Reoperative surgery for degenerated aortic bioprostheses: predictors for emergency surgery and reoperative mortality.

OBJECTIVE: The long-term outcome of patients with aortic bioprosthetic valves could be improved by decreasing the reoperative mortality rate.
METHODS: Predictors of emergency reoperation and reoperative mortality were identified retrospectively in 172 patients who had the first bioprosthetic aortic valve replacement between 1975 and 1988 (mean age 46+/-13 years) and were subjected to replacement of the degenerated bioprostheses between 1978 and 1997 (mean age 56+/-14 years). Emergency reoperation had to be performed in 31 patients (18%).
RESULTS: The operative mortality was 5.2% (9/172), 22.6% for emergency (odds ratio 11.17; 95%-confidence limit 4.33-28.85) and 1.4% for elective replacement of the degenerated aortic bioprosthesis (P<0.0001; OR=20.3). Patients who died at reoperation had higher transvalvular gradients before the primary aortic valve replacement (P=0.007), received smaller bioprostheses at the first operation (P=0.03), had later recurrence of symptoms after the first aortic valve replacement (P=0.04), a higher pre-reoperative New York Heart Association (NYHA) class (P=0.02), and a higher incidence of coronary artery disease (P=0.001) and pulmonary artery hypertension (P=0.009). Endocarditis before the primary aortic valve replacement (P=0.004), postoperative pneumonia at the first operation (P=0.005), pulmonary hypertension (P=0.0004) acquired during the interval, later recurrence of symptoms (P=0.04) after the first operation, a lower ejection fraction at the time of reoperation (P=0.03) and acute onset of bioprosthetic regurgitation (P=0.00002) were predictors for emergency surgery. Higher transvalvular gradients at the primary aortic valve replacement (P=0. 006), coronary artery disease (P=0.003) acquired during the interval, the need for concomitant coronary artery revascularization (P=0. 001), sex (P=0.02) and size (P=0.05) and type of the bioprostheses used (P=0.007) were incremental predictors for reoperative mortality which were independent of emergency surgery.
CONCLUSIONS: Elective replacement of failed aortic bioprostheses is safe. Patients undergoing emergency reoperation have a considerably higher mortality. They can be identified by a history of native aortic valve endocarditis, higher transvalvular gradients at primary aortic valve replacement, smaller bioprostheses, and pulmonary hypertension or coronary artery disease acquired during the interval. A failing bioprosthesis must be replaced at its first sign of dysfunction.