LONG-TERM RESULTS OF SURGERY FOR ACTIVE INFECTIVE ENDOCARDITIS

Objective: This paper was undertaken to determine the long-term outcom e of active infective endocarditis treated with antibiotic and radical excision of infected tissues by surgery. Methods: From October 1978 t o August 1993, 122 consecutive patients were operated on during the ac ute phase of infective endocarditis. There were 85 men and 37 women wh ose mean age was 50 years, ranging from 20 to 79. Surgery was needed b ecause of one or more of the following complications: cardiogenic/sept ic shock in 19 patients, congestive heart failure in 68, persistent se psis in 64, peripheral embolization in 20, and cerebral embolization i n 10. The offending microorganism was identified in 110 patients, stap hylococci were the most common ones. Seventy-six patients had native v alve endocarditis and 46 had prosthetic valve endocarditis. Simple val ve replacement or repair was performed in 60 patients and radical rese ction of the valve and surrounding tissues with reconstruction of the heart with either fresh autologous pericardium or glutaraldehyde-fixed bovine pericardium was performed in 62 with paravalvular abscess. Pul monary autograft and aortic homograft were used in only three patients , the remaining patients had either bioprostheses or mechanical heart valves if valve repair was not feasible. Results: There were nine deat hs, for an operative mortality of 7.4%. Logistic regression analysis i dentified preoperative shock and renal failure as predictors of operat ive mortality. Operative survivors were followed up from 4 to 173 mont hs, mean of 56.4. The actuarial survival at 10 years was 61 +/- 6%. Lo gistic regression analysis identified preoperative New York Heart Asso ciation functional class IV and perioperative renal failure as predict ors of late mortality. Eight patients developed recurrent endocarditis 10-102 months postoperatively. The freedom from recurrent endocarditi s at 10 years was 79 +/- 9%. All patients who developed this late comp lication had paravalvular abscess at the time of original operation. C onclusions: These data suggest that surgery for active infective endoc arditis yield a high probability of eradicating the infection with rel atively low operative mortality and good long-term results. Copyright (C) 1997 Elsevier Science B.V.