Influence of patient-prosthesis mismatch on long-term results after aortic valve replacement with a stented...

The study aim was to compare long-term results of bioprostheses implanted in the aortic position, with and without patient-prosthesis mismatch (defined as effective orifice area (EOA)/body surface area (BSA) < or =0.85 cm2/m2).

Between 1986 and 1990, 90 consecutive patients (mean age 72.6 years; mean BSA 1.77+/-0.2 m2) each received an aortic Medtronic Intact valve (19 mm, n = 35; 21 mm, n = 29; >23 mm, n = 26). Of these patients, 64 had a patient-prosthesis mismatch, and 26 had no mismatch. Median follow up was 7.3 years.

At 10 years postoperatively, there was no significant inter-group difference in actuarial freedom from thromboembolism (90.7% in mismatch group, 79.6% in no-mismatch group; p = 0.16), hemorrhage (86% versus 83.3%; p = 0.59), endocarditis (98.2% versus 86.7%; p = 0.1), structural valve deterioration (97% versus 100%; p = 0.57) and reoperation (96.4% versus 94.8%; p = 0.2). At the same time, overall actuarial survival was 42.1+/-6.5% in the mismatch group and 22.6+/-8.6% in the no-mismatch group (p = 0.08). By multivariate analysis, the main risk factor for late death was a preoperative left ventricular ejection fraction (LVEF) <50% (p = 0.001). Freedom from cardiac death was 70+/-6% and 35.7+/-11% in the mismatch and no-mismatch groups respectively (p = 0.005), but this was not significantly different when LVEFs were paired (LVEF >50% p = 0.33, LVEF <50% p = 0.28). The NYHA functional status of survivors showed 94.4% of the mismatch group and 100% of the no-mismatch group to be in NYHA classes I and II (p = 1).

Within this patient population it was not possible to demonstrate any negative effects of patient-prosthesis mismatch at 10 years after Intact aortic valve replacement; the LVEF was the only predictor for late death.