Surgery for Active Infective Mitral Valve Endocarditis: A 20-Year, Single-Center Experience

Background and aim of the study: A retrospective analysis was conducted of the early and long-term results of mitral valve repair (MVRep) and mitral valve replacement (MVR) in patients with isolated infective mitral valve endocarditis. Methods: Between May 1986 and December 2007, a total of 1,163 patients with active infective endocarditis (AIE) were operated on. Of these patients, 497 showed an endocarditic involvement of the mitral valve. Sixty-one of these patients underwent MVRep and 219 MVR, with 24% cases of prosthetic valve endocarditis (PVE). The patients' perioperative characteristics, cumulative survival, freedom from recurrence and reoperation and independent risk factors for early mortality were analyzed. Follow up (0-21 years) was complete in 96.5% of cases; the total follow up was 348 and 810 patient-years (pt-yr) in the MVRep and MVR groups, respectively. Results: Typically, the MVR patients were significantly older (p <= 0.001), preoperatively more often intubated (p = 0.008) and in cardiogenic shock (p = 0.045), and more often underwent emergency surgery (p = 0.023). MVRep was associated with a significantly better survival, with 30-day, one-, five- and 10-year survival rates of 90.1 +/- 3.9%, 83.2 +/- 4.8%, 77.0 +/- 5.7% and 60.5 +/- 8.0%, respectively (p = 0.002).Survival after MVR was significantly worse with abscess formation (p = 0.0002) and PVE (p = 0.038). Freedom from reoperation due to reinfection after 10 years was 89.4 +/- 7.0% after MVRep, with early endocarditis recurrence in two patients (3%), and 91.0 +/- 2.5% after MV, with early recurrence in four patients (2%) (p = 0.46). Multivariate analysis identified pre-operative ventilation (OR = 6.3), mitral valve abscess formation (OR = 5.3), PVE (OR = 3.1) and age >= 60 years (OR = 2.8) as independent risk factors for early mortality. Conclusion: Compared to the MVRep group, patients requiring MVR had more advanced endocarditis and were more critically ill. These results suggest that the early outcome might have been improved if patients had been operated on before either heart failure or the development of septic shock. MVRep for AIE showed a low operative mortality and provided satisfactory freedom from recurrent infection and repeat operation. If all infected material could be resected such that the remaining tissue would allow the re-shaping of a competent valve, then MVRep could be performed also in infective endocarditis, in line with the general recommendations for mitral valve surgery.