Aortic valve morphology and paravalvular leak regression after a self-expandable transcatheter aortic valve replacement

Abstract

Aims: The study aimed to compare paravalvular leak (PVL) changes after a transcatheter aortic valve replacement (TAVR) with self-expandable prosthesis between different aortic valve morphologies and evaluate the impact of paravalvular leak regression on clinical prognosis. Methods: Patients with aortic stenosis (AS) successfully treated with a self-expandable TAVR who were followed up for at least 1 year at our centre were consecutively enrolled from January 2016 to August 2019. Paired serial changes in paravalvular leak and other haemodynamic parameters by echocardiography were collected and compared between the bicuspid valve (BAV) and tricuspid aortic valve (TAV). A logistic regression model was used to explore the predictors of paravalvular leak regression (<1 grade) 1 year after transcatheter aortic valve replacement, while its impact on subsequent clinical outcomes (all-cause mortality and rehospitalisation for heart failure (HF)) was further evaluated using Kaplan–Meier analysis. Results: A total of 153 bicuspid valve and 114 tricuspid aortic valve patients were finally enrolled; haemodynamic parameters and paravalvular leak severity were comparable before the discharge between the two groups. The peak transaortic velocity, mean transvalvular gradient, and effective orifice area all significantly improved (p < 0.05) without intergroup differences at all follow-up timepoints. Significant paravalvular leak reduction was observed only in the TAV group (1.75% vs. 4.39%, p = 0.029), while moderate paravalular leak was still more prevalent in BAV (7.19% vs. 1.75%, p = 0.041) at the 1-year follow-up. Multivariable analyses identified the bicuspid valve, asymmetric calcification, and undersizing as independent predictors of failure of the 1-year paravalvular leak reduction in patients with mild or moderate paravalvular leak after discharge. Patients without a paravalvular leak reduction within 1 year showed a relatively higher 2-year all-cause mortality and HF (HR: 5.994, 95% CI: 1.691–21.240, and p = 0.053) rates thereafter. Conclusion: In AS patients after self-expandable transcatheter aortic valve replacement, paravalvular leak regression within 1 year was less prevalent in bicuspid valve morphology. The failure of paravalvular leak reduction might lead to an increased risk of poorer prognosis in the long run.