ABSTRACT NUMBER: ESOC2026A2268 PRIOR ANTICOAGULATION FAILURE AND LONG-TERM RECURRENCE AFTER ATRIAL FIBRILLATION–RELATED STROKE

Abstract Background and aims Patients with known atrial fibrillation (AF) remain at risk of late recurrent cerebrovascular events after a first ischemic stroke (IS). A clinically relevant subset experiences the index stroke while already receiving oral anticoagulation (“breakthrough” events). We investigated whether prior anticoagulation at the time of the index stroke is independently associated with long-term cerebrovascular recurrence (LTCR) in patients with known AF. Methods Retrospective analysis of a cohort of 817 patients (2005–2024) with IS and AF. The outcome was LTCR. Cumulative incidence functions (CIFs) were estimated overall and by prior anticoagulation and compared using Gray’s test. Multivariable Fine–Gray subdistribution hazard models were fitted for recurrence and for death. Models included stroke severity and clinical covariates (CHA₂DS₂-VASc score, dyslipidemia, valvulopathy, smoking, alcohol overuse, and post-stroke oral anticoagulation type: none, vitamin K antagonist, or direct oral anticoagulant). Results We observed 171 LTCR events and 354 deaths. Mean follow-up was 51.3 months. Overall CIF for LTCR was 4.6%, 11.8%, and 18.3% at 12, 36, and 60 months, respectively. Prior anticoagulation was present in 506 patients (61.9%) and was associated with higher LTCR(21.0% vs 14.3%; Gray p=0.00010). In adjusted Fine–Gray models, prior anticoagulation remained associated (subdistribution hazard ratio [sHR] 3.69; 95% confidence interval [CI] 2.55–5.33; p<0.001), while vitamin-K-antagonist-therapy (sHR 0.33; 95% CI 0.23–0.48; p<0.001) and direct-oral-anticoagulant-therapy (sHR 0.19; 95% CI 0.13–0.28; p<0.001) were associated with lower LTCR. Conclusions In patients with IS and known AF, prior anticoagulation identifies a higher-risk subgroup with increased LTCR. Conflict of interest All authors: nothing to disclose