Applicability of Transcatheter Aortic Valve Replacement Trials to Real-World Clinical Practice: Findings From EXTEND-CoreValve.

Objectives The aim of this study was to examine the applicability of pivotal transcatheter aortic valve replacement (TAVR) trials to the population of Medicare patients undergoing TAVR. Background It is unclear whether randomized controlled trial results of novel cardiovascular devices apply to patients encountered in clinical practice. Methods Characteristics of patients enrolled in the U.S. CoreValve pivotal trials were compared with those of the population of Medicare beneficiaries who underwent TAVR in U.S. clinical practice between November 2, 2011, and December 31, 2017. Inverse probability weighting was used to reweight the trial cohort on the basis of Medicare patient characteristics, and a real-world treatment effect was estimated. Results A total of 2,026 patients underwent TAVR in the U.S. CoreValve pivotal trials, and 135,112 patients underwent TAVR in the Medicare cohort. Trial patients were mostly similar to patients at baseline, though trial patients were more likely to have hypertension (50% vs 39%) and coagulopathy (25% vs 17%), whereas patients were more likely to have congestive heart failure (75% vs 68%) and frailty. The estimated treatment effect of TAVR was an 11.4% absolute reduction in death or stroke (95% CI: 7.50%-14.92%) and an 8.7% absolute reduction in death (95% CI: 5.20%-12.32%) at 1 year with TAVR compared with conventional therapy (surgical aortic valve replacement for intermediate- and high-risk patients and medical therapy for extreme-risk patients). Conclusions The trial and populations were mostly similar, with some notable differences. Nevertheless, the extrapolated treatment effect was at least as high as the observed trial treatment effect, suggesting that the absolute benefit of TAVR in clinical trials is similar to the benefit of TAVR in the U.S. setting.