Anticoagulation Therapy in Endovascular Thrombectomy Patients With Large‐Vessel Occlusion Caused by Cardioembolism

Background In patients with ischemic stroke, cardioembolism, predominantly caused by atrial fibrillation (AF), is a leading preventable cause of large‐vessel occlusion (LVO). Despite proven efficacy, inadequate oral anticoagulant (OAC) therapy continues to be a major problem in patients with AF, mechanical heart valves, and other potential sources of emboli. We aimed to determine rates of cardioembolic LVO, the adequacy of OAC, and the association of OACs with clinical outcomes in patients with LVO treated with endovascular thrombectomy (EVT). Methods Consecutive patients with cardioembolism treated with EVT were included and had pre‐stroke indication for OACs and the adequacy of anticoagulation determined. The primary outcome was 3‐month functional independence (modified Rankin Scale score 0–2). Secondary outcomes included early neurological recovery (reduction in National Institutes of Health Stroke Scale score ≥8 points, or score of 0–1 at 24 hours), symptomatic intracranial hemorrhage, and 3‐month mortality and modified Rankin Scale score. Results There were 784 patients treated with EVT, of whom 416 (53.1%; 231 men; mean±SD age, 67.1±15.9) had cardioembolic LVO. Of those with cardioembolism, 221 (53.1%) had prevalent AF, 99 (23.8%) incident AF, 48 (11.5%) mechanical heart valves, 10 (2.4%) left ventricular thrombus, and 38 (9.1%) other cardioembolic causes. A total of 242 patients were identified as having a pre‐stroke indication for OACs, and 67 (27.7%) of these patients were considered adequately anticoagulated at time of LVO. In all 242 patients with cardioembolism with a pre‐stroke indication for OACs, adequate anticoagulation was not associated with EVT outcomes. However, adequate anticoagulation in patients with prevalent AF was associated with lower rates of internal carotid artery occlusion (6.8% versus 18.9%; P =0.03) and higher rates of functional independence (odds ratio=1.94; 95% CI, 1.02–3.78]; P =0.04). Conclusions Over half of patients treated with EVT were assessed as having a cardioembolic cause. Just over a quarter of patients with an indication for OACs at the time of LVO were adequately anticoagulated. These figures suggest that a substantial number of EVT procedures may have been avoided if more patients had been adequately anticoagulated.