By Jeffrey Zimmet, MD, PhD
Associate Professor of Medicine, University of California, San Francisco; Director, Cardiac Catheterization Laboratory, San Francisco VA Medical Center
The PARTNER 3 trial randomized low-risk patients to transcatheter aortic valve replacement vs. surgical aortic valve replacement. Five-year data show no significant differences in the composite endpoint of death, stroke, or rehospitalization.
Mack MJ, Leon MB, Thourani VH; PARTNER 3 Investigators. Transcatheter aortic-valve replacement in low-risk patients at five years. N Engl J Med 2023; Oct 24. doi: 10.1056/NEJMoa2307447. [Online ahead of print].
Transcatheter aortic valve replacement (TAVR) initially was approved as an alternative to surgical aortic valve replacement (SAVR) for patients at high or prohibitive risk from surgery. Since TAVR now is generally approved for the treatment of severe aortic stenosis (AS) regardless of the degree of surgical risk, TAVR has surpassed SAVR in numbers in the United States as in much of the world.
The PARTNER 3 trial was a comparative study in which 1,000 patients at low surgical risk were randomly assigned to either TAVR (503 patients) or SAVR (497 patients). The short-term results of this trial have been published previously. The average age of enrolled patients was 73 years, the mean Society of Thoracic Surgeons mortality risk score was 1.9%, and 69% of patients were men. Concomitant coronary revascularization was performed in 6.5% and 12.8% of the TAVR and SAVR groups, respectively. A majority of patients remained in the group to which they were assigned — 495 patients in the TAVR group and 453 patients in the surgery group received the intended valve. Complete five-year follow-up data were available for almost 92% of patients, with more surgical than TAVR patients lost to follow-up (complete data were available in 94.6% of TAVR patients compared with 88.3% of SAVR patients).
The primary endpoint, a composite of death, stroke, or rehospitalization related to the valve, procedure, or congestive heart failure (CHF), occurred in 111 of 496 TAVR patients and in 117 of 454 patients in the SAVR group, which was not statistically different. Individual components of the primary endpoint also were not significantly different, although deaths were numerically higher in the TAVR group, and there was a nonsignificant trend toward higher rehospitalization in the SAVR group — 13.7% vs. 17.4% (hazard ratio, 0.75; 95% confidence interval, 0.54 to 1.05).
A review of secondary endpoints showed some expected differences. For example, serious bleeding was more frequent among SAVR patients, and new-onset atrial fibrillation was threefold higher in the SAVR group compared with the TAVR group. On the other hand, new requirement for a permanent pacemaker was higher among TAVR patients, but the absolute difference was small: 13.5% of the patients in the TAVR group and 10.4% of those in the surgery group required pacemakers. Likewise, clinically significant valve leaflet thrombosis was more frequent with TAVR, occurring in 12 patients (2.5%) in the TAVR group and in only one patient (0.2%) in the surgery group.
At five years, echocardiographic analysis showed similar aortic valve gradients and calculated aortic valve areas between the two groups. TAVR patients were more likely to have at least mild aortic regurgitation, which was seen in 24.5% of TAVR patients and in only 6% of SAVR patients. Five-year mortality was not different among patients with aortic regurgitation compared with those without. Bioprosthetic valve failure of any cause was estimated at 3.3% of the TAVR group and 3.8% of the surgery group, while “severe” structural or hemodynamic valve deterioration was present in 1.1% of the TAVR group and 1.0% of the surgery group. Functional status was very similar between the two groups, with nearly identical assessments by the Kansas City Cardiomyopathy Questionnaire.
The authors concluded that low-risk patients with severe AS treated with TAVR or SAVR had similar five-year outcomes with respect to survival, stroke, and rehospitalization, and had similar hemodynamic performance of the valves.
COMMENTARY
In recent years, we have seen the inevitable application of TAVR to lower-risk and younger patients, with increasingly lower numbers of patients being referred for open surgical valve replacement. In this setting, the reporting of longer-term results, both in terms of clinical outcomes and valve hemodynamics, is essential for helping our patients make clinical decisions for this common cardiac problem. At first glance, the five-year data reported here are encouraging, showing no significant between-group differences in the major clinical outcomes.
However, it is important to take note of what patients were studied in these trials, and to pay attention to the differences that are noted at this intermediate stage. First, although this was a trial of patients at low estimated surgical risk and included a younger cohort compared with prior trials, these patients still had a mean age of 73 years, with a standard deviation of approximately six years. Since TAVR is applied increasingly to not only low-risk patients but also to younger patients, it will be extremely important to understand how these valves perform over time. And while the summary data look encouraging, with similar valve gradients and similarly low rates of severe structural deterioration, there are notable differences where five-year data may not be enough to inform decision-making for younger patients. The rates of mild-or-greater aortic regurgitation clearly are much higher after TAVR compared with SAVR, occurring in approximately one-quarter of TAVR patients. We ultimately will need significantly longer follow-up to learn whether this has a significant effect on downstream outcomes. Similarly, the modest but significant increase in TAVR patients requiring permanent pacing may have more relevance among patients who potentially have decades of remaining lifespan.
The greater incidence of clinically significant valve thrombosis in the TAVR group mirrors prior investigations, but the reasons for this difference remain unclear. The numbers (12 patients in the TAVR group vs. one in the SAVR group) were small, and this finding did not appear to affect valve durability at this intermediate time point.
In summary, this analysis of five-year outcomes comparing TAVR and SAVR supports the current practice of offering TAVR to older low-risk patients. Longer-term data will be required to inform clinical decision-making as TAVR is applied to younger patients.
