Abstract
BACKGROUND/OBJECTIVE: In persons whose sole indication for anticoagulation is atrial fibrillation (AF), early therapeutic anticoagulation after acute ischemic stroke (AIS) may decrease ischemic risk without increasing hemorrhagic risk. However, literature to guide anticoagulation timing in patients with a non-AF indication remains extremely limited. METHODS: This retrospective cohort study compared outcomes of early (within ≤4 days of AIS) versus late anticoagulation (5-14 days) for persons with AIS and non-AF indications for anticoagulation. The primary outcome was a composite of intracranial hemorrhage or major extracranial bleeding while on therapeutic anticoagulation, within 30 days of the index event. The main secondary outcome was a composite of major bleeding events while on therapeutic anticoagulation, recurrent AIS, systemic embolism, and all-cause mortality, within 30 days of the index event. RESULTS: Eighty-one patients were included for analysis, with 65 patients in the early cohort and 16 patients in the late cohort; median time to anticoagulation was 1 day and 7 days, respectively. The most common indication for anticoagulation was deep vein thrombosis. The primary composite outcome occurred in 3 patients (4.6%) in the early cohort and 2 patients (12.5%) in the late cohort (p = 0.255). The secondary composite outcome occurred in 10 patients (15.4%) in the early cohort and 7 patients (43.8%) in the late cohort (p = 0.034). There were no statistical differences in any individual components of the composite outcomes, although recurrent AIS and mortality had numerically higher incidence in the late cohort. CONCLUSIONS: In this retrospective study, early anticoagulation was not associated with increased major bleeding risk, but late anticoagulation was associated with an increased composite risk of major bleeding, thrombotic events, and all-cause mortality, driven by increases in recurrent AIS and mortality. Further studies are warranted to expound on the optimal timing of anticoagulation in this patient population.