LONG-TERM FOLLOW-UP OF DIFFERENT MODELS OF MECHANICAL AND BIOLOGICAL MITRAL PROSTHESES

Three hundred eighty-five valve prostheses were implanted between 1974 and 1981 in patients with isolated mitral disease: 157 caged-ball val ves (156 Starr-Edwards; 1 Smeloff-Cutter) (group A), 107 tilting-disc valves (44 Bjork-Shiley, 49 Sorin, 14 Lillehei-Kaster) (group B), 121 porcine bioprostheses (45 Carpentier-Edwards, 66 Liotta, 10 Hancock) ( group C). Perioperative mortality was 9.5% in group A, 11.2% in group B and 6.6% in group C. The follow-up was 86% complete. Actuarial freed om from complications was calculated as follows (linearised rates in b rackets) in groups A, B and C, respectively: survival: 47.01% +/- 0.11 (3% patient/yr), 53.37% +/- 0.08 (1.8% patient/yr), 61.24% +/- 0.05 ( 2.2% patient/yr); thromboembolism: 67.94% +/- 0.09 (1.18% patient/yr); 73.07% +/- 0.06 (1% patient/yr); 97.43% +/- 0.02 (0.02% patient/yr); anticoagulation-related hemorrhage: 84.10% +/- 0.13 (0.18% patient/yr) , 97.21% +/- 0.01 (0.12% patient/yr), 100%; prosthetic valve endocardi tis: 100% in groups A and B, 95.76% +/- 0.02 (0.18% patient/yr) in gro up C; valve-related mortality: 87.52% +/- 0.03 (0.75% patient/yr), 87. 96% +/- 0.03 (0.56% patient/yr), 82.53% +/- 0.04 (0.93% patient/yr); v alve failure: 81.22% +/- 0.07 (0.56% patient/yr), 63.36% +/- 0.1 (1.06 % patient/yr), 14.31% +/- 0.05 (4% patient/yr); treatment failure: 78. 81% +/- 0.05 (1.12% patient/yr), 76.44% +/- 0.09 (0.62% patient/yr), 8 0.97% +/- 0.04 (1% patient/yr); all valve-related morbidity and mortal ity: 40.43% +/- 0.13 (1.93% patient/yr), 57.76% +/- 0.08 (1.43% patien t/yr), 14.96% +/- 0.05 (4.18% patient/yr); all valve-related morbidity and mortality at 5 years: 91.97% +/- 0.02 (7.8% patient/yr), 87.06% /- 0.03 (3.6% patient/yr), 90.27% +/- 0.03 (2.6% patient/yr); at 10 ye ars: 80.4% +/- 0.03 (4.6% patient/yr), 75.91% +/- 0.03 (2.6% patient/y r), 37.44% +/- 0.05 (4.18% patient/yr). The long-term results of the f irst generation valve prostheses are uniformly unsatisfactory. At leas t 15 years of follow-up are required for a reliable evaluation as all valve-related morbidity and mortality dramatically increase in the sec ond decade of follow-up.