ACASA-TAVI
2026
AntiCoagulation versus AcetylSalicylic Acid after Transcatheter Aortic Valve Implantation
In 360 patients after TAVI without an indication for anticoagulation, 12 months of a factor Xa inhibitor reduced leaflet thrombosis on CT and a safety composite at 12 months compared with aspirin.
Result
Efficacy: co-primary: (1) incidence of hypo-attenuated leaflet thickening (HALT) on 4-dimensional cardiac CT at 12 months; (2) safety composite of VARC-3 bleeding, thromboembolic events (myocardial infarction, stroke) and all-cause death at 12 monthsrisk ratio 0.55 (95% CI 0.37-0.82) · P = 0.004
Below 1 favours 12 months of a factor Xa inhibitor, above 1 favours aspirin.
- Absolute difference
- 15.4 points
- NNT
- 7
Absolute difference and NNT are derived from the event rates.
Also reported: The safety composite of VARC-3 bleeding, thromboembolism and death ran 7.5% vs 10.6% (risk difference −3.3 percentage points, 95% CI −9.5 to 2.8; P < 0.001 for non-inferiority); bleeding was low overall and numerically higher on aspirin, with life-threatening or fatal bleeding only in that arm. Deaths 2 vs 10.
- Subtopic
- TAVR antithrombotics
- Design
- investigator-sponsored, pragmatic, multicentre (Oslo and Bergen, Norway), prospective randomised open-label blinded-endpoint (PROBE) all-comers trial; 1:1 anti-factor Xa DOAC monotherapy vs aspirin; co-primary endpoints at 12 months; long-term follow-up to 5 and 10 years
- Relevance
- Important: one of several pillars.
- Source
- Dodgson S, Herstad J, Kløve SF, et al. JAMA. 2026; published online 30 August.doi 10.1001/jama.2026.17036
Discussion & critique
A factor Xa inhibitor halves subclinical leaflet thrombosis after TAVI without paying for it in bleeding, in patients aged 65 to 80 with no other reason to anticoagulate. The endpoint is a CT finding, not an event, and GALILEO already showed that anticoagulating this population can cost lives, so the two deaths versus ten here are reassuring but far too few to answer that. What the trial establishes is the imaging effect and the absence of an obvious safety price at 360 patients; whether leaflet thickening is worth preventing remains the open question.