Twenty-year clinical experience with porcine bioprostheses.

BACKGROUND

For the past 25 years, porcine valves have been the most widely implanted bioprosthesis, thereby becoming the standard for comparison with newer bioprosthetic valves.

METHODS

We retrospectively analyzed 2,879 patients who underwent aortic (AVR; n = 1,594) or mitral (MVR; n = 1,285) valve replacement between 1971 and 1990. Follow-up was 97% complete and extended to 20 years (total, 17,976 patient-years). Patient age ranged from 16 to 94 years; mean age in patients who underwent AVR was 60 +/- 15 (+/- standard deviation) years; that for patients who underwent MVR was 58 +/- 13 years.

RESULTS

The operative mortality rates were 7% +/- 1% (70% confidence limits) for AVR and 10% +/- 1% for MVR. Actuarial estimates of freedom from structural valve deterioration at 10 and 15 years were 78% +/- 2% (SE) and 49% +/- 4%, respectively, for the AVR subgroup; and 69% +/- 2% and 32% +/- 4%, respectively, for the MVR subgroup (AVR > MVR; p < 0.05). Estimates of freedom from reoperation at 10 and 15 years were 76% +/- 2% and 53% +/- 4%, respectively, for the AVR subgroup and 70% +/- 2% and 33% +/- 4%, respectively, for the MVR subgroup (AVR > MVR; p < 0.05). Estimates of freedom from thromboembolism at 10 and 15 years were 92% +/- 1% and 87% +/- 2%, respectively, for the AVR subgroup and 86% +/- 1% and 77% +/- 3%, respectively, for the MVR subgroup (AVR > MVR; p < 0.05). Estimates of freedom from anticoagulant-related hemorrhage at 10 and 15 years were both 96% +/- 1% for the AVR subgroup and 93% +/- 1% and 90% +/- 2%, respectively, for the MVR subgroup (AVR > MVR; p < 0.05). Estimates of freedom from valve-related mortality at 10 and 15 years were 86% +/- 1% and 78% +/- 3%, respectively, for the AVR subgroup and 84% +/- 2% and 70% +/- 4%, respectively, for the MVR subgroup (p = not significant). Multivariate analysis (Cox model) showed younger age, later year of operation, and valve site (MVR > AVR) to be significant risk factors for structural valve deterioration. Younger age, later year of operation, valve site (MVR > AVR), and renal insufficiency were the significant, independent risk factors for reoperation. Multivariate analysis revealed that higher New York Heart Association functional class, longer cardiopulmonary bypass time, congestive heart failure, renal insufficiency, and longer cross-clamp time were significant risk factors for valve-related mortality. Valve manufacturer did not emerge as a factor in any analysis.

CONCLUSIONS

These long-term results with porcine bioprostheses were satisfactory, particularly in older patients and those undergoing AVR. As expected, younger age was a significant risk factor for structural valve deterioration and reoperation in both groups. Surprisingly, the durability of porcine bioprosthetic valves has not improved over time, which possibly can be attributed to more enhanced postoperative surveillance and earlier reintervention. These first-generation Hancock and Carpentier-Edwards porcine bioprostheses achieved similar long-term performance.