This retrospective cohort study used the Epic Cosmos database to identify adults undergoing index transcatheter aortic valve replacement (TAVR) between February 2010 and May 2026. The cohort comprised 300,927 patients. Annual procedural volume increased more than tenfold between 2016 and 2025, reflecting rapid uptake of TAVR in clinical practice. Over the study interval, short-term procedural outcomes improved substantially: thirty-day major adverse cardiovascular events (MACE) declined from 31.8% in the pre-2017 period to 18.6% after 2022 (P<0.001), and thirty-day mortality declined from 3.0% to 1.4% (P<0.001).
Investigators performed a retrospective analysis using Epic Cosmos. Primary outcomes included aortic valve reintervention and thirty-day MACE. Reintervention incidence was estimated with competing-risk methods treating death as the competing event. The study used propensity-score matching to compare outcomes after redo-TAVR (TAVR-in-TAVR) versus surgical explantation and to compare TAVR with surgical bioprosthetic aortic valve replacement (SAVR). A prespecified one-year landmark analysis evaluated clinically significant valve failure requiring repeat intervention. The authors reported that the institutional ethics committee (St Luke’s University Health Network) waived ethical approval for the work (IRB-FY2026-117). Data availability was stated via the Epic Cosmos links reported in the source.
During follow-up, 3,315 patients underwent redo-TAVR and 347 underwent surgical explantation. The cumulative incidence of repeat aortic valve intervention after index TAVR was 1.1% at 3 years, 1.2% at 5 years, 1.5% at 7 years, and 2.7% at 10 years. Incidence of reintervention was significantly lower in more contemporary procedural eras (Gray test, P<0.001), indicating improvements in device durability, patient selection, procedural technique, or a combination of these factors. The study estimated reintervention rates using competing-risk methodology, explicitly accounting for death as a competing outcome.
In unadjusted counts, 3,315 patients received redo-TAVR while 347 underwent surgical explantation during the study period. Propensity-score matching was used to compare procedural and short-term outcomes between these strategies. Compared with surgical explantation, redo-TAVR was associated with lower thirty-day rates of mortality, stroke, acute kidney injury, and major bleeding, indicating a lower perioperative risk profile for a transcatheter approach in the real-world cohort.
However, when the analysis focused on longer-term survival among propensity-matched hospital survivors, surgical explantation was associated with superior long-term survival. The reported hazard ratio for long-term mortality with surgical explantation compared with redo-TAVR was 0.64 (95% CI: 0.44–0.93; P=0.018). This finding indicates a trade-off between lower perioperative morbidity and mortality with redo-TAVR and a possible long-term survival advantage for selected patients undergoing surgical explantation.
A prespecified one-year landmark analysis compared timing and incidence of clinically significant valve failure requiring repeat intervention after TAVR versus SAVR. The analysis found that clinically significant valve failure requiring repeat intervention tended to occur earlier after TAVR than after SAVR. Despite this earlier timing, the overall cumulative incidence of repeat intervention was lower following TAVR than after SAVR within the observed follow-up. The source reports these comparative timing phenomena but does not provide additional subgroup breakdowns or device-specific durability data in the abstract.
In this large national real-world cohort, contemporary TAVR practice was associated with steadily improving procedural outcomes and a low cumulative incidence of repeat aortic valve intervention over up to ten years. Redo-TAVR offers lower short-term perioperative risk compared with surgical explantation, but surgical explantation was associated with superior long-term survival among propensity-matched hospital survivors. The finding that clinically significant valve failure requiring repeat intervention occurred earlier after TAVR than after SAVR highlights the need to consider lifetime valve management strategies, particularly as TAVR expands into younger patients with longer life expectancy.
Limitations and additional details reported in the source: the report is a retrospective analysis of Epic Cosmos data; specific device types, patient-level covariates, and subgroup details beyond the high-level outcomes summarized here were not provided in the abstract. The authors declared no competing interests and noted that all relevant ethical guidelines were followed; ethical approval was waived by the reporting IRB. The dataset and methodological resources were indicated as available via the Epic Cosmos links provided in the source record.