Aprotinin in Primary Cardiac Surgery: Operative Outcome of Propensity Score-Matched Study

Background Some recent multicenter series have questioned the safety of aprotinin in primary cardiac operations. We report a large, single-center experience with aprotinin therapy in primary cardiac operations and discuss the limitations and potential confounders of current treatment strategies. Methods We compared myocardial infarction, neurologic events, renal insufficiency, and operative death after first-time coronary or valve procedures, or both, in 3334 patients treated with full-dose aprotinin with 3417 patients not treated with aprotinin who underwent operation between March 1998 and January 2007. Further analysis was performed for 341 propensity score-matched pairs. Results There were substantial differences between the groups. Aprotinin patients were higher risk on account of older age, unstable symptoms, poor ejection fraction, preoperative hemodynamic support, emergency/urgent operations, and combined coronary/valve operations. Postoperative bleeding and blood product transfusion were considerably reduced in aprotinin patients, as was median duration of mechanical ventilation. Aprotinin was neither a predictor of postoperative myocardial infarction, renal insufficiency, neurologic dysfunction, or operative death. Achieving parity between the groups by propensity score matching eliminated the elevated rates of postoperative renal insufficiency, neurologic dysfunction, and operative death observed in aprotinin patients in the unmatched comparison. These adverse outcomes were evenly distributed between matched groups. Conversely, blood transfusion had univariate associations with all adverse outcome measures. Conclusions Full-dose aprotinin use was not associated with myocardial infarction, neurologic dysfunction, renal insufficiency, or death after coronary or valve operations. We observed less postoperative bleeding and blood product transfusion, and early extubation with the use of aprotinin.