The 365 patients studied came from both an initial registry and a continued access registry, but were largely similar on baseline characteristics. All underwent valve-in-valve TAVR withtranscatheter heart valves.
Mortality in the cohort was 12.1% at 1 year, 22.2% at 2 years, and 32.7% at 3 years, according to results reported at the meeting, which is sponsored by the Cardiovascular Research Foundation. In contrast, the rate of stroke was stable over time, at 5.1%, 5.1%, and 6.2%, respectively.
Repeat valve replacement (either surgical or transcatheter) had been performed in 0.6% of patients at 1 year, 0.6% at 2 years, and 1.9% at 3 years. “I think this is comparable to [what is seen in] surgical series,” Dr. Webb commented.
Between 30 days and 3 years, there were insignificant decreases in total aortic regurgitation that was moderate or worse in severity (from 2.9% to 2.5%) and paravalvular aortic regurgitation of these severities (from 2.6% to 1.4%). Although the valve used was older, “still, we had excellent sealing and aortic insufficiency was not a problem with these patients,” he noted.
In “interesting” findings, prevalence of mitral regurgitation that was moderate or worse continued falling, from 17.2% at 30 days to 8.6% at 3 years, and prevalence of tricuspid regurgitation that was moderate or worse did as well, from 21.8% to 18.8%.
“I was a little suspicious this was just a survival issue, that patients with severe mitral or tricuspid regurgitation died and, consequently, the average patient was less likely to have [these findings]. But the analysis that’s being done is linear mixed-effects analysis, which accounts for the survival bias,” Dr. Webb said.
The reasons for these trends are unknown, but possibly improved left ventricular function led to functional (rather than structural) improvements in mitral and tricuspid regurgitation.
At 3 years, proportions of patients with various NYHA classes were much the same as they had been at 30 days: class I (51.4% vs. 53.9%), class II (34.6% vs. 35.7%), and class III (13.0% vs. 9.2%). Similarly, the mean KCCQ overall summary score at 30 days (70.8) was sustained at 3 years (73.1).
Risk of death did not differ significantly according to the surgical valve size as labeled, the surgical valve true internal dimensions, the mode of valve failure, the approach used (transfemoral vs. transthoracic), or the residual gradient after valve implantation.
Analysis of the registry data is ongoing. For example, the investigators will be looking more closely at determinants of outcomes, such as additional characteristics of the surgical valve alone and in combination with those of the new valve. “We are all very aware that a lot of the outcomes have to do with what surgical valves you had to begin with. I think that is really critical – what surgical valve is in there,” he said.
Dr. Webb reported that he receives grant/research support and honoraria from, and is on the steering committee for, Edwards Lifesciences. The registry is sponsored by Edwards Lifesciences.