ABSTRACT NUMBER: ESOC2026A370 OPTIMAL TIMING OF ANTICOAGULATION AFTER ISCHAEMIC STROKE AND ATRIAL FIBRILLATION: A BAYESIAN TIMING-RISK ANALYSIS OF THE “CATALYST” COLLABORATION

Abstract Background and aims In people with acute ischaemic stroke (AIS) and atrial fibrillation (AF), initiating direct oral anticoagulants (DOACs) within 4 days compared to a later start reduces the risk of recurrent ischaemic and haemorrhagic events. However, the optimal timing for DOAC initiation within this early window remains unclear. Methods We estimated the association between DOAC initiation timing and the risk of a composite outcome (recurrent ischaemic stroke, symptomatic intracranial hemorrhage, unclassified stroke) within 30 days in an individual patient-data meta-analysis (6619 participants). We implemented four synthesis models: two for the targeted randomized timing (“as randomized”) and two for the observed timing (“as observed”). We used a linear and a non-linear function to estimate the association between timing and outcome risk. Model parameters were synthesized across studies using random-effects meta-analysis fitted within a Bayesian framework. Results Delaying DOAC initiation was linked to a higher 30-day risk of the composite outcome. The time-risk analyses showed that each additional day of delay increased the probability of the composite outcome. The non-linear models suggested an almost linear relationship, indicating that even starting DOACs on day 1 did not appear to increase harm and might offer the greatest benefit. Conclusions In people with AF related AIS there is an approximately linear association between the timing of DOAC initiation and the risk of recurrent events. Our findings do not support delaying DOAC initiation even within the first four days after an ischaemic stroke, though data were limited for people with very large infarcts or severe haemorrhagic transformation. Conflict of interest Konstantina Chalkou: Nothing to disclose, Georgia Salanti: Nothing to disclose, Hakim-Moulay Dehbi: Nothing to disclose, Jesse Dawson: reports institutional research grants from the Stroke Association for the ELAN and OASIS trials, from National Institute for Health and Care Research for the CLASP study, and a Medicines Development Fellowship from the UK Medical Research Council, Jonas Oldgren: reports institutional research grants from Amgen, AstraZeneca, Bayer, and Roche Diagnostics; and institutional honoraria from Pfizer, Truman J Milling: reports consulting fees from AstraZeneca, CLS Behring, and Octapharma, Steven J Warach: reports institutional research grants from Lone Star Stroke Consortium and the State of Texas, Signild Åsberg: Nothing to disclose, David Werring: reports grant funding from the British Heart Foundation, Rosetrees Trust, National Brain Appeal, Stroke Association, and a National Institute for Health and Care Research Senior Investigator Award, Urs Fischer: reports research support from the Swiss National Science Foundation and the Swiss Heart Foundation.