We thank Drs Iqbal and Serruys for their interesting comments on our study.1 Transcatheter aortic valve replacement is a very good alternative to conventional surgical replacement in patients with aortic stenosis at increased risk for surgery. 2,3 As efficient as it is, transcatheter aortic valve replacement has important limitations, including postprocedural aortic regurgitation (AR). Our study was designed to (1) describe the rate of postprocedural AR evaluated at discharge in a large series of consecutive patients treated with balloon-expandable (BE) and self-expandable devices, (2) analyze the predictors of postprocedural ARs in the overall population for each device, and (3) analyze the impact of postprocedural ARs on clinical outcome.The results of our study 1 and others 2 demonstrate clearly that postprocedural AR≥moderate as evaluated by echocardiography at discharge is frequent (15%) and is an independent predictor of long-term mortality. Our study demonstrates also that the clinical significance of a postprocedural AR≥moderate is similar in patients treated with a BE or a self-expandable device, with an hazards ratio for 1-year mortality of 2.5 and 2.1, respectively.It was suggested that postprocedural AR could regress over time, in particular after the implantation of a self-expandable device, thus minimizing the clinical impact of postprocedural AR. Although a serial echocardiography analysis of the CoreValve-Extreme-RiskPivotal-Trial 4 suggested that such late regression could occur, the CoreValve-High-Risk-Pivotal-Trial 3 was less conclusive. In that study, postprocedural AR≥moderate was observed in 9.1% (33/363) at discharge, in 10% (36/356) at 1 month, and in 7.0% (21/291) at 1 year (Supplementary Table S10).3 These figures do not show any regression of postprocedural AR between discharge and 1 month, but the apparent trend observed between 1 month and 1 year could have been as likely attributable to the increased mortality in patients with AR as to a real regression of AR between 1 month and 1 year. The absence of regression of postprocedural AR during the first month 3,4 together with our finding that a postprocedural AR≥moderate at discharge is associated with the same extrarisk of mortality by 1 year for both devices demonstrates that the evaluation of AR at discharge is clinically relevant and suggests that the late regression of AR, if it occurs, is too late to impact mortality.As pointed out, despite a lower rate of postprocedural AR in a BE device, 1,5 there is no clear evidence that implantation of a BE rather than a self-expandable device is associated with a survival advantage. Although there is no definite explanation to this observation, it is likely that some other differences could obviate the benefit of a lower rate of postprocedural AR. Trans apical delivery, which is associated with a high 1-year mortality, and aortic annulus rupture, which is a deadly complication, are highly specific to BE devices and are among the potential candidates. This unanswered question should not mini...