The design of the Thoratec ventricular assist device

Transcription

1 Clinical Experience With 111 Thoratec Ventricular Assist Devices Lawrence R. McBride, MD, Keith S. Naunheim, MD, Andrew C. Fiore, MD, Debbie A. Moroney, BSN, and Marc T. Swartz, BA Division of Cardiothoracic Surgery, Department of Surgery, Saint Louis University, St. Louis, Missouri Background. Ventricular assist devices (VADs) have gained wider acceptance due to refinements in patient selection and management and device availability. Methods. To evaluate early and late results, we reviewed our first 111 patients with the Thoratec VAD. Results. Forty-four patients were supported for myocardial recovery. The mean age in the recovery group was 51.9 years. There were 18 left VADs (LVADs), 17 biventricular VADs (BVADs), and nine right VADs (RVADs). Complications included bleeding in 20 patients (45%) and device-related infection in 1 patient (2%). Nineteen were weaned from the VAD, with 12 survivors. Sixtyseven patients were supported as a bridge to cardiac transplantation. The mean age was 41.5 years. There were 39 LVADs and 28 BVADs. Complications included bleeding in 21 patients (31%) and device-related infection in 12 (18%). Three patients were weaned and 39 patients were transplanted from the assist device, for a total of 42 bridge survivors. Device-related thromboembolism occurred in 9 patients (8.1%), 7 of whom were bridge to transplantation. The duration of VAD support ranged from 0.1 to 27 days (mean 4.5 days) in the recovery group and 0.2 to 184 days (mean 40.7 days) in the bridge to transplantation group. The 10-year actuarial survival was 16% for the recovery group, 22% for the bridge group, and 33% for transplanted patients. Conclusions. Despite advances, VAD support remains associated with significant morbidity and operative mortality. (Ann Thorac Surg 1999;67:1233 9) 1999 by The Society of Thoracic Surgeons The design of the Thoratec ventricular assist device (VAD) dates back to the mid 1970s, with its origins based on the work of Pierce and Donachy at Pennsylvania State University [1]. For almost 20 years, Thoratec Laboratories Corporation (Pleasanton, CA) has been making improvements upon this initial design, and the current Thoratec VAD has evolved from this process. Changes to the device have included fabrication of the pumping chamber and cannulas from Thoratec s BPS- 215M polyurethane elastomer (Thoralon), improvements in the cannulation as well as the connector systems, and a microprocessor-based pneumatic control console. Saint Louis University has had the Thoratec VAD since 1981 as a result of participation in a National Heart, Lung, and Blood Institute multicentered trial to evaluate VAD support in patients with postcardiotomy shock. For this reason, our initial 5-year experience was heavily weighted towards postcardiotomy (recovery) patients. By 1985, it was apparent that several other groups could potentially benefit from this technology, including those who deteriorate while awaiting cardiac transplantation, acute myocardial infarction shock patients, and those who develop cardiogenic shock as a result of nonischemic cardiomyopathies. Presented at the Forty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Orlando, FL, Nov 12 14, Address reprint requests to Dr McBride, Department of Surgery, Saint Louis University, 3635 Vista Ave at Grand Blvd, St. Louis, MO ; mcbridlr@wpogate.slu.edu. Material and Methods From 1981 through 1996, the protocols for this study were reviewed and approved by the Food and Drug Administration and the Institutional Review Board of Saint Louis University. Since January 1997, the Thoratec VAD has been used in accordance with the premarketing approval labeling of the Food and Drug Administration. Descriptions of the Thoratec device, as well as techniques for insertion, have been previously published [2, 3]. All Thoratec VADs were inserted in beating hearts during normothermic cardiopulmonary bypass. Postcardiotomy patients who received a left VAD (LVAD) underwent left atrial cannulation through purse string sutures in the left atrial appendage, the dome of the left atrium, or anterior to the entrance of the right pulmonary veins. Patients who were considered candidates for cardiac transplantation had left ventricular apex cannulation. All right VADs (RVADs) had inflow cannulas placed in the right atrium with purse string sutures. VAD outflow cannulas were sutured to the ascending aorta or main pulmonary artery. In general, all patients met inclusion, exclusion, and hemodynamic criteria adopted by the National Heart, Lung, and Blood Institute clinical investigation group as well as the Food and Drug Administration. In brief, patients were candidates when optimal preload, maximum inotropic, and/or intraaortic balloon pump (IABP) support resulted in inadequate hemodynamic indices, defined as: (1) cardiac index 1.80 L/m 2 /min; (2) elevated 1999 by The Society of Thoracic Surgeons /99/$20.00 Published by Elsevier Science Inc PII S (99)

2 1234 McBRIDE ET AL Ann Thorac Surg THORATEC VENTRICULAR ASSIST DEVICES 1999;67: Table 1. Anticoagulation Protocol Insertion with CPB, full heparin Aprotinin at implant and removal (one-half Hammersmith) Reverse heparin No anticoagulation for 24 h At 24 h, dextran 25 cc/h or heparin 10 U/kg/h 24 h after starting heparin, increase PTT to 1.5 control Start warfarin as soon as taking oral medications INR (stop heparin when INR 2.5) Add ASA 81 mg/day at days if stable Supplemental heparin (10 U/kg/h) if INR 2.5 ASA aspirin; CPB cardiopulmonary bypass; INR international normalization ratio; PTT partial thromboplastin time. systemic vascular resistance; (3) systolic blood pressure 90 mm Hg; (4) right and/or left atrial pressures 20 mm Hg; and (5) urine output 20 ml/h. After the Thoratec VAD had been cleared by the Food and Drug Administration, these criteria were sometimes relaxed; however, all patients manifested cardiogenic shock, inability to maintain major organ function, and/or refractory ventricular tachyarrhythmias. The recovery group was composed of patients in whom recovery of the natural heart was anticipated. The bridge group consisted of patients thought to have irreversible myocardial damage and in whom cardiac transplantation would be required for survival. These categorizations are somewhat subjective since several patients jumped between groups depending on their ventricular function and clinical circumstances. Determination of perioperative myocardial infarction was made by analysis of serial electrocardiograms, lactate dehydrogenase and creatinine phosphokinase myocardial isoenzymes, troponin levels, and pathological characteristics (biopsy or autopsy). Ventricular function was evaluated by hemodynamic measurements, radionuclide scans, echocardiography, and, in selected patients, cardiac catheterization. Examination of the blood sacs and cannulas were done by gross inspection in all patients, and by light and electron microscopy in selected patients. Survivors were evaluated by echocardiography, nuclear ventriculography multigated acquisition, or cardiac catheterization approximately 1 month after removal of the VAD or heart transplantation. The anticoagulation protocol underwent several modifications during the period of this study. The current anticoagulation strategy is shown in Table 1. All VADs were implanted utilizing full heparinization (activated clotting time greater than 500 sec). In our early experience, postcardiotomy patients would not have postoperative heparin started until the weaning process was initiated. For the past 5 years we have used low-dose heparin initially rather than dextran. Since the majority of patients implanted during the second half of this study were bridge to transplant, we divided the bridge group (67 patients) into an early and late experience. The early experience included patients implanted between 1982 and 1991 (32 patients). The late experience included patients who had devices implanted between 1992 and 1998 (35 patients). The definitions of the complications have been previously published [4]. Data were analyzed with the Statview for Windows statistical software package (Version 4.53; Abacus Conce, Inc, Berkeley, CA). A 2 test was used to determine significance for discrete variables. Continuous variables were analyzed by a two-tailed Student s t test. A p value of less than 0.05 was considered significant. Actuarial analysis was computed with the method of Kaplan and Meier. A log-rank (Mantel-Cox) analysis was used to determine actuarial significance. Results Recovery Patients Forty-four patients ranging in age from 15 to 71 years (mean age years) were supported with the Thoratec VAD with the anticipation that myocardial recovery would occur. There were 36 males and 8 females, with body surface areas ranging from 1.5 to 2.35 m 2 (mean ). Thirty-five patients were supported after reparative cardiac procedures, 4 after cardiac transplantation, 4 after acute myocardial infarction, and 1 developed cardiogenic shock associated with viral myocarditis. The diagnoses for the 44 patients are shown in Table 2. Table 3 shows the type of support and results with each method of support. Thirty-five patients received LVADs, of which 30 had pump inflow via the left atrium and 5 had left ventricular apex cannulation. Mean maximal VAD flows ranged from 1.3 to 6.3 L/min (mean 4.43 L/min), which resulted in VAD flow indices of 0.94 to 2.91 L/m 2 /min (mean 2.08 L/m 2 /min). VAD support ranged from 0.1 to 27 days (mean 4.5 days). Nine patients (20%) died in the operating room, with 3 additional patients surviving less than 24 h. Twenty-two patients (50%) were supported from 1 to 7 days, with 10 survivors. Ten patients were supported longer than 7 days, with 2 survivors. Complications for the recovery group are included in Table 4. Bleeding was the most common complication, occurring in 45% of the patients. Infections of any type occurred in 39% of the patients, with 2% being device related. Renal and respiratory failure each occurred in 23% of the patients. Eleven percent of the patients had thrombus located in the device at the time of explantation or autopsy. Thromboembolism occurred in 9% of the patients, however, only 5% (2 patients) were thought to be device related. Mechanical failure occurred in 1 patient and was readily corrected. The causes of death in the 32 nonsurvivors were most commonly attributed to a combination of cardiac failure (biventricular failure), renal failure, bleeding, and sepsis. Twenty-five of the 32 nonsurvivors underwent complete postmortem examinations. Eighteen (72%) of these 25 patients showed pathological evidence of acute or evolving myocardial infarction. Of the 44 patients in the recovery group, 19 (43%) (18

3 Ann Thorac Surg McBRIDE ET AL 1999;67: THORATEC VENTRICULAR ASSIST DEVICES 1235 Table 2. Diagnoses Recovery (n 44) No. of Patients Bridge to Transplantation (n 67) No. of Patients Postcardiotomy Ischemic cardiomyopathy 26 Coronary artery disease 19 Idiopathic cardiomyopathy 25 Mitral valve and coronary disease 7 Postpartal cardiomyopathy 6 Left ventricular aneurysm and 3 Congenital heart defects 2 coronary disease Aortic valve disease 3 Retransplantation 2 Congenital heart defects 2 Valvular cardiomyopathy 3 Ascending aortic aneurysm 1 Viral cardiomyopathy 3 Acute myocardial infarction 4 Shock after heart transplantation 4 Viral myocarditis 1 postcardiotomy, 1 viral myocarditis) were weaned and 12 (27%) patients were discharged. The 12 survivors were followed for 6 to 186 months after hospital discharge. There were 6 late deaths occurring at 6, 21, 22, 54, 92, and 136 months. Only 1 of these 6 late deaths was not cardiac related. Six patients are alive (mean 140 months). The survivors are currently New York Heart Association class I or II. Sixty-seven percent of the recovery group died within 1 month of device implant. Actuarial survival for the recovery group (16% at 10 years) is shown in Figure 1. Bridge to Cardiac Transplantation Since 1982, 67 patients (21 female, 46 male), 9 to 66 years of age (mean age 41.5 years), have been supported with the Thoratec VAD as a bridge to cardiac transplantation. The diagnoses of these 67 patients are listed in Table 2. Table 3 shows the types of ventricular support. Sixty-two had a left ventricular apex cannulation and 5 had a left atrial cannulation. VAD flows range from 1.45 to 6.6 L/min (mean 5.1 L/min), with VAD flow indices ranging from 0.84 to 3.62 L/m 2 /min (mean 2.76 L/m 2 / min). The durations ranged from 0.2 to 184 days (mean 40.7 days). Complications for the bridge group are listed in Table 4. The most common complication was infection occurring in 49% (33 patients), with 15% (10 patients) having device-related infections. Other frequent complications included bleeding and respiratory failure. Twenty-four percent (16 patients) had thrombus visible in the device at the time of the device explantation. However, only 19% (13 patients) had evidence of thromboemboli documented by computed tomography scan, autopsy, or clinical observation of transient ischemic attacks. Ten percent (7 patients) were considered to have device-related thromboemboli; however, only 5 of these 7 patients had thrombus identified in the device at the time of explantation and had confirmation of a thromboembolic event by one of the previously mentioned methods. Two patients had a neurologic deficit lasting greater than 24 h as a result of thromboemboli. One patient s neurologic deficit resolved within 3 days and the other recovered completely within 1 month of transplantation. Hemolysis occurred in 16% of the patients; however, it was often difficult to positively identify the VAD as the cause. Mechanical failures, although occurring in 18% (12 patients), were usually minor inconveniences that resulted in no harm to any patient. Thirty-nine of 67 (58%) patients were successfully transplanted and discharged (Table 3). Three additional bridge patients were weaned from the devices and discharged. Twenty-five (37%) patients developed complications that excluded them from cardiac transplantation and died during VAD support. The 42 long-term survivors were followed from 1 to 163 months (mean 34.6 months). The actuarial survival of the bridge group is shown in Figure 1. Like the recovery group, a significant percentage of bridge patients (33%) died within 1 month of VAD implant. The 1-, 5-, and 10-year actuarial survival Table 3. Type of Support and Results Device Recovery (n 44) Weaned Survived Bridge to Transplantation (n 67) Weaned Transplantation Survived LVAD LVAD and centrifugal RVAD RVAD and IABP BVAD Total BVAD biventricular assist device; IABP intraaortic balloon pump; LVAD left ventricular assist device; RVAD right ventricular assist device.

4 1236 McBRIDE ET AL Ann Thorac Surg THORATEC VENTRICULAR ASSIST DEVICES 1999;67: Table 4. Complications Recovery (n 44) Bridge (%) (n 67) Variable Percent No. Percent No. p Value Bleeding NS Renal failure Infection NS Device related Respiratory failure NS Thrombus in device NS Thromboemboli NS Device related NS Neurologic deficit NS Hemolysis NS Mechanical failure NS not significant; for the 39 patients transplanted was 89%, 85%, and 33%. There were 11 late deaths in the bridge group at 4, 6, 12, 15, 50, 72, 76, 79, 85, 87, and 110 months. Most of these late deaths were attributed to accepted posttransplant processes. Chest tube drainage in the bridge to transplant group decreased from 1,489 1,080 cc/m 2 body surface area early in our experience, to cc/m 2 body surface area more recently ( p 0.05). For the same periods, the amount of packed red blood cell transfusions has decreased from to ( p 0.01) per patient (Table 5). Discussion patients. Since our first clinical Thoratec VAD implant in March 1982, there have been significant improvements in the field of mechanical circulatory support. Improved patient selection criteria has been one of the most important advances, and many of these guidelines were established early [5, 6]. Another key factor affecting success has been refinements in the prevention and treatment of complications. Improvements in anticoagulation and antibiotic protocols have decreased the incidence and severity of thromboembolism and infection and permitted longer durations of support [7 9]. We realized early that biventricular failure and the need for biventricular support was more common and more important than previously thought. Between 1982 and 1985, there were numerous postcardiotomy patients who received only LVADs and died of right heart failure [10]. Fifty-nine percent of the recovery group received biventricular support (biventricular VAD [BVAD] or RVAD and IABP). Forty-two percent of the bridge to transplant group received biventricular support in the form of BVADs. This reduction in biventricular support was due to different patient populations, decreased use of isolated RVADs, and improvements in the understanding and management of right ventricular dysfunction [11 13]. Fig 1. Actuarial survival for recovery and bridge to transplant groups. Whenever possible, we perform left ventricular apex cannulation. Left ventricular apex cannulation provides higher LVAD flows with lower preload levels. This results in superior washing of the blood sac and valves, while allowing more physiologic cardiac filling pressures [14]. We were able to successfully place a left ventricular apex cannula in a 40-kg 9-year-old boy and maintain VAD flows in excess of 4.5 to 5.0 L/min. The 3 patients weaned in the bridge to transplant group had left ventricular apex cannulation. The cannulation site was oversewn at the time of device removal, and left ventricular ejection fractions continued to improve to above 55% within 2 months of device removal. Bleeding is a common complication that occurred in 45% of the recovery group and 31% of the bridge group. Preoperative management of patients receiving anticoagulants, improvements in device insertion techniques, including better graft preclotting procedures, as well as the use of Aprotinin, and more fresh frozen plasma has led to a reduction in bleeding. While the actual incidence of bleeding (percentage of patients) has changed little over the past 16 years, chest tube drainage and the requirements for blood product transfusions have decreased. Device-related infections fell from 22% in the period 1982 to 1992, to 15% for 1993 to This decrease in the device-related infections occurred despite a significant increase in the average duration of support. The longer durations of support increase the risk of not only device- Table 5. Device-Related Morbidity: Bridge Group Only Variable Early ( ) n 32 Late ( ) n 35 p Value Bleeding (cc/m 2 ) 1, Infection 7 (22%) 5 (15%) NS Thromboembolism 2 (6%) 5 (14%) NS Mechanical problem 6 (19%) 6 (17%) NS Duration of support 24.6 days 57.3 days patients.

5 Ann Thorac Surg McBRIDE ET AL 1999;67: THORATEC VENTRICULAR ASSIST DEVICES 1237 related infections but also other infectious complications, such as line sepsis from catheters required for antibiotic or anticoagulant treatment [15]. Improved education and nursing care of the cannula exit sites along with routine use of antibiotic irrigations have decreased the incidence and severity of device-related infection. To further reduce infection, patients are extubated, mobilized, and have intravenous lines and chest tubes removed as soon as possible. Our experience and that of others suggest that controlled device-related infections do not have a significant impact on survival in the bridge to transplant population [15, 16]. At the time of device explant, thrombus was identified in 21 of the 149 (14%) Thoratec devices, resulting in a patient complication rate of 19% (21 of 111). Fifteen (71%) of the 21 patients who had thrombus identified in the device at the time of explantation had no evidence of thromboembolism, whereas 6 of 9 (66%) patients who had confirmed device-related thromboemboli had thrombus found in the device. It seems that the development of the thrombus within the device does not necessarily lead to embolization or, if embolization occurs, it is often clinically insignificant and undetectable. At the same time, if a documented thromboembolic event does occur, it is possible no residual evidence would be left within the device. All patients on our transplant list, whether at home, in the hospital receiving inotropic support, or those with assist devices, must meet similar criteria before cardiac transplantation will be undertaken. Patients must be ambulatory, free of significant infection, able to take adequate nutrition, and capable of maintaining normal major organ function. These criteria have resulted in a 100% posttransplant survival rate (30 day) for our Thoratec bridge patients, refuting any criticism that organs are being wasted on less than perfect candidates. The 30-day survival rate for patients transplanted without assist devices over the same time interval at our center was 94% (NS). The 10-year actuarial survival rate for the 39 patients undergoing cardiac transplantation was 33%. This is consistent with the 10-year actuarial survival of patients undergoing cardiac transplantation from Registry data [17]. This information supports the concept that patients undergoing bridge to cardiac transplantation are at no greater risk for early or late posttransplant death than patients undergoing transplantation who are not bridged. While there have been technical refinements in the Thoratec VAD over the duration of this study, the system being used today is not very different from that originally designed to be used in postcardiotomy patients for durations of 7 to 14 days. Fortunately, Pierce and Donachy were gifted engineers, who overdesigned their device well beyond its intended use. Most of the improvements in the field of mechanical circulatory support, however, are attributable not to technology but rather to advances in patient selection and management. The initiation of bridging to cardiac transplantation necessitated that patients be supported for extended durations. To reduce complications and cost, it became necessary to establish protocols for long-term management of anticoagulation, infection, and device operation. Noninvasive techniques to evaluate volume and perfusion status were perfected, and patients were eventually transferred to noncritical care areas or home. Our experience with the Thoratec VAD suggests that it is well suited for bridging to cardiac transplantation. Unlike implantable left ventricular assist systems, it can be placed in smaller patients and is appropriate for patients known to have severe biventricular failure. Our longest duration of support was 184 days; however, we had 2 other patients supported greater than 150 days, all of whom, from all indications, could have been supported indefinitely. While there has been a considerable amount of information gained as the result of this experience, because the study has been extended over a long duration with such a diverse patient population, it is difficult to show a positive trend in the learning curve. In a way, we are victims of our own success. As we learn to deal more effectively with problems, we extend our patient entry criteria to include sicker and more complicated cases. Twenty-three of 111 patients (21%) were on extracorporeal membrane oxygenation for acute cardiac decompensation before VAD placement. Fourteen of these 23 (61%) were during the past 3 years. Five additional patients were transferred to our center while being supported by either centrifugal pumps (3 patients) or an Abiomed (2 patients) during the same 3-year period ( ). Comparison of device-related morbidity for early versus late bridge groups is shown in Table 5. We have been able to show improvements in treating bleeding complications by reducing postoperative chest tube drainage. There has also been a reduction in the incidence and severity of device-related infections; however, this was not statistically significant. Unfortunately, thromboembolic complications actually increased, but this also was not statistically significant. Mechanical problems remain fairly constant. During this time the mean duration of support increased from 24.6 to 57.3 days. Since many device-related complications are strongly correlated with longer durations of support, we evaluated device-related morbidity for these two groups using an actuarial freedom from events analysis. This analysis is shown in Figure 2. Both groups (early and late) developed significant device-related complications; however, the more recent group (late) had fewer device-related complications than the early group ( p 0.02). Data from Table 5 and Figure 2 suggest that, despite experience and efforts to reduce morbidity, complications remained common occurrences. These data suggest that the clinical application of permanent VADs be reconsidered. Our similar experiences with the Novacor LVAD and the experiences of others using the Heartmate device support this hesitancy [18, 19]. The Thoratec VAD has been the workhorse of our mechanical circulatory support program for almost two decades. During that time it has proven repeatedly its versatility and durability in a complex clinical setting. In

6 1238 McBRIDE ET AL Ann Thorac Surg THORATEC VENTRICULAR ASSIST DEVICES 1999;67: Fig 2. Actuarial freedom from event analysis for device-related events in the bridge to transplant group (early vs late [recent] experiences). the near future a portable drive system will be available that will allow greater patient mobility and eventual hospital discharge. While the Thoratec device in its current configuration may never attain the status of a permanent device, its place in the history of cardiac replacement has been well established. Supported in part by the National Heart, Lung, and Blood Institute, Grant No. NO1-HV References 1. Donachy JH, Landis DL, Rosenberg G, et al. Design and evaluation of a left ventricular assist device: the angle port pump. In: Unger F, ed. Assisted circulation. Berlin: Springer- Verlag, 1979, Farrar DJ, Hill JD, Gray LA, et al. Heterotopic prosthetic ventricles as a bridge to cardiac transplantation. A multicenter study in 29 patients. N Engl J Med 1988;318: Ganzel BL, Gray LA, Slater AD, Mavroudis C. Surgical techniques for the implantation of heterotopic prosthetic ventricles. Ann Thorac Surg 1989;47: Pennington DG, McBride LR, Peigh PS, Miller LW, Swartz MT. Eight years experience with bridging to cardiac transplantation. J Thoracic Cardiovasc Surg 1994;107: Pennington DG, Joyce LD, Cabrol C, et al. Patient selection (panel discussion). Ann Thorac Surg 1989;47: Farrar DJ. Preoperative predictors of survival in patients with Thoratec ventricular assist devices as a bridge to heart transplantation. J Heart Lung Transplant 1994;13: Copeland J. The biomaterial-blood interface in circulatory support devices: a cardiac surgeon s view. In: Frazier DH, Graham T, Hill JD, Lewis T, Pennington DG, eds. Mechanical circulatory support. London: Edward Arnold, 1995: Szukalski EA, Reedy JE, Pennington DG, et al. Oral anticoagulation in patients with ventricular assist devices. ASAIO Trans 1990;36:M Didisheim P, Olsen DB, Farrar DJ, et al. Infections and thromboembolism with implantable cardiovascular devices. ASAIO Trans 1989;35: Pennington DG, McBride LR, Swartz MT, et al. Use of the Pierce-Donachy ventricular assist device in patients with cardiogenic shock after cardiac operations. Ann Thorac Surg 1989;47: Kormos RL, Gasior T, Antaki J, et al. Evaluation of right ventricular function during clinical left ventricular assistance. ASAIO Trans 1989;35: Farrar DJ. Ventricular interactions during mechanical circulatory support. Semin Thorac Cardiovasc Surg 1994;6: Woodard JC, Chow E, Farrar DJ. Isolated ventricular systolic interaction during transient reductions in left ventricular pressure. Circ Res 1992;70: Lohmann DP, Swartz MT, Pennington DG, et al. Left ventricular (LV) vs. left atrial (LA) cannulation for the Thoratec ventricular assist device (VAD). Trans Am Soc Artif Intern Organs 1990;36: McBride LR, Swartz MT, Reedy JE, Miller LW, Pennington DG. Device-related infections in patients supported with mechanical circulatory support devices greater than 30 days. Trans Am Soc Artif Intern Organs 1991;37:M Argenziano M, Catanese KA, Moazami N, et al. The influence of infection on survival and successful transplantation in patients with left ventricular assist devices. J Heart Lung Transplant 1997;16: Hosenpud JD, Bennett LE, Keck BM, Fiol B, and Novick RJ. The Registry of the International Society for Heart and Lung Transplantation: 14th Official Report J Heart Lung Transplant 1997;16: Pennington DG, Swartz MT, Lohmann DP, McBride LR. Cardiac assist devices. Surg Clin North Am 1998;78: McCarthy PM, Smedira NO, Vargo RL, et al. One hundred patients with the Heartmate left ventricular assist device: evolving conce and technology. J Thorac Cardiovasc Surg 1998;115: DISCUSSION DR D. GLENN PENNINGTON (Winston-Salem, NC): Larry, congratulations. That is really an outstanding study. In spite of my involvement early on, you were the primary driving force and have really been the one responsible for these good results. I just have a couple of questions. What was the incidence of need for biventricular support, particularly in the bridge group? And tell me also if you thought any of those patients with severe biventricular failure might have been better served with a total heart replacement, such as the Cardiowest. All of these patients were kept in the hospital during the time of VAD implantation, weren t they? DR MCBRIDE: Yes. DR PENNINGTON: Of course, now there is a portable pump for the Thoratec being used in Europe and Canada. I wonder how many of these patients you thought could have gone home if you had that portable device available? DR MICHAEL A. GREENE (Melbourne, FL): I wanted to ask you a question about the bridge patients. Did you identify any problem with elevated percent reactive antibodies (PRA) levels after a long period of support, and if so, do you do anything to intervene before you transplant? DR MCBRIDE: Yes. We have had some patients that have had elevated PRAs. In this particular group, 7 out of 67 patients had