BackgroundThe optimal timing of initiating oral anticoagulants after reperfusion therapy for ischemic stroke is unknown. Factors related to early initiation of rivaroxaban and differences in clinical outcomes of stroke patients with nonvalvular atrial fibrillation (NVAF) who underwent reperfusion therapy was investigated.MethodsFrom data of 1,333 NVAF patients with ischemic stroke or transient ischemic attack (TIA) in a prospective multicenter study, patients who started rivaroxaban after intravenous thrombolysis and/or mechanical thrombectomy were included. The clinical outcomes included the composite of ischemic events (recurrent ischemic stroke, TIA, or systemic embolism) and major bleeding at 3 months.ResultsAmong the 424 patients, the median time from index stroke to starting rivaroxaban was 3.2 days. On multivariable logistic regression analysis, infarct size (odds ratio [OR], 0.99; 95%CI, 0.99–1.00) was inversely and successful reperfusion (OR, 2.13; 95%CI, 1.24–3.72) was positively associated with initiation of rivaroxaban within 72 hours. 205 patients were assigned to the early group (< 72 hours) and 219 patients (≥ 72 hours) to the late group. Multivariable Cox regression models showed comparable hazard ratios between the two groups at 3 months for ischemic events (hazard ratio [HR], 0.18; 95%CI, 0.03–1.32) and major bleeding (HR, 1.80; 95%CI, 0.24–13.54).ConclusionsInfarct size and results of reperfusion therapy were associated with the timing of starting rivaroxaban. There were no significant differences in the rates of ischemic events and major bleeding between patients after reperfusion therapy who started rivaroxaban < 72 hours and ≥ 72 hours after the index stroke.Clinical trial registrationUnique identifier: NCT02129920; URL: https://www.clinicaltrials.gov.