Background:The current clinical practice standard is 10% to 20% oversizing of self-expanding valves in transcatheter aortic valve replacement. We aimed to determine whether >20% oversizing of self-expanding valves (Medtronic Evolut) would lead to better valve performance with similar or better outcomes.

Methods:From October 2011 to December 2016, we approached all transcatheter aortic valve replacement patients with a conscious attempt at large oversizing (>20%). The most common valve used, excluding those used in valve-in-valve patients, was the 29-mm Evolut R (29%). We used a retrospective chart review to compare moderate oversizing (group 1; 10% to 20%) with large oversizing (group 2; >20%).

Results:Of 556 patients, 45% were male; the overall mean Society of Thoracic Surgeons risk score was 5.8 ± 3.8. Eighty-five (15%) patients needed a pacemaker, and 21 (3.8%) developed significant paravalvular leak. Mean oversizing was 20.3% ± 6.0%, with 41.4% of patients included in group 1 and 54.5% in group 2. Incidences of complications in group 2 vs. group 1 were as follows: a) paravalvular leak (2.0 vs. 6.1%; odds ratio = 0.31, p = 0.01), b) pacemaker (15 vs. 14%), c) gastrointestinal bleed (n = 4 vs. 0; 1.3 vs. 0.0%; p = 0.03), d) annular dissection (n = 1 vs. 0; 0.3 vs. 0%; p = 0.29), e) mortality (n = 5 vs. 4; 1.6 vs. 1.7%). Incidence of paravalvular leak was higher in those who died than survivors (13 vs. 1.3%; p ≤ 0.0001).

Conclusions:These data suggest that, in current self-expanding valves, >20% oversizing delivers a significantly lower prevalence of paravalvular leak without an increase in other complications. Since paravalvular leak is associated with increased mortality, >20% oversizing may represent a superior prosthesis choice.