Eur J Cardiothorac Surg. 2025 Dec 1;67(12):ezaf348. doi: 10.1093/ejcts/ezaf348.
ABSTRACT
OBJECTIVES: To critically appraise the design of the ongoing multicentre, randomized START YOUNG trial comparing transcatheter aortic valve implantation (TAVI) with bioprosthetic surgical aortic valve replacement (SAVR) in patients aged 65-75 years with symptomatic severe aortic stenosis (AS).
METHODS: Publicly available trial information (ClinicalTrials.gov: NCT06861361) was reviewed, and the design was benchmarked against the minimum expected methodological standards for generating practice-changing evidence in patients with very long life expectancy (LE) following SAVR. The appraisal domains included objectives and study endpoints; selected non-inferiority (NI) margins; adequacy of follow-up for late events; statistical analysis plans and handling of crossovers/missing data; consistency of surgical comparators; and prespecified sub-groups.
RESULTS: Several major concerns were identified: (1) co-primary safety/efficacy composites mix hard adverse events (death, stroke) with softer outcomes, with rehospitalization likely to dominate; (2) limited duration primary follow-up (30 days/1 year), with only optional survaillance beyond 5 years, inadequate for durability and late hazards; (3) wide NI margins (7%-8%) that could accept clinically meaningful inferiority; (4) an unclear statistical analysis plan; (5) a device-specific TAVI arm versus heterogeneous SAVR comparator; and (6) up to 25% bicuspid anatomy without stratified randomization or powered analyses. Collectively, these choices markedly increase the risks of type I error, and misinterpretation in a population with likely crossing hazard curves and a long median LE comparable to that of the age-matched general population after SAVR.
CONCLUSIONS: In order to generate guideline-informing evidence for extending TAVI to younger patients with severe AS, the trial should prioritise hard adverse endpoints, tighten NI margins, mandate extended follow-up, standardise the surgical comparator, and adequately power the bicuspid subgroup. Without these revisions, continuing the study under the current protocol risks yielding biased, non-guideline informative results.
PMID:41351226 | DOI:10.1093/ejcts/ezaf348