Durability of porcine valves at fifteen years in a representative North American patient population.

Isolated aortic (n = 857) or mitral (n = 793) valve replacement with a porcine bioprosthesis was performed in 1650 patients between 1971 and 1980. Follow-up (total = 12,012 patient-years) extended to more than 15 years and was 96% complete. Patient age ranged from 16 to 87 years; mean age was 59 +/- 11 years (+/- 1 standard deviation) for the aortic valve replacement cohort and 56 +/- 12 years for the mitral valve replacement cohort. The operative mortality rates were 5% +/- 1% (+/- 70% confidence limits) and 8% +/- 1%, respectively, for the aortic and mitral subgroups. Estimated freedom from structural valve deterioration (+/- 1 standard error of the mean) after 10 and 15 years was significantly higher for the aortic than for the mitral valve replacement subgroup (85% +/- 0.4% and 63% +/- 3% versus 78% +/- 2% and 45% +/- 3%, respectively, p = 0.001). Reoperation-free actuarial estimates were also significantly greater for the aortic valve replacement cohort: 83% +/- 2% and 57% +/- 3% versus 78% +/- 2% and 43% +/- 3% for mitral valve replacement at 10 and 15 years, respectively. The mortality rate for reoperative aortic valve replacement was 11% +/- 1%; it was 8% +/- 1% for reoperative mitral valve replacement. Importantly, the estimates of freedom from valve-related death (including sudden, unexplained deaths) were relatively high at 10 and 15 years: 78% +/- 2% and 69% +/- 3% in the aortic cohort and 74% +/- 2% and 63% +/- 3% in the mitral cohort (p = not significant). Excluding sudden, unexplained deaths, these estimates were 81% +/- 3% (aortic) and 73% +/- 4% (mitral) at 15 years. Thromboembolism-free rates were 84% +/- 3% (aortic) and 78% +/- 6% (mitral) at 15 years, and freedom from anticoagulant-related hemorrhage was 96% +/- 1% and 89% +/- 2%, respectively. At the time of current follow-up, 13% of patients having aortic valve replacement and 50% of patients having mitral valve replacement were receiving warfarin sodium. The hazard functions for thromboembolism and prosthetic valve endocarditis were constant and remained less than 1%/pt-yr over the entire follow-up period.(ABSTRACT TRUNCATED AT 400 WORDS)