Anticoagulation in Atrial Fibrillation With Previous Intracranial Bleeding

BACKGROUND AND OBJECTIVES: Patients with atrial fibrillation (AF) and previous intracranial bleeding are at high risk of ischemic and bleeding events, and the benefits and risks of oral anticoagulation (OAC) remain uncertain. We investigated competing risks of recurrent intracranial bleeding, ischemic stroke (IS), and death from other causes according to anticoagulation strategy. METHODS: This nationwide Finnish cohort study using linked administrative registries included patients with first-ever AF diagnosed in 2010-2018 and intracranial bleeding at least 120 days earlier. Follow-up began 90 days after AF diagnosis. Patients were categorized as receiving direct oral anticoagulants (DOACs), warfarin, or no OAC based on drug purchases. Warfarin users were further stratified by time in therapeutic range (TTR). Inverse probability of treatment weighting balanced baseline characteristics. Competing risks were analyzed using adjusted cumulative incidence functions and cause-specific Cox regression. RESULTS: Among 1,502 patients (median age 78 years [IQR 71-85], 44% female, 41% with only traumatic intracranial bleeding), 15% received DOACs, 29% warfarin, and 56% no OAC. At 1 year, adjusted absolute risks of IS were 0.7%, 2.1%, and 4.1%; intracranial bleeding, 1.2%, 2.7%, and 1.9%; and death from other causes, 4.7%, 8.6%, and 14.7%, respectively. Compared with no OAC, DOAC use was associated with lower risks of IS (HR 0.17, 95% CI 0.04-0.71) and death (HR 0.20, 95% CI 0.09-0.46) without increased intracranial bleeding risk (HR 0.70, 95% CI 0.19-2.50). Warfarin use was associated with lower risks of IS (HR 0.49, 95% CI 0.26-0.93) and death (HR 0.60, 95% CI 0.44-0.80), but increased intracranial bleeding risk (HR 1.85, 95% CI 1.01-3.41). Among warfarin users, TTR ≥70% was associated with lower risks of IS (HR 0.24, 95% CI 0.08-0.79) and death (HR 0.19, 95% CI 0.10-0.36); associations were attenuated with poorer or missing TTR. DISCUSSION: In patients with AF and previous intracranial bleeding, OAC was associated with lower risks of IS and death. DOAC use was not associated with increased intracranial bleeding risk, whereas warfarin use was. Limited information on previous intracranial bleeding and potential residual confounding warrant cautious interpretation. These findings support individualized anticoagulation decisions in this high-risk population.