Bridging to Cardiac Transplantation with Pulsatile Ventricular Assist Devices

Transcription

1 ORIGINAL ARTICLES Bridging to Cardiac Transplantation with Pulsatile Ventricular Assist Devices Kirk R. Kanter, M.D., Lawrence R. McBride, M.D., D. Glenn Pennington, M.D., Marc T. Swartz, B.A., Shelly A. Ruzevich, R.N., Leslie W. Miller, M.D., and Vallee L. Willman, M.D. ABSTRACT As cardiac transplantation becomes more commonplace in the treatment of end-stage heart failure and as suitable donors become less available, an increasing number of patients will require mechanical circulatory assistance to bridge to transplantation. Since 1982, refractory hemodynamic instability requiring placement of pulsatile ventricular assist devices (VADs) has developed in 11 candidates for transplantation aged 24 to 54 years (mean, 39.6 years). A pneumatic Pierce-Donachy pump was used in 9 patients and an electrical Novacor pump in 2. The cause of the cardiomyopathy was ischemic in 6, postpartum in 2, idiopathic in 2, and doxorubicin hydrochloride toxicity in 1. Seven patients required left ventricular support (LVAD); 4 required biventricular mechanical support (BVAD). Duration of support ranged from 8 hours to 91 days with flows ranging from 4.1 to 8.5 L/min (mean, 5.5 Llmin). Although hemodynamic stability was achieved in all 11 patients, contraindications to transplantation developed in 5 patients during VAD support (renal failure in 4, sepsis in 3, disseminated intravascular coagulopathy in 1). The remaining 6 patients (4 with an LVAD, 2 with a BVAD) remained good candidates for transplantation despite major complications in 5 (mediastinal bleeding in 3, driveline infection in 3, development of preformed antibodies in 2, small embolic stroke caused by device malfunction in l). The 3 patients who were supported the longest (24, 75, and 91 days) were ambulatory while awaiting a donor heart. All 6 patients underwent successful transplanation after 8 hours to 91 days (mean, 24 days) of support. Other than one sternal wound infection, there were no major complications after transplantation. The 6 survivors have been followed 4 to 30 months (mean, 14 months) with 1 late death due to medical noncompliance; the remaining 5 patients are in New York Heart Association Functional Class I. These results indicate that pulsatile VADs can provide prolonged hemodynamic support prior to transplantation with good preservation of endorgan function. Using strict criteria for determining suit- From the Departments of Surgery and Medicine (Cardiology), St. Louis University Medical Center, St. Louis, MO. Presented at the Thirty-fourth Annual Meeting of the Southern Thoracic Surgical Association, Boca Raton, FL, Nov 5-7, Address reprint requests to Dr. Kanter, Department of Surgery (Cardiothoracic), Emory Clinic, 1365 Clifton Rd, NE, Atlanta, GA ability for transplantation, successful short-term outcome in this series was uniform. Cardiac transplantation has become a widely accepted therapy for the treatment of end-stage heart failure with more than 1,400 transplantations performed in 1986 [l]. However, approximately 20% of suitable transplant candidates die before a donor heart can be obtained [2]. Although many of these potential recipients die out of hospital, hemodynamic deterioration develops in a significant number. This condition potentially can be temporarily reversed with inotropic agents and mechanical circulatory support until a suitable donor is located. The concept of temporary mechanical circulatory support as a "bridge" to cardiac transplantation was introduced clinically by Cooley and co-workers [3] in 1969 with the use of the Liotta artificial heart prior to transplantation. The first use of a ventricular assist device (VAD) for mechanical support prior to transplantation is credited to Norman and colleagues [4], who supported a patient for 5 days with an abdominal left VAD (LVAD) prior to orthotopic transplantation. Both of these patients died of infection and multiorgan failure. To our knowledge, the first hospital survivors of mechanical assistance for bridging to transplantation were reported by Reemtsma and associates [5] in 1978; they supported 3 patients with intraaortic balloon counterpulsation prior to successful transplantation. Although the efficacy of intraaortic balloon counterpulsation for temporary hemodynamic support of the potential transplant recipient has been verified by others [6], it is clear that this method will be inadequate for a certain number of patients. Successful bridging to transplantation with VADs was performed almost simultaneously in September, 1984, by Oyer and colleagues using a Novacor implantable electrical left ventricular assist system [7, 81 and by Hill and associates [9] with a Pierce-Donachy pneumatic LVAD. Since then, there has been increasing experience with both VADs [lo-141 and the total artificial heart [1P18] for mechanical support prior to transplantation. In view of the decreasing availability of donor hearts, it is important periodically to analyze critically results with these devices to ascertain their effectiveness and to determine if the continued use of precious donor hearts in patients supported with mechanical devices is justified. To examine this question, this report outlines the experience at St. Louis University Medical Center with 134 Ann Thorac Surg , Aug Copyright by The Society of Thoracic Surgeons

2 135 Kanter et al: Bridge to Transplantation 11 patients supported with pulsatile VADs as a prelude to cardiac transplantation. Materials and Methods Description of Devices Two types of pulsatile VAD were used in this series, the Pierce-Donachy VAD and the Novacor left ventricular assist system. The Pierce-Donachy VAD (Thoratec Laboratories Corporation, Berkeley, CA) has been described in detail previously [19]. It is an externally driven pneumatic sactype pump that rests on the abdominal wall with inlet and outflow cannulas traversing the anterior abdominal wall inferior to the costal margin. Pump inlet cannulas (venous return) are available for cannulation of either the atrium or the left ventricular apex. The pump outflow cannula (arterial inflow) contains a woven Dacron graft, which is sewn either to the ascending aorta or to the pulmonary trunk. The Pierce-Donachy VAD can be used for left ventricular support, right ventricular support (RVAD), or two pumps can be used simultaneously for biventricular support (BVAD). The Novacor left ventricular assist system (Novacor Medical Corporation, Oakland, CA) is an electrically powered dual pusher plate sactype pump that is implanted in the abdominal wall posterior to the rectus abdominis muscle [20]. Pump inflow (venous return) is achieved with a large left ventricular apical cannula, and pump outflow (arterial inflow) is through a woven Dacron conduit sewn to the ascending aorta. Although designed for eventual use as a totally implantable left ventricular assist system [21], the device as used in this study had a cable exiting the abdominal wall through a long subcutaneous tunnel for use as a power line, for monitoring of device function, and for an air vent. The Novacor VAD is designed for implantation in the left ventricular apex only; it cannot be used with left atrial cannulation or for right ventricular support. Patients and Techniques Since April, 1982, 11 patients (9 men, 2 women) aged 24 to 54 years (mean, 39.6 years) have been supported with pulsatile VADs as a bridge to cardiac transplantation. Prior to device implantation, all patients were in profound cardiogenic shock despite maximal inotropic and vasodilator support (11 patients), endotracheal intubation (10 patients), and intraaortic balloon counterpulsation (8 patients). The cause of the cardiac failure was ischemic in 6, postpartum cardiomyopathy in 2, idiopathic cardiomyopathy in 2, and cardiomyopathy secondary to doxorubicin hydrochloride (Adriamycin) toxicity in 1 (Table 1). Four of the 6 patients with ischemic disease were in profound shock because of acute myocardial infarction. One of these 4 (Patient 8) had refractory cardiac arrest following failed percutaneous transluminal coronary angioplasty. He was resuscitated with extracorporeal membrane oxygenation (ECMO), which was maintained for 24 hours prior to VAD insertion. Only 1 patient in this series (Patient 4) required VAD insertion because of failure to be weaned from cardio- pulmonary bypass following cardiac operation (coronary artery bypass grafting in this instance). Seven patients were supported with an LVAD, a Pierce-Donachy in 5 and a Novacor in 2 (see Table 1). Patients 10 and 11 had both a Pierce-Donachy LVAD and a Pierce-Donachy RVAD implanted at the initial operation because of biventricular failure. The remaining 2 patients were managed initially with a Pierce-Donachy LVAD, but subsequently severe right ventricular failure developed and necessitated placement of a Medtronic centrifugal RVAD [22] 12 hours later in Patient 1 and placement of a Bio-Medicus centrifugal RVAD [23] 48 hours later in Patient 4. A left atrial inflow cannula was used for the LVAD in 3 patients; a left ventricular apical cannula was used in the remaining 8 patients. Mean pump flows for the LVADs ranged from 4.1 to 8.5 L/min (average, 5.5 L/min). In the patients with biventricular support, the LVAD flow was generally higher than the RVAD flow because of the bronchial circulation and the fact that venous drainage (pump inflow) is more complete in the LVAD with left ventricular apical cannulation compared with the RVAD with right atrial cannulation. In the patients who did not have major bleeding complications, low molecular weight dextran was initiated when the mediastinal drainage was less than 200 ml/hr. When the chest drains were removed, an intravenous infusion of sodium heparin was used for anticoagulation. A combination of orally administered warfarin sodium and dipyridamole was used for those patients who required prolonged support (longer than one week) prior to transplantation. Prophylactic antibiotics with good antistaphylococcal coverage (usually a second-generation cephalosporin) were used in all patients perioperatively. Long-term prophylactic antibiotics were not used, although a careful search was made for sites of infection and proven infections were treated aggressively. Strict reverse isolation was maintained because of concern about depressed T-cell function in patients with VADs [24]. Results Duration of support in this series ranged from 8 hours to 91 days (mean, 24 days) (see Table 1). Hemodynamic stability was achieved in all 11 patients. Major complications occurred in 10 of the 11 patients (Table 2). Complications were severe enough to contraindicate cardiac transplantation in 5 patients: renal failure, 4; septic complications, 3; and disseminated intravascular coagulopathy, 1. More than one complication occurred in 3 of the 5 patients. The most common complication in this series was bleeding requiring mediastinal exploration (55%). Renal failure developed in 4 patients, all of whom consequently were excluded from transplantation. Patient 2 had severe renal impairment prior to VAD insertion (blood urea nitrogen, 68 mg/dl; creatinine, 4.8 mg/dl), and Patient 8 had evidence of incipient renal failure with a doubling of the creatinine level from 1.0 to

3 136 The Annals of Thoracic Surgery Vol 46 No 2 August 1988 Table 1. Patient Characteristics Patient No., LVAD Age (yr), Type of Inflow Flow Rate" Duration of Sex Diagnosis Support Cannula Site (Llmin) Support (d) Complications Outcome 1. 24, F Postpartum car- P-D LVAD, Left atrium (4.6 (L) 3.2 (L), RV failure, bleeding, Died diomyopathy Medtronic (3.5) (R) 2.7 (R) renal failure, he- RVAD~ molysis 2. 32, M Idiopathic car- P-D LVAD Left atrium (4.2) 5.0 Renal failure, bleed- Died diomyopathy ing, hepatic failure 3. 54, M Acute MI P-D LVAD Left ventricle (6.1) 1.5 None Transplantation; discharged 15 d; NYHA Class I (30 mo) 4. 50, M Postcardiotomy P-D LVAD, Left atrium (5.7) (L) 22 (L), RV failure, bleeding, Weaned; died 12 shock BioMedicus (4.2) (R) 6 (R) mediastinitis, he- hr later of heart RVAD' molysis failure 5. 41, F Postpartum car- P-D LVAD Left ventricle (4.1) 0.3 B 1 e e d i n g Transplantation; diomy opa thy discharged 34 d; NYHA Class I (24 mo) 6. 27, M Doxorubicin Novacor Left ventricle (8.5) 29 RV failure, renal fail- Died cardiomyop- LVAD ure, wound infecathy tion, antibodies 7. 46, M Ischemic car- Novacor Left ventricle (6.8) 91 Antibodies, driveline Transplantation; diomyopathy LVAD infection, RV fail- discharged ure 11 d; NYHA Class I(11 mo) 8. 51, M Acute MI P-D LVAD Left ventricle 2.M1.8 (5.2) 6 Renal failure, stupor, Weaned; died 3 sepsis mo later of ventricular fibrillation 9. 34, M Acute MI P-D LVAD Left ventricle (5.7) 24 Driveline infection, Transplantation; device malfunc- discharged tion, stroke, he- 12 d; late death molysis (6 m ~ ) , M Acute MI P-D LVAD, Left ventricle (4.6) (L) 3.0 (L), Bleeding, arrhyth- Transplantation; P-D RVAD (4.1) (R) 3.0 (R) mias wound infection; discharged 39 d; NYHA Class I (9 mo) , M Idiopathic car- P-D LVAD, Left ventricle (6.3) (L) 75 (L), Bleeding, driveline Transplantation; diomyopathy P-D RVAD (4.0) (R) 75 (R) infection, antibod- discharged ies 13 d; NYHA Class I (4 mo) "Data are shown as the range with the mean in parentheses. bthis was instituted 12 hours after LVAD insertion. This was instituted 48 hours after LVAD insertion. LVAD = left ventricular assist device; P-D = Pierce-Donachy; RVAD = right ventricular assist device; L = left pump; R = right pump; RV = right ventricular; MI = myocardial infarction; NYHA = New York Heart Association. 2.0 mg/dl despite normal urine output during the 24 hours of ECMO perfusion prior to LVAD insertion. All three infections in the group having transplantation were driveline infections requiring debridement and parenteral antibiotics. Two of the infections in the patients not having transplantation were wound infections leading to sepsis. In Patient 4, purulent mediastinitis developed; he had required mediastinal exploration for bleeding on four occasions. In Patient 6, the pocket in the abdominal wall for the Novacor pump was difficult to dissect because of scarring from previous irradiation for lymphoma. The pocket became infected and dehisced, leading to sepsis, renal failure, and consequent exclusion from consideration for transplantation. Substantial right ventricular failure developed after implantation of an LVAD in 6 patients (see Tables 1, 2).

4 137 Kanter et a1 Bridge to Transplantation Table 2. Cornplicatioris Occurring during VAD Support No Trans- Transplantation plantation Complication (N = 5) (N = 6) Bleeding 3 3 Renal failure 4 0 Infection 3 3 RV failure 3 1 Antibodies 1 2 Hemolysis 2 1 Device malfunction 0 1 VAD = ventricular assist device; RV = right ventricular In Patients 10 and 11, a Pierce-Donachy RVAD was placed at the time of LVAD insertion. In Patients 6 and 7, who had a Novacor device, right ventricular failure responded to vigorous pharmacological support. In Patients l and 4, a centrifugal RVAD was placed 12 hours and 48 hours, respectively, after LVAD insertion. Severe bleeding developed in both patients, and neither received a transplant. After insertion of the RVAD, both these patients also had major hemolysis that was attributed to the centrifugal pump. The other instance of hemolysis occurred in Patient 9, but resolved with reduction of the rate of rise of left ventricular pressure of the control drive system (see Tables 1, 2). During the period of VAD support, antibodies developed in 3 patients presumably because of the multiple blood products transfused perioperatively. In Patient 6, antibodies to platelets developed and necessitated repeated platelet transfusions. Patients 7 and 11 had cytotoxic antibodies against random donor pools, which resulted in positive crossmatches to several potential donor hearts. The antibody levels decreased to less than 7% reactivity after 70 days and 58 days, respectively, of VAD support with subsequent negative crossmatches to potential donor hearts. Device malfunction occurred only once in 338 patientdays of pulsatile VAD support. It happened to Patient 9 whose Pierce-Donachy LVAD driven by a Thoratec control console inexplicably stopped pumping for a brief period without sounding of the appropriate alarms. The patient became hemodynamically unstable; this was reversed by simply switching to the backup drive console once the problem was recognized. Patients 7,9, and 11 were supported 91 days, 24 days, and 75 days, respectively, prior to transplantation. All 3 were extubated and had all intravenous catheters and monitoring lines removed. They were able to ambulate in the hall and even exercise on a stationary bicycle while awaiting a donor heart. Six of the 11 patients underwent orthotopic cardiac transplantation 8 hours to 91 days after VAD insertion (see Table 1). Transplantation was performed in the standard fashion [25] after the establishment of cardiopulmonary bypass and the removal of the VAD. The driveline insertion sites were packed with povidoneiodine gauze and allowed to slowly granulate postoperatively. Immunosuppression consisted of cyclosporine, prednisone, and azathioprine; this is the standard regimen used for heart transplant recipients at St. Louis University Medical Center. There were only two major complications during and after transplantation. Patient 9, whose Pierce-Donachy LVAD briefly malfunctioned, sustained a small stroke during transplantation. This was presumed to be due to embolization of a clot that was in the device itself and that formed during the brief period of stasis while the LVAD was not pumping. The episode of device malfunction had occurred just hours before transplantation. Examination of the explanted LVAD showed some small clots and fibrin near the valve housing and in the blood sac itself. All neurological deficits were completely resolved at the time of hospital discharge 12 days following transplantation. The other complication occurred in Patient 10 three weeks after transplantation when a sternal wound infection developed, requiring operative debridement and prolonged antibiotic irrigation of the mediastinum. The infection healed eventually, and he was discharged 39 days after transplantation. The remaining 4 patients undergoing transplantation did well and were discharged 11, 13, 15, and 34 days after transplaritation (see Table 1). The 6 patients receiving a transplant have been followed 4 to 30 months after transplantation (mean, 14 months). There has been 1 late death (Patient 9) at 6 months due to rejection caused by noncompliance with the immunosuppressive regimen. The other 5 patients are all in New York Heart Association Functional Class I. Two are working, 1 is a housewife, 1 is unemployed, and 1 is currently recuperating from the transplantation. Comment The results of this series indicate that critically ill patients with profound cardiac failure refractory to standard supportive measures (inotropic and vasodilator support, intraaortic balloon counterpulsation) can be stabilized hemodynamically with pulsatile VADs. A certain proportion will be suitable candidates for subsequent cardiac transplantation with high expectations of success. Several important issues pertaining to bridging to transplantation with VADs are worthy of comment. It is clear from this series that among the most crucial considerations determining the ultimate outcome of bridging to transplantation with mechanical circulatory support is the proper selection of patients not only for VAD implantation but also for subsequent consideration for transplantation. The current philosophy at St. Louis University Medical Center mandates that the patient being considered for support with a VAD prior to

5 138 The Annals of Thoracic Surgery Vol 46 No 2 August 1988 transplantation must be a good candidate for transplantation at the time of VAD implantation. Thus, for example, the patient who has decompensated hemodynamically and also has major underlying renal insufficiency would be excluded from consideration for VAD support rather than receiving implantation of a VAD in the hope that the improved circulation afforded by the device would improve the renal dysfunction, hepatic dysfunction, or pneumonia. This concept was borne out in this series by Patients 2 and 8, who had preexisting renal insufficiency that failed to improve with VAD support (see Table 1). In particular, the presence of renal failure in a patient supported with a VAD is a grave prognostic sign highly predictive of death [26]. Obviously, there are gradations in this concept, and at times it may be extraordinarily difficult to determine whether the profound oliguria associated with hemodynamic deterioration is simply a reflection of the underlying inadequate cardiac output and will improve with VAD support or whether the oliguria reflects incipient renal failure that will progress and end in ultimate failure despite adequate circulatory support. This is among the many questions concerning bridging to transplantation that need to be analyzed as more experience is gained with these procedures. Similarly, once the patient has been stabilized hemodynamically with a VAD, careful reevaluation of the patient with respect to suitability for transplantation must be performed on an ongoing basis. In 5 patients in this series, contraindications to cardiac transplantation while they were on VAD support (see Table l), and they were eliminated from consideration for transplantation. Although this policy at times requires very difficult decision-making, we think it is imperative to use the same strict criteria for suitability for transplantation in patients supported with VADs as in patients undergoing routine transplantation evaluation. An unerring adherence to this policy was responsible in large part for the uniform survival of patients having transplantation after VAD support in this series. The choice of device for mechanical circulatory assistance prior to transplantation also requires careful consideration, as it can greatly influence the ultimate clinical outcome. It is axiomatic that the condition of patients under consideration for VAD implantation be refractory to less aggressive measures including optimal pharmacological support and intraaortic balloon counterpulsation when appropriate. Having determined the necessity as well as the suitability for VAD support, the next consideration is the choice of device itself. There has been limited experience with centrifugal pumps or ECMO as methods for bridging to transplantation [lo, 12, 27, 281. In general, when these methods have been used successfully, the duration of support prior to transplantation has been less than 4 days. The current proliferation of cardiac transplantation centers with the attendant reduction in available donor hearts makes it less likely that a suitable donor heart will be procured within 4 days of initiation of mechanical circulatory support. Although 3 of the patients in this series underwent transplantation within 4 days of implantation of the VAD, 3 others were supported with pulsatile VADs for 24, 75, and 91 days before transplantation. Furthermore, the development of preformed cytotoxic antibodies in Patients 7 and 11 prolonged the interval of VAD support in this series. We currently use washed red blood cells and HLA type-specific platelets to minimize the development of antibodies. Although in this series these antibody levels eventually declined and allowed transplantation, this potential complication must be considered when selecting the device for circulatory support. It is unlikely that the less physiological centrifugal pumps can provide adequate support for prolonged periods in patients for whom a donor heart cannot be located readily. During this period of prolonged support in these 3 patients, they were ambulatory, free from invasive lines other than the pump drivelines, and even able to exercise on stationary bicycles. Undoubtedly this degree of mobility vastly reduced the potential pulmonary and infectious complications so frequently seen in patients who are at bed rest with mechanical circulatory support [29, 301. Furthermore, the paucity of device-related complications in these patients supported for extended periods was gratifying, underscoring the premise that pulsatile VADs as used in this series can provide adequate physiological support for prolonged intervals. In 4 of the 11 patients in this series, right ventricular failure developed requiring support with inotropic agents (Patients 6 and 7), or centrifugal RVADs (Patients 1 and 4). A Pierce-Donachy RVAD as well as an LVAD was implanted in Patients 10 and 11 when they were first seen. This high incidence of biventricular failure has been observed previously in patients requiring mechanical circulatory support for postcardiotomy cardiogenic shock [31, 321 as well as in patients undergoing subsequent cardiac transplantation [ The Pierce-Donachy VAD is ideally suited for this problem in that two pumps can be employed simultaneously for biventricular support, as was done in Patients 10 and 11. The Novacor left ventricular assist system, however, can be utilized only as an LVAD. Both Patients 6 and 7, in whom the Novacor was used, required temporary pharmacological support of the right ventricle. Obviously, the appearance of profound right ventricular failure refractory to pharmacological support in a patient with a Novacor LVAD would require some form of additional mechanical assistance of the right ventricle; in our institution, we would place a Pierce-Donachy RVAD. The Novacor LVAD has certain advantages over a Pierce-Donachy LVAD. In general, higher flows are achieved with the Novacor because of a larger left ventricular apical cannula. Second, there is only one cable violating the integrity of the skin as opposed to two cannulas with the Pierce-Donachy, thereby giving a theoretical reduction in the potential for driveline infection. Finally, the drive console for the Novacor is more sophisticated than that for the Pierce-Donachy, thereby

6 139 Kanter et al: Bridge to Transplantation allowing prolonged adjustment-free support. We therefore prefer to utilize a Novacor LVAD in a large candidate (to accommodate the pump itself in the abdominal wall) with predominant left ventricular failure and apparently reasonable right ventricular function. The current widespread enthusiasm for the total artificial heart for mechanical support prior to transplantation [ eliminates the concerns about biventricular failure. However, two recent multi-institutional reports [14, 181 have detailed less than 65% survival of patients supported with the total artificial heart who underwent subsequent transplantation. At this point, it is unclear whether this simply reflects a "learning curve" in the selection and management of patients or is, in fact, due to some inherent flaw in the device itself. Obviously, the comparison of biventricular support with the total artificial heart versus biventricular support with pulsatile VADs is beyond the scope of this study. This topic warrants intense scrutiny by investigators to elucidate the optimal management of the patient with biventricular failure who needs mechanical circulatory support prior to cardiac transplantation. In summary, this report has outlined an experience with 11 patients with profound cardiac failure who were supported with pulsatile VADs as a prelude to transplantation. Although hemodynamic stability was achieved in all, complications were common and precluded transplantation in 5. The remaining 6 underwent successful transplantation after periods of support up to 91 days. This indicates the feasibility of prolonged mechanical support as a bridge to transplantation, and demonstrates that with rigid selection criteria, the success rate in this difficult group of patients can be gratifyingly high. Clearly, this series supports the further use and study of pulsatile VADs as bridges to cardiac transplantation. References Kaye MP: The registry of the International Society for Heart Transplantation: fourth official report, J Heart Transplant 6:63, 1987 Copeland JG, Emery RW, Levinson MM, et al: The role of mechanical support and transplantation in treatment of patients with end-stage cardiomyopathy. Circulation 72: Suppl2:7, 1985 Cooley DA, Liotta D, Hallman GL, et al: Orthotopic cardiac prosthesis for two-staged cardiac replacement. Am J Cardiol 24:723, 1969 Norman JC, Cooley DA, Kahan BD, et al: Total support of the circulation of a patient with post-cardiotomy stoneheart syndrome by a partial artificial heart (ALVAD) for 5 days followed by heart and kidney transplantation. Lancet 1:1125, 1978 Reemtsma K, Krusin R, Edie R, et al: Cardiac transplantation in patients requiring mechanical circulatory support. N Engl J Med 298:670, 1978 Hardesty RL, Griffith BP, Trento A, et al: Mortally ill patients and excellent survival following cardiac transplantation. Ann Thorac Surg 41:126, 1986 Pennington DG: Circulatory support pre-transplant. In Wallwork J (ed): Cardiac and Cardiopulmonary Transplantation. London, Grune & Stratton, (in press, 1988) 8. Williams BA, Lough ME, Shinn JA: Left ventricular assist device as a bridge to heart transplantation: a case study. J Heart Transplant 6:23, Hill JD, Farrar DJ, Hershon JJ, et al: Use of a prosthetic ventricle as a bridge to cardiac transplantation for postinfarction cardiogenic shock. N Engl J Med 314:626, Pennington DG, Codd JE, Merjavy JP, et al: The expanded use of ventricular bypass systems for severe cardiac failure and as a bridge to cardiac transplantation. Heart Transplant 3:170, Oaks TE, Pae WE, Rosenberg G, et al: The use of a paracorporeal ventricular assist device as a bridge to cardiac transplantation. Trans Am SOC Artif Intern Organs 33:408, Bolman RM, Spray TL, Cox JL, et al: Heart transplantation in patients requiring preoperative mechanical support. J Heart Transplant 6:273, Hill JD, Farrar DJ, Hershon JJ, et al: Bridge to cardiac transplantation: successful use of prosthetic biventricular support in a patient awaiting a donor heart. Trans Am SOC Artif Intern Organs 32:233, Pae WE, Pierce WS: Combined registry for the clinical use of mechanical ventricular assist pumps and the total artificial heart: first official report, J Heart Transplant 6:68, Levinson MM, Smith RG, Cork R, et al: Three recent cases of the total artificial heart before transplantation. J Heart Transplant 5:215, Joyce LD, Pritzker MR, Kiser JC, et al: Use of the mini Jarvik-7 total artificial heart as a bridge to transplantation. J Heart Transplant 5:203, Griffith BP, Hardesty RL, Kormos RL, et al: Temporary use of the Jarvik-7 total artificial heart before transplantation. N Engl J Med 316:130, Olsen DB, Riebman JB, De Paulis R, et al: Registry and tabulations of orthotopic total artificial hearts in humans. Trans Am SOC Artif Intern Organs 33:182, Pennington DG, Samuels LD, Williams G, et al: Experience with the Pierce-Donachy ventricular assist device in postcardiotomy patients with cardiogenic shock. World J Surg 9:37, Oyer PE, Stinson EB, Portner PM, et al: Development of a totally implantable, electrically actuated left ventricular assist system. Am J Surg 140:17, , Portner PM, Oyer PE, Jassawalla JS, et al: A totally implantable ventricular assist device for end-stage heart disease. In Unger F (ed): Assisted Circulation 2. Berlin, Springer- Verlag, 1984, pp Pennington DG, Merjavy JP, Swartz MT, Willman VL: Clinical experience with a centrifugal pump ventricular assist device. Trans Am SOC Artif Intern Organs 28:93, , Magovern GJ, Park SB, Maher TD: Use of a centrifugal pump without anticoagulants for postoperative left ventricular assist. World J Surg 9:25, McBride LR, Pennington DG, Tsai CC, et al: Influence of ventricular assist devices on lymphocytes. J Heart Transplant 5:379, Lower RR, Stofer RC, Shumway NE: Homovital transplantation of the heart. J Thorac Surg - 41:196, Kanter KR, Swartz MT, Pennington DG, et al: Renal failure in patients with ventricular assist devices. Trans Am SOC Artif Intern Organs 33:426, 1987

7 140 The Annals of Thoracic Surgery Vol 46 No 2 August Gifford D, Wheeldon D, Wallwork J, Spratt P: Balloon counterpulsation together with Bio-Medicus pump closed circuit membrane oxygenation for left ventricular assist: a case report. J Extra-Corporeal Techno1 18:29, Phillips SJ, Zeff RH, Wickemeyer WJ, et al: Bridging circulatory support before heart transplant without invasion of the mediastinurn. J Heart Transplant 6:116, Zumbro GL, Kitchens WR, Shearer G, et al: Mechanical assistance for cardiogenic shock following cardiac surgery, myocardial infarction, and cardiac transplantation. Ann Thorac Surg 44:11, McBride LR, Ruzevich SA, Pennington DG, et al: Infectious complications associated with ventricular assist device support. Trans Am SOC Artif Intern Organs 33:201, Pennington DG, Merjavy JP, Swartz MT, et al: The importance of biventricular failure in patients with postoperative cardiogenic shock. Ann Thorac Surg 39:16, Young JN, Iverson LIG, Ennix CL, et al: Biventricular support is superior to univentricular support for mechanical circulatory assistance in patients after cardiotomy. J Heart Transplant 6:313, 1987 Notice from the American Board of Thoracic Surgery The Part I (written) examination will be held at the Hyatt- Regency, Dallas/Fort Worth Airport, Dallas, TX, on January 21, The closing date for registration is August 1, To be admissible for the Part I1 (oral) examination, a candidate must have successfully completed the Part I (written) examination. A candidate applying for admission to the certifying examination must fulfill all the requirements of the Board in force at the time the application is received. Please address all communications to the American Board of Thoracic Surgery, One Rotary Center, Suite 803, Evanston, IL