Infective endocarditis in opiate addicts: Analysis of 80 cases studied at necropsy

Abstract

Eighty opiate addicts were studied at necropsy. Fifty-nine patients had anatomic evidence of active infective endocarditis (IE); 11 had healed IE; and 10 had both. Of the 80 patients, the first episode of IE involved a single right-sided cardiac valve in 24 patients (30%); both a right- and a left-sided valve in 13 patients (16%); a single left-sided valve in 33 patients (41%); and both left-sided valves in 10 patients (13%). Of the 320 cardiac valves in the 80 patients, 103 were sites of vegetations, an average of 1.3 of the 4 valves. Of the 80 patients, the tricuspid valve was infected in 35 (44%), mitral in 34 (43%), aortic in 32 (40%) and pulmonic in 2 (3%). Of the 103 infected cardiac valves, the infection caused sufficient damage to cause dysfunction in 70 (68%): in 28 (88%) of 32 infected aortic valves; in 22 (63%) of 35 infected tricuspid valves; in 19 (56%) of the 34 infected mitral valves; and in 1 of the 2 infected pulmonic valves. Of the 80 patients, 57 (71%) had sufficient valvular damage to cause valvular dysfunction. Of the 80 patients, gross examination of the valves at necropsy indicated that the infected valve almost certainly had been anatomically normal in 65 patients (81%) and abnormal in 15 patients (19%) before the onset of IE. Of the 65 patients with previously anatomically normal valves, 86 (33%) of their 260 cardiac valves were sites of infection (average 1.3 valves/patient); of the 15 patients with infection superimposed on a previously abnormal valve, the infection in each involved previously abnormal valves (21 in the 15 patients) or 17 (28%) of their 60 cardiac valves were sites of infection (average 1.1 valve/patient). Of the 15 patients with abnormal cardiac valves before the infection, 7 had congenitally bicuspid aortic valves and 8 had diffuse fibrous thickening of the mitral valve typical of rheumatic heart disease with (6 patients) or without (2 patients) diffuse fibrous thickening of tricuspid aortic valves. Of the 80 patients, 42 (53%) died during their first episode of active IE, 17 (21%) underwent operative excision with or without valve replacement during the active IE, and in 21 patients (26%) the first episode of active IE healed. In 10 of the latter 21 patients, active IE recurred and was fatal. A total of 19 patients had cardiac valve excision with or without replacement, 17 during active IE and 2 after healing. Of the 19 patients, 5 died within 30 days of operation, 10 (56%) within a year of operation and 16 (80%) within 3 years. Of the 13 patients surviving > 30 days after valvular operation, 5 developed prosthetic or bioprosthetic endocarditis, which was fatal in each. Thus, the first episode of active IE appears to have a high mortality because the infection most often involves previously normal valves that are easily destroyed by vegetations. When the valvular vegetations are operatively excised, with or without replacement, prognosis remains poor because of the danger of prosthetic or bioprosthetic endocarditis. When the active IE is eradicated by antibiotics without operation, recurrence of IE at sites of previously healed IE is common.