Early and late survival following replacement of prosthetic heart valves

From May 1975 to December 1988, 62 patients underwent 66 replacements of prosthetic heart valves, due to structural valve failure in 37, prosthetic valve endocarditis (PVE) in 17, periprosthetic leak in 5, thrombosed valve in 3, hemolysis in 2, and prophylactic removal in 2. Before reoperation, 3 patients were in NYHA Class I, 24 in Class II, 21 in Class III, and 18 in Class IV. Replacements of valve prostheses were at the mitral position in 33 patients, at the aortic position in 16, at the aortic and mitral position in 9, at the mitral and tricuspid position in 5, at the tricuspid position in 2, and at the pulmonary position in 1. There were 6 early deaths (9.1%), and 13 late deaths (7.2 +/- 0.4%/patient-year). Clinical improvement of the current survivors was good. The follow-up duration was 4 months to 11.4 years (mean 3.5 years), and the cumulative follow-up was 180 patient-years (100% complete). Patients undergoing replacement of 2 prostheses showed a higher early mortality rate (35.7%) than those who received replacement of 1 prosthesis (1.9%) (p less than 0.01). The actuarial survival including early deaths was compared using the generalized Wilcoxon test. Survival was better in patients with replacement of one prosthesis than in those with replacement of 2 prostheses (p less than 0.001), and was better in NYHA Class I to III patients than in Class IV patients (p less than 0.05). Replacement of 1 prosthesis for patients with primary tissue failure of bioprostheses showed no operative mortality and a good 5 year survival rate (96 +/- 4%). High risk factors for replacement of prosthetic valves included double valve replacement, NYHA Class IV and PVE.