ABSTRACT NUMBER: ESOC2026A2431 PRESS START: A REAL-WORLD LOOK AT THE TIMING OF ANTICOAGULATION AFTER STROKE

Abstract Background and aims Current studies support infarct size–guided anticoagulation after cardioembolic stroke, although no widely accepted guidelines define an optimal timing. In 2023, the ELAN trial classified ischemic strokes according to infarct size and defined “early” and “late” anticoagulation strategies (minor: <2 and 3-4 days; moderate: <2 and 6-7 days; major: 6-7 and 12-14 days), showing no significant differences in complications. We aimed to evaluate these strategies in real-world experience. Methods Restrospective study of atrial fibrillation-related cardioembolic strokes admitted to our tertiary-care centre between 2023-2024. Patients meeting ELAN trial criteria for “early” or “late” strategies were included. Baseline and discharge characteristics and complications during follow-up were analysed. Results Among 168 patients with cardioembolic stroke, 96 met ELAN trial criteria for "early" (58.3%) or "late" (41.7%) anticoagulation strategies. Baseline demographic and clinical characteristics, NIHSS and modified Rankin Scale (mRS), were similar between groups. At discharge, “early” group showed significantly better outcomes, with lower median NIHSS (2 vs 5, p=0.009) and mRS (2 vs 3, p=0.035). After anticoagulation, no recurrent ischemic strokes were observed, with no significant differences in intracranial haemorrhage (p=0.305) or all-cause mortality at 1 (p=0.569) or 6 months (p=0.691). Conclusions In our real-world experience, early anticoagulation was significantly associated with better functional outcomes, differing from the ELAN trial. By contrast, no statistically significant differences were observed in haemorrhagic complications or mortality, consistent with the ELAN trial. These findings support infarct size–guided anticoagulation after cardioembolic stroke. However, differences in sample size and baseline characteristics highlight the need for larger studies to confirm these results. Conflict of interest All authors have nothing to disclose