Why the categorization of indexed effective orifice area is not justified for the classification of prosthesis-patient mismatch.

Objectives: Although the impact of prosthesis-patient mismatch (PPM) on survival has been widely studied, there has been little debate about whether the current definition of PPM truly reflects hemodynamic obstruction. This study aimed to validate the categorization of indexed effective orifice area (EOAi) for the classification of PPM.
Methods: In total, 2171 patients who underwent aortic valve replacement with a surgical stented bioprosthesis in 5 trials (CoreValve US High-Risk, SURTAVI [Surgical Replacement and Transcatheter Aortic Valve Implantation Trial], Evolut Low Risk, PERIGON [PERIcardial SurGical AOrtic Valve ReplacemeNt] Pivotal Trial for the Avalus valve, and PERIGON Japan) were used for this analysis. The echocardiographic images at the 1-year follow-up visit were evaluated to explore the association between EOAi and mean aortic gradient and its interaction with other patient characteristics, including obesity. In addition, different criteria of PPM were compared with reflect elevated mean aortic gradients (≥20 mm Hg).
Results: A relatively smaller exponential decay in mean aortic gradient was found for increasing EOAi, as the slope on the log scale was -0.83 versus -2.5 in the publication from which the current cut-offs for PPM originate. The accuracy of the American Society of Echocardiography, Valve Academic Research Consortium-2, and European Association of Cardiovascular Imaging definitions of PPM to reflect elevated mean aortic gradients was 49%, 57%, and 57%, respectively. The relation between EOAi and mean aortic gradient was not significantly different between obese and non-obese patients (P = .20).
Conclusions: The use of EOAi thresholds to classify patients with PPM is undermined by a less-pronounced exponential relationship between EOAi and mean aortic gradient than previously demonstrated. Moreover, recent adjustment for obesity in the definition of PPM is not supported by these data.
(Copyright © 2020 The Author(s). Published by Elsevier Inc. All rights reserved.)