A prospective, randomized comparison of 3 contemporary bioprosthetic aortic valves: Should hemodynamic performance influence device selection?

Abstract

Latest generation biologic aortic valve prostheses were designed to improve hemodynamic performance. We sought to determine whether there are clinically important early differences among these devices. Three hundred adults with severe aortic valve stenosis undergoing aortic valve replacement were randomized to receive the Edwards Magna, Sorin Mitroflow, or St. Jude Epic bioprostheses (n=100, n=101, n=99, respectively). Early hemodynamic performance was studied by echocardiography. Mean patient age was 76 ± 8 years and there were 203 men (68%). There were no significant differences in baseline characteristics among implant groups. Early mortality was 1.7%, and there were no differences in early adverse events. Postoperative echocardiography showed small but statistically significant differences overall between the Magna, Mitroflow, and Epic valves in mean gradient (14.2 mm Hg, 16.3mm Hg, 16.5 mm Hg, respectively; P=.011), aortic valve area (2.05 cm(2), 1.88 cm(2), 1.86 cm(2), respectively; P=.012), and indexed aortic valve area (1.05 cm(2)/m(2), 0.97 cm(2)/m(2), 0.95 cm(2)/m(2), respectively; P=.012). Prosthetic performance was similar among all with a small (≤21 mm) aortic annulus. Patients who received the Magna device with a 23-mm annulus had slightly greater indexed aortic valve area; those with >23 mm had a slightly lower transprosthetic gradient. Analogous trends were found when data were stratified by either commercial implant size or echocardiography-determined aortic annulus size. Severe patient-prosthesis mismatch was infrequent overall and was similarly low among devices (P value not significant). This prospective, randomized comparison reveals that there are small but consistent early postoperative hemodynamic differences among current third-generation porcine and pericardial aortic valve prostheses. The 3 valves studied performed equally well in patients with a small (≤21 mm) aortic annulus. The Magna valve had a slightly lower mean gradient in those with larger annular size (>23 mm). Longitudinal follow-up of these randomized cohorts is essential to determine late clinical implications of these early postoperative findings.