Impact of Frailty and Prefrailty on Outcomes of Transcatheter or Surgical Aortic Valve Replacement.

Abstract

Randomized trials have shown short- and mid-term benefits with transcatheter versus surgical aortic valve replacement (TAVR versus SAVR) for patients at intermediate or low-risk for surgery. Frailty and prefrailty could explain some of this benefit due to an impaired ability to recover fully from a major surgical procedure. We examined 2-year outcomes (survival and Kansas City Cardiomyopathy Questionnaire [KCCQ] scores) among patients at intermediate or low surgical risk treated with transfemoral-TAVR or SAVR within the PARTNER (Placement of Aortic Transcatheter Valves) 2A trial, SAPIEN 3 intermediate-risk registry, and PARTNER 3 trial. Frailty was examined as a continuous variable based on grip strength, gait speed, serum albumin, and activities of daily living. We tested the interaction of frailty markers by treatment (TAVR versus SAVR) in proportional hazards regression models (survival) and piecewise linear regression models (KCCQ), adjusting for patient demographic and clinical factors. Among the 3025 patients in the analytic cohort (2003 TAVR, 1022 SAVR; mean age 79.3 years, 61.6% men), 799 (26.4%) were nonfrail, 2041 (67.5%) were prefrail (1-2 frailty markers), and 185 (6.1%) were frail (3-4 frailty markers). Increasing frailty (none versus prefrail versus frail) was associated with higher 2-year mortality (5.5% versus 11.1% versus 22.8%; log-rank P<0.001) and worse 2-year health status among survivors (KCCQ scores adjusted for baseline: 84.8 versus 79.6 versus 77.4, P<0.001). In multivariable models, there were no significant interactions between frailty markers and treatment group for either survival (interaction P=0.39) or health status (interaction P>0.47 for all time points). In a cohort of older patients with severe aortic stenosis who were at low or intermediate surgical risk, increasing frailty markers were associated with worse 2-year mortality and greater health status impairment after either TAVR or SAVR, but there were no significant interactions between type of valve replacement and frailty with respect to either outcome.