In a pooled analysis from two randomized trials, transcatheter aortic valve implantation (TAVI) was associated with significantly less bioprosthetic valve dysfunction (BVD) than a surgical prosthetic implantation, according to data presented as a late-breaker at the Cardiovascular Research Technologies conference, sponsored by MedStar Heart & Vascular Institute.
“The difference in valve performance was driven by a twofold lower SVD [structural valve deterioration] and a 3-fold lower severe PPM [prothesis-patient mismatch] for TAVI versus surgery,” reported Steven J. Yakubov, MD.
The data were pooled from the CoreValve U.S. Pivotal and SURTAVI randomized trials. Of patients participating in these two trials, 5-year follow-up data were available for 1,128 randomized to the CoreValve/Evolut TAVI and 971 randomized to surgical prosthetic valve replacement.
The major focus of the study was on the cumulative incidence of BVD, but the study also included separate analyses on the relationship between BVD and clinical outcomes. Preprocedural indicators for BVD at 5 years were also analyzed.
SVD was defined as a mean gradient increase of at least 10 mm Hg from discharge to 30 days, along with at least 20 mm Hg at last echo or new-onset aortic regurgitation. Nonstructural valve deterioration (NSVD) was defined as severe PPM at discharge or 30 days or severe paravalvular regurgitation through 5 years. In addition to these two components, the BVD endpoint also included thrombosis and endocarditis.
Surgical valve deterioration high at 5 years
On the basis of these definitions, the rate of BVD at 5 years was 14.2% in the surgery group and 7.8% in the TAVI group, translating into a 50% risk reduction in favor of TAVI (hazard ratio, 0.50; P < .001).
Thrombosis or endocarditis occurred in low rates in both groups, but every other component of BVD favored TAVI significantly, not just numerically. This included SVD (2.2% vs. 4.4%; P = .004), and the two components of NSVD, PPM (3.7% vs. 11.8%; P < .001) and severe paravalvular regurgitation (0.2% vs. 1.2%; P = .02).
When stratified by annular diameter, the relative advantage of TAVI over surgery was greatest in those valves with diameters of up to 23 mm. In this group, the lower relative rate in the TAVI group (8.6% vs. 19.7%) represented a nearly 70% reduction in risk of valve deterioration at 5 years (HR, 0.31; P < .001).
However, the advantage at 5 years also remained substantial and significant in larger valves (8.1% vs. 12.6%), translating into a 40% risk reduction in favor of TAVI (HR, 0.60; P = .002).
Independent of type of valve replacement, BVD at 5 years was associated with worse outcomes, including significantly increased risks for all-cause mortality (HR, 1.46; P = .004), cardiovascular mortality (1.84; P < .001), and hospitalization for valve disease or worsening heart failure (HR, 1.67; P = .001).
The baseline characteristics that were statistically associated with BVD at 5 years on multivariate analysis in pooled data from both the TAVI and surgical groups included age (P = .02), a creatinine clearance less than 30 mL/min per 1.73 m2 (P = .006), and a low relative baseline left ventricular ejection fraction (P < .001).