-type="bibr">43</xref>], and facilitate central control over China operations. In 1992, total volumes shipped by the FESU to China were 22,889 million sticks, projected to increase to 31,634 million sticks (turnover of US$676.6 million) by 1994 [ <xref rid="pmed-0030228-b044" ref-type="bibr">44</xref>]. As part of an overall restructuring of BAT operations the FESU was renamed Asia Pacific North in 1993, and then subsequently absorbed under the so-called Project Battalion into the newly formed Asia Pacific Regional Business Unit [ <xref rid="pmed-0030228-b045" ref-type="bibr">45</xref>]. </p><p>In seeking to further “disassociate BAT China from the export business” and establish a “structure appropriate for the return of Hong Kong to Chinese control in 1997” [ <xref rid="pmed-0030228-b046" ref-type="bibr">46</xref>], the contraband trade was to be shifted from Hong Kong to Singapore, from where new supply routes into China were already being explored [ <xref rid="pmed-0030228-b047" ref-type="bibr">47</xref>]. This entailed BAT's establishment of a “consignment warehouse” in Singapore to enable more flexible and timely delivery to transit agents: </p><disp-quote><p>This is because we hope to be able to move the distribution of product currently being sold through Hong Kong exports to the Singapore warehouse which will act as a buffer warehouse in the supply chain. We can minimise costs if we can use ship-to-ship or container-to-container transhipment through the Free Trade Zone in Singapore [ <xref rid="pmed-0030228-b048" ref-type="bibr">48</xref>]. </p></disp-quote><p>In July 1993, the warehouse facility opened as “the export supply point for approximately 40% of the combined export sales volume into China”, producing tax savings of £1.5 million (US$998,283) in 1993 ([ <xref rid="pmed-0030228-b049" ref-type="bibr">49</xref>]; available as <xref ref-type="supplementary-material" rid="sd003">Text S3</xref>). </p></sec><sec id="s3c"><title>Controlling Prices to “enhance performance”</title><p>While official imports were subject to import quotas and tariffs, offering “relative poor corporate profitability” [ <xref rid="pmed-0030228-b051" ref-type="bibr">51</xref>], BAT stood to earn huge profits by careful pricing of contraband cigarettes. An important consideration was how price could be used to establish brand status within the market. These efforts were apparent for the premium brand State Express 555, which enjoyed “intrinsic exclusivity and prestige” ([ <xref rid="pmed-0030228-b052" ref-type="bibr">52</xref>]; available as <xref ref-type="supplementary-material" rid="sd005">Text S5</xref>). The company was keen to build market share, in competition with other “international brands”, but wanted to avoid the threat of “oversupply of SEFK [State Express Filter Kings] to erode price/image” [ <xref rid="pmed-0030228-b052" ref-type="bibr">52</xref>, <xref rid="pmed-0030228-b053" ref-type="bibr">53</xref>]. While RJ Reynolds's pricing strategy was closely monitored ([ <xref rid="pmed-0030228-b054" ref-type="bibr">54</xref>]; available as <xref ref-type="supplementary-material" rid="sd006">Text S6</xref>), BAT was especially aware of Philip Morris (PM): </p><disp-quote><p>As CNTC outlets act as a showcase to Chinese smokers, BAT has ensured that [State Express] 555′s price is positioned at the premium end and above the free market [contraband] price. This deliberate strategy is designed to enhance performance in the free market where the volume is significant. PM adopt a different approach, whereby Marlboro positioned in price both well below 555s [sic] price and below the free market price, has achieved brand leadership in the channel. Despite this, with three key brands BAT will still sell half as much again as PM, with a corporate share estimated at 27% in 1992 (compared to 18.8% for PM) ([ <xref rid="pmed-0030228-b055" ref-type="bibr">55</xref>]; available as <xref ref-type="supplementary-material" rid="sd007">Text S7</xref>). </p></disp-quote><p>There is evidence that BAT tried to “enhance performance” by controlling the prices of illegal imports ([ <xref rid="pmed-0030228-b050" ref-type="bibr">50</xref>]; Available as <xref ref-type="supplementary-material" rid="sd004">Text S4</xref>). BAT thus sought to price its leading brand at a carefully judged premium over those of its competitors, notably PM, intending that State Express 555 should “maintain a 50 cent premium over Marlboro in both the official (CNTC) and Free Market” [ <xref rid="pmed-0030228-b050" ref-type="bibr">50</xref>], the precise differential varying by channel and region. </p><p>The ability to influence pricing of both legal and illegal imports was predicated on control of supply. The challenge was to ensure sufficient supply of BAT brands to compete with domestic and other international brands, without exerting downward pressure on prices. Indeed, the company was aware of the “ability of free market to significantly undercut retail pricing of official duty paid outlets, particularly in the Southern provinces, thereby eroding our potential share performance in official outlets vs free markets” [ <xref rid="pmed-0030228-b053" ref-type="bibr">53</xref>]. For example, BAT became concerned that the price of Brown and Williamson's premium brand, Lucky Strike, needed to be raised on the “free market”, reportedly “42% below Marlboro” [ <xref rid="pmed-0030228-b056" ref-type="bibr">56</xref>]. In seeking to control prices, however, the company faced the potential problem of alternative sources of contraband supplying the market: </p><disp-quote><p>The CNTC stores are one small source for leakage into the free market….LSF [Lucky Strike Filter] is in the free market in Guangzhou where its price is half that of MARLBORO. The Brand Owner believes that any significant price increase will just suck in transit and that critical mass is needed before any major price repositioning can be embarked on…([ <xref rid="pmed-0030228-b057" ref-type="bibr">57</xref>]; available as <xref ref-type="supplementary-material" rid="sd008">Text S8</xref>). </p></disp-quote><p>This careful balance between supply and price is described by Kam Ka Ng (Research Manager for BAT China):</p><disp-quote><p>We could push as many cigarettes as possible, but to maintain a consistent growth in the long term would require a more strategic approach. We feel that we may be milking the cow, and not feeding it enough to maintain its premium positioning…([ <xref rid="pmed-0030228-b058" ref-type="bibr">58</xref>]; available as <xref ref-type="supplementary-material" rid="sd009">Text S9</xref>). </p></disp-quote><p>Another complication was the growing presence of counterfeits. By the mid 1990s, there were five sources of foreign cigarette brands in China—cigarettes manufactured locally under licence, legal imports, illegal imports by appointed agents, illegal imports “leaked” from legal or illegal suppliers, and counterfeits. A “significant increase in counterfeit products in China” [ <xref rid="pmed-0030228-b059" ref-type="bibr">59</xref>] was reported in the late 1990s, and was attributed to the decline in both legal and illegal imports following the Asian financial crisis, which pushed “infringers [counterfeiters] deeper underground” to make “use of unknown transit routes” [ <xref rid="pmed-0030228-b060" ref-type="bibr">60</xref>]. As Richard Duncan (BATHK) wrote: </p><disp-quote><p>Counterfeit has become a major problem in China, causing a loss to BAT in the order of 40 million pounds per annum. Increasingly, there is evidence that China is becoming a source of counterfeit for markets outside China, counterfeiting brands that are not necessarily sold in China [ <xref rid="pmed-0030228-b061" ref-type="bibr">61</xref>]. </p></disp-quote><p>It is estimated that around two million cartons of counterfeit cigarettes, both domestic and international brands, were sold in China in 1999, double the amount in 1997. The Chinese government believes that, in many cases, local officials are involved. Raids of factories and workshops during 1998 yielded 3.15 billion counterfeit cigarettes [ <xref rid="pmed-0030228-b062" ref-type="bibr">62</xref>]. </p></sec><sec id="s3d"><title>“It is important that BAT do not get left behind” ([ <xref rid="pmed-0030228-b063" ref-type="bibr">63</xref>]; Available as <xref ref-type="supplementary-material" rid="sd010">Text S10</xref>): Smuggling in Pursuit of Global Competitiveness </title><p>By the late 1980s, smuggling had become so lucrative that concerns were raised that the pursuit of a joint venture (JV) in China might be counterproductive:</p><disp-quote><p>On China, the current view was that the return from exports (including some transit via the USSR) was much higher than from investment. Therefore, a visit by the Chairman [Patrick Sheehy] to Beijing in October (with the ERT [European Round Table of Industrialists]) might actually be counterproductive if he was under pressure to commit to investment [ <xref rid="pmed-0030228-b064" ref-type="bibr">64</xref>]. </p></disp-quote><p>The company thus approached JV negotiations with relative caution [ <xref rid="pmed-0030228-b024" ref-type="bibr">24</xref>]. The strategic question to be answered, according to Paul Adams, was “why are we looking at JV's rather than a continuation of transit...?” [ <xref rid="pmed-0030228-b065" ref-type="bibr">65</xref>] Estimated sales figures show that “Hong Kong exports” comprised 22% of total sales volume and 27% of total profits for BAT in 1992. These proportions were expected to rise to 29% and 31%, respectively, by 1997 [ <xref rid="pmed-0030228-b016" ref-type="bibr">16</xref>]. </p><p>While Hong Kong exports were seen as the “key to the future [of BAT]”, BAT Head of Corporate Planning Graham Burgess recognised the trade's vulnerability given the “danger of serious action by the authorities” [ <xref rid="pmed-0030228-b016" ref-type="bibr">16</xref>]. In reviewing the BAT company plan for 1993–1997, BAT Group Planning Coordinator Hilary Barton questioned the sustainability of such rapid growth [ <xref rid="pmed-0030228-b066" ref-type="bibr">66</xref>], while Adams cautioned against excessive reliance on FESU sales: </p><disp-quote><p>It is clear that we have relied (and I suspect will continue to rely) on Hong Kong Exports/FESU to meet budget and drive BATCo. growth. I understand and accept the realities of this.</p></disp-quote><disp-quote><p>However, it occurs to me that our ability to “make our numbers” through Hong Kong Exports may have caused us to ease the performance pressures off other markets…[ <xref rid="pmed-0030228-b067" ref-type="bibr">67</xref>]. </p></disp-quote><p>Progress by competitors seeking JVs in China was also a concern. When PM reached agreement in 1993 for Marlboro to be produced locally, a memo from BAT China's Susan Osborne warned of “serious implications for our GT [general trade] business in China in the medium to longer term” [ <xref rid="pmed-0030228-b068" ref-type="bibr">68</xref>], while Adams feared “Philip Morris may encourage a G.T. clamp-down” ([ <xref rid="pmed-0030228-b069" ref-type="bibr">69</xref>]; available as <xref ref-type="supplementary-material" rid="sd011">Text S11</xref>). </p><p>Reliance on an increasingly vulnerable contraband trade in the region shaped the extensive reorganisation of BAT operations in 1993 under Projects Rubicon and Battalion. Given “the increasing ‘globality' of the tobacco business”, the need for structural change was recognised “if BATCo is to fulfil its prime objectives of: (i) growing the value of BAT's brand assets and; (ii) maximising the income stream from each region” [ <xref rid="pmed-0030228-b070" ref-type="bibr">70</xref>]. The reorganisation was “made necessary, largely because of the spectacular success of the export business in recent years and our expectations that this growth will continue” [ <xref rid="pmed-0030228-b071" ref-type="bibr">71</xref>]. Combined management was seen as “necessary if we are to remain dynamic and progressive and most importantly, successful in today's competitive business environment” [ <xref rid="pmed-0030228-b072" ref-type="bibr">72</xref>]. The plan to pursue a twin-track strategy to compete with other suppliers is detailed in a memo entitled “Distribution Initiatives within PRC”: </p><disp-quote><p>As long as free market sales remain dominant, alternative routes of distribution of unofficial imports need to be examined, evaluated and, if appropriate, maximised. It is recognised that distribution of our product in China is key to BATCo's long term success and a key objective within BAT China's JV initiative is to achieve widespread distribution networks….</p></disp-quote><disp-quote><p>It is important that BAT do not get left behind…BAT would benefit through a wider spread of distribution, particularly in secondary and tertiary cities, ahead of competition…It is envisaged that this initiative would be a short to medium term strategy prior to establishment of any JV operation and naturally any action taken would need to run in parallel with our JV plans…[ <xref rid="pmed-0030228-b063" ref-type="bibr">63</xref>]. </p></disp-quote><p>As part of BAT's ultimate goal of becoming “the Premier Tobacco Company in the World” [ <xref rid="pmed-0030228-b071" ref-type="bibr">71</xref>], the company sought to reach other regions within China, with stated plans to “investigate alternative export routes/customers that will improve direct penetration of UK brands in northern and central provinces” [ <xref rid="pmed-0030228-b073" ref-type="bibr">73</xref>]. </p></sec></sec><sec id="s4"><title>Discussion</title><p>In common with other TTCs, BAT has depicted smuggling as an inevitable consequence of “tax differentials, weak border controls, and import restrictions and bans” [ <xref rid="pmed-0030228-b074" ref-type="bibr">74</xref>]. In his speech to the BAT Annual General Meeting in 2003, then Chairman Martin Broughton urged governments to enter into partnership with the industry to combat illicit trade: </p><disp-quote><p>Governments must consider whether they want the industry to slip into the hands of traffickers and criminals, or to be run by legitimate, tax-paying companies working to manage a risky product responsibly. Many governments recognize that ‘Big Tobacco' is ‘Responsible Tobacco' [ <xref rid="pmed-0030228-b075" ref-type="bibr">75</xref>]. </p></disp-quote><p>The documents presented in this paper demonstrate that the separation between “Big Tobacco” and “traffickers and criminals” has been less stark than such posturing would suggest. While public statements by BAT have portrayed smuggling as “inimical to our long-term business interests” [ <xref rid="pmed-0030228-b076" ref-type="bibr">76</xref>], internal documents illustrate how the company's strategy in China, nonpareil among its future priorities, centred on the supply, oversight, and control of the illicit trade. </p><p>The documents demonstrate that contraband has been a hugely profitable and integral part of BAT operations in China over the past two decades. Initially a means of circumventing restricted access to the Chinese market, it became a hugely profitable income stream. Contraband was then used to build market presence, in competition with other international brands, with supply and price carefully managed. By the early 1990s, smuggling into China represented “the largest BATCo market” [ <xref rid="pmed-0030228-b048" ref-type="bibr">48</xref>], integral to the growth of BAT overall and its aspirations to supersede PM to become the world's leading tobacco company. For the Chinese government, this illicit trade undermined restrictions on imports, represented an enormous loss of tax revenue, and stimulated demand for premium brand cigarettes. </p><p>Alongside this lucrative trade, BAT China, deliberately insulated from such operations, sought to build relations with Chinese officials [ <xref rid="pmed-0030228-b043" ref-type="bibr">43</xref>]. Paul Bingham (Marketing Manager, BAT) noted that BAT China's planned medium-term strategy should address “the ability of BAT to advise the Government generally on tax revenue methods—VAT, excise, direct taxes, oil, alcohol, etc” [ <xref rid="pmed-0030228-b077" ref-type="bibr">77</xref>]. In 1994, Adams cited the smuggling issue in a discussion with the Chinese ambassador to Britain to gain support for a JV: </p><disp-quote><p>Demand for our international brands outstrips the supply via the CNTC Friendship Stores and Duty Free shops. This created a black market. Substantial domestic manufacture by a j.v. would provide China with a share in the profits hitherto taken by smugglers….China had much more to gain with us than without us ([ <xref rid="pmed-0030228-b078" ref-type="bibr">78</xref>]; available as <xref ref-type="supplementary-material" rid="sd012">Text S12</xref>). </p></disp-quote><p>Yet such lobbying disguises the extent to which BAT has been a critical part of the global smuggling problem, rather than an appropriate partner in its resolution.</p><p>Neither BAT nor its senior directors have yet been held accountable, whether via litigation or public inquiry. The United Kingdom has perhaps come closest to pursuing accountability via the concern of the House of Commons Health Select Committee that led to a lengthy investigation by the Department of Trade and Industry. The eventual abandonment of this closed investigation in 2004, and the decision, without publication of its findings, not to pursue a criminal investigation [ <xref rid="pmed-0030228-b079" ref-type="bibr">79</xref>], has come amid reports of undue political influence exercised on BAT's behalf at the highest levels of government [ <xref rid="pmed-0030228-b080" ref-type="bibr">80</xref>]. This, alongside the continued accumulation of documented evidence of complicity, raises questions that the Health Select Committee might usefully revisit. </p><p>Given the inherently transnational nature of smuggling, collective efforts across countries offer the most promising avenue for developing an effective policy response. The FCTC is particularly significant in this respect, particularly given China's ratification of the convention on 11 October 2005. The first international public health treaty negotiated by the World Health Organization, the FCTC both offers an institutional basis to support the broad development of tobacco control and incorporates a clear focus on tobacco smuggling as a core issue in health policy. Indeed, smuggling received a unique degree of attention via the International Conference on Illicit Tobacco Trade [ <xref rid="pmed-0030228-b081" ref-type="bibr">81</xref>], reflected in the comparatively detailed coverage afforded the illicit trade via Article 15 [ <xref rid="pmed-0030228-b082" ref-type="bibr">82</xref>]. Notwithstanding such progress, however, there is a clear need for the broad obligations of the FCTC to be augmented with more specific measures via the negotiation of a dedicated protocol on the illicit trade in tobacco [ <xref rid="pmed-0030228-b083" ref-type="bibr">83</xref>]. Such a protocol should incorporate, among other things, anti-money-laundering provisions, a mandatory tracking system of markings and codes, records of shipments, and a strict liability and compensation system [ <xref rid="pmed-0030228-b084" ref-type="bibr">84</xref>, <xref rid="pmed-0030228-b085" ref-type="bibr">85</xref>]. The basis for such a compensation system might be provided by the conditions accepted by Philip Morris International under its 2004 agreement with the European Union [ <xref rid="pmed-0030228-b086" ref-type="bibr">86</xref>], a system of binding obligations that contrasts with the weaker voluntary approach of “memoranda of understanding” advocated by BAT [ <xref rid="pmed-0030228-b087" ref-type="bibr">87</xref>]. </p></sec><sec sec-type="supplementary-material" id="s5"><title>Supporting Information</title><supplementary-material content-type="local-data" id="sd014"><label>Alternative Language Abstract S1</label><caption><title>Chinese Translation of the Abstract</title><p>(54 KB PDF)</p></caption><media xlink:href="pmed.0030228.sd014.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd001"><label>Text S1</label><caption><title>Reference [ <xref rid="pmed-0030228-b021" ref-type="bibr">21</xref>] </title><p>(443 KB PDF)</p></caption><media xlink:href="pmed.0030228.sd001.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd002"><label>Text S2</label><caption><title>Reference [ <xref rid="pmed-0030228-b022" ref-type="bibr">22</xref>] </title><p>(3.3 MB PDF)</p></caption><media xlink:href="pmed.0030228.sd002.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd003"><label>Text S3</label><caption><title>Reference [ <xref rid="pmed-0030228-b049" ref-type="bibr">49</xref>] </title><p>(341 KB PDF)</p></caption><media xlink:href="pmed.0030228.sd003.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd004"><label>Text S4</label><caption><title>Reference [ <xref rid="pmed-0030228-b050" ref-type="bibr">50</xref>] </title><p>(450 KB PDF)</p></caption><media xlink:href="pmed.0030228.sd004.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd005"><label>Text S5</label><caption><title>Reference [ <xref rid="pmed-0030228-b052" ref-type="bibr">52</xref>] </title><p>(1 MB PDF)</p></caption><media xlink:href="pmed.0030228.sd005.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd006"><label>Text S6</label><caption><title>Reference [ <xref rid="pmed-0030228-b054" ref-type="bibr">54</xref>] </title><p>(1.9 MB PDF)</p></caption><media xlink:href="pmed.0030228.sd006.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd007"><label>Text S7</label><caption><title>Reference [ <xref rid="pmed-0030228-b055" ref-type="bibr">55</xref>] </title><p>(2.8 MB PDF)</p></caption><media xlink:href="pmed.0030228.sd007.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd008"><label>Text S8</label><caption><title>Reference [ <xref rid="pmed-0030228-b057" ref-type="bibr">57</xref>] </title><p>(2.3 MB PDF)</p></caption><media xlink:href="pmed.0030228.sd008.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd009"><label>Text S9</label><caption><title>Reference [ <xref rid="pmed-0030228-b058" ref-type="bibr">58</xref>] </title><p>(1.1 MB PDF)</p></caption><media xlink:href="pmed.0030228.sd009.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd010"><label>Text S10</label><caption><title>Reference [ <xref rid="pmed-0030228-b063" ref-type="bibr">63</xref>] </title><p>(1.2 MB PDF)</p></caption><media xlink:href="pmed.0030228.sd010.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd011"><label>Text S11</label><caption><title>Reference [ <xref rid="pmed-0030228-b069" ref-type="bibr">69</xref>] </title><p>(812 KB PDF)</p></caption><media xlink:href="pmed.0030228.sd011.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd012"><label>Text S12</label><caption><title>Reference [ <xref rid="pmed-0030228-b078" ref-type="bibr">78</xref>] </title><p>(756 KB PDF)</p></caption><media xlink:href="pmed.0030228.sd012.pdf"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material><supplementary-material content-type="local-data" id="sd013"><label>Protocol S1</label><caption><title>Three Supply Routes for Smuggling Cigarettes into China</title><p>(739 KB DOC)</p></caption><media xlink:href="pmed.0030228.sd013.doc"><caption><p>Click here for additional data file.</p></caption></media></supplementary-material></sec>
Free radical theory of autoimmunity	<sec><title>Background</title><p>Despite great advances in clinical oncology, the molecular mechanisms underlying the failure of chemotherapeutic intervention in treating lymphoproliferative and related disorders are not well understood.</p></sec><sec><title>Hypothesis</title><p>A hypothetical scheme to explain the damage induced by chemotherapy and associated chronic oxidative stress is proposed on the basis of published literature, experimental data and anecdotal observations. Brief accounts of multidrug resistance, lymphoid malignancy, the cellular and molecular basis of autoimmunity and chronic oxidative stress are assembled to form a basis for the hypothesis and to indicate the likelihood that it is valid <italic>in vivo</italic>.</p></sec><sec><title>Conclusion</title><p>The argument set forward in this article suggests a possible mechanism for the development of autoimmunity. According to this view, the various sorts of damage induced by chemotherapy have a role in the pattern of drug resistance, which is associated with the initiation of autoimmunity.</p></sec>	<contrib id="A1" corresp="yes" contrib-type="author"><name><surname>Kannan</surname><given-names>Subburaj</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	Theoretical Biology & Medical Modelling	<sec><title>Background: review of the literature</title><sec><title>Multi-drug resistance: <italic>a multi-step process</italic></title><p>After exposure to chemotherapeutic drugs, lymphoid cells develop along two distinct pathways. First, a cell population susceptible to the drugs dies by apoptosis or necrosis, depending on the severity of treatment. Secondly, a few cells evolve one or more mechanisms for survival, resisting the damage inflicted by the drugs. It is well known that chemotherapeutic drugs induce tumor cell death via apoptosis through DNA damage, and, in particular, activation of proteolytic enzymes involved in programmed cell death. When one drug fails, various others are tried as parts of a therapeutic regimen. Such drugs kill cancer cells by increasing their sensitivity via alterations in internal mechanisms, a desired outcome for effective chemotherapy. Some tumor cells evolve mechanisms, as yet poorly understood, by which they acquire resistance to structurally and functionally unrelated drugs; this is referred to as multi-drug resistance.</p></sec><sec><title>Multi-drug resistance: <italic>a selective adaptation mechanism</italic></title><p>Distinct factors contributing to the formation of tumorigenic phenotypes ensure that each malignant cell is unique in terms of activation of oncogenes and inactivation of tumor suppressor genes. Drug-exposed tumor cells are subjected to sustained to oxidative stress and become tolerant to it. During this time window, selection pressure imposed by the chemotherapeutic drugs causes the selective overgrowth of cells that can withstand them. It is also possible that normal, but susceptible, cells may acquire drug resistance by cellular overgrowth in their neighborhood [<xref ref-type="bibr" rid="B1">1</xref>].</p></sec><sec><title>Multi-drug resistance: <italic>an intrinsic or acquired phenomenon</italic></title><p>Development of drug resistance could be either intrinsic or acquired during neoplasia formation. Intrinsic resistance is possibly an inherent property of the species, developed during the course of evolution. Acquired drug resistance possibly originates in the host because of one or more of the following factors: <bold>1. </bold>reduced absorption of the specific drug; <bold>2. </bold>delayed/expedited rate of metabolism by the specific organ involved; <bold>3. </bold>loss of drug accumulation mechanism (decreased import); <bold>4. </bold>increased drug elimination (increased export) (e.g. multi-drug resistance in cancer cells); <bold>5. </bold>conversion of active drug to an inactive form (e.g. penicillinase, insecticide resistance) or to a prodrug no longer converted to its active form (e.g. resistance to purine analogues in cancer cells); <bold>6. </bold>elimination of target (e.g. induction of alternative pathway) or alteration of target's affinity for the drug; <bold>7. </bold>overproduction of target (e.g. gene amplification); <bold>8. </bold>accumulation of metabolite antagonistic to drug (e.g. PABA overproduction by Pneumococci) [<xref ref-type="bibr" rid="B2">2</xref>].</p></sec><sec><title>Multi-drug resistance: <italic>evolution by inhibition of apoptosis</italic></title><p>These factors contribute towards reducing the level of the drug in the serum. Other factors contributing to the evolution of drug resistance and inhibition of apoptosis may include: tolerance to the drug effects; failure and/or lack of delivery of a given drug to the tumor site (owing to size or location of the tumor, or low absorption rate of a high molecular weight drug); and non- specific interactions of drugs with healthy cells [<xref ref-type="bibr" rid="B3">3</xref>,<xref ref-type="bibr" rid="B4">4</xref>]. As a result, each malignant cell is unique in terms of activation of oncogenes and inactivation of tumor suppressor genes and hence in the tumorigenic phenotypes to which it can give rise; any given tumor cell population becomes heterogeneous [<xref ref-type="bibr" rid="B5">5</xref>]. Although many studies have demonstrated the critical role of anti-apoptotic components including Bcl-2, Bcl-xL and Mcl-1, and proapoptotic components such as Bax, Bak and Bad, in the evolution of multi-drug resistance, the underlying molecular mechanism is not clear at present. Overexpression of Bcl-2, Bcl-xL or Mcl-1 has been shown to prevent drug-induced apoptosis in several cell lines [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B7">7</xref>].</p></sec><sec><title>Multi-drug resistance: <italic>role of epigenetic mechanism(s)</italic></title><p>It has been suggested that drug resistance is implicitly mediated via epigenetic changes in the form of altered gene expression induced by transacting factors, and is definitely not due to alteration of the tumor cell genome. However, the DNA double strand breaks (dsbs) are considered responsible for drug toxicity and are linked to cell death, mostly via apoptosis [<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B9">9</xref>]. Drug-sensitive cells exposed to alkylating agents manifest a sustained increase in reactive oxygen species (ROS) levels along with DNA dsbs. ROS and dsbs are suggested causes of drug sensitive tumor cell death via apoptosis. Furthermore, after a period of time without exposure to alkylating agents, drug resistant cells in culture become sensitive and die via apoptosis. It is tempting to speculate that the drug resistance observed <italic>in vitro </italic>is a transient or evanescent phenomenon. This may not be the case in the <italic>in vivo </italic>or clinical scenario, which would severely limit the ability to correlate <italic>in vitro </italic>findings with clinical manifestations <bold>(Kannan, unpublished data).</bold></p></sec><sec><title>Multi-drug resistance: <italic>possible role of DNA damage-dependent mechanisms</italic></title><p>In addition, acquired resistance to the alkylating family of drugs has been attributed to such factors as increased expression of glutathione-S-transferase (GST) and changes associated with signaling events upstream of the site of the action of Bcl-2 family members. It is implied that genes conferring protection from apoptosis are up-regulated in the surviving cells and confer drug resistance. The steps involved in the formation of resistance to alkylating agents involve the following sequence: cellular uptake of the drug [<xref ref-type="bibr" rid="B10">10</xref>], conversion to the active form, formation of DNA-mono adducts, cross-link formation, and detoxification of the free intracellular drug or its reactive metabolite, e.g. by conjugation with reduced glutathione by GST [<xref ref-type="bibr" rid="B4">4</xref>].</p></sec><sec><title>Multi-drug resistance: <italic>possible role of DNA damage-independent mechanisms</italic></title><p>However, it has also been proposed that drug resistance can evolve independently of DNA damage and repair [<xref ref-type="bibr" rid="B11">11</xref>]. Oxidative stress-induced DNA damage could be overcome in drug-sensitive tumor cells, but it probably occurs through the loss of appropriate antiapoptotic genes, so the sensitive cells undergo apoptosis. Drug-resistant cells overcome oxidative stress by efficiently repairing the as-yet-unknown extent of damage in genomic DNA, resulting in a drug resistance genotype. In terms of DNA repair, an alkylating agent-resistant phenotype of B-cell chronic lymphocytic leukemia has been attributed to rapid inter-strand, cross-link repair, and also to upregulation of double-strand repair proteins with increased formation of nuclear foci [<xref ref-type="bibr" rid="B12">12</xref>].</p><p>Owing to the sustained accumulation of DNA strand breaks and impairment of the DNA double-strand break repair machinery, such as homologous recombination (HR) and non-homologous endjoining (NHEJ), along with genomic instability to the alkylating agents, the resistant cells undergo apoptosis [<xref ref-type="bibr" rid="B13">13</xref>]. In a comparative study using variants of human ovarian carcinoma cells sensitive or insensitive to alkylating agents, it was found that increased levels of anti-apoptotic proteins prevent the drug-resistant cells dying from apoptosis. In contrast, by producing increased levels of pro-apoptotic proteins, the drug-sensitive cells were rendered apoptotic. The sensitive cells were shown to undergo more free radical formation and genomic DNA damage with the impairment of DNA damage repair mechanisms [[<xref ref-type="bibr" rid="B14">14</xref>-<xref ref-type="bibr" rid="B16">16</xref>]; Kannan, unpublished observations].</p></sec><sec><title>Malignant lymphocyte disorders</title><p>Lymphocyte malignancies encompass the development of lymphocyte neoplasms from cells at a stage before their differentiation to T- and B- cells; i.e. either primordial or differentiated stem cells. In particular, acute lymphocytic leukemias are known to originate from the primitive lymphoid stem cells that normally differentiate into T- or B-cell phenotypes, whereas chronic lymphocytic leukemias arise from a well- differentiated B-cell progenitor. It is also known that multiple myelomas are likely to originate from B cells at a later stage of maturation. Depending on the various regulatory molecules involved in stem cell development, the ensuing lymphocytic disease might be of several kinds: <bold>I</bold>. hairy-cell leukemia, <bold>II</bold>. prolymphocytic leukemia, <bold>III. </bold>natural killer-cell large granular lymphocytc leukemia and <bold>IV. </bold>plasmacytoma.</p><p>Both B-cell and T-cell neoplasms constitute a multiplicity of disorders depending on the developmental stage (see table 96-1 in[<xref ref-type="bibr" rid="B17">17</xref>]). Secretions of monoclonal proteins (a distinctive class of immunoglobulins) are an essential feature of both neoplastic transformation and clonal proliferation in B cells. Owing to the increase in secretory protein levels in the circulation, the viscosity of the blood is increased and erythrocytes aggregate, causing a hyper-viscosity syndrome [<xref ref-type="bibr" rid="B18">18</xref>]. Moreover, immunoglobulins that precipitate below 37°C are known to cause Raynaud syndrome, skin ulcerations, purpura, digital infarction and gangrene. Deposition of immune complexes in the glomerular tufts is a critical factor in the evolution of a spectrum of renal diseases. Formation and accumulation of excess immunoglobulin heavy chains in plasma cell myelomas causes amyloids, resulting in primary amyloidosis.</p><p>The generation of immunoglobulins that recognize self antigens, referred to as auto-reactive antibodies, during B-cell neoplasia is known to cause autoimmune thrombocytopenia and possibly autoimmune neutropenia. Antibodies directed against tissue proteins are factors in the etiopathogenesis of autoimmune thyroiditis, adrenalitis, encephalitis, and, conceivably, peripheral neuropathies. Although neuropathy is not a common sequela, organomegaly, endocrinopathy or POEMS syndrome are potential outcomes (reviewed in [<xref ref-type="bibr" rid="B17">17</xref>]). Malignant B-cell infiltration into bone marrow suppresses hemopoiesis, causing anemia, granulocytopenia and possibly thrombocytopenia. While combinations of proliferating and infiltrating malignant B-cells are a significant factor in the genesis of splenomegaly and lymphadenopathy of superficial or deep lymph nodes, malignant B-cells in prolymphocytic or hairy cell leukemia, after infiltrating the bone marrow and spleen, may cause spleen enlargement (reviewed in [<xref ref-type="bibr" rid="B17">17</xref>]).</p><p>Cutaneous T-cell lymphomas are associated with elevated levels of Th-2-type-associated cytokines such as IL-2, IL-4, IL-5, IL-10, IL13, IFN-γ, TNF-β, TNF-α and GM-CSF, which might compound the occurrence of eosinophilia and eosinophilic pneumonia [<xref ref-type="bibr" rid="B19">19</xref>]. The neoplastic plasma cells in multiple myeloma secrete IL-1, causing the stimulation of osteoclast proliferation, oteolysis, severe bone pain and pathological fractures [<xref ref-type="bibr" rid="B20">20</xref>]. Moreover, in lymphoma-associated hemo-phagocytic syndrome, excessive secretion of IL-1 might play a role in the inappropriate secretion of antidiuretic hormone [<xref ref-type="bibr" rid="B21">21</xref>]. It has been suggested that the uncontrolled extra-renal production of calcitriol is the major humoral mediator of hypercalcemia in both Hodgkin's disease and non-Hodgkin's lymphomas [<xref ref-type="bibr" rid="B22">22</xref>]. Lymphocytic malignancies are known to be sensitive to cytotoxic drugs, causing hyperuricosuria, hyperkalemia and hyperphosphatemia. These abnormalities result from metabolic disruption and are collectively referred to as "tumor lysis syndrome" [<xref ref-type="bibr" rid="B23">23</xref>].</p><p>Cutaneous T-cell lymphomas encompass malignant cells that home to skin, occasionally producing desquamating erythroderma, as observed in Sézary syndrome, nodular infiltrative lesions, and, in specific incidences, mycosis fungoides. Lymphocytic leukemia and lymphoblastic lymphomas of T-cell origin are associated with the enlargement of mediastinal regions. B-cell lymphomas have frequently been observed in bones, bowel, kidneys, lungs, heart, joints, endocrine and salivary glands and less frequently in the extra-nodal regions. Marginal zone B-cell lymphomas of the mucosa-associated lymphoid tissue (MALT) type have been recorded mostly in the stomach and also in the salivary glands, though extra-nodal origin is a possibility, as indicated by columnar or cuboidal epithelium in staging (reviewed in [<xref ref-type="bibr" rid="B17">17</xref>]).</p></sec><sec><title>Loss of self-tolerance and onset of autoimmunity: a muti-factorial phenomenon</title><p>Among the foremost risk factors for the development of autoimmune diseases are polymorphisms among genes regulating the onset of the self-tolerance and immune regulation – <italic>(autoimmune regulator (AIRE)</italic>, the T cell immunoglobulin and mucin-domain-containing (TIM) gene family, and cytolytic T lymphocyte-associated antigen 4 (CTLA-4)) – which have a significant role in sustaining self-tolerance and in the onset of autoimmunity (reviewed in [<xref ref-type="bibr" rid="B24">24</xref>-<xref ref-type="bibr" rid="B26">26</xref>]). <italic>AIRE </italic>null mice develop multi-organ failure because of the specific reduction in ectopic transcription of genes encoding peripheral antigens. On the basis of these findings it has been suggested that thymically-imposed "central" (self) tolerance plays a pivotal role in the in genesis of autoimmunity [<xref ref-type="bibr" rid="B27">27</xref>].</p></sec><sec><title>Cellular basis of antigen presentation and cross-presentation: role of MHC and T-cells</title><p>In particular, the MHC haplotype contributes either by enabling peptide epitopes to be presented in the periphery, increasing T cell activation, or by aborting the presentation of self(host)-antigens in the thymus. As a result, more aggressive T cells or fewer regulatory T (Treg) cells are formed in the host. It is known that Th1-type responses such as interferon-γ [IFN-γ] and interleukin 12 [IL-12] are associated with destructive autoimmune responses, while Th2-type responses (IL-4, IL-5, IL-13) counter-regulate cell-mediated autoimmune processes [<xref ref-type="bibr" rid="B28">28</xref>].</p><p>Helper T (CD4+) cells recognize peptides presented by MHC class II molecules, whereas cytotoxic T cells (CTLs/CD8) recognize peptides presented by MHC class I. MHC class II molecules present peptides originating from exogenous sources that enter the cell by endocytosis. MHC class I molecules present antigens of endogenous origin, which are synthesized within the cells. Dendritic cells also process exogenous antigens into the MHC class I pathway; this is referred to as "<italic>cross-presentation</italic>". As a result, host immune systems generate immunity to exogenous agents and develop tolerance to self antigens. Therefore, cross-presentation is a typical source of indiscriminate presentation of self and foreign antigens [<xref ref-type="bibr" rid="B29">29</xref>]. The thymus is the control center where potentially aggressive T cells specific for autoantigens are eliminated and CD4<sup>+</sup>CD25<sup>+ </sup>Treg cells that recognize autoantigens are selected. Observations on T cell antigen receptor (TCR) transgenic mice show that the thymus regulates the release of antigen-specific CD4<sup>+</sup>CD25<sup>+ </sup>Treg cells into the peripheral circulation [<xref ref-type="bibr" rid="B30">30</xref>].</p></sec><sec><title>Regulatory role of pro-inflammatory and regulatory cytokines in the onset of autoimmunity</title><p>As per the generally-accepted mechanism for the pathogenesis of autoimmune diseases, naive T cells upon activation by antigen produce IL-2 and then undergo clonal expansion and produce pro-inflammatory cytokines (tumor necrosis factor, TNF). CD4<sup>+ </sup>T cells differentiate into at least two subsets of helper cells, T helper 1 (Th1) and T helper 1 (Th2). Th1 produce IFN-γ and lymphotoxin-α under regulation by IL-12, which activates the lymphocyte transcription factor STAT4 (signal transducer and activator of transcription 4) and plays a significant role in the onset of autoimmunity. In contrast, Th2 cells produce IL-4, which constrains cell-mediated immunity (CMI) and possibly inhibits the onset of autoimmune disease. Attenuation of the IL-10 and transforming growth factor-β (TGF-β) effect has been shown to result in inflammation and the onset of autoimmune diseases. In the context of lymphoid malignancy and autoimmunity, the absence of STAT3 from myeloid cells results in the onset of autoimmune diseases [<xref ref-type="bibr" rid="B31">31</xref>].</p><p>Therefore, the balance between proinflammatory (IL-2, IFN-γ and TNF) and regulatory (IL-4, IL-10 and TGF-β) cytokines probably determines any predisposition to develop autoimmune disease. IL-2 is known to mediate apoptosis through two different pathways, passive cell death (PCD) [<xref ref-type="bibr" rid="B32">32</xref>] and activation-induced cell death (AICD) [<xref ref-type="bibr" rid="B33">33</xref>]. PCD occurs mostly because crucial pro-survival signals are absent. Lack of cytokine signaling has been shown to cause an increase in mitochondrial permeability and cytochrome <italic>c </italic>release; along with apoptotic protease-activating factor 1 (APAF1), cytochrome <italic>c </italic>activates caspase-9 and downstream effector caspases. AICD, which is essential for the pathogenesis of autoimmunity and autoimmune lymphoproliferative syndrome, occurs mainly because of IL-2 mediated signaling via the death-domain-containing receptor Fas (CD95).</p><p>After the Fas-associated death domain (FADD) is activated, caspase-8 becomes active and downstream signaling results in cell death. Fas/Fas ligand (FasL; CD178) are deficient in autoimmune <italic>lpr </italic>and <italic>gld </italic>mice, which develop profound lymphadenopathies. The autoimmune lymphoproliferative disease (ALPS) found in humans is also consequent on mutations in Fas [<xref ref-type="bibr" rid="B33">33</xref>,<xref ref-type="bibr" rid="B34">34</xref>]. Furthermore, IL-2 has been shown to upregulate CTLA-4 (cytotoxic T-lymphocyte-associated protein 4; CD152), and CTLA-4-deficienciy leads to a fatal lymphoproliferative disease that is more aggressive than the lymphoproliferative disorders caused by either IL-2 or Fas deficiency [<xref ref-type="bibr" rid="B35">35</xref>]. However, abrogation or attenuation of chemokine induction (CXCL10, IP-10, inflammatory cytokines) at an auxiliary location would be likely to impair the onset of autoimmune disorder[<xref ref-type="bibr" rid="B36">36</xref>].</p></sec><sec><title>Neo-antigens and molecular mimicry</title><p>The structures of host macromolecules and small molecules are markedly altered by acute or chronic oxidative stress and can behave as antigens ("neo-antigens"). Neo-antigens with sufficient homology or identity to host antigenic proteins prompt auto-reactivity. This phenomenon is referred to as "molecular mimicry". A detailed study demonstrating molecular mimicry linkages between viruses and host structures has been reported [<xref ref-type="bibr" rid="B37">37</xref>].</p><p>Drug metabolism is well known to generate neo-antigens in the form of protein adducts [<xref ref-type="bibr" rid="B38">38</xref>,<xref ref-type="bibr" rid="B39">39</xref>]. Potentially, chronic oxidative stress (COS) could be a slowly-evolving concomitant of the generation of mimetic neoantigens. Over time, COS could generate several adducted and/or non-adducted molecules that would essentially act as a "neo-antigens". This is consistent with the slow maturation of auto-antibodies in the evolution of autoimmune diseases. In practice, it is possible that more than one neo-antigens/autoantigens are involved in amplifying the autoaggressive lymphocytes by a process referred to as <italic>"antigen spreading"</italic>. This is an autoimmune reaction initially directed against a single autoantigen that spreads to other autoantigens, causing the T helper cells to recognize them [<xref ref-type="bibr" rid="B40">40</xref>].</p></sec><sec><title>Molecular mimicry is most important cause of autoimmunity after viral infection</title><p>Molecular mimicry results in self-reactive T cells that are activated by cross-reactive ligands originating from infectious pathogens. It is of pivotal significance that molecular mimicry alone, in the context of infection, could not initiate an autoimmune disorder. However, in combination with various cellular factors (epitope avidity and extent of CD8<sup>+ </sup>T cell activation), it may accelerate the process [<xref ref-type="bibr" rid="B41">41</xref>]. By utilizing distinct viral strains, it has been shown that T-cell receptor (TCR) affinity, and also negative regulatory molecules of host origin, are likely to play a crucial role in attenuating autoimmunity. It has further been suggested that autoimmune diseases <italic>per se </italic>are due to combinations of genetic and environmental factors. In particular, the affinity between the TCR and activating peptide-MHC ligand is essential; it might act as a limiting factor in eliciting an autoimmune response by molecular mimicry [<xref ref-type="bibr" rid="B42">42</xref>].</p><p>On the basis of these premises, it is argued that the amounts of genotoxic drugs and their adduct-forming metabolic derivatives probably play a pivotal role in accelerating autoimmune processes. Of course, they would act in combination with host-derived regulatory factors and the outcome would depend on the genetic predisposition of the individual. COS is highly likely to predispose an individual to an autoimmune disorder simply because the intrinsic defense mechanisms are depleted.</p></sec><sec><title>"Bystander effect" and its role in the breakdown of self-tolerance: <italic>a positive regulator of the onset of autoimmunity</italic></title><p>During COS, neo-antigens with target organ specificity potentially cause tissue damage and release a plethora of sequestered auto-antigens. This process is referred to as the "bystander effect". Such an outburst of autoantigens from the target tissue would potentially amplify the effect of the neo-antigens, leading to the breakdown of self-tolerance. To date, there is no definitive evidence that the host is either primed or programmed to bestow tolerance on the newly-evolving antigens resulting from COS.</p></sec><sec><title>Chemical immunomodulation</title><p>Although the drugs used in clinical oncology are of low molecular weight, they cannot activate T cells but sensitize specific lymphocytes. Subsequent contact with a similar chemical or a metabolic derivative induces neo-antigens, which in combination with sufficient co-stimulatory signals cause the release of proinflammtory cytokines (e.g. IL-2, IL-7) and co-stimulatory mediators. This in turn causes the activation of neutrophils, monocytes, macrophages and complement pathways. It is reasonable to envisage that chemical activation of a leukocyte population releases pro-inflammatory cytokines, which in turn determine whether the host develops sensitivity or tolerance. It should be pointed out that drug-induced immune derangements are similar to those in graft-vs-host diseases [<xref ref-type="bibr" rid="B43">43</xref>].</p></sec><sec><title>Immune response(s) to chemical stimulants</title><p>Owing to the overwhelming antigenic load in the host, naive T-cells (Th0) are activated and form Th1 or Th2 subpopulations. The Th1 response is characterized by the isotype specificity of the immunoglobulins formed (IgG2a and IgG2b), while the TH2 response is characterized by elevated levels of IgG1 and IgE. In most instances the initial APC is a dentritic cell (DC, CD80+, CD86+), but B-lymphocytes are the APCs for some chemicals that stimulate Th2 [<xref ref-type="bibr" rid="B44">44</xref>,<xref ref-type="bibr" rid="B45">45</xref>]. The immune modulation caused by a chemotherapeutic drug need not resemble the effects of its metabolic derivatives. It is therefore virtually impossible to determine the specific cause and effect relationship in a chemotherapeutic drug-induced autoimmune disease [<xref ref-type="bibr" rid="B46">46</xref>].</p></sec><sec><title>Chemically-induced co-stimulatory signals</title><p>The following evidence supports the notion that COS plays a role in autoimmunity. Protein adducts generated as a result of oxidative metabolism of 2-bromo-2-chloro-1,1,1,-trifluoroethane (TFA-protein adducts) in CYP450 2E1 induce fulminant autoimmune-mediated halothane hepatitis [<xref ref-type="bibr" rid="B47">47</xref>]. TFA-protein adducts are structurally similar to the pyruvate dehydrogenase complex (PDC), which was identified as an autoantigen causing primary biliary cirrhosis, resulting in progressive destruction of the bile ducts [<xref ref-type="bibr" rid="B48">48</xref>]. Since sub-clinical primary biliary cirrhosis (PBC) is not easily diagnosed, the pathogenesis of PBC subsequent to the immune response to TFA-protein adducts or PDC might be due to acceleration of a pre-existing sub-clinical PBC. Protein modifications that lead to the formation of ("non-self") neo-antigens induce halothane hepatitis. At present there is no direct evidence to implicate TFA-adducts in co-stimulation <italic>per se</italic>.</p></sec><sec><title>Role of metabolites/metabolic intermediates in the onset of autoimmunity</title><p>Aberrant induction of self-tolerance, chromatin-reactive T cells, auto-antibodies to chromatin, and inhibition of unresponsiveness to low-affinity auto-antigens (self-components) in the thymus during positive selection are characteristic effects of procainamide (procainamide-hydroxylamine) metabolites [<xref ref-type="bibr" rid="B49">49</xref>]. Administration of anti-CTLA-4 (blocking Ab) in mice that are susceptible to mercuric chloride (HgCl<sub>2</sub>) induced autoimmunity causes an increase in anti-nucleolar auto-antibodies. In DBA/2 mice, which are resistant to heavy metal-induced autoimmunity, similar treatment leads to the production of anti-nucleolar Abs, thus overcoming the genetic configuration of autoimmunity [<xref ref-type="bibr" rid="B50">50</xref>].</p></sec><sec><title>Antigen spreading and its role in autoimmunity</title><p>Antigen spreading is significant in autoimmunity induced in a mouse model by xenobiotics (procainamide, mercuric chloride and gold (I)). Adoptive transfer of CD4<sup>+</sup>CD25<sup>+ </sup>T cells from the xenobiotic-treated mice to untreated mice inhibits the formation of antinuclear autoantibodies. On the basis of these observations it was suggested that the T cell reactivity induced by the xenobiotic treatment may spread from xenobiotic-induced, nucleoprotein-related neoantigens to peptides of unaltered nucleoproteins [<xref ref-type="bibr" rid="B51">51</xref>].</p></sec><sec><title>Toll-like receptors and autoimmunity</title><p>The toll-like receptors (TLRs) are a germ-line-coded receptor family that plays a pivotal role in innate immunity in a wide spectrum of organisms from insects to mammals. The innate immune response mechanism is either initiated or activated by structures referred to as pathogen-associated molecular patterns (PAMP), which are recognized by corresponding pattern recognition receptors (PRR). The best-characterized PAMPs are lipopolysaccharides (LPS), peptidoglycans, mannans, bacterial DNA and double-stranded bacterial RNA. Macrophages, B-cells and dendritic cells (DC) express PRR. PRR are classified into three specific types: secreted, endocytic and signaling. Mannan binding lectin represents the secreted type, while the macrophage mannose receptor belongs to the endocytic class and the toll-like receptors (TLRs) are signaling types [<xref ref-type="bibr" rid="B52">52</xref>].</p><p>TLRs are essential for detecting PAMPs, and this has been identified as the first line of defense for pathogen recognition, for which a range of antimicrobial products and numerous proinflammatory cytokines are generated by the host. The <italic>Drosophila </italic>protein Toll that is required for mounting an effective immune response to <italic>Aspergillus fumigatus </italic>has been identified as a lipopolysaccharide (LPS) receptor. It plays a pivotal role in the primary recognition of infectious pathogens by mammals [<xref ref-type="bibr" rid="B53">53</xref>].</p><p>TLRs are divided into five subfamilies on the basis of amino acid sequence homology: TLR-1, 2, 6, 10; TLR-3; TLR-4; TLR-5; and TLR-7, 8, 9. Structurally, the extracellular region of a TLR contains leucine-rich repeats flanked by cysteine-rich motifs; a TOLL/IL-1 receptor (TIR) homology domain in the cytoplasmic region is critical for signaling. Given the sequence similarities between the TIR domain and the cytoplasmic tails of IL-1 and IL-18 receptors, it has been suggested that their signaling sequences are similar (see box 12-1 and Figure 12-3 of [<xref ref-type="bibr" rid="B54">54</xref>].</p><p>Antigen (epitope-specific) recognition by B-cell receptors (BCR) induces signals that cause B cell proliferation and antibody production. Concurrent recognition by CD4 (T-helper) cells generates specific cytokines that are essential for antigen-specific antibody production by B-cells. It has been conclusively demonstrated that loss of tolerance to a given antigen by both B and T-cells is a primary cause of autoimmune reactions [<xref ref-type="bibr" rid="B55">55</xref>]. B-cells are known to generate anti-self IgG2a antibodies of low affinity. However, these IgG2a can be recognized as immunogenic by B-cells and as PAMP by TLRs, thus inducing autoantibodies to nuclear antigens [<xref ref-type="bibr" rid="B56">56</xref>].</p><p>Concurrent activation of BCR and B-cell TLR-9 by such IgG2a is due to recognition as non-self and results in the formation of a self DNA (autoantigen)-IgG2a (autoantibody) immune complex. Binding of such a complex to BCR triggers endocytosis, causing effective delivery of the denatured chromatin fragments to endosome-associated TLR-9. Activation of TLR-9 by exogenous or endogenous CpG-DNA in MLR-Fas <italic>lpr/lpr </italic>mice induces the progression of renal diseases [<xref ref-type="bibr" rid="B57">57</xref>]. It should be pointed out that MyD88-dependent receptor activation is required for the formation of autoantibody-autoantigen immune complexes in adaptive immune responses [<xref ref-type="bibr" rid="B58">58</xref>].</p><p>Dendritic cells (DCs) engorged with a cardiac muscle-specific self (autoantigen) peptide caused CD4<sup>+</sup>-cell-mediated autoimmune myocarditis, which progresses to dilated cardiomyopathy and heart failure. It has been suggested that this is a TLR-dependent process. Formation of the self-peptide-loaded DCs may have been provoked by various microbial epitopes acting via TLRs during chronic infection [<xref ref-type="bibr" rid="B59">59</xref>]. One such factor may be uric acid. Studies on DC maturation have shown that uric acid is a major endogenous danger signal from injured cells. It induces DC maturation in the presence of antigen and significantly up-regulates the generation of responses originating from CD8<sup>+ </sup>T cells [<xref ref-type="bibr" rid="B60">60</xref>].</p><p>Thus, in the context of the hypothesis proposed here (see below), it is possible that a host overloaded with one or more antigenic determinants (epitopes) from one or more infectious agents causes stress that in turn activates TLRs. It is contended that by a yet uncharacterized mechanism, this could lead to multi-drug resistance during the course of treatment for these infections, and this could develop into autoimmunity.</p><p>However, the free radical theory of autoimmunity proposed here differs markedly from the "danger theory" of immune activation proposed by Matzinger. Essentially, this states that the host immune system does not differentiate between self- and non-self- but mounts an effective immune response only to danger signals originating from necrotic or stressed cells [<xref ref-type="bibr" rid="B61">61</xref>].</p><p>In support of my contention, it has been proposed that macrophages and DCs express TLR on their surfaces so they can recognize PAMP and initiate appropriate signals for inducing reactive oxygen and nitrogen intermediates. Because this would also activate APCs by inducing pro-inflammatory cytokines and up-regulate co-stimulatory molecules for the activation of TLRs (see Figure 1, 2 and Table 1 of [<xref ref-type="bibr" rid="B62">62</xref>]), it has been proposed that one possible mechanism for the onset of autoimmunity is mediation of the breakdown of peripheral tolerance by hyperactive APCs, causing the activation of autoreactive cells [<xref ref-type="bibr" rid="B63">63</xref>].</p><p>I contend that autoimmunity resulting from COS is not an all-or-none response but an evolutionary process that engulfs the host immune system over a period of time. Arguably, a danger signal could play a pivotal role in the onset of autoimmune disorders, but this alone could never account for the collapse of the host immune response. For an effect of such magnitude, the host must endure diverse stress signals that eventually lead to the collapse of tolerance and trigger autoimmune reactions.</p><p>Although a particular TLR is responsible for ligand-induced signaling, the TLR repertoire that confers ligand-binding and signaling specificities results from heterodimerization and the participation of diverse non-TLR adaptor molecules. Several adaptor molecules have been identified and prominent among these are myeloid differentiation factor 88 (MyD88), MyD88-adaptor-like/TIR-associated protein (MAL/TIRAP), Toll-receptor-associated activator of interferon (TRIF) and Toll-receptor-associated molecule (TRAM). They transduce signals from all regions homologous to the Toll/interleukin-1 receptor (IL-1R) (TIR) domain. Such signals activate intracellular protein kinases, which in turn activate transcription factors that up-regulate inflammatory cytokine and fibrotic genes (see Figure 1 in [<xref ref-type="bibr" rid="B64">64</xref>]). The function of a fifth adaptor, SARM (sterile alpha and HEAT/Armadillo motif protein), has yet to be defined [<xref ref-type="bibr" rid="B64">64</xref>]. The molecular basis of the TLR-associated signaling cascade has been discussed in detail elsewhere [<xref ref-type="bibr" rid="B65">65</xref>].</p><p>TLR4/MD-2 and RP105/MD-1 signal the presence of LPS in the host. RP105 is expressed on virtually all mature B-lymphocytes. RP105/MD-1-mediated signaling induces B-cell proliferation and the pathogenesis of systemic lupus erythematosus (SLE) [<xref ref-type="bibr" rid="B66">66</xref>]. It has been suggested that TLR9 activation triggers systemic autoimmunity and contributes through adaptive and innate immune mechanisms to the CpG-DNA-induced succession of lupus nephritis [<xref ref-type="bibr" rid="B67">67</xref>].</p><p>Activation of naïve polyclonal B cell proliferation by TLR7 ligands such as <italic>resiquimod (R848) </italic>and <italic>loxoribine </italic>requires the presence of plasmacytoid dendritic cells (PDCs), whereas similar activation via the TLR9 ligand CpG is independent of PDCs. Also, in the presence of type I interferon (IFN), ligation of TLR7 triggers the multiplication of polyclonal B cells and their subsequent differentiation toward Ig-producing plasma cells. This process occurs independently of T and B cell Ag [<xref ref-type="bibr" rid="B68">68</xref>].</p><p>TLR-9 induced lupus B cell activation has been shown to modulate T-cell mediated inflammatory reactions through IL-10. Furthermore, B-cell mediated lupus pathogenesis can be mediated by B cells acting as APCs for auto-antigens and autoantibody-producing effector cells. B-cells are also sources of IL-10 [<xref ref-type="bibr" rid="B69">69</xref>].</p><p>Cross-presentation of peripheral self-Ags by DC can induce deletion of auto-reactive CTL by cross-tolerance. Activation of tolerogenic DC may cause autoimmunity by stimulating autoreactive CTL. It was concluded that DC activation by TLR ligands is insufficient to break peripheral cross-tolerance in the absence of specific CD4<sup>+ </sup>T helper cells, so autoimmunity is promoted by stimulating the early effector phase of autoreactive CTL only when their precursor frequency is extremely high [<xref ref-type="bibr" rid="B70">70</xref>]</p></sec><sec><title>Cause and effect of COS in the onset of autoimmunity</title><p>Substantial improvement in the therapeutic regimen increases the survival rate for cancer patients (62% of adult and 77% of pediatric cancer patients survive beyond 5 years). It is therefore theoretically possible that the lingering effects of radiation therapy and chemotherapeutic drugs, or a combination of both, impart slow but finite damage to non-cancerous tissues or cells. This condition could therefore be described as a chronic disease, and the late effects of radiation and chemotherapy on normal tissues/cells remain a significant health risk [<xref ref-type="bibr" rid="B71">71</xref>].</p><p>Upon exposure to genotoxic agents, bioactive metabolites are formed <italic>in vivo</italic>. These metabolites as well as the parent compound damage subcellular components in a target organ-specific manner. The extent of such damage depends on the concentrations of metabolic intermediates or parent compounds and/or both. Exposure to toxic chemicals can cause increased iron accumulation in the spleen as a result of erythrocyte damage and free iron release in the target organ. Accumulation of free iron has been shown to cause the activation of phagocytic cells (macrophages, neutrophils) and subsequent release of reactive oxygen and nitrogen species <italic>in vivo </italic>[<xref ref-type="bibr" rid="B72">72</xref>-<xref ref-type="bibr" rid="B74">74</xref>].</p><p>The late effects of radiation and chemotherapeutic drugs (and genotoxic compounds) on normal cells would lead to activation of stress-response kinases (mitogen-activated stress kinases – MAPK) and redox-sensitive transcription factors, and up-regulate pro-inflammatory cytokines [<xref ref-type="bibr" rid="B75">75</xref>,<xref ref-type="bibr" rid="B76">76</xref>]. The continued and ever-increasing presence of chemotherapeutic drug metabolites, in combination with substances released from the damaged tissues, are potential sources of as yet uncharacterized toxins. It is possible that these compounds could cause an aberrant (chronic inflammatory) response <italic>in vivo </italic>[<xref ref-type="bibr" rid="B77">77</xref>].</p><p>In such pathophysiological states, peripheral blood leukocytes would probably be activated, leading to a sequential respiratory burst releasing ROS (H<sup>•</sup>, <sup>•</sup>OH, O<sub>2 </sub><sup>•- </sup>and H<sub>2</sub>O<sub>2</sub>). Owing to its inherent instability and reactivity, <sup>•</sup>OH reacts with biological molecules in its vicinity within 10<sup>-9 </sup>s of its formation [<xref ref-type="bibr" rid="B78">78</xref>]. In addition to ROS in pathophysiological states, iNOS, the inducible or calcium-independent isoform of nitric oxide synthase, mediates the formation of di-nitrogen trioxide (N<sub>2</sub>O<sub>3</sub>) and peroxynitrite (O = NOO-) (RNOS), causing nitrosative and oxidative stress <italic>in vivo </italic>[<xref ref-type="bibr" rid="B79">79</xref>] and altering signaling cascades with effects on the regulation of gene expression [<xref ref-type="bibr" rid="B80">80</xref>,<xref ref-type="bibr" rid="B81">81</xref>].</p><p>Protein nitration occurs as a result of oxidative stress. Tyrosine residues are nitrated via the peroxynitrite-mediated pathway, and haem-containing peroxidase catalyzed reactions also occur. Nitrite (NO<sub>2 </sub><sup>-</sup>) is an end-product of NO metabolism that can be oxidized by haem peroxidases (e.g. horseradish peroxidase, lactoperoxidase and myeloperoxidase) forming the reactive nitrogen species NO<sub>2</sub>. Nitrite also reacts with tyrosine residues causing the nitration of proteins. NO<sub>2 </sub>reacts with HOCl via a myeloperoxidase-catalyzed reaction between H<sub>2</sub>O<sub>2 </sub>and Cl<sup>-</sup>, forming nitryl chloride (NO<sub>2</sub>Cl). NO<sub>2</sub>Cl also nitrates protein tyrosine residues.</p><p>S-Nitrosylated proteins are formed when cysteine thiol groups react with nitric oxide (NO) in the presence of an electron acceptor to form an S-NO bond. Nitrated proteins are protected from reductive or transnitrosative degradation by storage in membranous structures (e.g. lipophilic protein folds, vesicles and interstitial spaces). Caspases are typically sequestered in an S-nitrosylated (inactive) form within the intermembrane spaces of mitochondria. Appropriate apoptotic stimuli (e.g. Fas-Fas ligand binding) release caspases into cytosol where they are denitrosylated to initiate apoptosis. Hypoxia-inducible factor I (HIF-1), stimulating proteins 1 and 3 (Sp1 and Sp3), nuclear factor-κB (NF-κ B) and the prokaryotic transcription factor OxyR are also affected by S- nitrosylation.</p><p>ROS/RNOS are transient, necessitating the use of reaction products as biomarkers or indices of nitrosative and oxidative stress <italic>in vivo</italic>. Oxidized products such as thiobarbituric acid reaction products (TBARs), 4-hydroxynonenal (4-HNE) and hexane are routinely considered markers of lipid peroxidation in both animal models and patients [<xref ref-type="bibr" rid="B82">82</xref>]. Furthermore, sustained nitrosative stress (increased nitrotyrosine formation) has been observed post-irradiation [<xref ref-type="bibr" rid="B83">83</xref>,<xref ref-type="bibr" rid="B84">84</xref>]. Increased lipid peroxidation has also been reported in patients developing radiation pneumonitis [<xref ref-type="bibr" rid="B85">85</xref>]. Hypoxia was identified in the rat lung 6 weeks after a single dose of 28 Gy using the hypoxia marker pimonidazole [<xref ref-type="bibr" rid="B86">86</xref>]. During hypoxia, increased ROS/RNOS production and reduced antioxidant and antioxidant enzyme production have been observed [<xref ref-type="bibr" rid="B87">87</xref>-<xref ref-type="bibr" rid="B89">89</xref>].</p><p>Sustained ROS/RNOS accumulation leads to COS <italic>in vivo</italic>. Although the precise mechanisms involved are not known at present, possibilities include reduced levels of the antioxidant vitamins C and E [<xref ref-type="bibr" rid="B90">90</xref>,<xref ref-type="bibr" rid="B91">91</xref>], differential regulation of the xanthine oxidoreductase system [<xref ref-type="bibr" rid="B92">92</xref>] and/or aberrant arachidonic acid metabolism [<xref ref-type="bibr" rid="B93">93</xref>].</p><p>Changes in protein conformation can cause aggregation and accumulation in tissues <italic>in vivo</italic>. Metabolic activities (possibly oxidative stress/nitrosative stress) can lead to sustained changes in the levels of metal ions, chaperone proteins and pH, leading to macromolecular crowding and increasing the concentration of misfolded proteins in the intracellular milieu. These changes have a significant role in protein aggregation and consequent loss of functional properties. It is equally possible that proteins with altered conformation have toxic effects in the intracellular milieu [<xref ref-type="bibr" rid="B94">94</xref>].</p><p>Despite the lack of direct evidence, it is feasible that proteins with altered conformations may present sequences similar to host self-antigenic determinants (epitopes) and therefore play a role in neo-antigen formation, thus perturbing the host immune system and potentially contributing to the evolution of multidrug resistance-induced autoimmunity.</p><p>As suggested earlier, pro-inflammatory cytokines (IL-2) probably contribute to the loss of self-tolerance and possibly to the formation of neo-antigens. It is conceivable that in malignant disorders, cells that are subjected to stress (oxidative/nitrosative) become unresponsive to ever-increasing doses of ionizing radiation or chemotherapeutic drugs because of their altered state. Thus, the patient would lose pre-existing antioxidant defense mechanisms, causing limited or no antioxidant defense. This would pave the way for the loss of drug sensitivity and multi-drug resistance, and also the generation of COS.</p><p>In summary, the aforementioned circumstances create a state of imbalance between ROS/RNOS generation and removal <italic>in vivo</italic>, thereby causing oxidative/nitrosative stress. My contention is that the intracellular environment, as outlined above, would be suitable for deregulating the immune mechanism(s) in the cancer patient. Autoimmune mechanisms(s) would subsequently be triggered, causing loss of self-tolerance and the development of autoimmune disorders.</p></sec><sec><title>Activation of antioxidant mechanisms as cause of drug resistance: a critical appraisal</title><p>In essence, the multi-drug resistance (MDR) phenotype of a given clinical tumor burden results from a combination of factors including decreased uptake of cytotoxic anti-neoplastic drugs and alterations in intracellular metabolism impairing the capacity of such drugs to kill cells. As a result, tumor cells do not adequately control the cell cycle, and there is increased repair of DNA damage, decline in apoptotic cell death and deregulated energy-dependent efflux of cytotoxic drugs across the plasma membrane [<xref ref-type="bibr" rid="B95">95</xref>].</p><p>Overexpression of Pgp, MDR-associated-protein 1 (MRP1/ABCC1) or ABCG2 has been positively correlated with the evolution of MDR as well as cross-resistance to structurally unrelated anti-neoplastic drugs in a clinical tumor burden [<xref ref-type="bibr" rid="B96">96</xref>]. Of the several mechanisms suggested to underlie the onset of MDR, ABCC1 (MRP1) and its homologues ABCC2 (MRP2), ABCC3 (MRP3), ABCC6 (MRP6) and ABCC10 (MRP7) mediate the transport of glutathione (GSH), glucuronate or sulfate conjugates of organic anions [<xref ref-type="bibr" rid="B97">97</xref>]. Inhibitors of Pgp have been evaluated as effective therapeutic measures for blocking the efflux of chemotherapeutics used in clinical oncology.</p><p>Owing to the heterogeneity of a tumor burden (in particular colon, kidney or adrenocortex) endowed with both Pgp- and non-Pgp-dependent mechanisms that cause MDR, inhibitors of Pgp are of limited scope as adjuncts to chemotherapy. Furthermore, the unexpectedly high mortality rate has been attributed to intolerable tissue toxicity that is partly due to the elevated plasma concentrations of Pgp inhibitors [<xref ref-type="bibr" rid="B98">98</xref>]. An ideal chemotherapeutic transporter antagonist with high transporter affinity and low pharmacokinetic interaction with non-related drugs, which would restore the efficiency of treatment in MDR tumor burdens with no or minimum cytotoxicity, has yet to be formulated [<xref ref-type="bibr" rid="B99">99</xref>]. On the basis of these observation it is suggested that in a given clinical tumor burden, "intrinsic MDR" and "acquired MDR" might coexist, making existing therapeutic interventions unable to prevent relapse.</p></sec><sec><title>Oxidative stress and MDR</title><p>Chemosensitizers as an adjunctive components in combination therapy (chemotherapy) have a variable therapeutic index. This approach is also known to generate oxidative stress owing to the accumulation of ROS and RNS in the intracellular milieu of a given MDR tumor. It has been suggested that oxidative stress-induced apoptosis is a plausible cytotoxic effect of chemotherapeutics in a MDR tumor burden [<xref ref-type="bibr" rid="B100">100</xref>-<xref ref-type="bibr" rid="B104">104</xref>].</p></sec><sec><title>Anti-oxidative (enzymatic and non-enzymatic) system and MDR</title><p>Antioxidants confer protection against oxidative stress by quenching free radicals, chelating redox metals and interacting with (and regenerating) other antioxidants within the "antioxidant network". When optimal concentrations are sustained in tissues and biofluids they can function in both the aqueous and membrane domains. The most efficient enzymatic antioxidants are superoxide dismutase, catalase and glutathione peroxidase [<xref ref-type="bibr" rid="B105">105</xref>]. Non-enzymatic antioxidants include vitamins C and E, carotenoids, thiols (glutathione, thioredoxin and lipoic acid), natural flavonoids and melatonin [<xref ref-type="bibr" rid="B106">106</xref>]. Few antioxidants can regenerate other antioxidants to restore the reduced intracellular state via an "antioxidant network". The redox cycles of vitamins E and C have this capacity, driven by the redox potentials of the [Red/Ox] couple [<xref ref-type="bibr" rid="B107">107</xref>]. Antioxidants attenuate ROS by binding to transition metal-containing proteins, transferrin or ceruloplasmin, inhibiting cellular reactions (Vitamin E) and detoxifying ROS and RNS, (GSH, SODs, catalase) [<xref ref-type="bibr" rid="B108">108</xref>].</p></sec><sec><title>Role of thioredoxin (Trx)/thioredoxin reductase (TrxR) system in multi-drug resistance</title><p>Overexpression of thioredoxin reductase (TrxR) increases growth rates and resistance to cytotoxic agents that induce oxidative stress [<xref ref-type="bibr" rid="B109">109</xref>,<xref ref-type="bibr" rid="B110">110</xref>]. Detoxification of ROS and up-regulation of antioxidant genes are plausible mechanisms [<xref ref-type="bibr" rid="B111">111</xref>]. In tumor cells, the thioredoxin (Trx)/thioredoxin reductase (TrxR) couple produces a reduced form of extracellular Trx, which acts as a growth factor conferring protection from the NK-lysin, tumor necrosis factor-α and the T-lymphocyte respiratory burst [<xref ref-type="bibr" rid="B112">112</xref>]. Effective inhibition of the Trx/TrxR system increases the sensitivity of tumor cells to chemotherapeutic compounds [<xref ref-type="bibr" rid="B113">113</xref>-<xref ref-type="bibr" rid="B115">115</xref>]. Thioredoxin and peroxiredoxin 1 are up-regulated in drug-resistant breast cancer patients who are clinical non-responders to docetaxel [<xref ref-type="bibr" rid="B116">116</xref>].</p></sec><sec><title>Role of the glutaredoxin reductase (GR)/glutaredoxin (GRX) system in drug resistance</title><p>Increased activity of antioxidants such as catalase, glutathione peroxidase and DT-diaphorase, and increase in glutathione level, are associated with the onset of resistance in Chinese hamster cells chronically exposed to menadione [<xref ref-type="bibr" rid="B117">117</xref>]. Increased expression the α,μ and π isoenzymes of GST might confer a multidrug-resistant phenotype on rat hepatic preneoplastic nodules [<xref ref-type="bibr" rid="B118">118</xref>]. Increases in gamma-glutamyl transpeptidase (GGT) and gamma-glutamylcysteine synthetase (GCS) have a role in the acquired resistance to quinone toxicity in rat lung epithelial cells [<xref ref-type="bibr" rid="B119">119</xref>]. Elevated GST activity might also contribute to the conversion of breast tumors to a tamoxifen-resistant phenotype [<xref ref-type="bibr" rid="B120">120</xref>].</p><p>Development of resistance to doxorubicin in human erythroleukemia cells is correlated with an increase in GSH and GST and related enzymes (glutathione peroxidase, glutathione reductase) [<xref ref-type="bibr" rid="B121">121</xref>]. Genomic amplification resulting in an increase in GST-π expression has been observed in head and neck squamous cell carcinoma cell lines models and also in clinical tumor burdens resistant to cisplatin. Clinical reports confirm mortality in patients with GST-π amplification in the entire tumor burden subsequent to chemotherapy [<xref ref-type="bibr" rid="B122">122</xref>]. GST-π been shown to attenuate the formation of the 7-(2-oxo-hepyl)-substituted 1, N(2)-etheno-2'-deoxyguanosine adduct with 2'-deoxyguanosine in human colonic cancer cells that are resistant to anticancer drugs [<xref ref-type="bibr" rid="B123">123</xref>].</p><p>siRNA-mediated down-regulation of GRX-2 in HeLa cells induces an increased sensitivity to doxorubicin and phenylarsine oxide. In normal HeLa cells, exposure to non-lethal oxidative stress causes an increase in endogenous GRX-2. This has been implicated in protection against toxic compounds that induce oxidative stress [<xref ref-type="bibr" rid="B124">124</xref>].</p><p>CAL1 human melanoma cells overexpress GST-μ1 after exposure to anticancer drugs (vincristine, chlorambucil). A concurrent increase in both GST-μ1 and MRP-1 might have a role in protection against vicristine-mediated cytotoxicity [<xref ref-type="bibr" rid="B125">125</xref>]. Also, it has been suggested that GRX-2 is pivotal in the glutathionylation and deglutathionylation of proteins via multiple signaling pathways in a wide range of GSH/GSSG ratios associated with different cellular redox states [<xref ref-type="bibr" rid="B126">126</xref>]. The transcription factor Nrf2 mediates the up-regulation of γ-GCS and GSH synthesis, and induces resistance to Imatinib (BCR/ABL tyrosine kinase inhibitor) in chronic myelogenous leukemia [<xref ref-type="bibr" rid="B127">127</xref>].</p><p>Activation of glutathione peroxidase and glutathione reductase was observed in the onset of radio resistance and cross-resistance to chemotherapeutic agents in glioblastoma [<xref ref-type="bibr" rid="B128">128</xref>]. Increased binding of AP-1 (activator protein-1) activity, observed in an electrophoretic mobility shift assay, was suggested as the molecular mechanism by which GST is overexpressed, in turn conferring resistance to doxorubicin in leukemia [<xref ref-type="bibr" rid="B129">129</xref>].</p><p>Overexpression of GRX-2 in HeLa cells confers a significant antiapoptotic and pro-survival effect upon exposure to doxorubicin and phenylarsine oxide, possibly via inhibition of cytochrome c release [<xref ref-type="bibr" rid="B130">130</xref>]. Pancreatic cancer cells exposed to triterpenoid 2-cyano-3,12-dioxooleyl-1,9-diene-28-imidazolide (CDDO-Im) caused the depletion of mitochondrial glutathione, leading to apoptosis [<xref ref-type="bibr" rid="B131">131</xref>].</p><p>A suggested downstream target of redox-sensitive signaling is ribonucleotide reductase, which is likely to play a significant role in the antioxidant-mediated protection of tumor cells [<xref ref-type="bibr" rid="B132">132</xref>]. Glutathione peroxidase 1 and GST-π1 are up-regulated in breast cancer patients who are clinical non-responders to docetaxel but show drug resistance [<xref ref-type="bibr" rid="B133">133</xref>].</p></sec><sec><title>Role of redox-sensitive signaling proteins and MDR</title><sec><title>Metallothioneins (MTs)</title><p>MTs are zinc-binding protein thiols with antioxidant attributes that increase in tumor cells. This observation suggests that overexpression of MTs is significant in the acquisition of drug resistance in human tumor cells [<xref ref-type="bibr" rid="B134">134</xref>].</p></sec><sec><title>Superoxide dismutase (SOD)</title><p>A significant increase in SOD may be involved in the conversion of breast tumors to a tamoxifen-resistant phenotype [<xref ref-type="bibr" rid="B120">120</xref>]. Mn-SOD-dependent activation of the zinc-dependent matrix metalloproteinase family (MMP-1,-2) has been positively correlated with an increased incidence of metastasis in gastrointestinal tumors [<xref ref-type="bibr" rid="B135">135</xref>]. Also, increased Mn-SOD activity confers resistance to doxorubicin in human erythroleukemia cells [<xref ref-type="bibr" rid="B121">121</xref>]. It has been suggested that up-regulation of SOD is significant in the onset of radioresistance and cross-resistance to chemotherapeutic agents in glioblastoma [<xref ref-type="bibr" rid="B136">136</xref>].</p></sec><sec><title>Vitamin C</title><p>The role of Vitamin C in decreasing the incidence of stomach, lung and colorectal cancer may be attributable to the inhibition of <italic>N</italic>-nitroso compound generation [<xref ref-type="bibr" rid="B137">137</xref>].</p></sec><sec><title>Flavonoids</title><p>A regular intake of flavonoids such as polyphenols and quercetin is linked to lower incidences of gastrointestinal and also lung and breast cancer [<xref ref-type="bibr" rid="B138">138</xref>].</p></sec><sec><title>Selenium</title><p>Se (200μg/day) seems to reduce the incidences of lung, colon and prostate cancer [<xref ref-type="bibr" rid="B139">139</xref>,<xref ref-type="bibr" rid="B140">140</xref>]. Increased levels of antioxidant enzymes (SOD, catalase, glutathione peroxidase) and non-enzymatic antioxidants (GSH, vitamin C, thioredoxin) are significant in several clinical tumor burdens [<xref ref-type="bibr" rid="B141">141</xref>].</p><p>Collectively, antioxidants confer a reducing intracellular environment enabling tumor burdens with MDR to evade apoptosis and acquire growth advantage with increased cell survival signals [<xref ref-type="bibr" rid="B142">142</xref>]. For these reasons, the "redox buffering" capacity of a given tumor burden is a potential therapeutic target for effective cancer-preventive and therapeutic drug design [<xref ref-type="bibr" rid="B143">143</xref>].</p></sec><sec><title>Summary</title><p>Oxidative stress induces conformational changes in intracellular proteins containing cysteine residues, triggering ionization of the sulfhydryl moiety (-CH<sub>2</sub>SH to -CH<sub>2</sub>S<sup>-</sup>). Pro-survival genes are up-regulated in a reduced intracellular state. It is suggested that increased expression of pro-survival genes enables tumor cells to evade chemotherapeutically-induced cytotoxicity, conferring an adaptive growth advantage that aggravates the tendency of tumor cells with MDR to repopulate the tumor burden. This leads to relapse and the subsequent development of autoimmunity via several chronic redox state-dependent reaction cascades.</p></sec></sec><sec><title>Hypothesis</title><p>The following proposal (illustrated in Figure <xref ref-type="fig" rid="F1">1</xref>) accounts for the role of COS as an essential element in the evolution of drug resistance-mediated induction of autoimmunity. The hypothetical scheme is based on the premises outlined above, and defines a possible sequence of events beginning from radiation therapy/chemotherapy and leading to the evolution of autoimmune disorders.</p><fig position="float" id="F1"><label>Figure 1</label><graphic xlink:href="1742-4682-3-22-1"/></fig><p>As shown in Figure <xref ref-type="fig" rid="F1">1</xref>, the toxicity consequent on radiation therapy/chemotherapy is due to the metabolic derivatives and/or ionic overload (imbalance). These, in turn, induce the activation of phagocytes and the associated respiratory burst causing the release of ROS and RNS. Lymphocytes (T and B cells) that are subjected to oxidative stress are more sensitive to such stress. Lymphocytes exposed to H<sub>2</sub>O<sub>2 </sub>exhibit extensive genomic damage (DNA strand breaks). Oxidative/nitrosative stress-induced alteration in the host lymphocyte genome is likely to engender aberrant cellular pathways that could lead to rapid cell lysis (necrosis) or programmed cell death (apoptosis). Owing to the overwhelming ROS/RNS, the antioxidant reserve may be depleted or inactivated by one or more mechanism(s). Such an intracellular environment would be conducive to lipid peroxidation and accumulation of oxidized lipid-derived aldehydes.</p><p>Lipid peroxidation-derived aldehydes are a potential source of protein modification (oxidation resulting in carbonylation and nitrosation). This phase could be classed as a state of acute oxidative stress. It is recognized that acute oxidative stress causes genomic and/or mitochondrial DNA damage. As a result, it is conceivable that pro-apoptotic gene expression would be up-regulated. Should this become overwhelming, the tumor cells are programmed to die by apoptosis <bold>(First Selection)</bold>. These processes would be exacerbated if the DNA repair mechanisms <italic>in vivo </italic>failed or were defective.</p><p>It is possible that during acute oxidative stress, anti-apoptotic proteins/drug resistance genes may be overexpressed in a select but limited number of tumor cells. It is also possible that those cells may have endured minimal or no loss of anti-oxidant defense. The select cell population would therefore survive and eventually cope with the ensuing COS. Such a pathophysiological state would obviously lead to inflammation, engendered by inflammatory cytokines and tumorigenic chemokines, which in turn prompt deregulation of the immune defense mechanisms.</p><p>Should there be a loss of antioxidant defense or failure of adaptive mechanisms, or both, vulnerable tumor cells would be programmed to undergo apoptosis <bold>(Second Selection)</bold>. Essentially, COS acts as a regulator for selecting the most tolerant tumor cell population, which is adapted to survive radiation/genotoxic insults. But there are no adequate signaling mechanism(s) in these aberrant cells, impairing their ability to sustain homeostasis.</p><p>The COS-induced impairment of the intracellular metabolic machinery again causes sustained genomic DNA damage and potentially depletes the antioxidant reserve. This may occur either concurrently or sequentially and lead to the multi-drug resistant genotype, as well as facilitating the formation of neo-antigens. It has previously been hypothesized that, owing to the loss of tolerance to autoantigens and sustained accumulation of neo-antigen(s), an autoimmune response would be likely to evolve [<xref ref-type="bibr" rid="B144">144</xref>,<xref ref-type="bibr" rid="B145">145</xref>].</p><p>A factor that has frequently been overlooked in the <italic>in vivo </italic>context is that cells undergoing apoptosis or necrosis release considerable amounts of proteolytic enzymes and other mediators. It is conceivable that these factors would be detremental to normal as well as multi-drug-resistant tumor cells. They would cause havoc in intracellular signaling, contributing to the alteration in gene expression. It is also possible that these factors would lead to cell death via either apoptosis or necrosis <bold>(Second Selection)</bold>.</p><p>Those normal cells that acquire tumor-resistant genotypes are successful in passing through all the selection processes and would eventually emerge as multidrug-resistant cells. Under these circumstances, accumulation of neo-antigens or other inflammatory cytokines would also decimate the host immune defenses. These events lead to an intracellular milieu conducive to a lethal breakdown in host self-tolerance. This, in turn, causes a surge in immune disorders, leading to the sequence of events that result in autoimmunity.</p><p>Should this effect prevail, it would be a potential cause of death in a cancer patient. Therefore, an effective means of treating the lymphoproliferative disorders that result from failure of therapeutic measures is to delineate the precise molecular mechanisms leading to COS and the consequent genomic DNA damage and expression and regulation of repair proteins. Only a combination of therapeutic measures can attenuate or disengage the pathophysiological consequences of these deleterious COS-induced effects on tumor cells.</p><p>In summary, these COS-induced effects would directly (i) alter signal transduction and (ii) induce epigenetic mechanisms (hypo/hyper-methylation; hypo/hyper-acetylation of the elements that regulate anti-apoptotic or pro-apoptotic genes) [<xref ref-type="bibr" rid="B146">146</xref>-<xref ref-type="bibr" rid="B151">151</xref>]. In the context of COS as a regulatory element in the evolution of autoimmunity, alterations in cytokines (IL-2 expression) would exert the most significant impact next to the neoantigens. Therapeutic measures directed at modulating such epigenetic mechanism(s) in a development stage-specific manner (lymphoproliferative disorder) would potentially attenuate the impact of drugs and the subsequent evolution of autoimmune disorders.</p></sec></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Note</title><p>*<bold><italic>In honor of my mother Srimathi Kannika Kannan</italic></bold></p><p>Chronic oxidative stress is an essential regulatory element in the evolution of drug resistance-mediated induction of autoimmunity</p></sec>
Identification of novel functional sequence variants in the gene for peptidase inhibitor 3	<sec><title>Background</title><p>Peptidase inhibitor 3 (PI3) inhibits neutrophil elastase and proteinase-3, and has a potential role in skin and lung diseases as well as in cancer. Genome-wide expression profiling of chorioamniotic membranes revealed decreased expression of PI3 in women with preterm premature rupture of membranes. To elucidate the molecular mechanisms contributing to the decreased expression in amniotic membranes, the PI3 gene was searched for sequence variations and the functional significance of the identified promoter variants was studied.</p></sec><sec sec-type="methods"><title>Methods</title><p>Single nucleotide polymorphisms (SNPs) were identified by direct sequencing of PCR products spanning a region from 1,173 bp upstream to 1,266 bp downstream of the translation start site. Fourteen SNPs were genotyped from 112 and nine SNPs from 24 unrelated individuals. Putative transcription factor binding sites as detected by <italic>in silico </italic>search were verified by electrophoretic mobility shift assay (EMSA) using nuclear extract from Hela and amnion cell nuclear extract. Deviation from Hardy-Weinberg equilibrium (HWE) was tested by χ<sup>2 </sup>goodness-of-fit test. Haplotypes were estimated using expectation maximization (EM) algorithm.</p></sec><sec><title>Results</title><p>Twenty-three sequence variations were identified by direct sequencing of polymerase chain reaction (PCR) products covering 2,439 nt of the PI3 gene (-1,173 nt of promoter sequences and all three exons). Analysis of 112 unrelated individuals showed that 20 variants had minor allele frequencies (MAF) ranging from 0.02 to 0.46 representing "true polymorphisms", while three had MAF ≤ 0.01. Eleven variants were in the promoter region; several putative transcription factor binding sites were found at these sites by database searches. Differential binding of transcription factors was demonstrated at two polymorphic sites by electrophoretic mobility shift assays, both in amniotic and HeLa cell nuclear extracts. Differential binding of the transcription factor GATA1 at -689C>G site was confirmed by a supershift.</p></sec><sec><title>Conclusion</title><p>The promoter sequences of PI3 have a high degree of variability. Functional promoter variants provide a possible mechanism for explaining the differences in PI3 mRNA expression levels in the chorioamniotic membranes, and are also likely to be useful in elucidating the role of PI3 in other diseases.</p></sec>	<contrib id="A1" contrib-type="author"><name><surname>Chowdhury</surname><given-names>Mahboob A</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Kuivaniemi</surname><given-names>Helena</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" corresp="yes" contrib-type="author"><name><surname>Romero</surname><given-names>Roberto</given-names></name><xref ref-type="aff" rid="I1">1</xref><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A4" contrib-type="author"><name><surname>Edwin</surname><given-names>Samuel</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A5" contrib-type="author"><name><surname>Chaiworapongsa</surname><given-names>Tinnakorn</given-names></name><xref ref-type="aff" rid="I2">2</xref><xref ref-type="aff" rid="I3">3</xref><email>[email protected]</email></contrib><contrib id="A6" corresp="yes" contrib-type="author"><name><surname>Tromp</surname><given-names>Gerard</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	BMC Medical Genetics	<sec><title>Background</title><p>PI3 [Gene ID: 5266] is a member of the 'trappin' gene family [<xref ref-type="bibr" rid="B1">1</xref>]. The trappin gene family members are defined by an amino-terminal transglutaminase substrate domain consisting of hexapeptide repeats with the consensus sequence of GQDPVK and a carboxy-terminal four-disulphide bond core. PI3, also known as trappin-2, elafin, elastase specific inhibitor and skin-derived antileukoproteinase (SKALP), is a low-molecular weight, 6 kDa serine protease inhibitor [<xref ref-type="bibr" rid="B2">2</xref>], that is capable of inhibiting neutrophil elastase (also known as elastase 2; ELA2; [GeneID: 1991]) and proteinase 3 (PRTN3; [GeneID: 5657]; also known as the Wegener autoantigen, P29). PI3 has been mapped to chromosome 20q12-13.1 [<xref ref-type="bibr" rid="B3">3</xref>], and this locus contains 14 genes expressing protease inhibitor domains with homology to whey acidic protein (WAP). Human PI3 gene spans about 11,620 bp and consists of three exons [<xref ref-type="bibr" rid="B2">2</xref>,<xref ref-type="bibr" rid="B4">4</xref>]. The gene has multiple transcription start sites and the mRNA has been reported to have an unusually short 5'-UTR (5'-untranslated region) [<xref ref-type="bibr" rid="B5">5</xref>].</p><p>Initially, PI3 was identified in human epidermis of psoriatic patients [<xref ref-type="bibr" rid="B6">6</xref>], and later in bronchial secretions from patients with bronchial carcinoma [<xref ref-type="bibr" rid="B7">7</xref>] and chronic obstructive pulmonary disease [<xref ref-type="bibr" rid="B2">2</xref>], as well as in epidermal [<xref ref-type="bibr" rid="B8">8</xref>] and breast tumors [<xref ref-type="bibr" rid="B9">9</xref>]. In addition to its antipeptidase role, PI3 has antimicrobial activity and is a component of the innate immune system to protect epithelial surfaces from infection [<xref ref-type="bibr" rid="B10">10</xref>-<xref ref-type="bibr" rid="B13">13</xref>]. Expression of PI3 can be induced by inflammatory mediators such as tumor necrosis factor (TNF) and interleukin 1 beta (IL1B) [<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B15">15</xref>].</p><p>In our previous report we identified PI3 as a down-regulated gene in the chorioamnionitic membranes of patients with preterm premature rupture of membranes (PPROM) [<xref ref-type="bibr" rid="B16">16</xref>]. In this study, we investigated the possible molecular mechanisms that control the expression of PI3 by carrying out a detailed analysis of the PI3 gene sequences.</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Genomic DNA isolation</title><p>Blood samples were obtained from 112 healthy unrelated African-American individuals after written informed consent. The collection of samples, and their utilization for research purposes, was approved by the Institutional Review Boards of Wayne State University and the National Institute of Child Health and Human Development, NIH. Genomic DNA was extracted from blood samples using QIAGEN<sup>® </sup>DNA Blood BioRobot<sup>® </sup>9604 kit (QIAGEN Inc., Valencia, CA.).</p></sec><sec><title>Direct sequencing of PCR products</title><p>Genomic DNA was used as a template to generate three overlapping PCR products of 724 bp, 717 bp and 1,328 bp in size extending from 1,173 bp upstream to 1,266 bp downstream of the translation start site of the PI3 gene [GenBank: NT_011362]. Primers are listed in Table <xref ref-type="table" rid="T1">1</xref>. All PCRs were carried out in 100-μl volumes containing 1.5 mM of MgCl<sub>2</sub>, 0.2 mM dNTPs, 0.4 μM of each primer, 3 U of Taq DNA polymerase (Roche Molecular Systems, Inc., Branchburg, NJ) and 100 ng of genomic DNA. A 10 minute initial denaturation at 94°C was followed by 40 cycles consisting of 30 s denaturation at 94°C, 30 s annealing at 50°C to 55°C, and 1 minute extension at 72°C. PCR products were analyzed on 2% agarose gels. PCR products were purified by ultrafiltration (Centricon Centrifugal Filter Devices, Millipore, Bedford, MA), and sequenced by cycle sequencing and dye terminator labeling (ABI<sup>® </sup>BigDye™ Terminator v1.1 Cycle Sequencing kit, Applied Biosystems, Foster City, CA). Sequencing reactions were purified using gel filtration columns (CENTRI-SEP, Princeton Separation, Adelphia, NJ) and run on 310 or 3700 Genetic Analyzer (Applied Biosystems). Sequences were edited using BioEdit [<xref ref-type="bibr" rid="B17">17</xref>]. Fourteen SNPs were genotyped from 112 unrelated individuals and nine SNPs from 24 unrelated individuals (Table <xref ref-type="table" rid="T2">2</xref>).</p><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>Oligonucleotide primers used in the study.</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left"><bold>Primer Code</bold></td><td align="left"><bold>Sequence<sup>a</sup></bold></td><td align="left"><bold>Purpose</bold></td><td align="center"><bold>PCR product (bp)</bold></td><td align="center"><bold>Annealing temperature (°C)</bold></td></tr></thead><tbody><tr><td align="left">01F_PI3</td><td align="left">tgagaagggtgtgtgaaggaa</td><td align="left">PCR and sequencing</td><td align="center">724</td><td align="center">55</td></tr><tr><td align="left">01R_PI3</td><td align="left">accactcccagcatcaa</td><td align="left">PCR and sequencing</td><td align="center">724</td><td align="center">55</td></tr><tr><td align="left">02F_PI3</td><td align="left">gagttttttgcaggaccagg</td><td align="left">PCR and sequencing</td><td align="center">717</td><td align="center">52</td></tr><tr><td align="left">02R_PI3</td><td align="left">gaacagaaagctgaaatctg</td><td align="left">PCR and sequencing</td><td align="center">717</td><td align="center">50</td></tr><tr><td align="left">Seq_P13_1328bp_F</td><td align="left">caagctggactgcataaaga</td><td align="left">PCR</td><td align="center">1328</td><td align="center">54</td></tr><tr><td align="left">Seq_P13_1328bp_R</td><td align="left">cagccttcttttgtgtcttc</td><td align="left">PCR</td><td align="center">1328</td><td align="center">53</td></tr><tr><td align="left">Seq_P13_Int1_F</td><td align="left">tgcataaagattggtatggc</td><td align="left">sequencing</td><td align="center">-</td><td align="center">52</td></tr><tr><td align="left">Seq_PI3_Int2_F</td><td align="left">tttaaaccttgggtgtggac</td><td align="left">sequencing</td><td align="center">-</td><td align="center">54</td></tr><tr><td align="left">Seq_PI3_Int3_F</td><td align="left">gaggtgtaccttccctactc</td><td align="left">sequencing</td><td align="center">-</td><td align="center">54</td></tr><tr><td align="left">-1077_A_F</td><td align="left">ctctccttgtctcAgtgtattagagtc</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-1077_G_F</td><td align="left">ctctccttgtctcGgtgtattagagtc</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-1067_A_ F</td><td align="left">ctcagtgtattagAgtcgtttttctca</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-1067_G_F</td><td align="left">ctcagtgtattaggGtcgtttttctca</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">+1063_A_F</td><td align="left">gtgtattagagtcAtttttctcagaca</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">+1063_G_F</td><td align="left">gtgtattagagtcGtttttctcagaca</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-960_T_F</td><td align="left">ggaacccccgtttTcccctttcattactt</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-960_D_F</td><td align="left">ggaacccccgtttcccctttcattactt</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-911_A_F</td><td align="left">gttaatagaccagaccaaAtctcacac</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-911_G_F</td><td align="left">gttaatagaccagaccaaGtctcacac</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-689_C_F</td><td align="left">tgtatacatgataCatgttttctacta</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-689_G_F</td><td align="left">tgtatacatgataGatgttttctacta</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-675_C_F</td><td align="left">atgttttctactaCtttctgattattt</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-675_T_F</td><td align="left">atgttttctactaTtttctgattattt</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-453_T_F</td><td align="left">ttgatgctgggagTggtaaaatgataa</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-453_G_F</td><td align="left">ttgatgctgggagGggtaaaatgataa</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-338_G_F</td><td align="left">gaataaccttcgGtgattcctttctcttct</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-338_A_F</td><td align="left">gaataaccttcgAtgattcctttctcttct</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-258_A_F</td><td align="left">taataagtgagccAgcacttctactct</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr><tr><td align="left">-258_G_F</td><td align="left">taataagtgagccGgcacttctactct</td><td align="left">gel shift assay</td><td align="center">-</td><td align="center">-</td></tr></tbody></table><table-wrap-foot><p><sup>a</sup>The nucleotide in upper case is the variant nucleotide.</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T2"><label>Table 2</label><caption><p>Minor allele frequencies of the 23 SNPs detected in the PI3 gene.</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="center" colspan="2"><bold>Location</bold></td><td></td><td></td><td></td><td></td></tr><tr><td></td><td colspan="2"><hr></hr></td><td></td><td></td><td></td><td></td></tr><tr><td align="left"><bold>GenBank entry position<sup>a</sup></bold></td><td align="left"><bold>nt<sup>b</sup></bold></td><td align="left"><bold>region</bold></td><td align="left"><bold>Non-synonymous substitution</bold></td><td align="left"><bold>MAF</bold></td><td align="left"><bold>rs number</bold></td><td align="left"><bold>MAF in dbSNP</bold></td></tr></thead><tbody><tr><td align="left">48460A>G</td><td align="left">-1077</td><td align="left">promoter</td><td></td><td align="left">0.107</td><td></td><td></td></tr><tr><td align="left">48470A>G</td><td align="left">-1067</td><td align="left">promoter</td><td></td><td align="left">0.107</td><td></td><td></td></tr><tr><td align="left">48474G>A</td><td align="left">-1063</td><td align="left">promoter</td><td></td><td align="left">0.121</td><td></td><td></td></tr><tr><td align="left">48577T>Del</td><td align="left">-960</td><td align="left">promoter</td><td></td><td align="left">0.103</td><td></td><td></td></tr><tr><td align="left">48626G>A</td><td align="left">-911</td><td align="left">promoter</td><td></td><td align="left">0.009</td><td></td><td></td></tr><tr><td align="left">48669C>G</td><td align="left">-868</td><td align="left">promoter</td><td></td><td align="left">0.138</td><td align="left">2267863</td><td></td></tr><tr><td align="left">48848C>G</td><td align="left">-689</td><td align="left">promoter</td><td></td><td align="left">0.107</td><td></td><td></td></tr><tr><td align="left">48862C>T</td><td align="left">-675</td><td align="left">promoter</td><td></td><td align="left">0.107</td><td></td><td></td></tr><tr><td align="left">49084T>G</td><td align="left">-453</td><td align="left">promoter</td><td></td><td align="left">0.005</td><td></td><td></td></tr><tr><td align="left">49199G>A</td><td align="left">-338</td><td align="left">promoter</td><td></td><td align="left">0.107</td><td></td><td></td></tr><tr><td align="left">49279A>G</td><td align="left">-258</td><td align="left">promoter</td><td></td><td align="left">0.005</td><td></td><td></td></tr><tr><td align="left">49586C>T</td><td align="left">+50</td><td align="left">exon 1</td><td align="left">T17M</td><td align="left">0.107</td><td align="left">17333103</td><td align="left">0.169</td></tr><tr><td align="left">49681C>A</td><td align="left">+145</td><td align="left">IVS 1</td><td></td><td align="left">0.107</td><td align="left">17333180</td><td align="left">0.169</td></tr><tr><td align="left">49698T>A</td><td align="left">+162</td><td align="left">IVS 1</td><td></td><td align="left">0.455</td><td align="left">1983649</td><td align="left">0.471</td></tr><tr><td align="left">49940C>G</td><td align="left">+404</td><td align="left">IVS 1</td><td></td><td align="left">0.083<sup>c</sup></td><td></td><td></td></tr><tr><td align="left">49944C>T</td><td align="left">+408</td><td align="left">IVS 1</td><td></td><td align="left">0.125<sup>c</sup></td><td></td><td></td></tr><tr><td align="left">50105C>G</td><td align="left">+569</td><td align="left">IVS 1</td><td></td><td align="left">0.146<sup>c</sup></td><td align="left">16989785</td><td align="left">0.056</td></tr><tr><td align="left">50163A>G</td><td align="left">+627</td><td align="left">IVS 1</td><td></td><td align="left">0.146<sup>c</sup></td><td align="left">17424356</td><td align="left">0.169</td></tr><tr><td align="left">50287T>A</td><td align="left">+751</td><td align="left">IVS 1</td><td></td><td align="left">0.063<sup>c</sup></td><td align="left">6032040</td><td align="left">0.176</td></tr><tr><td align="left">50495A>C</td><td align="left">+959</td><td align="left">exon 2</td><td align="left">T34P</td><td align="left">0.125<sup>c</sup></td><td align="left">2664581</td><td align="left">0.156</td></tr><tr><td align="left">50659C>T</td><td align="left">+1123</td><td align="left">exon 2</td><td></td><td align="left">0.020<sup>c</sup></td><td></td><td></td></tr><tr><td align="left">50762C>A</td><td align="left">+1226</td><td align="left">IVS 2</td><td></td><td align="left">0.125<sup>c</sup></td><td></td><td></td></tr><tr><td align="left">50770C>A</td><td align="left">+1234</td><td align="left">IVS 2</td><td></td><td align="left">0.125<sup>c</sup></td><td align="left">17424474</td><td align="left">0.152</td></tr></tbody></table><table-wrap-foot><p><sup>a</sup>GenBank accession No. AL049767.12.</p><p><sup>b</sup>Location is with respect to translation start site.</p><p><sup>c</sup>Frequency estimated from 24 unrelated individuals.</p></table-wrap-foot></table-wrap></sec><sec><title><italic>In silico </italic>search for transcription factor binding sites</title><p>The sequences in and around the SNP sites in the promoter region were searched for putative transcription factor binding sites using three different computer programs: TESS [<xref ref-type="bibr" rid="B18">18</xref>,<xref ref-type="bibr" rid="B19">19</xref>], Alibaba 2.1 [<xref ref-type="bibr" rid="B20">20</xref>,<xref ref-type="bibr" rid="B21">21</xref>], and MatInspector [<xref ref-type="bibr" rid="B22">22</xref>,<xref ref-type="bibr" rid="B23">23</xref>]. Default parameters were used as search criteria.</p></sec><sec><title>Electrophoretic mobility shift assays (EMSA)</title><p>Oligonucleotides (Table <xref ref-type="table" rid="T1">1</xref>) and their complementary strands were designed and purchased as gel purified (IDT, Coraville, IA). Complementary oligonucleotides were annealed to each other to generate double-stranded probes. EMSAs were performed using commercially available HeLa cell nuclear extracts (Promega, Madison, WI) and nuclear extracts prepared from primary amnion cell cultures as previously described [<xref ref-type="bibr" rid="B24">24</xref>] since we had previously demonstrated that PI3 protein was produced by a variety of chorioamniotic membrane cell types with the highest amount produced by the amniotic epithelial cells [<xref ref-type="bibr" rid="B16">16</xref>]. Primary amnion cell cultures were established using amniotic membranes obtained from women not in labor at term who underwent elective cesarean deliveries for obstetrical indications. All other reagents were purchased from a commercial source and used according to the manufacturer's protocol (Promega, Madison, WI). The concentration of poly(dI-dC) (Amersham Biosciences Corp., Piscataway, NJ) in the reaction was optimized to 0.05 μg/μl to minimize non-specific binding. The concentrations of components in 10 μl reaction mixtures were as follows: 1X binding buffer [without poly(dI-dC)], 3.75 μg of HeLa or amnion cell nuclear extract, 0.05 μg/μl poly(dI-dC) and 50 fmole of <sup>32</sup>P-labeled double-stranded probe (>50,000 cpm). All the components, except <sup>32</sup>P-probe, were added to the reaction and incubated for 15 min on ice and 10 min at 20°C, followed by the addition of the <sup>32</sup>P-labelled probe, and incubation for 20 min at 20°C. For competition experiments, a 100-fold molar excess of unlabeled double-stranded oligonucleotides was added to the reaction mixture prior to the addition of the labeled probe. For supershift experiments, polyclonal antibodies against AP1 (Cat. No. sc-253X and sc-44X; Santa Cruz Biotechnology, Santa Cruz, CA), and GATA1 (Active Motif, Carlsbad, CA) were used. For AP1, after 20 min of incubation at 20°C with <sup>32</sup>P-labelled probe, 400 ng of corresponding antibodies were added to the reaction and incubated for another 15 min at 20°C. For the GATA1 assay, antibodies were added and incubated for 20 min at 20°C before adding the labeled probe [<xref ref-type="bibr" rid="B25">25</xref>]. Samples were run on non-denaturing 6% polyacrylamide gels in 0.5X TBE buffer, at 100 V for 80 min. X-ray film (Kodak, Rochester, NY) was exposed to dried gels 2 to 5 h at -80°C depending on signal intensity.</p></sec><sec><title>Nomenclature for sequence variants and genes</title><p>The variants and nucleotides are described following the guidelines of the Human Genome Variation Society (HGVS) [<xref ref-type="bibr" rid="B26">26</xref>]. SNPs are described using the genomic sequence AL049767.12 as a reference and numbered relative to the translation start site. Official gene symbols provided by Human Genome Organization (HUGO) Nomenclature Committee (HGNC) were used [<xref ref-type="bibr" rid="B27">27</xref>].</p></sec><sec><title>Statistical analyses</title><p>Tests for deviations from HWE were performed by using the χ<sup>2 </sup>goodness-of-fit test. Haplotypes were estimated following expectation maximization (EM) algorithm as implemented in the software Arlequin [<xref ref-type="bibr" rid="B28">28</xref>].</p></sec></sec><sec><title>Results</title><sec><title>SNP genotyping and haplotype construction</title><p>When this study was initiated, only two polymorphisms were known to exist in the PI3 gene, neither of them in the promoter region. We identified 23 SNPs (Table <xref ref-type="table" rid="T2">2</xref> and Fig <xref ref-type="fig" rid="F1">1</xref>) in the PI3 gene sequences of 24 unrelated individuals by direct sequencing of PCR products that spanned the region from 1,173 bp upstream to 1,266 bp downstream of the translation start site. Of the 23 SNPs, nine are in the dbSNP database [<xref ref-type="bibr" rid="B29">29</xref>] (Table <xref ref-type="table" rid="T2">2</xref>). Eleven SNPs were located in the promoter region, one in exon 1, seven in intron 1, two in exon 2 and two in intron 2. To obtain more reliable allele frequency estimates, a larger sample of 112 unrelated individuals was genotyped for 14 SNPs. Of the 23 SNPs, three (-911G>A, -453T>G and -258A>G) had minor allele frequencies (MAF) of <0.01 and 20 SNPs had a MAF between 0.02 and 0.46 (Table <xref ref-type="table" rid="T2">2</xref>). The genotype frequencies of all SNPs were in HWE. Three SNPs (+50C>T, +959A>C and +1123C>T) altered the codons (Table <xref ref-type="table" rid="T2">2</xref>). However, only two SNPs (+50C>T and +959A>C) altered amino acid (Table <xref ref-type="table" rid="T2">2</xref>). The T allele at +50 altered the 17<sup>th </sup>amino acid from threonine to methionine and C allele at +959 altered the 34<sup>th </sup>amino acid from threonine to proline. The 17<sup>th </sup>amino acid is part of the signal peptide sequence, whereas the 34<sup>th </sup>amino acid is part of the proprotein.</p><fig position="float" id="F1"><label>Figure 1</label><caption><p>Locations of the 23 SNPs detected in the region from 1,173 bp upstream to 1,266 bp downstream of the translation start site of the PI3 gene. Dark shaded boxes represent the three exons and light shaded boxes represent introns. For more information on the SNPs, see Table 2.</p></caption><graphic xlink:href="1471-2350-7-49-1"/></fig><p>Thirteen of the 23 SNPs were in complete linkage disequilibrium (Table <xref ref-type="table" rid="T3">3</xref>). Altogether 16 haplotypes were identified (Table <xref ref-type="table" rid="T3">3</xref>). PI3_F was the most common haplotype followed by PI3_H and PI3_K.</p><table-wrap position="float" id="T3"><label>Table 3</label><caption><p>Haplotypes constructed with 23 SNPs in the PI3 gene.</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left">Haplotype</td><td align="center">Frequency</td><td align="center">SD<sup>a</sup></td><td align="center">-1077A>G<sup>b</sup></td><td align="center">-1067A>G<sup>b</sup></td><td align="center">-1063G>A</td><td align="center">-960T>Del<sup>b</sup></td><td align="center">-911G>A</td><td align="center">-868C>G</td><td align="center">-689C>G<sup>b</sup></td><td align="center">-675C>T<sup>b</sup></td><td align="center">-453T>G</td><td align="center">-338G>A<sup>b</sup></td><td align="center">-258A>G</td><td align="center">+50C>T<sup>b</sup></td><td align="center">+145C>A<sup>b</sup></td><td align="center">+162T>A</td><td align="center">+404C>G</td><td align="center">+408C>T<sup>b</sup></td><td align="center">+569C>G</td><td align="center">+627A>G<sup>b</sup></td><td align="center">+751T>A</td><td align="center">+959A>C<sup>b</sup></td><td align="center">+1123C>T</td><td align="center">+1226C>A<sup>b</sup></td><td align="center">+1234C>A<sup>b</sup></td></tr></thead><tbody><tr><td align="left">PI3_A</td><td align="center">0.042</td><td align="center">0.0318</td><td align="center">A</td><td align="center">A</td><td align="center">A</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">A</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_B</td><td align="center">0.021</td><td align="center">0.0258</td><td align="center">A</td><td align="center">A</td><td align="center">A</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_C</td><td align="center">0.021</td><td align="center">0.0154</td><td align="center">A</td><td align="center">A</td><td align="center">A</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">A</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_D</td><td align="center">0.021</td><td align="center">0.0132</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">G</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_E</td><td align="center">0.021</td><td align="center">0.0188</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">G</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_F</td><td align="center">0.437</td><td align="center">0.1006</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_G</td><td align="center">0.021</td><td align="center">0.0209</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">G</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_H</td><td align="center">0.125</td><td align="center">0.0896</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_I</td><td align="center">0.042</td><td align="center">0.0355</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_J</td><td align="center">0.021</td><td align="center">0.0199</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">G</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_K</td><td align="center">0.083</td><td align="center">0.0458</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">G</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">G</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_L</td><td align="center">0.021</td><td align="center">0.0156</td><td align="center">A</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">G</td><td align="center">G</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">G</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">G</td><td align="center">G</td><td align="center">T</td><td align="center">A</td><td align="center">C</td><td align="center">C</td><td align="center">C</td></tr><tr><td align="left">PI3_M</td><td align="center">0.063</td><td align="center">0.0328</td><td align="center">G</td><td align="center">G</td><td align="center">G</td><td align="center">Del</td><td align="center">G</td><td align="center">C</td><td align="center">G</td><td align="center">T</td><td align="center">T</td><td align="center">A</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">T</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">G</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">A</td></tr><tr><td align="left">PI3_N</td><td align="center">0.021</td><td align="center">0.0177</td><td align="center">G</td><td align="center">G</td><td align="center">G</td><td align="center">Del</td><td align="center">G</td><td align="center">C</td><td align="center">G</td><td align="center">T</td><td align="center">T</td><td align="center">A</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">T</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">G</td><td align="center">T</td><td align="center">C</td><td align="center">T</td><td align="center">A</td><td align="center">A</td></tr><tr><td align="left">PI3_O</td><td align="center">0.021</td><td align="center">0.0201</td><td align="center">G</td><td align="center">G</td><td align="center">G</td><td align="center">Del</td><td align="center">G</td><td align="center">C</td><td align="center">G</td><td align="center">T</td><td align="center">T</td><td align="center">A</td><td align="center">A</td><td align="center">T</td><td align="center">A</td><td align="center">T</td><td align="center">G</td><td align="center">T</td><td align="center">C</td><td align="center">G</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">A</td></tr><tr><td align="left">PI3_P</td><td align="center">0.021</td><td align="center">0.0176</td><td align="center">G</td><td align="center">G</td><td align="center">G</td><td align="center">Del</td><td align="center">G</td><td align="center">C</td><td align="center">G</td><td align="center">T</td><td align="center">T</td><td align="center">A</td><td align="center">G</td><td align="center">T</td><td align="center">A</td><td align="center">T</td><td align="center">C</td><td align="center">T</td><td align="center">C</td><td align="center">G</td><td align="center">T</td><td align="center">C</td><td align="center">C</td><td align="center">A</td><td align="center">A</td></tr></tbody></table><table-wrap-foot><p><sup>a</sup>Standard deviation.</p><p><sup>b</sup>These SNP loci are in complete linkage disequilibrium with each other.</p></table-wrap-foot></table-wrap></sec><sec><title>Effect of SNPs on protein binding</title><p>We performed <italic>in silico </italic>searches for putative transcription factor binding sites at the 11 SNP sites in the promoter region of the PI3 gene. Except for one SNP (-675C>T), all other sites showed potential differential binding for at least one transcription factor (Table <xref ref-type="table" rid="T4">4</xref>). In other words, the transcription factor was predicted to bind to one of the alleles, but not the other for these 10 SNPs.</p><table-wrap position="float" id="T4"><label>Table 4</label><caption><p>Results from <italic>in silico </italic>searches for putative transcription factor binding sites.</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="center" colspan="2"><bold>Allele</bold></td><td align="center" colspan="2"><bold>Predicted transcription factor(s)<sup>a</sup></bold></td></tr><tr><td></td><td colspan="2"><hr></hr></td><td colspan="2"><hr></hr></td></tr><tr><td align="left"><bold>Location</bold></td><td align="center"><bold>Major</bold></td><td align="center"><bold>Minor</bold></td><td align="left"><bold>Major allele</bold></td><td align="left"><bold>Minor allele</bold></td></tr></thead><tbody><tr><td align="left">-1077</td><td align="center">A</td><td align="center">G</td><td></td><td align="left">Adf-2a</td></tr><tr><td align="left">-1067</td><td align="center">A</td><td align="center">G</td><td></td><td align="left">TBF1</td></tr><tr><td align="left">-1063</td><td align="center">G</td><td align="center">A</td><td align="left">GCN4</td><td align="left">AP1</td></tr><tr><td align="left">-960</td><td align="center">T</td><td align="center">Deletion</td><td align="left">NFATC2</td><td></td></tr><tr><td align="left">-911</td><td align="center">G</td><td align="center">A</td><td align="left">SP1, AP1</td><td></td></tr><tr><td align="left">-868</td><td align="center">C</td><td align="center">G</td><td></td><td align="left">NRC3C1</td></tr><tr><td align="left">-689</td><td align="center">C</td><td align="center">G</td><td></td><td align="left">GATA1</td></tr><tr><td align="left">-675</td><td align="center">C</td><td align="center">T</td><td></td><td></td></tr><tr><td align="left">-453</td><td align="center">T</td><td align="center">G</td><td></td><td align="left">MAZ</td></tr><tr><td align="left">-338</td><td align="center">G</td><td align="center">A</td><td></td><td align="left">AP1</td></tr><tr><td align="left">-258</td><td align="center">A</td><td align="center">G</td><td align="left">NF1, NFE2, Zta</td><td></td></tr></tbody></table><table-wrap-foot><p><sup>a</sup>The transcription factors, whose binding site is predicted to change by the SNP, are listed here.</p></table-wrap-foot></table-wrap><p>To verify experimentally the differential binding of transcription factors, we conducted electrophoretic mobility shift assays (EMSA) for 10 SNP sites located in the promoter region. Due to the presence of a long stretch of AC-repeats, EMSA was not carried out for the SNP -868C>G. Of the 10 putative sites, six (-1077A>G, -1067A>G, -1063G>A, -960T>Del, -689C>G, -338G>A) showed differential binding by transcription factors in nuclear extracts derived from HeLa cells (not shown), while only two (-1063G>A and -689A>G) showed differential binding using amnion cell nuclear extract (Fig <xref ref-type="fig" rid="F2">2</xref>). Those SNP sites that did not show differential binding with HeLa cell nuclear extract, also did not show differential binding with amnion cell nuclear extracts. A transcription factor in HeLa cell nuclear extract bound to -960T>Del. There was however, no binding by a transcription factor in amnion cell nuclear extract to this same site. No transcription factor in either HeLa or amnion cell nuclear extract bound to the -258A>G site. Our interest was the transcriptional regulation of the PI3 gene in amnion cells [<xref ref-type="bibr" rid="B16">16</xref>]. We, therefore, focused on the two SNPs (-1063G>A and -689C>G, Fig <xref ref-type="fig" rid="F2">2</xref>) that showed differential binding by transcription factors derived from the amnion cell nuclear extract. For -1063A>G and -689C>G, the banding patterns representing the protein-DNA complexes were similar when using HeLa and amnion cell nuclear extracts, although the band intensities were lower with the latter, probably due to a lower concentration of functional proteins (Fig <xref ref-type="fig" rid="F2">2</xref>). To determine the specificity of the binding, we used a competition assay (Fig <xref ref-type="fig" rid="F3">3</xref>). The differential binding that persisted after cross-competition (100-fold) was considered to be due to the SNP. For -1063A>G, one protein-DNA complex persisted after a labeled double-stranded A-probe was competed with double-stranded sequence differing only at the SNP (G instead of A). For -689C>G, two protein-DNA complexes persisted after labeled double-stranded G-probe was competed with a double-stranded sequence differing only at the SNP (C instead of G) (Fig <xref ref-type="fig" rid="F3">3</xref>).</p><fig position="float" id="F2"><label>Figure 2</label><caption><p>EMSA showing the banding patterns with HeLa and amniotic cell nuclear extracts for -1063A>G and -689C>G sites. The arrows indicate protein-DNA complexes formed when transcription factors bind to their target sites.</p></caption><graphic xlink:href="1471-2350-7-49-2"/></fig><fig position="float" id="F3"><label>Figure 3</label><caption><p>EMSA showing self- and cross-competition for differential binding for -1063A>G and -689C>G sites with amniotic or HeLa cell nuclear extracts. The arrows indicate the differential binding that consistently persisted after cross-competition.</p></caption><graphic xlink:href="1471-2350-7-49-3"/></fig><p>Our <italic>in silico </italic>search predicted that AP1 was the transcription factor that would bind differentially at the -1063 SNP. To investigate this we used 100-fold excess of a competitor with the consensus sequence for AP1 binding or the anti-AP1 antibody in the reaction. No change in the banding pattern was observed in the competition assay (Fig <xref ref-type="fig" rid="F4">4</xref>). Similarly, no supershift with anti-c-jun or anti-c-fos antibody was observed for -1063G>A polymorphism using amnion or HeLa cell nuclear extracts (Fig <xref ref-type="fig" rid="F4">4</xref>). Since a positive control, using consensus AP1 binding sequence, demonstrated a supershift against anti-c-jun and anti-c-fos antibodies (Fig <xref ref-type="fig" rid="F4">4</xref>), a failure in the supershift was unlikely to be due to technical problems. We, therefore, concluded that the protein that binds to the A probe at nt -1063 does not contain the AP1 epitope.</p><fig position="float" id="F4"><label>Figure 4</label><caption><p>Results of competition and supershift experiments for the -1063A>G site using HeLa cell nuclear extract. S, shift; SS, supershift. Arrow on the left indicates a protein-DNA complex specific to the transcription factor binding to the A-allele.</p></caption><graphic xlink:href="1471-2350-7-49-4"/></fig><p>For the SNP at nt -689, the transcription factor, whose binding was predicted to change due to the SNP, was GATA1 (Table <xref ref-type="table" rid="T5">5</xref>). As shown in Fig <xref ref-type="fig" rid="F5">5</xref>, a consensus sequence containing the GATA1 binding site was able to compete with the -689G probe (Fig <xref ref-type="fig" rid="F5">5A</xref>) and a supershift was observed with anti-GATA1 antibody when using amniotic cell nuclear extract (Fig <xref ref-type="fig" rid="F5">5B</xref>) indicating that GATA1 binds to the G-allele of the -689C>G polymorphism in the promoter region of PI3 gene.</p><table-wrap position="float" id="T5"><label>Table 5</label><caption><p>Previously reported transcription factor binding sites in the PI3 promoter.</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left" colspan="2"><bold>Transcription factor</bold></td><td></td><td></td><td></td><td></td><td align="center" colspan="3"><bold>Evidence<sup>c</sup></bold></td></tr><tr><td colspan="2"><hr></hr></td><td></td><td></td><td></td><td></td><td colspan="3"><hr></hr></td></tr><tr><td align="left"><bold>Gene Symbol</bold></td><td align="left"><bold>Alias</bold></td><td align="left"><bold>Position<sup>a</sup></bold></td><td align="left"><bold>AL049767.12<sup>b</sup></bold></td><td align="left"><bold>Author</bold></td><td align="left"><bold>PubMedID</bold></td><td align="left"><bold>IS</bold></td><td align="left"><bold>E</bold></td><td align="left"><bold>DM</bold></td></tr></thead><tbody><tr><td align="left">NFKB1</td><td align="left">NFκB</td><td></td><td align="left">48564–48581</td><td align="left">King et al. 2003</td><td align="left">14521952</td><td align="left">+</td><td></td><td></td></tr><tr><td></td><td></td><td></td><td></td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td align="left">-479 – -470</td><td align="left">49058–49067</td><td align="left">King et al. 2003</td><td align="left">14521952</td><td align="left">+</td><td></td><td></td></tr><tr><td></td><td></td><td align="left">-340 – -331</td><td align="left">49197–49206</td><td align="left">King et al. 2003</td><td align="left">14521952</td><td align="left">+</td><td></td><td></td></tr><tr><td></td><td></td><td align="left">-164 – -153</td><td align="left">49373–49384</td><td align="left">Zhang 1995</td><td align="left">7780965</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td></td><td></td><td align="left">Bingle et al. 2001</td><td align="left">11472979</td><td></td><td align="left">+</td><td align="left">+</td></tr><tr><td></td><td></td><td></td><td></td><td align="left">King et al. 2003</td><td align="left">14521952</td><td align="left">+</td><td></td><td></td></tr><tr><td></td><td></td><td></td><td></td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td align="left">JUN</td><td align="left">AP1</td><td align="left">-545 – -537</td><td align="left">48992–49000</td><td align="left">Zhang 1995</td><td align="left">7780965</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td></td><td></td><td align="left">Zhang 1997</td><td align="left">9377579</td><td></td><td align="left">+</td><td align="left">+</td></tr><tr><td></td><td></td><td></td><td></td><td align="left">King et al. 2003</td><td align="left">14521952</td><td align="left">+</td><td></td><td></td></tr><tr><td></td><td></td><td></td><td></td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td align="left">-356 – -345</td><td align="left">49181–49192</td><td align="left">Sallenave et al. 1994</td><td align="left">7946401</td><td align="left">+</td><td></td><td></td></tr><tr><td></td><td></td><td></td><td></td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td align="left">SP1</td><td></td><td align="left">-82 – -74</td><td align="left">49455–49463</td><td align="left">Zhang 1995</td><td align="left">7780965</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td></td><td></td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td align="left">CEBPB</td><td align="left">NFIL6</td><td align="left">-386 – -378</td><td align="left">49151–49159</td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td align="left">-356 – -345</td><td align="left">49181–49192</td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td align="left">-307 – -299</td><td align="left">49230–49238</td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td align="left">-203 – -194</td><td align="left">49334–49343</td><td align="left">Sallenave et al. 1994</td><td align="left">7946401</td><td align="left">+</td><td></td><td></td></tr><tr><td></td><td></td><td></td><td></td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td></td><td></td><td align="left">-126 – -117</td><td align="left">49411–49420</td><td align="left">Pol et al. 2003</td><td align="left">12542536</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td align="left">OCT1</td><td></td><td align="left">-590 – -582</td><td align="left">48947–48955</td><td align="left">Zhang 1995</td><td align="left">7780965</td><td align="left">+</td><td></td><td align="left">+</td></tr><tr><td align="left">EST1</td><td align="left">PEA-3</td><td align="left">-484 – -479</td><td align="left">49053–49058</td><td align="left">Sallenave et al. 1994</td><td align="left">7946401</td><td align="left">+</td><td></td><td></td></tr></tbody></table><table-wrap-foot><p><sup>a</sup>Position is given relative to ATG.</p><p><sup>b</sup>We have mapped the location used in the studies to the current GenBank sequence to standardize the numbering between all studies.</p><p><sup>c</sup>IS, <italic>in silico</italic>; E, EMSA; DM, deletion mapping.</p></table-wrap-foot></table-wrap><fig position="float" id="F5"><label>Figure 5</label><caption><p>EMSA with amniotic cell nuclear extract for the nt -689C>G SNP site. (A) Competition with -689C, -689G and GATA consensus sequences. The arrows indicate protein-DNA complexes which consistently persisted after cross-competition. (B) Supershift experiment. S, shift; SS, supershift.</p></caption><graphic xlink:href="1471-2350-7-49-5"/></fig></sec></sec><sec><title>Discussion</title><p>We observed a high degree of polymorphism within the PI3 gene with 23 SNPs detected, 11 of which were located in the promoter region. We found an amino acid substitution, T34P, in the 4<sup>th </sup>amino acid of the amino terminal-transglutaminase substrate domain, GQDPVK, of PI3. To determine if this SNP has a significant effect on the function of this domain, we searched for the consensus sequence of the transglutaminase substrate in other mammals. A similar sequence domain was identified in seminal vesicle protein I (Semg1) repeats in guinea pig (PROSITE documentation PDOC000282). Semg1 is a clotting protein that serves as the substrate in the formation of the copulatory plug [<xref ref-type="bibr" rid="B30">30</xref>]. Covalent clotting of this protein is catalyzed by a transglutaminase and involves the formation of γ-glutamyl-ε-lysine crosslinks. The consensus signature of this consensus sequence was [IVM]-X-G-Q-D-X-V-K-X<sub>5</sub>- [KN]-G-X<sub>3</sub>- [STLV]. The 6<sup>th </sup>amino acid from the left of this pattern, X, corresponds to the 34<sup>th </sup>amino acid of the PI3 protein, and appears not to be conserved.</p><p>Putative binding of several transcription factors in the promoter region of PI3 gene have been previously reported (Table <xref ref-type="table" rid="T5">5</xref>) [<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B14">14</xref>,<xref ref-type="bibr" rid="B15">15</xref>,<xref ref-type="bibr" rid="B31">31</xref>]. Except for two, none of these sites were polymorphic in our study. The two consensus sequences of NFKB1 at nt -964 to -956 and nt -340 to -331 identified in previous studies [<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B9">9</xref>] contained the -960T>Del and -338G>A SNP sites, respectively, in our study. Based on our <italic>in silico </italic>analysis, the binding of transcription factors NFATC2 and AP1 could be altered by the sequence changes at -960T>Del and -338G>A (Table <xref ref-type="table" rid="T4">4</xref>), but with EMSA we did not observe any differences between the two alleles using amnion cell nuclear extract. For -960T>Del we did not observe any binding to a transcription factor using amnion cell extract.</p><p>We identified 10 SNPs with alleles that were predicted to have different binding sites for one or more transcription factors by <italic>in silico </italic>searches (Table <xref ref-type="table" rid="T4">4</xref>). We tested the predicted differential binding at these sites by EMSA with both HeLa and amnion cell nuclear extracts. HeLa cells are used widely for studying the functionality of promoter polymorphisms. Since PI3 mRNA is down-regulated in chorioamniotic membranes of patients with PPROM [<xref ref-type="bibr" rid="B16">16</xref>], we also used a nuclear extract derived from amnion cells. Six of the 10 sites exhibited differential binding to transcription factors with the HeLa extract, in contrast to only two with amnion cell nuclear extract. Since we did not observe supershift at -1063G>A using antibody against AP1 and nuclear extract from either HeLa or amniotic cell line, it is likely that the differential binding was to a transcription factor other than AP1 or the antibody did not have the specific epitope. The presence of AP1 in the nuclear extract of both cell lines was confirmed by the supershift seen when using AP1 consensus probe. We demonstrated the binding of GATA1 to the G allele at the -689C>G site by supershift with an antibody against GATA1.</p><p>These findings suggest the involvement of GATA1 in the transcriptional regulation of PI3 gene in amnion cells and provide a possible genetic explanation for the downregulation of PI3 in chorioamniotic membranes from PPROM cases. We have previously demonstrated that the levels of neutrophil elastase (ELA2, [LocusID: 1991]) are increased in the amniotic fluid of patients with PPROM and acute chorioamnionitis [<xref ref-type="bibr" rid="B32">32</xref>]. It is therefore plausible to speculate that the production of PI3 in the fetal membranes is to protect the tissue from the damage that could be caused by increased amounts of neutrophil elastase. Our recent study [<xref ref-type="bibr" rid="B16">16</xref>] showing decreased expression of PI3 in the chorioamniotic membranes from patients with PPROM supports our hypothesis that patients who are not capable of producing adequate amounts of PI3 may be predisposed to PPROM.</p><p>It has been suggested that PI3 is involved in the pathophysiology of many clinical conditions. For example, PI3 was found in the epidermis of patients with psoriasis, but not in normal human epidermis [<xref ref-type="bibr" rid="B33">33</xref>]. Higher levels of PI3 were also observed in bronchial secretions from patients with chronic obstructive pulmonary disease [<xref ref-type="bibr" rid="B2">2</xref>] and bronchial carcinoma [<xref ref-type="bibr" rid="B7">7</xref>], and the expression of PI3 was decreased in breast [<xref ref-type="bibr" rid="B9">9</xref>] and in epidermal tumors [<xref ref-type="bibr" rid="B8">8</xref>,<xref ref-type="bibr" rid="B9">9</xref>]. The SNPs identified here will likely be useful for studying the molecular mechanisms of these diseases.</p></sec><sec><title>Conclusion</title><p>A high degree of polymorphism was detected in the PI3 gene with 23 SNPs, 11 of which are in the promoter region. Two SNP sites (-1063G>A and -689C>G) showed differential binding of transcription factors in nuclear extracts derived from both amnion and HeLa cells suggesting possible involvement of these two SNPs in the expression of PI3 gene. As the SNP site at -1063G>A did not bind to the transcription factor AP1 as suggested by <italic>in silico </italic>search, the bound transcription factor may not be in current database and needs to be characterized. Binding of GATA1 to the G allele at the -689C>G site suggests the involvement of GATA1 in the transcriptional regulation of PI3 gene in amnion cells. We have performed a genetic association study with PI3 variants, including the -689C>G variant, and found that it is associated with PPROM [manuscript in preparation]. We also previously demonstrated by immunohistochemistry that many cell types of the chorioamniotic membranes produce PI3 and that PI3 protein is decreased in chorioamniotic membranes from PPROM cases [<xref ref-type="bibr" rid="B16">16</xref>]. Together, these lines of evidence provide a plausible genetic explanation for the down regulation of PI3 in chorioamniotic membranes from PPROM cases. Previously the involvement of PI3 in the pathophysiology of many clinical conditions was suggested. The SNPs identified here provide the tools for studying the molecular mechanism of these diseases.</p></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>Mahboob A. Chowdhury designed and interpreted experiments; carried out sequencing, EMSAs and <italic>in silico </italic>searches; drafted the manuscript and approved the final version of it.</p><p>Helena Kuivaniemi designed and interpreted experiments; carried out PCRs; drafted the manuscript and approved the final version of it.</p><p>Roberto Romero provided overall direction to the project; was responsible for the clinical data, revised manuscript and approved the final version of it.</p><p>Samuel Edwin established amniotic cell cultures; revised manuscript and approved the final version of it.</p><p>Tinnakorn Chaiworapongsa provided clinical data and approved the final version of the manuscript.</p><p>Gerard Tromp designed and interpreted experiments; carried out all statistical analyses; provided overall direction to the project; drafted and revised the manuscript as well as approved the final version of it.</p></sec><sec><title>Pre-publication history</title><p>The pre-publication history for this paper can be accessed here:</p><p><ext-link ext-link-type="uri" xlink:href="http://www.biomedcentral.com/1471-2350/7/49/prepub"/></p></sec>
Do distribution volumes and clearances relate to tissue volumes and blood flows? A computer simulation	<sec><title>Background</title><p>Kinetics of inhaled agents are often described by physiological models. However, many pharmacokinetic concepts, such as context-sensitive half-times, have been developed for drugs described by classical compartmental models. We derived classical compartmental models that describe the course of the alveolar concentrations (F<sub>A</sub>) generated by the physiological uptake and distribution models used by the Gas Man<sup>® </sup>program, and describe how distribution volumes and clearances relate to tissue volumes and blood flows.</p></sec><sec sec-type="methods"><title>Methods</title><p>Gas Man<sup>® </sup>was used to generate F<sub>A </sub>vs. time curves during the wash-in and wash-out period of 115 min each with a high fresh gas flow (8 L.min<sup>-1</sup>), a constant alveolar minute ventilation (4 L.min<sup>-1</sup>), and a constant inspired concentration (F<sub>I</sub>) of halothane (0.75%), isoflurane (1.15%), sevoflurane (2%), or desflurane (6%). With each of these F<sub>I</sub>, simulations were ran for a 70 kg patient with 5 different cardiac outputs (CO) (2, 3, 5, 8 and 10 L.min<sup>-1</sup>) and for 5 patients with different weights (40, 55, 70, 85, and 100 kg) but the same CO (5 L.min<sup>-1</sup>). Two and three compartmental models were fitted to F<sub>A </sub>of the individual 9 runs using NONMEM. After testing for parsimony, goodness of fit was evaluated using median prediction error (MDPE) and median absolute prediction error (MDAPE). The model was tested prospectively for a virtual 62 kg patient with a cardiac output of 4.5 L.min<sup>-1 </sup>for three different durations (wash-in and wash-out period of 10, 60, and 180 min each) with an F<sub>I </sub>of 1.5% halothane, 1.5% isoflurane, sevoflurane 4%, or desflurane 12%.</p></sec><sec><title>Results</title><p>A three-compartment model fitted the data best (MDPE = 0% and MDAPE ≤ 0.074%) and performed equally well when tested prospectively (MDPE ≤ 0.51% and MDAPE ≤ 1.51%). The relationship between CO and body weight and the distribution volumes and clearances is complex.</p></sec><sec><title>Conclusion</title><p>The kinetics of anesthetic gases can be adequately described e by a mammilary compartmental model. Therefore, concepts that are traditionally thought of as being applicable to the kinetics of intravenous agents can be equally well applied to anesthetic gases. Distribution volumes and clearances cannot be equated to tissue volumes and blood flows respectively.</p></sec>	<contrib id="A1" equal-contrib="yes" corresp="yes" contrib-type="author"><name><surname>Hendrickx</surname><given-names>Jan FA</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" equal-contrib="yes" contrib-type="author"><name><surname>Lemmens</surname><given-names>Hendrikus JM</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" equal-contrib="yes" contrib-type="author"><name><surname>Shafer</surname><given-names>Steven L</given-names></name><xref ref-type="aff" rid="I1">1</xref><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib>	BMC Anesthesiology	<sec><title>Background</title><p>Like intravenous anesthetics, inhaled anesthetic agents can and are being safely administered without the knowledge of any pharmacokinetic model. In addition, the MAC and MAC awake concept link the end-expired concentrations (F<sub>A</sub>) with the clinically two most relevant endpoints, immobility and unconsciousness. Kinetic models can nevertheless be useful. Models are used to explain differences in wash-in and wash-out characteristics of different agents that may be relevant to clinical practice. Uptake models have also been used to try to predict the relationship between fresh gas flow, and delivered, inspired (F<sub>I</sub>), and end-expired concentrations of anesthetic agents in the commonly used circle breathing system [<xref ref-type="bibr" rid="B1">1</xref>]. Once a kinetic model has been developed, it can be linked to clinically relevant endpoints (immobility and unconsciousness) by incorporating an effect site compartment.</p><p>Physiological models are appealing because they can be understood in terms of anatomy, physiology, and physics. The kinetics of inhaled agents are commonly described using physiological modeling because of the availability of values for tissue partition coefficients, tissue volumes and organ blood flows. Yet many assumptions underlie these models, such as the absence of an arterial-end-expired gradient or the assumption that uptake and distribution in an organ is uniform and instantaneous as implicitly assumed by flow limited uptake and the use of a single value for tissue partition coefficients [<xref ref-type="bibr" rid="B2">2</xref>]. In addition, it is mathematically impossible to accurately fit all these parameters to "uptake" as it is clinically available in the operating room: the product of inspired concentration (F<sub>I</sub>) and ventilation minus the product of the end-expired concentration (F<sub>A</sub>) and ventilation. A parsimonious empirical polyexponential model ("classical compartmental modeling") similar to that used to describe the course of the concentration over time of intravenous agents may be better suited to analyze the partial pressure (or infusion rate) of potent inhaled anesthetics over time [<xref ref-type="bibr" rid="B3">3</xref>-<xref ref-type="bibr" rid="B5">5</xref>]. The first goal of this manuscript is to examine whether the kinetics of four different anesthetic agents generated by a physiological uptake model (used by the Gas Man<sup>® </sup>program, Med Man Simulations, Inc., Chestnut Hill, MA) can be equally well described by a mammilary compartmental model (derived using non-linear mixed effect modeling or NONMEM) [<xref ref-type="bibr" rid="B6">6</xref>]. If so, concepts applied to intravenous agents, like the context-sensitive half time, would also be applicable to inhaled agents. The derived compartmental model will be prospectively tested against the physiological model used by the Gas Man<sup>® </sup>program. More detail about both models is provided in the Appendix. The second goal of this manuscript is to address the controversial issue whether or to what an extent distribution volumes and clearances relate to tissue volumes and blood flows. We therefore examined whether distribution volumes and clearances match the tissue volumes and blood flows from the physiological model used to generate the data from which the parameters of the compartmental model were derived.</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Part I: Deriving the kinetic parameters of the compartmental model</title><p>Gas Man<sup>® </sup>was used to generate F<sub>A </sub>for 4 different agents (halothane, isoflurane, sevoflurane, and desflurane) during a wash-in and washout period of 115 min each with a high fresh gas flow (8 L.min-1) and a constant alveolar minute ventilation (4 L.min<sup>-1</sup>, regardless of patient weight). The default settings were used for the circuit and tissue composition. The inspired concentrations of halothane, isoflurane, sevoflurane, and desflurane were maintained constant during the wash-in period at 0.75; 1.15; 2; and 6% respectively. With each of these inspired concentrations, simulations were run for a 70 kg patient with 5 different cardiac outputs (2, 3, 5, 8 and 10 L.min<sup>-1</sup>) and for 5 patients with different weights (40, 55, 70, 85, and 100 kg) but the same cardiac output (5 L.min<sup>-1</sup>). This yielded 9 unique simulations for each agent (note that each group includes a 70 kg patient with a cardiac output of 5 L.min<sup>-1</sup>, which is why there were not 10 unique simulations). Two and three compartmental models were fitted to F<sub>A </sub>of the individual 9 simulations using NONMEM (the original data and the NONMEM control code are available in the Web supplement). The total administered dose was entered in the NONMEM data file as the product of the inspired concentration and duration. A naïve pooled data fit was done to obtain starting values for the structural parameters (distribution volumes and clearances) for each of the four agents. The effect of weight and cardiac output on each of the structural parameters (covariate analysis) was assessed using the minimum objective function of NONMEM, -2 × Log Likelihood (-2LL), to decide whether to accept or omit a covariate. The results were then tested for parsimony by sequentially deleting each of the covariate parameters using the NONMEM objective function as the criteria to accept or delete any covariate parameter. Goodness of fit was evaluated using median prediction error (MDPE) and median absolute prediction error (MDAPE) as a measure for bias and accuracy respectively [<xref ref-type="bibr" rid="B7">7</xref>]. MDPE and MDAPE were calculated as (F<sub>A</sub>Gas Man - F<sub>A</sub>3compartmental model)/F<sub>A</sub>3compartmental model) and \|(F<sub>A</sub>Gas Man - F<sub>A</sub>3compartmental model)/F<sub>A</sub>3compartmental model\| respectively, and are expressed in %.</p></sec><sec><title>Part II: Prospective testing of the compartmental model</title><p>Using the previously derived parameters with their covariates, the compartmental model was tested prospectively by having it generate F<sub>A </sub>for a patient with parameters different from those from which the model was derived. The F<sub>A </sub>output for a 62 kg patient with a cardiac output of 4.5 L.min<sup>-1 </sup>for three different durations (wash-in and wash-out period of 10, 60, and 180 min each) with an F<sub>I </sub>of 1.5% halothane, 1.5% isoflurane, sevoflurane 4%, or desflurane 12% was calculated by converting the volume and clearance parameters, including any covariate effects, into the coefficients and exponents defining the three exponential functions that together describe the F<sub>A </sub>course. Conversions were done with the Excel spreadsheet CONVERT.XLS [<xref ref-type="bibr" rid="B8">8</xref>]. The same patient parameter set was entered into Gas Man<sup>® </sup>and the respective end-expired concentrations were obtained. The predictive performance of the compartmental model was tested using MDPE and MDAPE as described above. MDPE and MDAPE less than 20% have been considered acceptable. All data and control files are available upon simple request to the authors (JHX).</p></sec></sec><sec><title>Results</title><sec><title>Part I</title><p>A three compartment model best described the F<sub>A </sub>course. The distribution volumes and clearances for each of the initial 9 simulation runs are presented in figure <xref ref-type="fig" rid="F1">1</xref> and <xref ref-type="fig" rid="F2">2</xref>. The pharmacokinetic parameters for the three compartmental model with cardiac output and weight as covariates (tested for parsimony) are presented in (Table <xref ref-type="table" rid="T1">1</xref> and <xref ref-type="table" rid="T2">2</xref>). As shown in table <xref ref-type="table" rid="T2">2</xref>, the models described the data almost perfectly (MDPE = 0% and MDAPE ≤ 0.074%), indicating that a 3 compartment mammilary model is sufficient to capture the pharmacokinetic behavior of the anesthetic gases in a high flow open circuit.</p><fig position="float" id="F1"><label>Figure 1</label><caption><p><bold>Structural parameters for each of the four agents in a 70 kg patient with a cardiac output ranging from 2 to 10 L.min-1 for a 115 min wash-in and wash-out period</bold>. * : denotes use of a second Y-axis for halothane (for clarity).</p></caption><graphic xlink:href="1471-2253-6-7-1"/></fig><fig position="float" id="F2"><label>Figure 2</label><caption><p>Structural parameters for each of the four agents in a patient with a cardiac output of 5 L.min-1 and weight ranging from 40 to 100 kg for a 115 min wash-in and wash-out period.</p></caption><graphic xlink:href="1471-2253-6-7-2"/></fig><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>The pharmacokinetic parameters for the three compartmental model with cardiac output (CO, L.min-1) and weight (WT, kg) as covariates (tested for parsimony).</p></caption><table frame="hsides" rules="groups"><thead><tr><td align="left"><bold>Halothane</bold></td></tr></thead><tbody><tr><td align="left"> V1 = 5.65 + 0.020(WT-70) + 0.68(CO-5)</td></tr><tr><td align="left"> V2 = 8.35 + 0.15(WT-70) - 0.49(CO-5)</td></tr><tr><td align="left"> V3 = 134.09 + 1.91(WT-70)</td></tr><tr><td align="left"> CL1 = 1.23 + 0.050(CO-5)</td></tr><tr><td align="left"> CL2 = 2.57 + 0.010(WT-70) + 0.47(CO-5)</td></tr><tr><td align="left"> CL3 = 0.87 + 0.17(CO-5)</td></tr><tr><td colspan="1"><hr></hr></td></tr><tr><td align="left"><bold>Isoflurane</bold></td></tr><tr><td colspan="1"><hr></hr></td></tr><tr><td align="left"> V1 = 4.21 + 0.020(WT-70) + 0.39(CO-5)</td></tr><tr><td align="left"> V2 = 3.16 + 0.065(WT-70) - 0.31(CO-5)</td></tr><tr><td align="left"> V3 = 61.40 + 0.87(WT-70)</td></tr><tr><td align="left"> CL1 = 1.12 + 0.024(CO-5)</td></tr><tr><td align="left"> CL2 = 1.05 + 0.0037(WT-70) + 0.17(CO-5)</td></tr><tr><td align="left"> CL3 = 0.46 + 0.092(CO-5)</td></tr><tr><td colspan="1"><hr></hr></td></tr><tr><td align="left"><bold>Sevoflurane</bold></td></tr><tr><td colspan="1"><hr></hr></td></tr><tr><td align="left"> V1 = 3.33 + 0.018(WT-70) + 0.21(CO-5)</td></tr><tr><td align="left"> V2 = 1.51 + 0.033(WT-70) - 0.18(CO-5)</td></tr><tr><td align="left"> V3 = 33.69 + 0.47(WT-70)</td></tr><tr><td align="left"> CL1 = 1.05 + 0.011(CO-5)</td></tr><tr><td align="left"> CL2 = 0.37 + 0.0034(WT-70) + 0.053(CO-5)</td></tr><tr><td align="left"> CL3 = 0.23 + 0.046(CO-5)</td></tr><tr><td colspan="1"><hr></hr></td></tr><tr><td align="left"><bold>Desflurane</bold></td></tr><tr><td colspan="1"><hr></hr></td></tr><tr><td align="left"> V1 = 2.94 + 0.017(WT-70) + 0.14(CO-5)</td></tr><tr><td align="left"> V2 = 1.56 + 0.030(WT-70) - 0.14(CO-5)</td></tr><tr><td align="left"> V3 = 14.08 + 0.19(WT-70)</td></tr><tr><td align="left"> CL1 = 1.03 + 0.0066(CO-5)</td></tr><tr><td align="left"> CL2 = 0.25 + 0.0022(WT-70) + 0.036(CO-5)</td></tr><tr><td align="left"> CL3 = 0.14 + 0.029(CO-5)</td></tr></tbody></table><table-wrap-foot><p>V = distribution volume (L), CL = clearance (L.min-1), with numbers 1, 2, and 3 referring to the first, second, and third compartment.</p></table-wrap-foot></table-wrap><table-wrap position="float" id="T2"><label>Table 2</label><caption><p>Goodness of fit for the predictions by three compartment model predictions during model derivation and prospective testing.</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="left" colspan="2">Model derivation</td><td align="left" colspan="2">Prospective testing</td></tr></thead><tbody><tr><td></td><td align="left">MDPE</td><td align="left">MDAPE</td><td align="left">MDPE</td><td align="left">MDAPE</td></tr><tr><td colspan="5"><hr></hr></td></tr><tr><td align="left">Halothane</td><td align="left">0.00</td><td align="left">0.18</td><td align="left">0.31</td><td align="left">0.38</td></tr><tr><td align="left">Isoflurane</td><td align="left">0.00</td><td align="left">0.29</td><td align="left">0.38</td><td align="left">0.52</td></tr><tr><td align="left">Sevoflurane</td><td align="left">0.00</td><td align="left">0.63</td><td align="left">0.27</td><td align="left">0.73</td></tr><tr><td align="left">Desflurane</td><td align="left">0.00</td><td align="left">0.74</td><td align="left">0.51</td><td align="left">1.51</td></tr></tbody></table><table-wrap-foot><p>MDPE = median prediction error; MDAPE = median absolute prediction error. Values are expressed in %.</p></table-wrap-foot></table-wrap></sec><sec><title>Part II</title><p>F<sub>A </sub>of each of the four agents generated by the compartmental model and the corresponding values from Gas Man<sup>® </sup>are illustrated in figure <xref ref-type="fig" rid="F3">3</xref>. Goodness of fit was excellent (MDPE ≤ 0.51% and MDAPE ≤ 1.51%) (Table <xref ref-type="table" rid="T2">2</xref>).</p><fig position="float" id="F3"><label>Figure 3</label><caption><p>Prospective testing of the model: the end-expired concentrations (FA) predicted by the compartmental model (full line) and the corresponding values simulated by Gas Man<sup>® </sup>(open circles) for halothane (A), isoflurane (B), sevoflurane (C), and desflurane (D).</p></caption><graphic xlink:href="1471-2253-6-7-3"/></fig></sec></sec><sec><title>Discussion</title><p>The time course of F<sub>A </sub>of a potent inhaled anesthetic generated by a physiological uptake model (Gas Man<sup>®</sup>) can be equally well described using a classical three compartmental model derived from data generated by the very same physiological model (Gas Man<sup>®</sup>). While the relationship between the distribution volumes and clearances and organ capacities and tissue blood flows is complex, it is nevertheless intuitively tempting to try to explain e.g. the associations between a decrease in cardiac output and the changes in the volumes of distribution in terms of alterations in tissue blood flow of distributions. Yasuda and colleagues used NONMEM derived classical compartmental parameters from end-expired concentrations and related the parameters of the exponential curves (exponents and time constants) to known physiological processes and anatomical compartments [<xref ref-type="bibr" rid="B9">9</xref>]. These parameters were subsequently used to help delineate several physiological compartments: the first, second, third, fourth, and fifth compartment were interpreted as representing lungs, vessel rich group, muscle group, intertissue diffusion, and the fat group. Hull, however, acclaims that while it is often suggested that some particular tissue or organ (such as the brain) be 'in' one compartment or another, such suggestions are ill-founded because parameters of the fit to the uptake data contain no information that might support such assumptions [<xref ref-type="bibr" rid="B10">10</xref>,<xref ref-type="bibr" rid="B11">11</xref>]. Similarly, Wissing argues that a precise allocation of several hypothetical peripheral compartments to anatomical defined tissues is hardly feasible [<xref ref-type="bibr" rid="B2">2</xref>,<xref ref-type="bibr" rid="B3">3</xref>].</p><p>If it is true that distribution volumes and clearances can be directly interpreted in terms of tissue volumes and tissue blood flows, a change in weight or cardiac output that is known to cause proportional changes in tissue volumes or tissue blood flow in the (physiological) model of Gas Man<sup>® </sup>should also cause proportional changes in the distribution volumes and clearances of the empirical compartmental (NONMEM) model. In our current simulations, an increase in cardiac output and weight does indeed increase clearance and distribution volume (Figures <xref ref-type="fig" rid="F1">1</xref> and <xref ref-type="fig" rid="F2">2</xref>). However, this type of relationship is not consistent. An increase in cardiac output is associated with a decrease in the clearance of V2 (the second compartment) with sevoflurane and isoflurane. In addition, the non-linear (almost hyperbolic) relationship between V3 (the third compartment) and cardiac output for all four gases cannot be interpreted in terms of an underlying physiological model or process. Therefore, the parameters of the underlying physiological model (organ capacities and blood flows) and those of an empirical model (distribution volumes and clearances) therefore cannot be interpreted in terms of one another.</p><p>We have shown that the kinetics of anesthetic gases can be described by a mammilary model. Therefore, concepts that are traditionally thought of as being applicable to the kinetics of intravenous agents can be equally well applied to anesthetic gases. One example of one of these concepts are context sensitive half-times, which interestingly had been described for inhaled agents before the term context sensitive half-time was introduced but were labeled "coasting times" by Lowe and Ernst [<xref ref-type="bibr" rid="B12">12</xref>,<xref ref-type="bibr" rid="B13">13</xref>].</p><p>This study does not address when or whether to use physiological modeling or classical compartmental modeling. Either modeling approach adequately describes and predicts the course of F<sub>A</sub>, but their parameters cannot be interpreted in terms of one another. A three compartment model can be a good fit of a concentration that is effectively the sum of a multi-organ (physiological) system, where each organ displays first order kinetics [<xref ref-type="bibr" rid="B14">14</xref>]. This study does not address the validity of the GasMan<sup>® </sup>model either. Even though GasMan<sup>® </sup>can be used to gain additional insights into the kinetic of inhaled agents, clinical validation of the GasMan<sup>® </sup>model is lacking.</p></sec><sec><title>Conclusion</title><p>We describe how non-linear mixed-effect modeling can be used to derive a three compartmental empirical model to describe the kinetic behavior of potent inhaled anesthetic generated by a physiological model (Gas Man<sup>®</sup>). Therefore, concepts that are traditionally thought of as being applicable to the kinetics of intravenous agents can be equally well applied to anesthetic gases. The relationship between the distribution volumes and clearances of the empirical model and tissue capacities and blood flow of the physiological model are complex, however. Our study reinforces the fact that the compartmental models (even with co-variate models) are an empirical description of a data set, and that the parameters have virtually no physiological interpretation.</p></sec><sec><title>Appendix. Graphical display and formulas of the empirical three-compartmental model (NONMEM) and the physiological model of Gas Man<sup>®</sup></title><sec><title>The empirical three-compartmental model (NONMEM)</title><p>With a three compartmental model, the time course of the drug's concentration can be described in three mathematically equivalent ways: (a) three volumes and three clearances; (b) five rate constants and a scaling factor; or (c) a tri-exponential equation. The volume-clearance scheme is useful to visualize how the drug moves throughout the body (figure <xref ref-type="fig" rid="F4">4A</xref>). In the rate constant-scaling factor scheme (figure <xref ref-type="fig" rid="F4">4B</xref>) each "micro-rate constant" k<sub>ij </sub>defines the rate of drug transfer from one compartment i to another compartment j. V<sub>1 </sub>is the scaling factor; V<sub>2 </sub>and V<sub>3 </sub>are not independent parameters here. The mathematical form of the three-compartment exponential equation is most commonly used (figure <xref ref-type="fig" rid="F4">4C</xref>). A, B, and C are called "coefficients"; α, β, and γ (occasionally called λ<sub>1</sub>, λ<sub>2</sub>, and λ<sub>3</sub>) are "exponents" or "hybrid rate constants". While the volume-clearance and the rate constant-scaling factor scheme parameters are mathematically easily converted into one another [<xref ref-type="bibr" rid="B8">8</xref>], the coefficients and exponentials of the tri-exponential fit are related to the volumes and clearances and to the rate constants and scaling factor in a slightly more complex manner.</p><fig position="float" id="F4"><label>Figure 4</label><caption><p>A. The three compartment empirical model: the volume-clearance scheme. B. The three compartment empirical model: the rate constant-scaling factor scheme. C. The three compartment empirical model: the mathematical three exponential scheme.</p></caption><graphic xlink:href="1471-2253-6-7-4"/></fig><p>By using a sum of least squares method (the minimum objective function or MOF), NONMEM determines whether a one, two or three compartmental model best describes the concentration course. Depending on the type of control file used, NONMEM analysis will yield either the exponents and coefficients or distribution volumes and clearances. The best fitting model is then optimized by mathematically linking patient covariates (e.g. weight, height, cardiac output) to those parameters, again using the MOF to decide whether or not a covariate significantly improves the fit.</p></sec><sec><title>The physiological model of Gas Man<sup>®</sup></title><p>The building blocks for the physiological model of Gas Man<sup>® </sup>are also labeled "compartments". In Gas Man<sup>®</sup>, the patient is modeled as four compartments: the lung, vessel rich group, muscle and fat; the circuit is a fifth. Each patient compartment refers to a group of tissues for which uptake can be described by a differential equation if tissue volume, tissue flow, and tissue solubility are known. The model assumes that anesthetic uptake by an organ is a perfusion-limited, exponential process. Organ vapor capacity depends on the size of the organ and the solubility of the agent in that particular organ. The rate at which the partial pressure in the organ increases, and eventually saturates (=equals the partial pressure in the arterial blood), depends both on the organ's capacity (the more can be stored at the same partial pressure, the longer this will take), and on organ blood flow (the higher blood flow, the faster it will equilibrate with the arterial partial pressure, and therefore the faster it saturates). Based on both tissue storage capacity and blood flow, organs and tissue groups are grouped into the VRG, MG, and FG group (vessel rich group, muscle group, and fat group, respectively). The Gas Man<sup>® </sup>simulation uses Euler's method of solution for the simultaneous differential equations, with linear coefficients that govern the five-compartment system. The model utilizes standard values for organ volumes, anesthetic solubilities, and regional blood flows. An example of the actual display is depicted in figure <xref ref-type="fig" rid="F5">5</xref>.</p><fig position="float" id="F5"><label>Figure 5</label><caption><p>The Gas Man<sup>® </sup>graphical interface.</p></caption><graphic xlink:href="1471-2253-6-7-5"/></fig></sec></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>All authors have contributed equally to all parts of the manuscript.</p></sec><sec><title>Pre-publication history</title><p>The pre-publication history for this paper can be accessed here:</p><p><ext-link ext-link-type="uri" xlink:href="http://www.biomedcentral.com/1471-2253/6/7/prepub"/></p></sec>
Heterogeneous dermatitis complaints after change in drinking water treatment: a case report	<sec><title>Background</title><p>The disinfectant monochloramine minimizes the formation of potentially hazardous and regulated byproducts, and many drinking water utilities are shifting to its use.</p></sec><sec><title>Case presentation</title><p>After a drinking water utility serving 2.4 million people switched to monochloramine for residual disinfection, a small number of residents complained of dermatitis reactions. We interviewed 17 people about their symptoms. Skin appearance, symptoms, and exposures were heterogeneous. Five respondents had history of hives or rash that preceded the switch to monochloramine.</p></sec><sec><title>Conclusion</title><p>The complaints described were heterogeneous, and many of the respondents had underlying or preexisting conditions that would offer plausible alternative explanations for their symptoms. We did not recommend further study of these complaints.</p></sec>	<contrib id="A1" equal-contrib="yes" corresp="yes" contrib-type="author"><name><surname>Weintraub</surname><given-names>June M</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" equal-contrib="yes" contrib-type="author"><name><surname>Berger</surname><given-names>Magdalena</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" contrib-type="author"><name><surname>Bhatia</surname><given-names>Rajiv</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib>	Environmental Health	<sec><title>Background</title><p>Disinfection of public water supplies protects public health by inactivating microbial pathogens. Byproducts of disinfection with chlorine have been associated with bladder and rectal cancers and to adverse reproductive outcomes [<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B2">2</xref>]. Because the disinfectant monochloramine minimizes the formation of potentially hazardous and regulated byproducts, many drinking water utilities are shifting to its use [<xref ref-type="bibr" rid="B3">3</xref>].</p><p>In February 2004 a water utility serving 2.4 million people in northern California replaced chlorine with monochloramine for secondary disinfection. Subsequently a small number of water customers raised concerns about skin rashes, attributing these rashes to the change in disinfection method. Skin complaints associated with water are not uncommon [<xref ref-type="bibr" rid="B4">4</xref>,<xref ref-type="bibr" rid="B5">5</xref>]. We are not aware of any previous work investigating this type of reaction as a specific response to the presence of monochloramine in the water supply. Dermatitis relating to water treatment is reported in two studies; one used a broad case definition [<xref ref-type="bibr" rid="B6">6</xref>], and the other revealed that the perception of change in water treatment was principally responsible for the symptoms, rather than any actual change in the water treatment [<xref ref-type="bibr" rid="B7">7</xref>]. Neither was specific to monochloramine.</p><p>In this context, we identified several possible explanations for the skin complaints that we received, including the following: (1) the symptoms were the result of underlying or preexisting conditions; (2) patients developed aquagenic pruritis or aquagenic pruritus of the elderly [<xref ref-type="bibr" rid="B4">4</xref>] independent of the change in water treatment and patients reported their symptoms knowing about a reported change in water treatment; and/or (3) reported symptoms were indeed the result of the change in water treatment. In order to gain insight into these hypotheses and evaluate the need for an epidemiologic investigation, local public health departments cooperated to develop a questionnaire to assess the homogeneity of the complaints. We hypothesized that homogeneity among the complaints might provide justification for a cross-sectional study of water customers; alternatively if we could not identify homogeneity, this might indicate the lack of a common cause, reducing the pressure for further investigation.</p><p>The questionnaire was administered between September 2004 and January 31 2005 to individuals who initiated calls to the health department. The public was made aware of the availability of the questionnaire through media reports and community meetings.</p></sec><sec><title>Case presentation</title><p>Seventeen respondents called the health department and were administered the questionnaire by telephone. The average age was 65 (range 45–87). Fourteen respondents were female. Almost half were retired or disabled (n = 8). Eight respondents lived alone; nine had two or more people living in their households.</p><p>Ten respondents said their skin problems started in February 2004, five reported an onset date of March 2004, and two reported an onset date later than April 2004. Itchiness was reported by 15 respondents. Symptoms reported included dry skin (n = 8), bumps on the skin (n = 7), burning skin (n = 7), and red skin (n = 6). Four or fewer respondents each reported hives or welts, soreness, rash, flaky skin, pins and needles or tingling sensations and purple bumps. Most respondents reported the skin problem was on the arms and legs (n = 11) and torso (n = 10); four or fewer reported the problem was on the head, eyelids, shoulders, fingers, toes, or "all over".</p><p>Seven respondents had no previous skin problems other than poison ivy, poison oak, or acne. The remaining respondents reported history of hives or rash (n = 5), shingles, eczema, cracking of skin, skin cancer, psoriasis, burning sensations, lice or scabies (three or fewer respondents each). Thirteen respondents indicated that their problems were ongoing and eight felt that they were worse after contact with water. Two respondents had taken time off from work for doctor appointments as a result of the skin problem. A total of fourteen respondents had visited their doctor because of their skin problem, none remembered being given a diagnosis. Most respondents showered at least every other day (n = 11), and had previous allergies (n = 11). There were no common (n > 3) exposures to specific brands of cosmetics, body/bath products, laundry products, or medications.</p></sec><sec><title>Conclusion</title><p>Our investigation indicates that the reported complaints were heterogeneous. Many of the respondents had underlying or preexisting conditions offering an alternative plausible explanation for their symptoms. Overall, results did not support the need for a wider study.</p><p>Our investigation was subject to several biases, and our findings should be interpreted with caution. The respondents were a convenience sample, and none were examined by a dermatologist as part of the investigation. The questionnaire was not validated. Most importantly, the investigation, the complaints, and speculation that these types of symptoms might be related to the change in water treatment were widely publicized in the local media.</p><p>Even with the widespread publicity, only 17 people volunteered to participate in the questionnaire in the five month period that it was open. Including seven who completed the questionnaire, a total of 48 calls from citizens with questions or complaints about chloramine were received by our health department between April 2004 and March 2006. Thirty-six of these callers were from outside of our health department jurisdiction, but within the water supply service area. The total population in the service area is 2.4 million.</p><p>This case investigation was designed to explore the complaints received by the health department. Although we recognized that the approach would not be sufficient to establish or disprove a causal relationship between the skin complaints and the presence of monochloramine in the water, we believe that our investigation was an appropriate step to determine the need for further investigation of these relationships. Nonetheless, clinicians working with populations served by utilities that are switching to monochloramine should be aware of our findings so that they are able to appropriately assess the timing, nature and alternative explanations for the symptoms.</p><p>Worldwide, many public drinking water providers are shifting to the use of monochloramine. In California, approximately 58% of the population in the 25 largest cities received water disinfected with monochloramine in 2005 [<xref ref-type="bibr" rid="B8">8</xref>]. Monochloramine is an effective disinfectant that minimizes the formation of trihalomethanes, for which there is strong evidence of a relationship with adverse health effects. We do not believe that the current evidence about the potential association between skin complaints and the presence of monochloramine in the water is a compelling reason to reconsider the use of monochloramine for residual water disinfection.</p></sec><sec><title>Competing interests</title><p>All authors are paid employees of the City and County of San Francisco Department of Public Health. Ms. Berger's and Dr. Weintraub's positions are funded under a work order from the San Francisco Public Utilities Commission, the agency that provides the drinking water discussed in this case report. However, the San Francisco Public Utilities Commission did not have any role in the design and conduct of the investigation, in the collection, analysis, and interpretation of the data, nor in the preparation, review, or approval of the manuscript.</p></sec><sec><title>Authors' contributions</title><p>JMW conceived the investigation, designed the questionnaire, participated in the data analysis, drafted parts of the manuscript, and critically revised the entire manuscript. MB designed and administered the questionnaire, performed the statistical analysis, drafted parts of the manuscript, and critically revised the entire manuscript. RB participated in the design of the investigation and questionnaire, and critically revised the manuscript. All authors read and approved the final manuscript.</p></sec>
Shoulder muscle EMG activity during push up variations on and off a Swiss ball	<sec><title>Background</title><p>Surface instability is a common addition to traditional rehabilitation and strength exercises with the aim of increasing muscle activity, increasing exercise difficulty and improving joint proprioception. The aim of the current study was to determine if performing upper body closed kinetic chain exercises on a labile surface (Swiss ball) influences myoelectric amplitude when compared with a stable surface.</p></sec><sec sec-type="methods"><title>Methods</title><p>Thirteen males were recruited from a convenience sample of college students. Surface electromyograms were recorded from the triceps, pectoralis major, latissimus dorsi, rectus abdominis and external oblique while performing push up exercises with the feet or hands placed on a bench and separately on a Swiss ball. A push up plus exercise was also evaluated with hands on the support surface.</p></sec><sec><title>Results and discussion</title><p>Not all muscles responded with an increase in muscle activity. The pectoralis major muscle was not influenced by surface stability. The triceps and rectus abdominis muscles showed increases in muscle activity only when the hands were on the unstable surface. The external oblique muscle was only influenced by surface stability during the performance of the push up plus exercise. No muscle showed a change in activation level when the legs were supported by the Swiss ball instead of the bench.</p></sec><sec><title>Conclusion</title><p>Muscle activity can be influenced by the addition of surface instability however an increase in muscle activity does not influence all muscles in all conditions. The relationship between the participant's center of mass, the location of the unstable surface and the body part contacting the Swiss ball may be important factors in determining the muscle activation changes following changes in surface stability.</p></sec>	<contrib id="A1" corresp="yes" contrib-type="author"><name><surname>Lehman</surname><given-names>Gregory J</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>MacMillan</surname><given-names>Brandon</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib><contrib id="A3" contrib-type="author"><name><surname>MacIntyre</surname><given-names>Ian</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A4" contrib-type="author"><name><surname>Chivers</surname><given-names>Michael</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A5" contrib-type="author"><name><surname>Fluter</surname><given-names>Mark</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib>	Dynamic Medicine	<sec><title>Introduction</title><p>Exercise balls, wobble boards and other labile surfaces commonly replace stable surfaces during the performance of resistance training exercises for both injury management and performance improvement. A common assumption is that an unstable surface places an increased demand on the neuromuscular system to stabilize articular joints which may have been created more unstable due to the labile surface. The purported benefits of training with this instability are improvements in joint proprioception and greater muscle activation requirements. Research investigating ankle disc training has shown improvements in proprioception and muscle reflex latency time [<xref ref-type="bibr" rid="B1">1</xref>,<xref ref-type="bibr" rid="B2">2</xref>] following the training regime and reduction in injury prevalence [<xref ref-type="bibr" rid="B3">3</xref>]. Similar improvements in joint proprioception have been documented in unstable shoulders following rehabilitation therapy using unstable surfaces [<xref ref-type="bibr" rid="B4">4</xref>].</p><p>It is often assumed that performing a resistance exercise on unstable surfaces results in greater muscle activity in an attempt to achieve joint stability. This assumption has mixed and somewhat sparse support. The majority of research has investigated the influence of labile surfaces on trunk muscle activity during trunk training exercises. Some research shows a consistent increase in selected (not all) trunk muscles during curl ups on an exercise ball [<xref ref-type="bibr" rid="B5">5</xref>], upper body exercises while seated on an exercise ball [<xref ref-type="bibr" rid="B6">6</xref>] and during unstable weighted squat movements while standing on semi inflated wobble discs [<xref ref-type="bibr" rid="B7">7</xref>]. Others have shown inconsistent changes across subjects with no statistical increase in muscle activity when replacing a Swiss ball for an exercise bench during resistance exercises for the upper body [<xref ref-type="bibr" rid="B6">6</xref>,<xref ref-type="bibr" rid="B8">8</xref>] and changes dependent upon centre of gravity location relative to the unstable surface during bridging/core stability exercises [<xref ref-type="bibr" rid="B9">9</xref>,<xref ref-type="bibr" rid="B10">10</xref>].</p><p>Anderson & Behm [<xref ref-type="bibr" rid="B11">11</xref>] documented the muscle activity of the primary movers and the force output during a chest press while lying on a Swiss ball (labile surface) and on a bench. The authors found a decrease in absolute force production on the labile surface when compared with the force produced on the stable surface; however, there was no difference in the amount of muscle activity between the two conditions. Suggesting that more muscle activation was required to achieve the same amount of force production on an unstable surface compared with a stable surface. This suggests that for the same amount of force production on an unstable surface compared with a stable surface requires a greater amount of muscle activation. An exercise similar to the chest press is the push up exercise. A review of the literature failed to provide any work documenting the influence of an unstable surface on the myoelectric activity of the shoulder and trunk muscles during push up exercises. This exercise assumes that the required force production during the push up will be consistent (due to gravity and body weight) regardless of the stability of the support surface. Therefore investigating this exercise allows us to determine if an unstable support surface necessitates greater muscle activation when the same force requirements are demanded across the same exercise performed on an unstable (Swiss ball) and stable surface (bench). It is the purpose of the current study to determine the effect of an unstable surface under the hands or under the feet during the push up and push up plus exercise on shoulder and trunk muscle activation levels.</p></sec><sec sec-type="methods"><title>Methods</title><sec><title>Patient characteristics</title><p>Thirteen healthy males (average age in years (standard deviation) 26.3 (1.5), average height (standard deviation) 176.7 cm (4.99) and average weight (standard deviation) 79.6 kg (7.34) with greater than 6 months of weight training experience, without back pain or upper limb injuries were recruited from a convenience sample of college students. Participants were required to sign an information and informed consent form prior to the study approved by the institution's Research Ethics Board.</p></sec><sec><title>Study protocol</title><p>To optimize EMG signal collection participants from a college population were recruited because of their athletic abilities and low subcutaneous fat. The myoelectric activity of the pectoralis major, latissimus dorsi, triceps, rectus abdomins and external oblique muscles were recorded during a series of different variations of the classic push up exercise.</p></sec><sec><title>Data collection hardware characteristics</title><p>Disposable bipolar Ag-AgCI disc surface electrodes with a diameter of one cm were adhered over the muscle groups parallel to their fiber orientation in the muscle belly. Before the application of the electrodes the skin was shaved with a disposable razor and abraded with a cotton swab and alcohol. The electrodes were attached to 5 leads which were connected to an EMG data collection system (Bortec, Calgary AB). The myoelectric activity of the muscles was collected with customized software (Delys EMGWorks, Boston, MA, USA). Raw EMG was amplified between 1000 and 20,000 times depending on the subject. The amplifier had a CMRR of 10,000:1 (Bortec EMG, Calgary AB, Canada). Raw EMG was band pass filtered (10 and 1000 Hz) and A/D converted at 2048 Hz using a National Instruments data acquisition system.</p></sec><sec><title>Maximum Voluntary Contractions (MVC)</title><p>In order to compare muscle activity across subjects and give biologically meaningful data maximal normalization contractions were performed for each muscle. This required the subject to maximally contract each muscle against the manual resistance provided by the experimenter. A maximal isometric contraction occurred twice for each muscle to ensure that an acceptable signal was recorded for each subject. The maximum muscle activity was calculated and recorded from a suitable maximum contraction and all subsequent muscle activity was expressed as a percentage of this maximum voluntary contraction (MVC).</p></sec><sec><title>Electrode placement and MVC testing procedure</title><p>The triceps electrode was placed on the long head (middle of the muscle belly) between origin and insertion. The MVC saw the shoulder and elbow flexed to 90 degrees while the EMG was recorded during resisted elbow and shoulder extension. The pectoralis major electrode was placed four fingerbreadths below the clavical, medial to the anterior axillary border. With the elbow flexed 90 degrees and the shoulder abducted 75 degrees the subject performed a maximal palm press while the muscle activity was recorded. The latissimus dorsi electrodes were 3 finger fingerbreadths distal to and along the posterior axillary fold, parallel to the lateral border of the scapula. With the elbow extended and the arm abducted 30 degrees in the coronal plane and internally rotated, attempted maximum shoulder extension, adduction and internal rotation was resisted with the muscle activity recorded. The rectus abdominis electrode was placed 3 cm lateral to the umbilicus in a vertical orientation. With the participant supine, trunk flexion was resisted and the MVC was recorded. The external oblique electrode was placed 15 cm lateral to the umbilicus on a 45 degree inferior angle. With the subject lying supine with knees flexed 90 degrees, resisted side bending was recorded.</p></sec><sec><title>Exercise protocol</title><p>Following the maximal voluntary contractions the participants were required to perform the following exercises in random order (arbitrarily determined by the experimenters). Participants performed the exercises identically across exercises. Three repetitions occurred for each exercise at the same rate. Participants began in the upright position when the EMG collection began. This position was held for 4 seconds. The eccentric (lowering) portion lasted 2 seconds. The patient then held the lowered position for 4 seconds. The concentric (raising) portion of the movement lasted for 2 seconds with a 4 second holding position once again in the upright position. Three repetitions were recorded during a 40 second collection period. An electrical trigger (foot switch) was used to mark the beginning of the first descent and the finish of the last repetition.</p></sec><sec><title>Movement tasks</title><p>All movements were completed in a random order in a standardized position with the hands shoulder width apart with the subject's middle finger under the acromioclavicular joint. The bench height and exercise ball height were standardized and identical to each other. A minimum of 3 minutes of rest occurred between exercises to prevent the influence of fatigue on myoelectric amplitude changes. This rest period is similar and exceeds other studies investigating similar phenoma [<xref ref-type="bibr" rid="B5">5</xref>,<xref ref-type="bibr" rid="B7">7</xref>,<xref ref-type="bibr" rid="B9">9</xref>].</p><p>1. Push up with feet on an exercise bench.</p><p>2. Push up with feet on exercise ball.</p><p>3. Push up with hands on exercise bench.</p><p>4. Push up with hands on exercise ball.</p><p>5. Push up plus with hands on exercise bench: Starting in the push up position the participant rolls the shoulders forward (scapular protraction) and then lowers their body while allowing the shoulder blades to approximate (scapular retraction).</p><p>6. Push up plus with hands on exercise ball. Same movement as #5.</p></sec><sec><title>EMG analysis</title><p>Both MVC data and myoeletric data from the exercises tasks were processed in the identical manner. Using EMG analysis software (EMGWorks, Delsys, Boston, MA) the myoelectrical was first demeaned (bias removed therefore the signal alternates around 0), then a root mean square technique (window of 150 ms and an overlap of 75 ms) was used to smooth the data thus providing a linear envelop of EMG activity. Using the electrical markings left by the foot switch trigger at the start and end of the movement the mean activity for the 3 repetitions was calculated. This mean activity was then expressed as a percentage of the peak activity found during the maximum voluntary contraction (MVC) for the corresponding muscle</p></sec><sec><title>Statistical analysis</title><p>A series of paired <italic>t-tests </italic>were used to assess for the influence of the Swiss ball on muscle amplitude. Differences across the different exercises was not examined-only between the stable and unstable conditions.</p></sec></sec><sec><title>Results</title><p>Table <xref ref-type="table" rid="T1">1</xref> shows the average muscle activity and standard deviations for the 6 different exercises studied. The latissimus dorsi values were removed from the study as the first six participants analyzed showed primarily noise and little muscle activity during the performance of the push up. The triceps muscle was significantly influenced by the addition of the Swiss ball during the performance of the push up with the hands on the ball surface and during the push up plus with the hands on the ball surface. Both exercises showed a statistically significant increase in muscle activity when performed on the Swiss ball. Conversely, the triceps activity was not influenced when the Swiss ball replaced the bench under the participant's feet. The addition of the Swiss ball did not influence the muscle activity of the pectoralis major during all exercises. Similar to the triceps the Swiss ball addition resulted in a significant increase in the myoelectric activity of the rectus abdominis during push ups with the hands on the Swiss ball and the push up plus. There was no change in muscle activity when the feet were on the unstable surface. The external oblique was not influenced by the addition of the Swiss ball for either standard push up exercise. The external oblique showed a statistically significant increase when the Swiss ball was added during the push up plus exercise.</p><table-wrap position="float" id="T1"><label>Table 1</label><caption><p>Average trunk and shoulder muscle activity during the push up and push up plus exercises when performed on a Swiss ball and an exercise bench.</p></caption><table frame="hsides" rules="groups"><thead><tr><td></td><td align="left" colspan="2"><bold>Triceps</bold></td><td align="left" colspan="2"><bold>Pectoralis Major</bold></td><td align="left" colspan="2"><bold>Rectus Abdominis</bold></td><td align="left" colspan="2"><bold>External Oblique</bold></td></tr><tr><td></td><td align="left"><bold>Bench</bold></td><td align="left"><bold>Ball</bold></td><td align="left"><bold>Bench</bold></td><td align="left"><bold>Ball</bold></td><td align="left"><bold>Bench</bold></td><td align="left"><bold>Ball</bold></td><td align="left"><bold>Bench</bold></td><td align="left"><bold>Ball</bold></td></tr></thead><tbody><tr><td align="left"><bold>PUHands</bold></td><td align="left">22.2 ± 8.8</td><td align="left">43.1 ± 17.3<break/>p = .002</td><td align="left">21.4 ± 11.8</td><td align="left">26.65 ± 14.5<break/>p = .341</td><td align="left">13.4 ± 5.4</td><td align="left">22.6 ± 8.6<break/>p = .001</td><td align="left">20.9 ± 15.6</td><td align="left">24.07 ± 11.9<break/>p = .1191</td></tr><tr><td align="left"><bold>PUFeet</bold></td><td align="left">32.7 ± 17.5</td><td align="left">29.9 ± 10.1<break/>p = .355</td><td align="left">30.4 ± 9.7</td><td align="left">29.9 ± 7.6<break/>p = .717</td><td align="left">19.2 ± 6.9</td><td align="left">19.6 ± 7.14<break/>p = .742</td><td align="left">27.79 ± 13.2</td><td align="left">26.7 ± 9.8<break/>p = .575</td></tr><tr><td align="left"><bold>PUPHands</bold></td><td align="left">18.7 ± 10.3</td><td align="left">29.6 ± 14.7<break/>p = .024</td><td align="left">14 ± 7.2</td><td align="left">21.4 ± 18.5<break/>p = .079</td><td align="left">14.9 ± 5.5</td><td align="left">25.4 ± 10.2<break/>p = .011</td><td align="left">19.2 ± 11.9</td><td align="left">23.9 ± 11.9<break/>p = .0069</td></tr></tbody></table><table-wrap-foot><p>Muscle activity is expressed as a percentage of a maximum voluntary contraction (%MVC). A * indicates a statistically significant difference (p < .05) between the exercises using a Swiss ball or bench as the support surface (PUHands = Push up with hands on bench/ball, PUFeet = push up with feet on ball/bench and PUPHands = push up plus exercise with hands on ball/bench).</p></table-wrap-foot></table-wrap></sec><sec><title>Discussion</title><p>Replacing an exercise bench for a Swiss ball can increase muscle activity however, the effect is both task and muscle dependent. The triceps and rectus abdominis were the two muscles most affected by the addition of an unstable surface. The pectoralis major muscle – the primary mover – was not influenced by the addition of the Swiss ball during any push up variation.</p><p>It is not possible to conclude that surface instability automatically results in increases in muscle activity. Merely adding an unstable surface is insufficient to influence all muscles. Of note is that placing the feet on the unstable surface resulted in no changes in muscle activity in all of the muscles studied. Our findings suggest that the unstable surface needs to be under the hands in order to result in an assumed destabilization effect and a subsequent increase in muscle activity during push up variations. It also appears that the further the distance the centre of mass is above the base of support (when unstable) can also influence the trunk muscle activity. While we did not measure the position of the centre of mass in this study we believe it is safe to assume that the centre of mass of the subject is higher relative to the Swiss ball when the hands are on the Swiss ball than when the feet are on the Swiss ball considering that the feet are the highest body part during the push up when the feet are on the Swiss ball. A different result may occur if the feet are placed on the Swiss ball and the hands are on a bench of equal height to the Swiss ball. This vertical distance from the Swiss ball may be an important factor in determining which exercises will see changes in myoelectric activity with the addition of a Swiss ball. Of interest was the finding that the external oblique was not influenced by the Swiss ball during the standard push up; however, when the participants performed the push up plus which finds the arms extended and the participant's chest farther away from the Swiss ball the external oblique showed greater activity during the unstable condition. We assume in our study that the centre of mass is higher during the push up plus with the arms extended than during the push up exercise that saw the elbows flex and extend with no scapular protraction at the top of the push up. The finding that the distance between chest and hands on a Swiss ball influences trunk muscle activity has previously been reported by Marshall and Murphy [<xref ref-type="bibr" rid="B9">9</xref>] during bridging/front support exercises.</p><p>The lack of change in the pectoralis major muscle during the push ups on the different stability surfaces is interesting considering the dramatic change in the triceps muscle. This may be due to the differences between the joints and associated movements that the two muscles cross. There is greater redundancy in the motor control of muscles crossing the anterior shoulder. The joint is stabilized by a multitude of muscles (biceps brachii, anterior deltoid, rotator cuff) and shoulder adduction torque is also created by a number of muscles in addition to the pectoralis major. There is also a smaller range of motion compared with the elbow joint. The pectoralis major is a single joint muscle whereas the triceps brachii is a two joint muscle which has stability and movement demands both at the elbow and the shoulder possibly resulting in such a dramatic change in muscle activity when the Swiss ball replaced the bench during the push up. The pectoralis major may only be concerned with its primary movement and have a smaller role in responding to changes in stability which may be controlled by other muscles which influence the shoulder joint. In contrast, the triceps brachii is the main extensor of the elbow with little help from the anconeus. Due to its mechanical advantage relative to the length of the forearm it may also have difficulty in responding to changes in stability compared with pectoralis major. It should also be noted that there is often a range of responses as seen in previous research [<xref ref-type="bibr" rid="B8">8</xref>]. Not every individual responded in the same manner to a change in surface stability. It is possible that there are individual factors that modulate the response to surface stability which also suggests that training may be influence the response to instability. We don't know if the increase in muscle activity or lack of change is involuntary or can be subject to change with training and feedback.</p><p>A limitation to explaining our results is the lack of kinematic and kinetic information describing the push up variations in our study. Having information on the centre of mass and its relationship to the bases of support (feet and hands) in the different exercises may help explain the different results found when the feet were placed on the Swiss ball compared with the hands on the Swiss ball. An attempt was made in this study to control for joint posture and the influence of gravity on the body's centre of mass. The exercise conditions were assumed to be identical except for the addition of the Swiss ball. An attempt was made to control for the speed of movement as well. These controls were imparted visually and with simple measurements. It is possible that subtle variations did exist between the conditions but we feel that these differences would not affect the results as they would be overlapped by the natural variability that is seen in EMG recording during any well controlled exercise.</p><p>Of practical interest are the low values for muscle activity during the exercises. For the population studied it implies that these exercises are insufficient to produce improvements in strength. Adaptations in terms of endurance or motor control are possible. A less athletic population may achieve strength benefits from these exercises.</p></sec><sec><title>Conclusion</title><p>The muscular response during push up exercises on unstable surfaces is task and muscle dependent. When the hands are supported by the Swiss ball (but not the feet) increases in muscle activity can be seen with a greater number of muscles affected.</p></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>GL: Conception, design, data collection, data analysis, manuscript preparation</p><p>BM, IM, MC, MF: design, data collection, manuscript preparation</p><fig position="float" id="F1"><label>Figure 1</label><caption><p>EMG linear envelopes during a push ups with the hands on an exercise bench and a Swiss ball. EMG activity is non-normalized and is therefore in arbitrary EMG units. A bias was added to the activity of the pectoralis major, rectus abdominis and external oblique for ease of viewing to prevent overlap of the muscle's linear envelopes. Three pushups occurred over the course of 30 seconds.</p></caption><graphic xlink:href="1476-5918-5-7-1"/></fig><fig position="float" id="F2"><label>Figure 2</label><caption><p>Raw EMG during three push ups with the hands on an exercise bench or a swiss ball for the triceps and pectoralis major muscle. PMBench = Pectoralis Major EMG on bench, PMBall = Pectoralis Major EMG on ball. Bias added to triceps activity for ease of viewing. Trial was 30 seconds in length.</p></caption><graphic xlink:href="1476-5918-5-7-2"/></fig><fig position="float" id="F3"><label>Figure 3</label><caption><p>Raw EMG for Rectus Abdominis and External Oblique muscles during three push ups with the hands on or off a Swiss ball. EO = External Oblique, RA = Rectus Abdominis.</p></caption><graphic xlink:href="1476-5918-5-7-3"/></fig></sec>
Assessing effects of a media campaign on HIV/AIDS awareness and prevention in Nigeria: results from the VISION Project	<sec><title>Background</title><p>In response to the growing HIV epidemic in Nigeria, the U.S. Agency for International Development (USAID) initiated the VISION Project, which aimed to increase use of family planning, child survival, and HIV/AIDS services. The VISION Project used a mass-media campaign that focused on reproductive health and HIV/AIDS prevention. This paper assesses to what extent program exposure translates into increased awareness and prevention of HIV/AIDS.</p></sec><sec sec-type="methods"><title>Methods</title><p>This analysis is based on data from the 2002 and 2004 Nigeria (Bauchi, Enugu, and Oyo) Family Planning and Reproductive Health Surveys, which were conducted among adults living in the VISION Project areas. To correct for endogeneity, two-stage logistic regression is used to investigate the effect of program exposure on 1) discussion of HIV/AIDS with a partner, 2) awareness that consistent condom use reduces HIV risk, and 3) condom use at last intercourse.</p></sec><sec><title>Results</title><p>Exposure to the VISION mass media campaign was high: 59%, 47%, and 24% were exposed to at least 1 VISION radio, printed advertisement, or TV program about reproductive health, respectively. The differences in outcome variables between 2002 baseline data and the 2004 follow-up data were small. However, those with high program exposure were almost one and a half (Odds Ratio [O.R.] = 1.47, 95% Confidence Interval [C.I.] 1.01–2.16) times more likely than those with no exposure to have discussed HIV/AIDS with a partner. Those with high program exposure were over twice (O.R. = 2.20, C.I. 1.49–3.25) as likely as those with low exposure to know that condom use can reduce risk of HIV infection. Program exposure had no effect on condom use at last sex.</p></sec><sec><title>Conclusion</title><p>The VISION Project reached a large portion of the population and exposure to mass media programs about reproductive health and HIV prevention topics can help increase HIV/AIDS awareness. Programs that target rural populations, females, and unmarried individuals, and disseminate information on where to obtain condoms, are needed to reduce barriers to condom use. Improvements in HIV/AIDS prevention behaviour are likely to require that these programmatic efforts be continued, scaled up, done in conjunction with other interventions, and targeted towards individuals with specific socio-demographic characteristics.</p></sec>	<contrib id="A1" corresp="yes" contrib-type="author"><name><surname>Keating</surname><given-names>Joseph</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A2" contrib-type="author"><name><surname>Meekers</surname><given-names>Dominique</given-names></name><xref ref-type="aff" rid="I1">1</xref><email>[email protected]</email></contrib><contrib id="A3" contrib-type="author"><name><surname>Adewuyi</surname><given-names>Alfred</given-names></name><xref ref-type="aff" rid="I2">2</xref><email>[email protected]</email></contrib>	BMC Public Health	<sec><title>Background</title><p>Like many African countries, Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome (HIV/AIDS) is a serious concern in Nigeria. By the turn of the century it was estimated that the HIV prevalence in Nigeria had exceeded 5%, which corresponds with approximately four million infected people [<xref ref-type="bibr" rid="B1">1</xref>]. In 2003, the HIV prevalence had exceeded 5.4%, with some estimates putting the number of infected Nigerians around 5 million [<xref ref-type="bibr" rid="B2">2</xref>]. The estimated annual deaths as a result of HIV/AIDS has increased from 50,000 in 1999 to over 350,000 in 2004, at enormous cost to the economic and health sectors in the country [<xref ref-type="bibr" rid="B3">3</xref>]. Although many Nigerians are aware that condom use can prevent HIV infection, and condoms are widely available, consistent condom use has remained relatively low over the years [<xref ref-type="bibr" rid="B4">4</xref>-<xref ref-type="bibr" rid="B8">8</xref>].</p><sec><title>The VISION Project</title><p>In response to the HIV/AIDS epidemic in Nigeria, the U.S. Agency for International Development (USAID) mission in Nigeria initiated the VISION Project, a three-year project designed to maximize the use of family planning services, HIV/AIDS services, and child survival services in Nigeria [<xref ref-type="bibr" rid="B9">9</xref>-<xref ref-type="bibr" rid="B13">13</xref>]. The project also sought to develop models of high-impact, high-performing family planning and reproductive health (FP/RH) service-delivery networks, built on public-private partnerships, to use as models for the delivery of similar interventions elsewhere. Mass media campaigns about HIV/AIDS and the repositioning of community-based health distribution centers were two of the primary activities undertaken by the VISION Project to improve access and awareness of FP/RH related issues and services. Training curricula and capacity-building development, and the development of training sessions for clinical trainers at university-based hospitals, were also important project activities for increasing FP/RH awareness and service capacity.</p><p>The VISION Project was implemented by EngenderHealth in partnership with IntraHealth, Johns Hopkins University Center for Communication Programs, Population Services International, and local non-governmental organizations (NGO) working in VISION Project target areas. The project targeted 15 local government areas (LGAs) in Bauchi, Enugu and Oyo states, and was active from September 2001 through September 2004 [<xref ref-type="bibr" rid="B9">9</xref>-<xref ref-type="bibr" rid="B13">13</xref>]. The project aimed to increase the demand for and use of FP/RH services through behavior change communication (BCC) activities and community mobilization efforts.</p><p>The VISION Project worked with local NGOs active in the selected LGAs to develop "informed clients." BCC activities aimed to increase knowledge and to empower individuals and communities to demand quality family planning, STI/HIV/AIDS, and reproductive health information and services. The project developed information and education print materials and radio public service announcements (PSA) for women with an unmet need for family planning, as well as messages aimed at men and youth that focused on male involvement and responsibility in reproductive decision-making. In collaboration with local sports clubs, the VISION Project also developed an outreach strategy primarily aimed at youth called <italic>Sports for Life </italic>to promote healthy lifestyles and spread information on family planning and HIV/AIDS prevention during football competitions.</p><p>The VISION Project media campaign also included a set of weekly radio programs in target areas to disseminate family planning, HIV/AIDS, and other reproductive health information to the general public. VISION, with its partner NGOs, organized radio discussion clubs in the VISION Project target areas and provided these clubs with radios. In addition to these activities, the VISION partners conducted several complementary mass media activities (including radio, TV, and print). For example, the Society for Family Health (SFH) implemented several radio dramas aimed at increasing awareness of HIV/AIDS, and its causes and consequences. These radio dramas included <italic>One Thing at a Time</italic>, Garin <italic>Muna Fata (Town of Hope)</italic>, <italic>Odenjiji</italic>, and <italic>Abule Oloke Merin</italic>. As well, SFH implemented a television to promote safe sex that featured Nigerian soccer superstar Sunday Oliseh. A billboard campaign to increase awareness that someone who is HIV positive may not have symptoms supplemented the radio and television campaign [<xref ref-type="bibr" rid="B14">14</xref>].</p><p>The purpose of this paper is twofold: 1) to identify determinants of FP/RH program exposure, and 2) to assess the effect of a FP/RH mass media campaign on HIV/AIDS awareness and condom use in the VISION Project target areas. Specifically, this paper tests whether exposure to VISION-related mass media programs (TV, radio, and printed advertisements about FP/RH and HIV/AIDS) had an effect on individual level willingness to discuss HIV/AIDS (awareness), whether said programs had an effect on individual level perception about condom use for reducing the risk of HIV infection (awareness), and whether exposure to mass media programs translates into increased condom use (HIV/AIDS prevention).</p></sec></sec><sec sec-type="methods"><title>Methods</title><sec><title>Study areas</title><p>Data were collected from a representative sample of households in the 15 Local Government Areas (LGA) targeted by the VISION Project. These include 5 LGAs in each of 3 states: Bauchi State located in the North-East, which is comprised mainly of Gerawa, Ningawa, Hausa, Fulani, and the Tangale ethnic groups; Eungu State located in the East, which is comprised mainly of the Igbo ethnic group; and Oyo State located in the South-Western part of Nigeria, which is comprised mainly of the Yoruba ethnic group. The selected LGAs in each of the 3 states represent VISION Project areas. LGAs were considered for selection into the VISION Project if: 1) the population was over 100,000, 2) the population contained at least 20,000 women of reproductive age, 3) the LGA contained private and public health facilities at the primary and secondary levels, 4) the LGA has access to media resources such as radio, television and newspapers, and 5) the LGA has support from Nigeria's federal Ministry of Health (MOH) and the relevant state-level Ministry of Health (MOH). VISION partners visited the LGAs and made the selections [<xref ref-type="bibr" rid="B10">10</xref>].</p></sec><sec><title>Data</title><p>This paper analyses data from the 2002 and 2004 household survey waves of the VISION Project evaluation [<xref ref-type="bibr" rid="B10">10</xref>,<xref ref-type="bibr" rid="B13">13</xref>]. The surveys were implemented by the Center for Research, Evaluation, and Resource Development (CRERD). The Tulane University Institutional Review Board approved this study. A two-stage cluster design was used to obtain a probability sample of respondents for this study. Sample size calculations were based on National contraception prevalence rates of 16% for females and 27% for males, and a design effect of 2. The target sample size was 1,100 respondents per state (after adding 10% for potential non-response and incomplete questionnaires). This sample size will enable us to detect a change of 10 percentage points in a variety of key indicators in the VISION Project areas in each state with 90% power and a probability of committing a type-I error set at 5% (two-sided test). In each state's VISION project area, 40 enumeration areas (EA) were randomly selected with probability of selection proportional to the population size (PPS) of the selected LGA. The State Office of the National Population Commission (NPC) provided a list of EAs in the project LGAs of each state. Within each selected EA, a household enumeration exercise was completed; between 27 and 28 households were then selected using systematic random sampling. All adults aged 15–49 years old in the household were listed, and one eligible person per household was selected using a table of random numbers. Interviewers obtained verbal informed consent from the selected participants. A total of 3,279 respondents across all three states completed the questionnaire (≅ 1% refusal/non-response rate) in 2004.</p><p>Data collection was conducted by trained interviewers. All fieldwork supervisors first participated in a 5-day centralized training. Subsequent to this, supervisors and interviewers jointly participated in 5-day regional trainings. The household questionnaire was field tested prior to the onset of the survey to check for errors, and to evaluate the readiness of the interviewers.</p><p>The survey questionnaire was based on the Demographic and Health Survey questionnaire. In addition to standard demographic and fertility questions, questions related to family planning, sexual activity and behavior, and exposure to various media campaigns were also asked. Specifically, respondents were asked if they had listened to the following programs on the radio: <italic>Kusaurara, Dunniya J'atau</italic>, <italic>A New Dawn (Ayedotun)</italic>, <italic>One thing At A Time</italic>, <italic>Gari Muna Fati</italic>, <italic>Abule Olokemerin</italic>, or <italic>Odenjinjin</italic>. Respondents were also asked if they had seen the <italic>Femi Kuti </italic>or <italic>Fati Mohammed </italic>television campaigns, if they had seen any HIV/AIDS or reproductive health advertisements in the newspaper, or had received any information from clinics or community health workers about HIV/AIDS or reproductive health.</p><p>The outcome measures for this analysis are as follows: 1) Have you ever talked with a partner about ways to prevent getting the virus that causes AIDS (yes/no)? 2) Can people reduce their chances of getting the AIDS virus by using a condom every time they have sex (yes/no)? 3) Did you use a condom during your last sexual encounter (yes/no)?</p><p>The respondent's age (continuous variable), education (categorized as none, primary, or secondary), religion (categorized as catholic, protestant or other Christian denomination, Muslim, or traditionalist), gender, marital status, place of residence (dichotomized as urban and peri-urban, or rural), knowledge of a condom source (yes/no), state of residence (categorized as Bauchi, Enugu, or Oyo), and whether or not the respondent has had at least 1 partner in the past 12 months, served as control variables in all regression models (described below). Based on the distribution of the data during preliminary analyses, exposure to media was measured by dichotomizing whether a respondent reads the newspaper at least once a week (yes/no), watches television at least once a week (yes/no), or listens to the radio at least once a week (yes/no). Media exposure variables served as control variables in the first-stage Poisson regression models only (see below).</p><p>Our indicators of program exposure included a total count of the number of FP/RH radio programs heard, television programs seen, or printed FP/RH advertisements seen over the past 6 months (0 – 10). A separate count of the number of radio programs exposed to (0 – 7), and the number of TV programs exposed to (0 – 2), were also used as indicators of program exposure in separate models. All values, whether part of the cumulative number of programs exposed to or the individual programs exposed to for the respective media types, were then categorized as low (none), medium (one), or high (2 or more). The decision to use the same categories for each program exposure model was based on the distribution of program exposure data for each media type. These analyses showed that the proportion of respondents exposed to each category (i.e. low, medium, high) of program exposure was similar between total program exposure counts, radio program exposure counts only, and TV program exposure counts only. No model was run for the printed advertisement media, as the questionnaire only asked about exposure to any printed advertisement (yes/no).</p></sec><sec><title>Data analysis</title><p>Data analyses were done using STATA™ 7.0. Comparisons of the outcome variables were made between the baseline 2002 data and the follow-up 2004 data. Chi-square statistics and two-stage logistic regression models were used to analyze the data [<xref ref-type="bibr" rid="B15">15</xref>]. Using data from the 2004 follow-on survey, the Durbin-Wu-Hausman test (augmented regression test) was performed to test for endogeneity between media exposure and the respective outcomes [<xref ref-type="bibr" rid="B16">16</xref>]. Since the results showed evidence of endogeneity (i.e. value of one independent variable is dependent on the value of other predictor variables), two-stage logistic regressions were performed using instrumental variables of program exposure. At the first stage, Poisson regression was used to estimate the total number of exposures to VISION programs using a set of exogenous variables. Because estimated exposure is a continuous variable, respondents were recoded has having low estimated exposure if the estimated exposure was less than 1, medium if estimated exposure was between 1 and 2, and high if estimated exposure was two or more, and used as indicators of program effectiveness in all logistic regressions. Wald statistics and log-likelihood ratios were used to identify variable significance and model fit, with alpha set at 0.05. To control for the effect of clustering within enumeration areas (EA), the Huber-White-Sandwich estimator of variance was used to obtain empirically estimated standard errors, with the EA defined as the <italic>cluster</italic>. Standardized state-level probability weights, based on the relative population size of program LGAs in each State (<italic>pweight</italic>) were applied to regressions, and to all between state estimates.</p></sec></sec><sec><title>Results</title><p>Socio-demographic characteristics of the population are presented in Table <xref ref-type="table" rid="T1">1</xref>. The average age (28.1) of the respondents and the mean number of sexual partners in the past year (1.2) were similar across states. In Bauchi, 73% of the respondents were married; 45% reported being married in Enugu and 60% reported being married in Oyo. Seventy-four percent of respondents in Oyo reported living in urban or peri-urban areas; 35% of respondents in Bauchi and 36% of respondents in Enugu reported living in urban or peri-urban areas. The states also differed in educational attainment, with Bauchi reporting the lowest proportion of respondents reporting secondary or higher education (20.4%) and Enugu reporting the highest proportion of respondents reporting secondary or higher education (60%). Oyo had the highest proportion of Protestants/other Christians (55%), Bauchi had the highest number of Muslims (89%), and Enugu had the highest proportion of Catholics (71%). Knowledge of a condom source also varied by state, with 57% of respondents in Oyo, and 51% and 26% of respondents in Enugu and Bauchi, respectively, reporting that they know where to get condoms.</p><sec><title>Exposure to media programs by the VISION partners</title><p>Figure <xref ref-type="fig" rid="F1">1</xref> illustrates the percentage of respondents exposed to the various VISION media programs by gender. In general, more males were exposed to programs than females, although more females were exposed to clinic-based information. VISION program exposure statistics are presented in Table <xref ref-type="table" rid="T2">2</xref>. The mean number of radio program exposures was highest in Bauchi (1.38) and lowest in Enugu (0.65). Bauchi also reported the highest mean number of TV program exposures (0.36). Seventy-seven percent of all respondents reported listening to the radio at least once a week and 47% reported watching TV at least once a week. On average, 59% of respondents heard at least 1 FP/RH radio program in the past 6 months, and 24% saw a FP/RH TV program. Forty-seven percent were exposed to at least 1 advertisement about HIV/AIDS, sexual abstinence or condom use in the last 6 months. Over 96% of all respondents thought it was acceptable to discuss HIV/AIDS on TV, the radio, or in the newspaper.</p><p>Table <xref ref-type="table" rid="T3">3</xref> presents the results of the first stage Poisson regressions. The socio-demographic variables included were almost all significantly related to the total number of program exposures, as well as the number of radio and TV program exposures, respectively. Male respondents were more likely to be exposed to media programs than female respondents, as indicated by a positive beta coefficient, and secondary education was positively associated with the number of programs exposures; low education level was negatively associated with the number of program exposures. Respondents located in urban and peri-urban areas were also more likely to have been exposed to programs than those located in rural areas. Because socioeconomic data were not available, it is difficult to capture whether location is also serving as a proxy for wealth or access to resources. No significant relationship was detected between whether the respondent has been sexually active in past year and the number of program exposures. As expected, significant positive relationships exist between whether respondents watch TV, read the newspaper, or listen to the radio, and the number of program exposures.</p></sec><sec><title>Effect of program exposure on HIV/AIDS awareness, perception, and prevention behaviour</title><p>Figures <xref ref-type="fig" rid="F2">2</xref>, <xref ref-type="fig" rid="F3">3</xref>, <xref ref-type="fig" rid="F4">4</xref> illustrate the outcome measures used in this analysis (2004 data), relative to the 2002 baseline data. The percentage of respondents that used a condom at last sex increased in all 3 states (Figure <xref ref-type="fig" rid="F2">2</xref>). Overall, the percentage increased from 13.9% in 2002 to 15.8% in 2004, although this increase was not statistically significant (χ<sup>2 </sup>= 3.2, P = 0.07). The percentage of respondents that used a condom at last sex was lowest in Bauchi and greatest in Enugu in both 2002 and 2004. The percentage of respondents that have discussed ways to prevent getting HIV/AIDS with a partner increased from 44.7% in 2002 to 48.1% in 2004 (χ<sup>2 </sup>= 3.7, P = 0.05) (Figure <xref ref-type="fig" rid="F3">3</xref>). The percentage of respondents that have discussed ways to prevent getting HIV/AIDS with a partner was over 50% for Enugu and Oyo in both 2002 and 2004. In Bauchi, the percent increased from 24% in 2002 to 38% in 2004. The percentage of respondents that believe consistent condom use reduces risk of HIV infection was well over 40% in all 3 states for both survey rounds (Figure <xref ref-type="fig" rid="F4">4</xref>); overall, the percent decreased from 57% in 2002 to 54% in 2004 (χ<sup>2 </sup>= 5.2, P = 0.02). This decrease largely stems from a reduction in perceived condom efficacy in Enugu.</p><p>The effect of exposure on the relative odds that respondents were willing to discuss HIV/AIDS with their partner is shown in Table <xref ref-type="table" rid="T4">4</xref>. Those with high program exposure were almost one and a half (O.R. = 1.47, C.I. 1.01–2.16) times more likely than those with no exposure to have discussed HIV/AIDS with their partner, after controlling for the potential confounding effects of socio-demographic variables. The results for models testing the effect of individual programs (i.e. radio programs and TV programs) on the odds that respondents discussed HIV/AIDS with their partners were not significant. Strong positive associations were detected between education and the relative odds of the outcome occurring, as evidenced by highly significant odds ratios for both primary and secondary education categories for all models. Married respondents, or respondents reportedly living with a partner, where nearly two times more likely than unmarried respondents to have discussed HIV/AIDS [O.R. = 2.01 (C.I. 1.58–2.55), 1.98 (C.I. 1.55–2.51), 2.00 (C.I. 1.57–2.54) for models 1–3, respectively]. Lastly, there appears to be a strong positive association between those reporting that they know where to obtain condoms, and willingness to discuss HIV/AIDS with their partners [OR = 2.83 (C.I. 2.22–3.61), 2.84 (C.I. 2.23–3.62), 2.85 (C.I. 2.24–3.64) for models 1–3, respectively].</p><p>The effect of program exposure on the belief that consistent condom use can reduce the risk of HIV infection is shown in table <xref ref-type="table" rid="T5">5</xref>. Those with high program exposure were over twice (O.R. = 2.20, C.I. 1.49–3.25) as likely as those unexposed to know that consistent condom use can reduce risk of HIV infection. Those with moderate program exposure were almost 1 and a half (O.R. = 1.42, C.I. 1.07–1.88) times more likely than those unexposed to know that condom use reduces risk of HIV infection. Model 2 indicates that exposure to at least 1 radio program is also significantly associated (O.R. = 1.33, C.I. 1.02–1.73) with the odds of the outcome occurring, as compared to those not exposed to radio programs. Exposure to TV programs alone was not associated with the outcome. Also noteworthy is the strong significant association between secondary or higher education, and knowing that condoms reduce risk of HIV infection [O.R. = 1.49 (C.I. 1.00–2.22), 1.85 (C.I. 1.27–2.70), 2.00 (C.I. 1.41–2.83) for models 1–3, respectively]. Those reporting knowledge of a condom source were almost four times more likely to also know that condoms can reduce the risk of HIV infection in all models. Those respondents located in Enugu were about half as likely as those in Oyo to know that condoms reduce risk of HIV.</p><p>The effect of exposure on condom use at last sexual intercourse is shown in table <xref ref-type="table" rid="T6">6</xref>. Program exposure had no significant impact on reported condom use at last sex. Similarly, radio program exposure and TV program exposure had no significant effect. Male respondents were also more likely than female respondents to have reportedly used a condom at last sex, for all program exposure models; those reporting knowledge of a condom source were approximately 9 times more likely than those without such knowledge to have used a condom at last sex for all program exposure models. Married respondents and respondents reportedly living with a partner were consistently less likely than the unmarried to have used a condom at last sex; those respondents located in Bauchi, were also less likely than those in the other two States to have used a condom at last sexual intercourse.</p></sec></sec><sec><title>Discussion</title><p>Nigeria is one of several sub-Saharan African countries trying to deal with the current HIV/AIDS epidemic. In response to this need, the USAID-funded VISION Project launched a mass media campaign focusing on HIV/AIDS awareness, perception, and behavior modification. The purpose of this paper was to assess the determinants of program exposure, and test whether exposure to FP/RH/HIV/AIDS program information translates into increased knowledge, awareness, and prevention of HIV/AIDS.</p><p>While the VISION project communication strategy predominantly focused on radio communications and community health workers, some of the VISION partners also used other media. The results from our study show that in combination these media activities had high reach among the target population. Exposure to radio communication was highest, with 59% of the population reporting to have heard at least one advertisement or radio program. Exposure to TV programs was the lowest media source, with 24% of the population reporting having seen at least one VISION program. Exposure to printed advertisements was also lower than radio program exposure, with approximately 47% having seen an advertisement about reproductive health or HIV/AIDS. Community health workers and visits to the clinic were also important information venues, with over 25% of the population having been exposed to one or the other.</p><p>Despite the relatively high percentage of respondents that believe consistent condom use can reduce the risk of HIV infection, and the high percentage of respondents that have discussed HIV/AIDS with a partner, relatively low numbers of respondents reported using a condom at last sex (<16%). This result is consistent with other surveys done in Nigeria and suggests that different strategies for reaching subgroups within the population may be needed to increase condom use. Programs that specifically target rural populations, females, and unmarried individuals, as well as disseminate information on where to obtain condoms, are needed to increase condom use in the VISION Project areas. The overall differences between the 2002 baseline data and the 2004 follow-up data in the three outcome indicators are small, suggesting that more needs to be done to increase HIV/AIDS awareness and influence behaviour change. However, the differentials by sub-groups of the population suggest that changes are starting to occur among smaller subgroups within the population. This is consistent with theories of behaviour change that posit that some subgroups may have a more advanced stage of readiness for behaviour change. It is therefore important to investigate if exposure to media campaign is facilitating behaviour change.</p><p>Also of significance is the detected association between education and HIV/AIDS awareness and perceptions about the use of condoms for reducing the risk of HIV transmission, and the lack of detected significance between education and condom use. In all models of this analysis, those knowing where to obtain condoms was positively associated with increased HIV awareness and perception, and the use of condoms. These important outcomes lend insight into how and where future programs should be developed and implemented. Paramount to effecting behaviour change is the development of programs that create educational opportunities to enhance awareness, as well as the development of programs that create awareness about where condoms can be found. Concerted efforts must be made to reach populations that do not readily access media sources, or that are incapable of reading important public health announcements to disseminate information on the importance of using condoms consistently to reduce the risk of HIV transmission. This conclusion is further supported by results that suggest a majority of the community believes it is O.K. to discuss HIV/AIDS on the television and radio. Future efforts should also focus on increasing awareness of, and access to, condoms within the community.</p><p>As with any data collection and evaluation exercise, there exists limitations in both the data and study design that may distort the true effect of the program on the outcomes investigated. The VISION Project was designed as a full-coverage program; as such, it was not possible to evaluate the effect of the program in relation to an equivalent comparison group. That is, we were unable to find a control group that would not be exposed to the intervention; because many of the partner NGOs were also active outside of the VISION Project target areas, contamination of the control areas was unavoidable. Hence, we decided on the use of a post-test design, and statistical techniques, to control for the potential limiting effects of extraneous variables in the analysis. That being said, in the absence of additional evidence, readers should be cautious when trying to generalize these results to other, and possibly dissimilar, areas.</p><p>There is evidence to suggest that program exposure is increasing the communities' willingness to discuss HIV/AIDS with partners, as well as increasing the communities' knowledge about the benefits of consistent use of condoms for reducing HIV/AIDS risk. Less clear however, is the effect of the respective programs on condom use, as evidenced by insignificant estimated values, and differences in the magnitude of the effect between models. One possible explanation is that it may take much longer for the radio and TV programs to effect condom use behaviour change within the community. As state above, a second possibility is that our study design failed to capture the true effect, as a quasi-experimental design with a comparison group is better than a post-test design for teasing out the controlled effects of program exposure on behaviour-related outcomes. In either case, this analysis suggests that a necessary next step is to specifically target groups within the population that are not currently aware of the dangers of HIV/AIDS, or not aware of the availability and location of prevention methods, and then scale up the coverage, content, and distribution of media campaigns in these populations to ensure that HIV/AIDS prevention information is readily available and understandable to the majority of the population.</p></sec><sec><title>Conclusion</title><p>The data shows that the FP/RH media campaign by VISION and its partners is reaching a large portion of the target population, and exposure to these mass media programs can help increase HIV/AIDS awareness. Programs that target rural populations, females, and unmarried individuals, as well as disseminate information on where to obtain condoms, are needed to reduce barriers to condom use. Although the overall differences in the outcome variables between 2002 baseline data and 2004 follow-up data were small, differences in the outcome variables observed in the respective project areas for sub-groups of the population suggest that change is starting to occur. Improvements in HIV/AIDS prevention behaviour are likely to require that these programmatic efforts be continued, scaled up, done in conjunction with other interventions, and targeted towards individuals or areas with specific socio-demographic characteristics.</p></sec><sec><title>Competing interests</title><p>The author(s) declare that they have no competing interests.</p></sec><sec><title>Authors' contributions</title><p>JK analyzed and interpreted the data, and drafted the manuscript. DM designed the study, assisted with the analysis and interpretation of data, and drafted sections of the manuscript. AA led data collection activities, designed the study, and provided substantial content for subsequent revisions to the manuscript. All authors read and approved the final manuscript.</p></sec><sec><title>Pre-publication history</title><p>The pre-publication history for this paper can be accessed here:</p><p><ext-link ext-link-type="uri" xlink:href="http://www.biomedcentral.com/1471-2458/6/123/prepub"/></p></sec>