Impact of valve prosthesis-patient mismatch estimated by echocardiographic-determined effective orifice area on long-term outcome after aortic valve replacement

Abstract

The impact of valve prosthesis-patient mismatch on long-term outcome after aortic valve replacement estimated by various variables such as projected indexed effective orifice area and internal geometric orifice area obtained from in vivo or in vitro published data is still controversial. The effective orifice area was measured by echocardiography in 533 patients. The mean age of the patients was 71 +/- 9 years; mean follow-up time was 4.7 +/- 2.2 years. The impact of severe (indexed effective orifice area <or=0.6 cm(2)/m(2)) and moderate mismatch (0.6 cm(2)/m(2) < indexed effective orifice area <or=0.85 cm(2)/m(2)) on survival was evaluated by Cox regression. Severe mismatch (hazard ratio: 1.9 [1.08-3.21]) was a significant predictor of survival time after adjustment for age, left ventricular ejection fraction, atrial fibrillation, New York Heart Association class, serum creatinine, and hemoglobin level. The 5- and 7-year survival rates were 71% +/- 4% and 54% +/- 8% for patients with severe mismatch and 83% +/- 4% and 80% +/- 8% for patients with mild mismatch, respectively. The correlation between projected and measured indexed effective orifice area was of medium strength (r = 0.49), and the frequency of observed mismatch depended linearly on the projected indexed effective orifice area. Although projected indexed effective orifice area and indexed internal geometric orifice area were significant predictors of severe mismatch, the sensitivity and specificity for severe prosthesis-patient mismatch were only 75% and 52%, using an optimal threshold of projected indexed effective orifice area defined by the Youden index. Severe prosthesis-patient mismatch estimated by effective orifice area measured within 10 days was an independent risk factor of survival time. Projected indexed effective orifice area determined at surgery does not sufficiently predict mismatch.