T sors in the following aspects: the porcine aortic valve

Transcription

1 Clinical and Hemodynamic Assessment of the Hancock I1 Bioprosthesis Tirone E. David, MD, Susan Armstrong, MSc, and Zhao Sun, MA Division of Cardiovascular Surgery, The Toronto Hospital and University of Toronto, Toronto, Ontario, Canada The Hancock I1 bioprosthesis was used for heart valve replacement in 614 patients from 1982 to 199. Aortic valve replacement (AVR) was performed in 376 patients, mitral valve replacement (MVR) in 195, and aortic and mitral valve replacement (DVR) in 43. The mean age was 62.7 years, and 78% of all patients were in New York Heart Association functional class I11 or IV before operation. Coronary artery bypass graft was necessary in 232 patients and replacement of ascending aorta in 55. There were 31 operative deaths (AVR, 4%; MVR, 6%; DVR, 9%). Follow-up was complete in 98.5% of the patients and extended from 12 to 13 months, with a mean of 49 months. At the last follow-up, 85% of the patients were in New York Heart Association class I or 11. The actuarial survival at 8 years was 79% k 3% for AVR, 68% & 4% for MVR, and 65% & 1% for DVR. The freedom from stroke at 8 years was 93% f 2% for AVR, 83% k 5% for MVR, and 9% 5% for DVR. At the end of 8 years 96% k 1% of all patients were free from endocarditis, 92% f 1% were free from primary tissue failure, and 89% & 3% were free from reoperation. The actuarial freedom from valve-related death at 8 years was 98% k 1% for AVR, 86% & 5% for MVR, and 91% f 6% for DVR. Hemodynamic assessment was obtained by Doppler echocardiography in all operative survivors and demonstrated satisfactorily effective valve orifices and transvalvular gradients. The clinical results obtained with the Hancock I1 bioprosthesis have been gratifying, particularly in the aortic position. This bioprosthesis is our biological valve of choice. ( ) he Hancock I1 bioprosthesis differs from its predeces- T sors in the following aspects: the porcine aortic valve is fixed with a buffered solution of.625% glutaraldehyde in two stages, an initial stage of low pressure and a late stage of physiologic pressure; the stent is made of Delrin; and the tissue is treated with sodium dodecyl-sulfate to retard calcification [l]. The Hancock I1 bioprosthesis used for aortic valve replacement is designed to be implanted in a supraannular position. The first Hancock I1 bioprosthesis was implanted at The Toronto Hospital in September This article is a review of all patients who received this bioprosthesis at that hospital. Patients and Methods From September 1982 to May 199, the Hancock I1 bioprosthesis was used for heart valve replacement in 614 patients: 397 men and 217 women. The mean age of the patients was 62.7 years and ranged from 18 to 86 years. Patients were divided into three groups: 376 patients had the aortic valve replaced with a Hancock I1 bioprosthesis (AVR), 195 had the mitral valve replaced (MVR), and 43 had both the aortic and mitral valve replaced (DVR). The clinical profiles of these patients are summarized in Table 1. Patients who had one valve repaired and one replaced with a Hancock I1 bioprosthesis were included under single valve replacement. Three patients had tricuspid Presented at the Twenty-eighth Annual Meeting of The Society of Thoracic Surgeons, Orlando, FL, Feb >5, Address reprint requests to Dr David, 2 Elizabeth St, 13EN219, Toronto, Ont M5G 2C4, Canada. valve replacement and were included in one of the above groups. One patient had aortic, pulmonary, and tricuspid valve replacement and was included in the AVR group. The operative procedures performed on these patients are listed in Table 2. The left atrial appendage orifice was oversewn from inside the left atrium in patients who had MVR. An effort was always made to match the size of the bioprosthesis to the body surface area of the patient to minimize transvalvular gradients. The guidelines for this matching were developed early in our experience with this bioprosthesis and are shown in Table 3. All operative survivors were discharged from the hospital on a regimen of warfarin sodium and were instructed to discontinue it after 3 months. After 1984 all patients were told to take 325 mg of aspirin a day after discontinuing warfarin administration. Operative survivors were followed up from October 1982 to September With the exception of 9 patients (1.5%) who were lost to follow-up, all survivors were contacted by telephone or questionnaire in The mean follow-up was 49 * 24 months (range, 12 to 13 months). Of the 9 patients lost to follow-up, 5 had had AVR, 3 MVR, and 1 DVR. Postoperative complications were prospectively compiled and analyzed according to the Guidelines for reporting morbidity and mortality after cardiac valvular operations approved by The Society of Thoracic Surgeons [2]. The cause of death was established from information obtained from hospital or attending physician s records or autopsy reports. Bioprosthetic valve failure was defined as any clinically significant valve 1992 by The Society of Thoracic Surgeons /92/$5.

2 662 DAVID ET fil HANCOCK I1 BIOFROSTHESIS Table 1. Clinical Profile of the Patients" Table 2. Operative Data" Variable AVR MVR DVR Variable AVR MVR DVR No. of patients Age (years, mean * standard deviation) Sex Male Female ECG Sinus rhythm Atrial fibrillation Heart block NYHA functional classification I I1 111 IV Previous valve operation AVWrepair MVWrepair TV repair Active infective endocarditis Aortic valve lesion Stenosis Insufficiency Mixed Mitral valve lesion Stenosis Insufficiency Mixed Tricuspid valve lesion Stenosis Insufficiency Coronary artery disease None One-vessel Two-vessel Three-vessel Left ventricular ejection fraction > <.2 Undetermined (74) 89 (46) 29 (67) 97 (26) 16 (54) 14 (33) 334 (89) 91 (47) 21 (49) 33 (9) 93 (48) 18 (48) 9 (2) 11 (5) 4 (9) 9 (2) 2 (1) 1 (2) 13 (27) 14 (7) 6 (14) 143 (38) 79 (4) 9 (2) 121 (32) 1 (51) 27 (63) 17 (4) 2 (1) 11 (26) 39 (2) 11 (26) 1 (.5) l(2) 14 (4) 11 (6) 6 (14) 16 (42)... 8 (19) 1 (27)... 2 (46) 116 (31) (35) 2 (.5) 42 (21) 11 (25) 25 (7) 19 (56) 17 (39) 3 (.8) 44 (22) 15 (34) 3 (2) 1 (2) 29 (15) 7 (16) 226 (6) 18 (55) 28 (65) 44 (12) 27 (14) 7 (16) 56 (15) 2 (1) 5 (11) 5 (13) 4 (2) 3 (7) 53 (14) 61 (31) 3 (7) 117 (31) 7 (36) 11 (25) 6 (16) 35 (18) 15 (35) 28 (8) 12 (6) 2 (5) 118 (3) 17 (9) 12 (28) a Values in parentheses are percentages. replacement; ECG = electrocardiogram; MVR = mitral valve replacement; NYHA = New York Heart Association; TV = tricuspid valve. stenosis or insufficiency documented by Doppler echocardiography, reoperation, or autopsy. All operative survivors had at least one echocardiographic study during the first year after the operation and whenever they experienced a cardiac or valvular complication. Statistical analyses were performed with SAS (Statistical Aortic valve replacement Mitral valve replacement Tricuspid valve replacement Pulmonary valve replacement Mitral valve repair Tricuspid valve repair Replacement of the ascending aorta Patch enlargement of aortic root Reconstruction of the mitral annulus Coronary artery bypass Repair of LV aneurysm Abdominal aortic aneurysm repair Size of aortic bioprosthesis 21 mm 23 mm 25 mm 27 mm 29 mm Size of mitral bioprosthesis 27 mm 29 mm 31 mm 33 mm (8) 49 (13) 79 (21) 143 (38) (6) 81 (21) 15 (28) 133 (35) 35 (9) (16) 7 (16) (38) 17 (39) (12) 7 (16) 15 (35) 12 (28) 4 (9) 16 (8) 3 (7) 64 (33) 4 (9) 9 (46) 26 (6) 25 (13) 1 (23) a Values in parentheses are percentages,. replacement; LV = left ventricular; MVR = mitral valve replacement. Analysis System Institute Incorporated, Cary, NC) and BMDP (BMDP Statistical Software, Los Angeles, CA) software. Postoperative events were characterized by actuarial statistics with the Kaplan-Meier method [3]. Results were expressed in percentages of the means and percentages of the standard errors of the means. Fisher's exact test or contingency table analysis was used to compare the thromboembolic rates in certain subgroups of patients. Results Patient Survival The operative mortality and morbidity are shown in Table 4. There were 31 operative and 74 late deaths. The causes of late deaths are listed in Table 5. Table 3. Guidelines for Matching Size of Bioprosthesis With Patient's Size BSA (m2) Aortic Valve (mm) Mitral Valve (mm) < or larger 27 or larger or larger 27 or larger or larger 29 or larger > or larger 31 or larger BSA = body surface area.

3 DAVID ET AL 663 Table 4. Operative Mortality and Morbidity Variable AVR MVR DVR Operative mortality Causes of death Myocardial infarction Low cardiac output syndrome Stroke Multiple organ failure Operative morbidity Reexploration for bleeding Reexploration for cardiac arrest Low cardiac output syndrome Perioperative myocardial infarction Perioperative stroke or TIA Sternal infection Permanent pacemaker 15 (4%) 12 (6%) 4 (9%) replacement; MVR = mitral valve replacement; TIA = transient ischemic attack. Figure 1 shows the actuarial survival for patients who had AVR, MVR, and DVR. The survival at 8 years was 79% f. 3% for AVR, 64% 2 5% for MVR, and 64% * 11% for DVR. The actuarial freedom from cardiac and valve-related deaths at 5 and 8 years is shown in Table 6. Anticoagulant and Thromboembolic Complications There were no hemorrhagic complications during the first 3 postoperative months while all operative survivors were receiving warfarin sodium. A number of patients continued to take oral anticoagulant beyond the third month, and 3 patients suffered serious hemorrhagic complications with the resulting death of 1. At the last follow-up, 98 patients (28 AVR, 57 MVR, and 13 DVR) were taking warfarin, 312 patients (228 AVR, 69 MVR, and 15 DVR) were taking aspirin, and 84 patients were not taking either warfarin or aspirin. There were no cases of bioprosthetic valve thrombosis. There were 23 thromboembolic events in the AVR group: eight strokes (one was fatal) and 15 transient Table 5. Causes of Late Deaths Cause AVR MVR DVR Valve-rela ted Stroke Endocarditis Anticoagulant-related hemorrhage Valve dehiscence Cardiac Myocardial infarctionkudden death Congestive heart failure Noncardiac Unknown replacement; MVR = mitral valve replacement. Table 6. Freedom From Cardiac and Valve-Related Death Event AVR MVR DVR Freedom from cardiac death At 5 years 9% 2 2% 81% f 3% 86% f 6% At 8 years 86% f 3% 74% 2 5% 86% f 6% Freedom from valverelated death At 5 years 99% +-.6% 94% f 2% 91% -C 6% At 8 years 98% -C 1% 86% f 5% 91% % 6% replacement; MVR = mitral valve replacement. cerebral ischemic attacks. The anticoagulant status of these patients at the time of the thromboembolic event was as follows: 7 patients were not taking any drugs, 6 were taking warfarin, and 7 were taking aspirin. Patients taking aspirin had a significantly lower thromboembolic risk than those taking warfarin or no medication (p <.1). Patients who had AVR and coronary artery bypass had a higher incidence of thromboembolic complications than those who had isolated AVR (p <.2). Patients who had associated operations such as mitral valve repair (3 patients) or replacement of the ascending aorta (49 patients) had the same risk of thromboembolic complication as patients who had only AVR. The actuarial freedom from thromboembolic events in patients who had AVR is shown in Figure 2. At 8 years, 8% * 5% of the patients had not had any thromboembolic complication and 93% * 2% had not had strokes. There were 19 thromboembolic events in the MVR group: 14 strokes (six were fatal) and five transient cerebral ischemic attacks. The anticoagulation regimen of these patients at the time of the thromboembolic event was the following: 7 patients were taking warfarin, 6 were taking aspirin, and 6 had no anticoagulants. Chronic atrial fibrillation was not associated with a higher incidence of thromboembolic events, but all patients in atrial fibrillation were taking either aspirin or warfarin. The actuarial freedom from thromboembolic complications in patients who had MVR is shown in Figure 3. At 8 years, 74% * 7% of the patients had not had any thromboembolic event and 83% +- 5% had not had strokes. Patients who had MVR had a higher rate of strokes than patients who had AVR (p <.5), but the overall rates of thromboembolic complications in patients who had AVR and MVR were not significantly different. In the DVR group, 3 patients suffered a stroke. The actuarial freedom from thromboembolic complications in patients who had DVR was 9% & 5% at 8 years. Bioprosthetic Valve Endocarditis Twelve patients had bioprosthetic valve endocarditis (4 AVR, 5 MVR, and 3 DVR). Four patients were treated with antibiotics and operation: 3 survived and 1 died. Eight patients were treated with antibiotics alone: 3 survived and 5 died. The actuarial freedom from bioprosthetic valve endocarditis for all patients at 5 and 8 years

4 664 DAVID ET AL Fig 7. Actuarial suvvival curves in patients who had aortic valve replacement (AVR), mitral valve replacement (MVR), and both aortic and mitral valve replacement (DVR). 1 9 % Pts at Risk: 8 7 6o 5 % 11% ii AVR I I I I I I I I 3% was 97% * 1% and 96% f 1%, respectively. The actuarial freedom from bioprosthetic valve endocarditis in patients who had AVR, MVR, and DVR is shown in Figure 4. Bioprosthetic Valve Failure Six patients experienced bioprosthetic valve incompetence because of primary tissue failure in 5 (3 AVR, 1 MVR, and 1 DVR) and tissue overgrowth on one of the leaflets with resulting mitral incompetence in 1. All 6 patients were successfully reoperated on. The five valves that failed because of primary tissue failure displayed tears in one or more leaflets in the commissural areas, and only mild calcification. The actuarial freedom from primary tissue failure for all patients at 5 and 8 years was 99% 2.4% and 92% 2 3%, respectively. The actuarial freedom from primary tissue failure in patients who had AVR, MVR, and DVR is shown in Figure 5. Reopera tions Fifteen patients underwent reoperation: 7 in the AVR group, 4 in the MVR group, and 4 in the DVR group. The indications for reoperation were bioprosthetic valve failure in 6 patients, endocarditis in 4, progression of coronary artery disease in 2, progression of mitral valve disease in a patient who had AVR, paravalvular leak in 1, and acute aortic dissection in 1. The bioprosthetic valve was explanted in 2 of the latter 5 patients because it had been in for more than 5 years though it was still functioning normally. There was only one death among these 15 patients reoperated on; it occurred after endocarditis and septic shock developed in a patient who had had DVR. The actuarial freedom from reoperation in all patients at 5 and 8 years was 97% * 1% and 89% f 3%, respectively. The actuarial freedom from reoperation in patients who had AVR, MVR, and DVR is shown in Figure 6. Hemodynamic Assessment Although 574 operative survivors had a Doppler echocardiographic study performed during the first year after the operation, only 152 had the study done at our institution and following the same protocol. These 152 studies formed the base for the hemodynamic assessment of the Hancock I1 bioprosthesis. The hemodynamic data obtained by Doppler echocardiography are shown in Table 7. Paravalvular leak was detected in only 1 patient who had AVR and was reoperated on. Late Functional Classification At the last follow-up, 491 patients were alive and had their original Hancock I1 bioprosthesis: 318 AVR, 143 MVR, and 3 DVR. Of the patients who had AVR, 197 (62%) were in New York Heart Association functional class I, 87 (27%) in class 11, and 34 (11%) in class 111. Of the patients who had MVR, 74 (54 %) were in functional class I, 38 (26%) in class 11, 3 (21%) in class 111, and 1 in class 5 Fig 2. Actuarial freedom from thromboembolic (TE) 1 n complications in patients who had aortic valve replace- 93% 2 3% ment. 9 8% 2 5% 8 - % 7 Pts at Risk: 6o 5o - AII TE events - Stroke :: I I I I JzLLi-

5 DAVID ET AL 665 Fig 3. Actuarial freedom from thromboembolic (TE) complications in patients who had mitral valve replacement. % 7 t 61 Ail TE events 5o Stroke PIS at Risk: I I I I I I I I Time (years) IV. Of the patients who had DVR, 8 (27%) were in class I, 16 (53%) in class 11, 5 (17%) in class 111, and 1 in class IV. Of the 72 patients who were in New York Heart Association class I11 or IV, 11 were disabled because of angina pectoris and 61 because of shortness of breath or fatigue despite satisfactory bioprosthetic valve function as assessed by echocardiography. Comment The Hancock bioprosthesis became commercially available in 197. The porcine aortic valve was fixed in a closed position at near-physiologic pressure with a stabilized solution of glutaraldehyde, and it was mounted in a Dacron cloth-covered rigid metal stent [4]. To reduce the mechanical stress on the leaflets during closure of the valve, a flexible polypropylene stent was developed and replaced the rigid metal stent in 1971 [5]. As clinical experience with this bioprosthesis increased, certain problems became apparent and modifications were made to improve its performance. Its obstructive nature was largely due to the physical properties of the fixed tissue, the stent, and the muscle bar to which the right cusp is attached. This problem was partially resolved by better tissue selection and further improved by excising the right cusp and replacing it with a larger noncoronary cusp of another valve. This resulting composite valve is referred to as the Hancock modified orifice valve and was intro- duced for clinical use in Although the clinical performance of the Hancock modified orifice valve appears to be similar to that of the standard Hancock valve, its effective orifice is larger [6, 71. Porcine bioprostheses have a limited durability because of tissue degeneration. Mechanical stress and host reactions to the porcine valve are the main causes of tissue degeneration. It has been shown that the pressure that the leaflets are exposed to during fixation with glutaraldehyde affects their dynamic features as well as their resistance to fatigue [8]. Calcification of the leaflets is also part of the degenerative process, and it is possible to decrease mineralization of the porcine valve leaflets by treating them with certain calcium-mitigating agents [9, 11. Finally, the stent also plays a role in the mechanical stress that the leaflets are subjected to during the cardiac cycle [5]. These factors were incorporated in the development of the Hancock I1 bioprosthesis in an attempt to improve its hemodynamic performance and durability. The Hancock I1 bioprosthesis has not been available long enough to determine whether it is more durable than other porcine valves, but by comparing our results with those reported by other investigators on the standard Hancock valve and the standard and supraannular Carpentier-Edwards bioprostheses [11-171, we found a lower rate of bioprosthetic valve failure at 8 years. However, the differences may be explained by other factors that affect the durability of bioprostheses such as the age of the Fig 4. Actuarial freedom from bioprosthetic endocarditis in patients who had aortic valve replacement (AVR), mitral valve replacement (MVR), and both aortic and mitral valve replacement (DVR). % 7 Pts at Risk: 6 5J DVR A MVR i;: AVR Time (years)

6 666 DAVID ET AL Fig 5. Actuarial freedom from bioprosthetic valve failure in patients who had aortic valve replacement (AVR), mitral valve replacement (MVR), and both aortic and mitral valve 1 9 f 93% 2 5% 93% * 4% Pts at Risk: DVR A MVRO AVR Time (years) patients. Our patients were older than in other reported series [ll-131. Although many factors affect the long-term survival of patients after heart valve replacement, an actuarial survival of 79% at 8 years after AVR is excellent, particularly because the mean age of our patients was 63 years and more than one-third of them had coronary artery disease. This high survival in our patients is partly due to the freedom from valve-related deaths among patients who had AVR. The actuarial survival in patients who had MVR or DVR was similar to those reported for other porcine bioprostheses The rate of thromboembolic complications observed in our patients was higher than that reported for other bioprostheses [ll-171. The difference lies largely in the number of transient cerebral ischemic attacks. We believe that high thromboembolic incidence is not a problem associated with the Hancock I1 bioprosthesis but is rather due to our interpretation of the patients symptoms, as we found similar rates with other bioprostheses [MI. Most thromboembolic events in patients who had AVR were transient cerebral ischemic attacks, whereas most events in patients who had MVR were completed strokes. The freedom from all thromboembolic events at 8 years was 8% for AVR and 75% for MVR, whereas the freedom from stroke was 93% for AVR and 83% for MVR. Patients who were receiving aspirin had a lower thromboembolic rate, particularly if they had had AVR. Patients who had isolated AVR were found to have a lower thromboembolic rate than those who had AVR and coronary artery bypass. The risk of endocarditis in patients with the Hancock I1 bioprosthesis is similar to that reported for other bioprostheses [ll-171. The freedom from endocarditis at 8 years was 96% for all patients. Echocardiographic assessment of the Hancock I1 bioprosthesis revealed that its effective orifice is satisfactory and compares favorably with that of other porcine bioprostheses [19,2]. It is important to match the size of the patient to the size of the bioprosthesis as shown in Tables 3 and 7. Small valves should be implanted in small patients. This is the reason why we had to enlarge the aortic root of 79 patients (21%)1 who had AVR. Patients must be matched not only to the correct size of heart valve prosthesis but also to the type. The decision as to whether to use a biological or a mechanical heart valve can usually be made before the operation, and the patient, the primary physician, and the attending cardiologist should be part of the decision-making. Valve replacement with a porcine bioprosthesis offers a great opportunity to restore the patient s health and normal life-style, particularly in older patients who need AVR. However, there may be no advantage in using a porcine valve in patients Fig 6. Actuarial freedom from reoperation in patients who had aortic valve replacement (AVR), mitral valve replacement (MVR), and both aortic and mitral valve replacement (DVR) % 7 6 t DVR A MVR PIS at Risk: AVR I I I I I I I I Time (years)

7 ~~ ~~ ~~ 1992; DAVID ET AL 667 Table 7. Results of Hemodunamic Assessment of the Hancock 11 Bioprosthesis by Doppler Echocardiograph4 Aortic Bioprosthesis Mitral Bioprosthesis Variable Body surface area (m ) (.9) (.8) (.1) (.7) (.7) (.1) (.11) (.9) (.8) Peak systolic gradient (mm Hg) (4) (3) (2) (3) (3) Valve area (cm ) (.11) (.16) (.15) (.18) (.15) (.14) (.11) (.1) (.22) Values are expressed as mean; standard deviation is shown in parentheses. who need MVR because of the high incidence of atrial fibrillation and consequent need for permanent anticoagula tion. This study was supported by a grant from the Heart and Stroke Foundation of Ontario. We are indebted to Dr A. Kerwin and Ms J. David for their assistance in the preparation of the manuscript. References 1. Wright JTM, Eberhardt CE, Gibbs ML, Saul M, Gilpin CB. Hancock 11-an improved bioprosthesis. In: Cohn LH, Gallucci V, eds. Cardiac bioprostheses. New York: Yorke Medical, Edmunds LH Jr, Clark RE, Cohn LH, Miller DC, Weisel RD. Guidelines for reporting morbidity and mortality after cardiac valvular operations. 1988;46: Kaplan EL, Meier. Non-parametric estimations from incomplete observations. J Am Stat Assoc 1958;53: Kaiser GA, Hancock WD, Lukban SB, Litwak RS. Clinical use of a new design stented xenograft heart valve prosthesis. Surg Forum 1969;2: Reis RL, Hancock WD, Yarbrough JW, Glancy DL, Morrow AG. The flexible stent. A new concept in the fabrication of tissue heart valve prostheses. J Thorac Cardiovasc Surg 1971; Cohn LH, DiSesa VJ, Collins JJ Jr. The Hancock modifiedorifice porcine bioprosthetic valve: Ann Thorac Surg 1989;48:S Zusman DR, Levine FH, Carter JE, Buckley MJ. Hemodynamic and clinical evaluation of the Hancock modified orifice aortic bioprosthesis. Circulation 1981;64(Suppl 2): Broom ND, Thomson FJ. Influence of fixation conditions on the performance of glutaraldehyde-treated porcine aortic valves. Towards a more scientific basis. Thorax 1979; Lentz DJ, Pollock EM, Olsen DB, Andrews EJ, Murashita J, Hastings WL. Inhibition of mineralization of glutaraldehyde- fixed Hancock bioprosthetic heart valves. In: Cohn LH, Gallucci V, eds. Cardiac bioprostheses. New York Yorke Medical, 1982: Carpentier A, Nashef A, Carpentier S, Ahmed A, Goussef N. Techniques for prevention of calcification of valvular bioprostheses. Circulation 1984;7O(Suppl 1): Magilligan DJ Jr, Lewis JW Jr, Stein P, Alam M. The porcine bioprosthetic heart valve: experience at 15 years. Ann Thorac Surg 1989;48:32& Perier P, Deloche A, Chauvaud S, et al. A 1-year comparison of mitral valve replacement with Carpentier-Edwards and Hancock porcine bioprostheses. 1989; 48: Hartz RS, Fisher EB, Finkelmeier B, et al. An eight-year experience with porcine bioprosthetic cardiac valves. J Thorac Cardiovasc 1986;91:91@ Jamieson WRE, Allen P, Miyagishima RT, et al. The Carpentier-Edwards standard porcine bioprosthesis. A first-generation tissue valve with excellent long-term clinical performance. J Thorac Cardiovasc Surg 199;99:54= Jamieson WRE, Tyers GF, Miyagishima RT, Germann E, Janusz MT, Ling H. Carpentier-Edwards porcine bioprostheses. Comparison of standard and supra-annular prostheses at 7 years. Circulation 1991;84(Suppl 3): Akins CW, Carroll DL, Buckley MJ, Daggett WM, Hilgenberg AD, Austen WG. Late results with Carpentier-Edwards porcine bioprosthesis. Circulation 199;82(Suppl 4): Bernal JM, Rabasa JM, Cagigas JC, Echevarria JR, Carrion MF, Revuelta JM. Valve-related complications with the Hancock I porcine bioprosthesis. A twelve- to fourteen-year follow-up study. J Thorac Cardiovasc Surg 1991;11: David TE, Ho WIC, Christakis GT. Thromboembolism in patients with aortic porcine bioprostheses. 1985;4: Levine FH, Carter JE, Buckley MJ, et al. Hemodynamic evaluation of Hancock and Carpentier-Edwards bioprostheses. Circulation 1981;64(Suppl2): Khan S, Mitchell RS, Derby GC, Oyer PE, Miller DC. Hemodynamic comparison of Hancock and Carpentier- Edwards mitral bioprosthetic valves. Circulation 199; 82(Suppl 4):7581. DISCUSSION DR JAVIER FERNANDEZ (Browns Mills, NJ): Dr David, I know that you do a lot of valve repairs. How do you compare your results of valve repair with the results of these Hancock valves in terms of valve-related complications and reoperation? DR DAVID: I have repaired aortic valves only in the past 3 or 4 years, SO I cannot make any comments on the long-term complications. I have been repairing mitral valves for more than a decade, and valve-related complications are uncommon. I do not know if it is reasonable to compare patients who had mitral valve repair with those who had mitral valve replacement because the pathologic process in the mitral valve is often different. I am able to repair the mitral valve in most patients with myxomatous disease, less than half of the patients with rheumatic disease, and only a few with ischemic mitral insufficiency. I should tell you that we have an ongoing cohort study of patients who had mitral

8 668 DAVID ET AL 1992; valve repair or replacement with and without preservation of chordae tendineae. I do not know details of valve-related complications, but the actuarial survival after mitral valve replacement with preservation of chordae tendineae is much better than after mitral valve replacement with excision of the mitral apparatus, and it is not much different than that of mitral valve repair. We will present these data in the near future. DR W. R. E. JAMIESON (Vancouver, BC, Canada): Dr David, we congratulate you on the very excellent evaluation of the Hancock I1 porcine bioprosthesis. As Dr Miller has stated, we have been anticipating your evaluation of this new prosthesis for some time. We, at the University of British Columbia, have experience with the Carpentier-Edwards Supra-Annular porcine bioprosthesis, which is the other second-generation prosthesis that has low-pressure glutaraldehyde tissue fixation. We have documented our experience at 8 years indicating very comparable results. These two new porcine bioprostheses have provided satisfactory clinical performance. I have one question for your consideration, related to the freedom from stroke of 9%. Is this freedom from permanent neurological deficit, and have you considered the composite evaluation of valve-related mortality and residual morbidity related to valve-related neurological events, which provides an indication of the quality of life of patients with bioprostheses? DR DAVID: I used your article on the performance of the Carpentier-Edwards bioprostheses for comparison with our results. You were the first to describe this parameter of valverelated complication, namely, the valve-related mortality and residual morbidity. We did not look into the residual morbidity in our analysis. We defined a transient ischemic attack as a neuro- logic deficit that resolved completely within 3 minutes, and stroke as a neurologic deficit that lasted longer than 3 minutes and may have been permanent. There was a high mortality among these patients; in the aortic valve replacement group there were 8 strokes with one death, and in the mitral valve replacement group there were 14 strokes with six deaths. That was one of the reasons why I presented transient ischemic attacks and strokes separately. DR JOHN E. HUTCHINSON 111 (Hackensack, NJ): I am concerned because I do so many reoperations on people who have biological prostheses. The number of reoperations keeps going up and up and up. I have always tried, in my practice, to pick a valve that suits the patient, and I have always believed that in patients who are going to be on chronic anticoagulation therapy, there is no specific advantage in using a bioprosthesis with the greater probability of reoperation. In your patient population, do you consider the need for chronic anticoagulation therapy as a relative contraindication for use of a tissue valve? DR DAVID: I agree with you that if' a patient is in chronic atrial fibrillation and is going to require life-long anticoagulation with warfarin, it may be inappropriate to use a bioprosthesis. As far as the type of valve a patient should have, I believe that the patient, his or her family, the family doctor, and the cardiologist should participate in the decision-making. I have always tried to match the patient to the heart valve prosthesis as far as types and sizes are concerned. Like you, I frequently operate on patients with failed bioprostheses, and often I implant another bioprosthesis because the patient wishes to have this type of valve. I guess warfarin interferes too much with their life-style.