[Reoperations on heart valve prostheses. Apropos of 58 cases]

In the period from 1968 to November 1980, 1 023 patients underwent surgery for single or multiple valve replacement. Fifty three patients (6,6% of the follow-up population) had to be reoperated, including 5 patients who had to be reoperated twice, giving a total of 58 reoperations. The average interval before reoperation was 30 months. The incidence was similar in monovalvular (7,5 p. 100 mitral valves, 5 p. 100 aortic valves) and polyvalvular (7 p. 100) cases. On the other hand, the incidence of reoperation of tricuspid prostheses (17 p. 100) was significantly superior to that of mitral valve (5,3 p. 100) or aortic valve (3,8 p. 100) prostheses. In 91 p. 100 of cases, the indication for reoperation was prosthetic valve dysfunction related to endocarditis in over a third of cases (21). In 32 cases, reoperation was required in the absence of any infectious process: 13 spontaneous perivalvular leaks, 10 thromboses, and 9 stenosing prostheses. There were no reoperations for wear of the prosthetic material. Only 9 p. 100 of patients were reoperated for uncorrected valvular disease. The prognosis of these reoperations was poor; hospital mortality being 42,5 p. 100. This high mortality rate is explained by the frequency of reoperation for infective endocarditis (36 p. 100) in our series, the mortality of which was 73,6 p. 100 and even higher when reoperation was an emergency for infectious or hemodynamic reasons. There was also a high mortality rate with reoperation for thrombosis (30 p. 100) because of the severe myocardial dysfunction in thrombosis of tricuspid prostheses and the emergency situation associated with mitral prosthetic valve thrombosis. Excluding these two complications, the average mortality was 21 p. 100. Although the surgical indications are relatively easy for thrombosis, perivalvular leak and stenosing prostheses, they are particularly difficult in infectious endocarditis especially with regards to the timing of reoperation. We believe that, ideally, reoperation should be delayed as long as possible to allow the antibiotic therapy the maximum time to take effect. Surgery can then be performed after controlling the infection and before the installation of severe hemodynamic distress.