Abstract

Patient-prosthesis mismatch is usually accepted to be associated with poor outcomes in patients with aortic valve replacement (AVR). This study aims to evaluate prevalence, sensitivity and specificity of mismatch measured with in vivo and in vitro methods, look for a relationship between mismatch and obesity, and investigate the effect of mismatch on left ventricle mass index (LVMI) regression and mortality. A total of 72 consecutive patients who underwent mechanical AVR between December 2011 and May 2013, were prospectively evaluated. EOA was measured with echocardiography in all patients on the 6th postoperative month and an Indexed Effective Orifice Area (EOA) ≤0.85cm2/m2 was accepted as mismatch with the in vivo measurement method. For the in vitro measurement method, charts provided by the valve manufacturers were used for EOA prediction. LVMI was also evaluated on the 6th and 12th postoperative months. Postoperative follow-up is 100% complete with 68.5±14.4 months. In vivo and in vitro mismatch prevalences were 43.5% and 25.0% with slight concordance (kappa=0.172). Sensitivity of in vitro measurements was poor (35.7%), but specificity was 80.5%. LVMI regressions were significant with both mismatch methods (p0.05 in both). In vitro EOA measurements have a poor sensitivity to predict mismatch preoperatively. Left ventricular mass regressions were significant in all groups with no difference in early and late mortality.

Key words: Aortic valve replacement, prosthesis-patient mismatch, in vivo indexed EOA, in vitro indexed EOA