Novel Minimally Invasive, Intrapericardial Implantable Cardioverter Defibrillator Coil System: A Useful Approach to Arrhythmia Therapy in Children

Transcription

1 Novel Minimally Invasive, Intrapericardial Implantable Cardioverter Defibrillator Coil System: A Useful Approach to Arrhythmia Therapy in Children Tain-Yen Hsia, MD, Scott M. Bradley, MD, Martin J. LaPage, MD, Sean Whelan, MD, J. Philip Saul, MD, Jeremy M. Ringewald, MD, and John H. Reed, MD Section of Pediatric Cardiothoracic Surgery, Division of Pediatric Cardiology, Medical University of South Carolina, Charleston, South Carolina Background. Current approaches to implantable cardioverter defibrillator (ICD) implantation in children remain challenging. Transvenous access may be limited due to patient size or anatomy, while epicardial patches require sternotomy or thoracotomy. We present an alternative approach; minimally invasive placement of a transvenous ICD coil within the pericardial space with active fixation. Methods. Between August 2005 and October 2008, 7 children meeting indications for ICD therapy for ventricular tachyarrhythmias underwent intrapericardial placement of an ICD coil system. Median age was 5 years (range, 1 to 17), weight was 14 kg (range, 8 to 46), and 4 patients weighed less than 20 kg. The ICD system was composed of a single or dual defibrillation coil, an active can, and either ventricular or dual chamber epicardial sense-pace leads. All implantations were performed through a small subxiphoid incision and pericardial window without sternotomy. The coil lead was actively fixated in the transverse sinus under fluoroscopic guidance, and the generator placed in a subrectus pocket in the upper abdomen through the same incision. Results. There were no perioperative complications, and no early or late deaths. All implants had acceptable defibrillation energies (range, 5 to 15 J) that were successfully tested intraoperatively, and none required energy increase or lead revision during follow-up (range, 1 to 39 months; median, 20 months). Impedance between the active can and the defibrillation coil remained stable in all. There were no inappropriate discharges. Thirty-six successful ICD discharges in 4 patients were recorded. Two patients underwent subsequent orthotopic heart transplantation and ICD system removal. Conclusions. Intrapericardial placement of an ICD coil system can be safely and successfully carried out through a minimally invasive subxiphoid approach in small infants and children. This novel ICD configuration demonstrates excellent performance, and provides a particularly efficacious approach to ventricular tachyarrhythmia therapy in pediatric patients. (Ann Thorac Surg 2009;87:1234 9) 2009 by The Society of Thoracic Surgeons Accepted for publication Jan 7, Presented at the Fifty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 5 8, Address correspondence to Dr Hsia, Cardiothoracic Surgery, Medical University of South Carolina, 96 Jonathan Lucas St, CSB 409 Charleston, SC 29425: hsiaty@musc.edu. Implantable cardioverter-defibrillator (ICD) therapy has been adopted increasingly in the pediatric population for prevention of sudden cardiac death associated with congenital heart disease (CHD), inherited arrhythmogenic disease, and cardiomyopathy [1, 2]. Despite improved transvenous technology and device miniaturization, ICD placement in children and patients with CHD continues to be challenging, and carries increased risks of malfunction and complications when compared with placement in adults [3]. The use of standard transvenous ICD lead systems suitable for older children is limited in younger children and infants due to small size, and an increased potential for venous obstruction [4, 5]. Furthermore, in children with CHD (such as single-ventricle defects palliated with Fontan connections), cardiac and venous anatomy may prohibit a transvenous approach. Endocardial ICD leads may also carry a risk of paradoxical emboli in patients with intracardiac shunts. Epicardial patch and subcutaneous ICD systems are the two most common alternatives used in children when a transvenous approach is undesirable or not feasible [6]. Epicardial patch ICD systems require a full sternotomy or thoracotomy, and can lead to constrictive pericarditis [4, 7]. In addition, crinkling and failure of the epicardial patch can result in increases in the defibrillation threshold that require lead revision in up to 25% of patients [8, 9]. Subcutaneous placement of an ICD lead or array in the chest wall avoids extensive operative exposure, but may require at least a partial sternotomy for epicardial sense-pace electrode placement, and another incision for generator placement [10 12] by The Society of Thoracic Surgeons /09/$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg HSIA ET AL 2009;87: INTRAPERICARDIAL ICD SYSTEM 1235 Table 1. Patient and Operative Characteristics Patient Age (years) Weight (kg) Diagnosis Sensing-Pacing Configuration (Chamber) ICD Coil Implant Lowest Defibrillation Energy (Joules) Brugada syndrome Single Fidelis 6931, 52 cm Idiopathic dilated cardiomyopathy Single Fidelis 6931, 52 cm Long QT syndrome Single Fidelis 6931, 52 cm Hypertrophic cardiomyopathy Dual Fidelis 6931, 52 cm Tricuspid atresia, s/p Dual Fidelis 6931, 52 cm 15 Atriopulmonary Fontan Long QT syndrome Single Fidelis 6947, 58 cm Idiopathic dilated cardiomyopathy Single Durata 7122, 60 cm 5 ICD implantable cardioverter defibrillator. Because a subcutaneous lead is no longer within the chest wall, there is concern of a heightened likelihood of external lead trauma and fracture [13, 14]. Finally, initial defibrillation thresholds tend to be higher with a subcutaneous system, and can increase with time, presumably due to the distance and mass between the lead array and the heart both at implantation and with growth [14, 15]. Due to these limitations, we propose a novel ICD configuration wherein a transvenous defibrillation coil is actively fixed within the transverse sinus, under fluoroscopic guidance through a subxiphoid pericardial window. Using the same minimally invasive incision, both the epicardial single or dual chamber sense-pace electrodes, and the abdominal ICD generator, are placed. The aim of this report is to assess the early results of adopting this intrapericardial ICD strategy in small children and patients in whom a transvenous approach is limited or not feasible. Fig 1. Single transvenous defibrillation coil with active fixation helix screw lead (Durata, St. Jude Medical, St. Paul, MN). Patients and Methods Patients Between August 2005 and October 2008, 7 patients with indications for ICD therapy as primary or secondary prevention of sudden cardiac death underwent placement of an intrapericardial ICD system through a minimally invasive subxiphoid approach. Patients selected for this approach were not amenable to transvenous ICD lead placement due to small size, poor venous access, or having a Fontan circulation. Patient diagnoses were long QT syndrome in 2, idiopathic dilated cardiomyopathy in 2, Brugada syndrome, hypertrophic cardiomyopathy, and tricuspid atresia palliated with atriopulmonary Fontan connection (Table 1). Four patients had inducible ventricular tachyarrhythmia elicited during electrophysiologic study, and 3 had unexplained syncope related to inherited arrhythmogenic diseases. Only 1 patient, with dilated cardiomyopathy, had a major preimplantation ventricular tachycardia requiring electrical cardioversion and intravenous lidocaine drip therapy. The Institutional Review Board of the Medical University of South Carolina approved this review and waived the need for parental consent for the study. Surgical Technique All patients underwent general endotracheal anesthesia, and the procedures were performed in the operating room. Through a subxiphoid incision, the xiphoid process was divided, and a limited pericardial window was created. Under fluoroscopic guidance, a transvenous ICD coil was directed posteriorly behind the heart, and superiorly toward the transverse sinus. In one patient with a previous sternotomy, moderate intrapericardial adhesions were taken down. Five patients received the Fidelis single coil (model 6931; Medtronic Inc, Minneapolis, MN), 1 the Fidelis dual coil (model 6947), and 1 the Durata single coil (model 7122; St. Jude Medical, St. Paul, MN). In the patient in whom the dual coil was used (due to manufacturer recall of the single coil), the distal coil was implanted as described above, and the proximal coil was capped. Once resistance to further advancement was met, the active fixation helix was screwed into the pericardium by turning the connector pin with the fixation tool until the screw was fully extended (Fig 1). Gentle tugging on the coil confirmed successful fixation to the pericardium. The coil was further secured to the diaphragmatic pericardium by two interrupted sutures placed through a suture sleeve. Atrial and ventricular epicardial steroid eluting unipolar or bipolar electrodes (Medtronic Inc) were placed for either single or dual

3 1236 HSIA ET AL Ann Thorac Surg INTRAPERICARDIAL ICD SYSTEM 2009;87: placed within the pocket. Modified defibrillation threshold testing was performed demonstrating at least two successful conversions from ventricular fibrillation to sinus rhythm. A pericardial drainage tube was not routinely used, and a postoperative chest X-ray confirmed adequate coil location (Fig 2). The ICD system was interrogated prior to patient discharge to ensure stable coil to generator impedance. Data Collection and Follow-Up All records were retrospectively reviewed. Details of postimplant defibrillation testing, ICD device interrogation, and appropriate-inappropriate shocks during follow-up were analyzed. All patients were followed within the Children s Heart Program of South Carolina, a pediatric cardiology-cardiac surgery network providing coordinated congenital cardiac care to the state. Results At the time of the ICD system placement, the median patient age was 5 years, with a range of 1 to 17 years. A median patient weight was 14 kg, with a range of 8 to 46 kg; 4 patients weighed less than 20 kg. Median follow-up was 20 months, ranging from 1 to 39 months. There were no perioperative complications, and no early or late deaths. All patients were extubated on the day of surgery. No patient had recorded premature ventricular beats intraoperatively or prior to discharge. All ICD systems had acceptable defibrillation thresholds, with lowest energy tested at 5 J in 1,10 J in 2,and 15 J in 4. No patient required lead revision or device energy increase. One of the patients with lowest energy tested at 15 J was brought back for repeat defibrillation threshold testing and was successful defibrillated at 5 J twice. There were no increases in the impedance between the coil and the generator; two patients had decreased impedance (65 to 59 and 70 to 59 ). There were no inappropriate discharges. Four patients had a total of 36 appropriate shocks. Each shock successfully converted the rhythm to sinus. The hypertrophic cardiomyopathy patient was shocked once, and the long QT syndrome patients were shocked 13 times and 5 times, respectively. One patient with incessant ventricular tachycardia-fibrillation secondary to end-stage idiopathic dilated cardiomyopathy was successfully shocked 17 times prior to initiation of extracorporeal membrane oxygenation support. He and another patient underwent subsequent successful orthotopic heart transplantation and ICD system removal. Fig 2. Post intrapericardial implantable cardioverter defibrillator placement (A) anteroposterior, and (B) lateral chest X-ray demonstrating location of the single defibrillation coil in the transverse sinus. chamber pacing-sensing. Through the same incision, a subrectus pocket was made in the right upper abdominal wall to accommodate the ICD generator (EnTrust D154VR; Medtronic Inc). After the generator was connected to the ICD coil and pace-sense electrodes, it was Comment This report shows that an intrapericardial placement of an ICD coil system can be safely and successfully carried out through a minimally invasive subxiphoid approach in small infants and children. In this small series of 7 patients, including 2 patients less than 1 year of age and 1 with previous sternotomy, this novel ICD configuration was employed without perioperative complications, and provided excellent performance during early fol-

4 Ann Thorac Surg HSIA ET AL 2009;87: INTRAPERICARDIAL ICD SYSTEM 1237 low-up. It successfully prevented sudden cardiac death in 4 patients, and none had inappropriate discharges. Our early experience was also free of ICD malfunction and lead failure, which have plagued other alternative ICD configurations. A recent review of a multiinstitutional ICD registry of 443 pediatric and congenital heart disease patients reported 64 perioperative (30- day) complications in 55 patients; 21% of the patients had inappropriate shocks with a mean of 6 per patient [2]. Lead failure and malfunction accounted for the majority of these complications. The indications for ICD therapy in children as primary prevention remains controversial, and the decision to implant an ICD in an asymptomatic child with increased risk for sudden cardiac death is often made more difficult by the lack of an efficacious and reliable ICD configuration [6]. While placement in children and congenital heart disease patients only comprise 1% of all ICD placements, several studies have documented a disproportionately high rate of complications and technical difficulties [16 19]. Link and colleagues [3] found a significantly higher rate of infection and lead malfunction in children receiving ICDs, when compared with adults in the same institution. Alexander and colleagues [19] reported a 38% complication rate in a 2-year follow-up of 76 children and congenital heart disease patients. They concluded that growth was strongly associated with lead failure, with a change in body surface area having the highest hazard ratio. The increased morbidity of ICD therapy in children affects the risk-benefit ratio, and may influence the decision to employ ICD implantation as primary prevention in small children. Several creative approaches to ICD placement in children have been introduced over the last few years, due to dissatisfaction with epicardial patch and subcutaneous lead strategies. Fischbach and colleagues [20] first introduced the use of a single defibrillation coil with an abdominally placed generator in children. However, the coil was transvenously introduced into the right ventricle, and the patient size was limited to over 20 kg. Kantoch and colleagues [21] directly placed a single active fixation coil into the right ventricle through a right atrial appendage purse string in a 3.5 year old with hypertrophic cardiomyopathy during concurrent surgical myectomy. Independently, Molina and Benditt [22] and Snyder and colleagues [23] proposed epicardial defibrillation leads through a subxiphoid exposure. In the former report, a transvenously placed right atrial shocking lead was needed, in addition to a halo-shaped coil lead placed on the diaphragmatic surface. In the latter report, a single coil was looped around both the right and left ventricles and fixed to the epicardium. This coil configuration can cause cardiac constriction, and may result in lead migration and failure with growth. Cannon and colleagues [4] described placement of a subcutaneous coil within the pericardium, without active fixation, through a full median sternotomy in 6 patients. Five of these patients were redo sternotomies, and one had an unacceptably high defibrillation threshold, requiring immediate addition of a second coil. Perhaps due to the lack of active fixation, the defibrillation thresholds were higher than in our experience (11 to 35 J), and 5 inappropriate device discharges were recorded in 3 patients during follow-up. More recently, Stephenson and colleagues [15] reported a multiinstitutional review of both subcutaneous and epicardial ICD systems. However, exposure was through a thoracotomy in the epicardial configuration, and the details of the surgical technique were unclear, including the intrapericardial location of the defibrillation coil. The authors reported difficulties in inducing ventricular tachyarrhythmia for device testing, high initial defibrillation thresholds, and a high number of inappropriate postimplant shocks. Compared with other ICD systems designed for children, the current intrapericardial ICD configuration and implantation technique has some advantages. All coils, ICD generators, and pacing-sensing electrodes were implanted through a single upper abdominal incision without full or partial sternotomy. In contrast, subcutaneous coil-lead strategy usually requires at least one separate incision for placement of epicardial pacing-sensing electrodes and the ICD generator. In a 4 patient case report from Kaltman and colleagues [24], where a single incision was applied, the subcutaneous ICD system failed to cardiovert one ventricular fibrillation episode, and falsely detected another. Because upper chest wall placement of the ICD generator in small children is limited, positioning the coil in the transverse sinus and the ICD generator in the upper abdomen provides an effective defibrillation vector and cardioversion energy field. This is supported by the general stability of both defibrillation energy and generator-coil impedance in our patients during postimplant ICD interrogations. Utilizing active coil fixation with the helix screw electrode and additional pericardial sutures, ICD malfunction or lead failure due to coil migration may be minimized, even with growth. By securing a posterosuperior location within the pericardium, the potential complication of cardiac strangulation associated with epicardial pacing electrodes can be avoided [25]. An intrapericardial ICD system also has some disadvantages. Placing the coil in the pericardial space can lead to pericarditis and adhesions. While no patient developed constrictive pericarditis in our series, 1 patient had dense pericardial adhesions encountered at postimplant heart transplantation. Except for a transvenous ICD system, all alternative ICD lead implantation techniques require at least a small pericardiotomy to place an epicardial ventricular sensing electrode, and thus carry the potential for adhesion formation and pericarditis. While proximity of the ICD coil to the epicardium may lead to irritation of diseased myocardium, and risk inducing ventricular arrhythmias, none of the patients had observed premature ventricular beats or tachyarrhythmias prior to discharge. The presence of dense pericardial adhesions from a previous sternotomy is likely the only relative anatomic exclusion criterion for the use of an intrapericardial ICD coil. However, as demonstrated by the successful coil implantation in a patient with a

5 1238 HSIA ET AL Ann Thorac Surg INTRAPERICARDIAL ICD SYSTEM 2009;87: previous atriopulmonary Fontan, having a previous sternotomy is not an absolute contraindication. In conclusion, intrapericardial placement of a defibrillation coil provides a safe and efficacious approach to ICD therapy in small children and congenital heart disease patients in whom a transvenous system is not feasible or desirable. It is applicable in infants, and in patients with a previous sternotomy. The minimally invasive subxiphoid approach, with fluoroscopic guidance of the coil into the transverse sinus and active lead fixation, combined with an abdominal ICD generator, allows an efficient defibrillation vector. This novel ICD configuration has demonstrated excellent performance with low and stable defibrillator energy and impedance, and good reliability with no inappropriate discharges or lead failures. The current results compare favorably with other nontransvenous ICD strategies. References 1. Friedman RA, Garson A, Jr. Implantable defibrillators in children: from whence to shock. J Cardiovasc Electrophysiol 2001;12: Berul CI, Van Hare GF, Kertesz NJ, et al. Results of a multicenter retrospective implantable cardioverter-defibrillator registry of pediatric and congenital heart disease patients. J Am Coll Cardiol 2008;51: Link MS, Hill SL, Cliff DL, et al. Comparison of frequency of complications of implantable cardioverter- defibrillators in children versus adults. Am J Cardiol 1999;83:263 6, A Cannon BC, Friedman RA, Fenrich AL, Fraser CD, McKenzie ED, Kertesz NJ. Innovative techniques for placement of implantable cardioverter-defibrillator leads in patients with limited venous access to the heart. Pacing Clin Electrophysiol 2006;29: Bar-Cohen Y, Berul CI, Alexander ME, et al. Age, size, and lead factors alone do not predict venous obstruction in children and young adults with transvenous lead systems. J Cardiovasc Electrophysiol 2006;17: Blom NA. Implantable cardioverter-defibrillators in children. Pacing Clin Electrophysiol 2008;31(suppl 1):S Chevalier P, Moncada E, Canu G, Claudel JP, Bellon C, Kirkorian G, et al. Symptomatic pericardial disease associated with patch electrodes of the automatic implantable cardioverter defibrillator: an underestimated complication? Pacing Clin Electrophysiol 1996;19(12 pt 1): Tummala RV, Riggio DR, Peters RW, Shorofsky SR, Gold MR. Chronic rise in defibrillation threshold with a hybrid lead system. Am J Cardiol 1996;78: Molina JE, Benditt DG, Adler S. Crinkling of epicardial defibrillator patches. A common and serious problem. J Thorac Cardiovasc Surg 1995;110: Schreiber C, Kostolny M, Eicken A, Lange R. Nonthoracotomy cardioverter defibrillator implantation in a 2-year-old infant with long QT syndrome. Ann Thorac Surg 2006;81:e Greene AE, Berger JT, Heshmat Y, Kuehl KS. Transcutaneous implantation of an internal cardioverter defibrillator in a small infant with recurrent myocardial ischemia and cardiac arrest simulating sudden infant death syndrome. Pacing Clin Electrophysiol 2004;27: Berul CI, Triedman JK, Forbess J, et al. Minimally invasive cardioverter defibrillator implantation for children: an animal model and pediatric case report. Pacing Clin Electrophysiol 2001;24: Kettering K, Mewis C, Dornberger V, et al. Long-term experience with subcutaneous ICD leads: a comparison among three different types of subcutaneous leads. Pacing Clin Electrophysiol 2004;27: Thogersen AM, Helvind M, Jensen T, Andersen JH, Jacobsen JR, Chen X. Implantable cardioverter defibrillator in a 4-month-old infant with cardiac arrest associated with a vascular heart tumor. Pacing Clin Electrophysiol 2001;24: Stephenson EA, Batra AS, Knilans TK, et al. A multicenter experience with novel implantable cardioverter defibrillator configurations in the pediatric and congenital heart disease population. J Cardiovasc Electrophysiol 2006;17: Kaski JP, Tome-Esteban MT, Mead-Regan S, et al. B-type natriuretic peptide predicts disease severity in children with hypertrophic cardiomyopathy. Heart 2008;94: Goel AK, Berger S, Pelech A, Dhala A. Implantable cardioverter defibrillator therapy in children with long QT syndrome. Pediatr Cardiol 2004;25: Eicken A, Kolb C, Lange S, et al. Implantable cardioverter defibrillator (ICD) in children. Int J Cardiol 2006;107: Alexander ME, Cecchin F, Walsh EP, Triedman JK, Bevilacqua LM, Berul CI. Implications of implantable cardioverter defibrillator therapy in congenital heart disease and pediatrics. J Cardiovasc Electrophysiol 2004;15: Fischbach PS, Law IH, Dick M 2nd, Leroy S, Mosca RS, Serwer GA. Use of a single coil transvenous electrode with an abdominally placed implantable cardioverter defibrillator in children. Pacing Clin Electrophysiol 2000;23: Kantoch MJ, Rebeyka IM, Houlden LA, Dyck JD. Direct intracardiac placement of an automatic implantable cardioverter defibrillator coil lead in a small child. Europace 2007;9: Molina JE, Benditt DG. An epicardial subxiphoid implantable defibrillator lead: superior effectiveness after failure of standard implants. Pacing Clin Electrophysiol 2004;27: Snyder CS, Lucas V, Young T, Darling R, Dalal G, Davis JE. Minimally invasive implantation of a cardioverterdefibrillator in a small patient. J Thorac Cardiovasc Surg 2007;133: Kaltman JR, Gaynor JW, Rhodes LA, et al. Subcutaneous array with active can implantable cardioverter defibrillator configuration: a follow-up study. Congenit Heart Dis 2007;2: Eyskens B, Mertens L, Moerman P, Ector H, Daenen W, Gewilling M. Cardiac strangulation, a rare complication of epicardial pacemaker leads during growth. Heart 1997;77: DISCUSSION DR ANDREW J. LODGE (Durham, NC): That was a nice presentation and a nice series with excellent results. We have used a similar approach in four children, all of whom were less than 13 kg. The smallest of these was a 5.5 kg child who was three months old. We had similar success with the exception of one patient that had high pacing and defibrillation thresholds and required the addition of a screw in ventricular lead as well as a subcutaneous patch. In redo patients we have taken the approach, as you have mentioned, of using an epicardial patch in the subcutaneous location with good results, and we haven t seen increases in the defibrillation thresholds. I have a few questions for you, though, mostly related to the operative approach. In your manuscript as well as your talk you made a point that you use a small subxiphoid incision for the entire procedure, but then you also mentioned alternative approaches such as a

6 Ann Thorac Surg HSIA ET AL 2009;87: INTRAPERICARDIAL ICD SYSTEM 1239 subcutaneous approach would require an additional partial sternotomy incision for epicardial lead placement as well as an incision for the generator placement. Why the distinction for that? In other words, why would the alternative approach require different incisions for the leads and the generator when you just use one for yours? And then on a related topic, I have had some difficulty placing dual chamber pacing leads through a subxiphoid incision alone and usually have to divide at least a small part of the lower sternum to place those leads on adequately. I was wondering if you could elaborate a little bit on the operative approach, in particular with relation to the one redo patient that you did, as it seems like from the incision that you describe, taking down adhesions posteriorly all the way up to the transverse sinus may be fairly difficult. Finally, we have avoided your technique in the redo patients, primarily for those reasons. So how feasible do you think it is in general in those patients, because, again, or do you think most redo patients would require an alternative approach? DR HSIA: Those are good questions, Andy, and I appreciate your comments. Regarding your first question, in most of the published papers on subcutaneous ICD [implantable cardioverter defibrillator] lead or array systems, multiple incisions were needed for the ICD lead, epicardial pacing electrodes, and the generator or active CAN [abdominal pulse generator]. In our series of seven patients, a single subxiphoid incision was used. In the talk and manuscript, we were trying to make the point that with the subcutaneous approach, intrapericardial access is still needed for epicardial sense-pace electrode placement. So we believe that our approach is an improvement than the subcutaneous or the epicardial patch technique in that we avoid extensive operation in the chest, and avoid additional incisions that may be required for the subcutaneous approach. This is in addition to better ICD performance and reliability. DR LODGE: But you could use your incision for the epicardial lead placement even if you needed an additional incision for the subcutaneous patch, right, the subxiphoid? DR HSIA: That would be correct. However, the subcutaneous lead placement usually involves tunneling a track in the subcutaneous space along the posterior chest wall, which can be quite uncomfortable and cosmetically unpleasing in a thin infant or child, or one who is failing to thrive. Our point is that the subcutaneous approach does not avoid the mediastinum, and access to the heart is still required. With the intrapericardial approach reported here, the ICD lead can be placed quickly along with the pace-sense electrodes and the generator, through one incision with very good device efficacy. Just a quick answer to your question about redos, on the one patient with previous sternotomy in our series, we were fairly fortunate in that her adhesions on the diaphragmatic level were not prohibitive. We were able to find some openings posteriorly, and track the lead up to the posterior aspect of the heart. This may not be possible in all patients with previous sternotomies. Therefore, as we reported, having previous sternotomy may be a relative contraindication for this technique. However, since placement of sense-pace electrodes mandates exploration within the pericardial space, one can first attempt placing an ICD coil first before resorting to a subcutaneous or other move invasive approaches. DR J. WILLIAM GAYNOR (Philadelphia, PA): I enjoyed that, T Y. I would say one thing. You can perform the subcutaneous implantation. We have reported the technique, both the technique and subsequent intermediate term follow-up. With a small subxiphoid incision, you position the RV [right ventricular] lead and place the device in the right upper quadrant. Using the tunneler you can position a single array that passes around the chest and all the way to the back. You get a very nice defibrillation vector from here to the back. And we have shown that the DFTs [defibrillation thresholds] are very stable, and it has worked very well in small kids. It can be done from one small subxiphoid incision. You don t necessarily need additional incisions when you use a subcutaneous array. DR HSIA: Well, that is a very good point. There are other issues, of course, associated with subcutaneous leads. In small children who are active, and prone to falling or playing contact sports, there is a risk for lead fracture. In most series on subcutaneous ICDs, there are nontrivial complication rates. DR GAYNOR: We have reported a couple years follow-up and haven t had any lead fractures or migration or perforation. So it has been very successful, even in small kids. DR CARL L. BACKER (Chicago, IL): That was a nice presentation, T Y, and I want to rise to agree with your technique. I think this approach is actually a significant step forward. Placement of the defibrillator coil in the transverse sinus is really a tremendous technical advance. We have a very busy EP service and with both the subcutaneous patch and with the coils in the subcutaneous layer we have had some complications. These include bleeding into the pocket and the coils poking out of the skin. This is really a great solution to both of these problems. I recently put an AICD [automatic ICD] in a three-month-old baby. We did not want to send him home until he had the device, and we used this technique and it worked very well. The question I wanted to ask you relates to the choice of where to put the device. In the two babies that I have done I have put the device in the middle and taken the rectus muscle down on both sides. In older patients you can put it under either the right rectus muscle or the left rectus muscle. What is your preference for the location of the device? DR HSIA: Thank you very much, Dr Backer. The simple answer is wherever we can find a space to fit the generator. In the two children who were one year of age or below, we, like you, wound up placing the generator in the midline, and they still had very good defibrillation thresholds and worked fine. So I don t think it is necessary to have the box placed in the right upper quadrant, although that would seem to be the preferred vector. Within the midline the device can still achieve pretty good defibrillation. DR BACKER: For some of the patients you don t have a choice, but, like I said, we have actually had a discussion with our cardiologists regarding the older patient in whom you can fit the device under one rectus muscle or the other. Some of our cardiologists think it should be under the right rectus muscle and some think the left side. What is your preferred approach? DR HSIA: We haven t had any patient where we went to the left side. DR BACKER: So you put them all on the right side? DR HSIA: On the right, or the midline if they are small.