Author + information
- Mohamed Abdel-Wahab, MD∗ ( and )
- Gert Richardt, MD
- ↵∗Heart Center, Segeberger Kliniken GmbH, Academic Teaching Hospital of the Universities of Kiel, Hamburg, and Lübeck, Am Kurpark 1, Bad Segeberg 23795, Germany
We appreciate the thoughtful comments of Dr. Montone and colleagues concerning the 1-year clinical outcome of the CHOICE trial, but do not share their concerns about a differential impact of baseline characteristics on the observed mortality rates in a randomized setting. Among all baseline clinical and echocardiographic characteristics of the CHOICE population, only sex was statistically significantly different between the balloon-expandable (BE) and self-expandable (SE) groups (age and baseline left ventricular function were not significantly different) (1,2). As mentioned in the paper, we performed a logistic regression analysis to adjust for sex, and the results are essentially unchanged if sex is taken into account (2) (unadjusted p value for all-cause mortality at 1-year using the Fisher exact test = 0.37, adjusted p value for all-cause mortality at 1-year using logistic regression = 0.33).
The assumption of Dr. Montone et al. that more than mild prosthetic valve regurgitation was not associated with higher mortality in the CHOICE population is not correct. As briefly mentioned in the discussion section of our paper (2), device success (which was mainly driven by the absence of more than mild paravalvular leaks) was independently associated with improved survival at 1 year (adjusted odds ratio calculated by logistic regression = 0.16, 95% confidence interval: 0.04 to 0.67, p = 0.01). As previously discussed, the lack of a mortality difference between both devices despite differences in device success could be partially explained by the moderate sample size of this study as well as the numerically higher rate of thromboembolic events in the BE group, although this remains speculative.
Finally, a potential association between the numerically higher incidence of baseline atrial fibrillation (AF) and the occurrence of stroke in the BE group cannot be entirely excluded. In fact, 6 of 11 stroke events in the BE group occurred in patients with AF, but the rate of new-onset AF was nearly identical in the BE group (9.8%) and SE group (9.4%, p = 1.0) (2).
Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2016 American College of Cardiology Foundation