Article Citation: Lansberg MG, Wintermark M, Chen H, Howard G, Cassarly C, Pauls Q, Kemp S, Harris TL, Krishnaiah B, Stanton RJ, Lyerly MJ, Miller BR, Smith EE, Tirschwell DL, Sheth KN, Kamel H, Longstreth WT Jr, Elkind MSV, Broderick JP, Lazar RM. Apixaban to Prevent Covert Infarcts After Cryptogenic Stroke in Patients With Atrial Cardiopathy: A Secondary Analysis of the ARCADIA Randomized Clinical Trial. JAMA Neurol. 2025 Mar 1;82(3):220-227.

Background: The recently published ARCADIA trial, which enrolled patients with ESUS and signs of atrial cardiopathy, failed to show that treatment with anticoagulation prevented recurrent ischemic stroke. However, covert infarcts, which are clinically asymptomatic but seen on MRI, occur commonly in the aging population and may lead to worsened cognition and future stroke. Whether anticoagulation prevented covert infarcts seen on MRI was the basis of the ARCADIA-MRI substudy, reviewed here.

Methods: Enrollment in ARCADIA-MRI occurred concurrently with enrollment in the main ARCADIA study (2019-2022) as a secondary analysis. ARCADIA was a randomized, multicenter, double-blind trial that enrolled patients with cryptogenic stroke and signs of atrial cardiopathy to receive apixaban or aspirin. Patients were enrolled in ARCADIA-MRI if they were able to undergo a baseline and follow-up MRI. Although 799 patients were screened, only 174 were included (79 in the apixaban arm and 95 in the aspirin arm). The primary outcome was the percentage of patients with the presence of 1 or more new non-lacunar infarcts on follow-up MRI that were determined to be covert (i.e., not causing clinical symptoms).

Results: The median time to follow-up MRI was 27 months in each arm. The primary outcome occurred in 5.1% of patients in the apixaban arm and 17.9% of patients in the aspirin arm (weighted RR 0.29, P=0.02). In contrast, the risk of covert lacunar infarcts was not decreased in patients treated with apixaban.

Commentary: Despite the negative results of ARCADIA, ARCADIA-MRI suggests that anticoagulation may indeed reduce the risk of covert non-lacunar infarcts in patients who present with ESUS and signs of atrial cardiopathy. Generalizability may be limited, however, by a relatively small number of enrolled patients compared to the main ARCADIA study. The main barriers appeared to be lower site participation in the substudy and difficulty in obtaining baseline and follow-up MRI scans, and the lower sample size is reflected by wider confidence intervals. Additionally, patients in this study were healthier than the overall ARCADIA cohort and had lower rates of self-discontinuation of the study drug. Supporting literature elsewhere is limited and thus far mixed, although some evidence focusing on embolic strokes has pointed towards a treatment benefit in patients treated empirically with anticoagulation.

Overall, while the results of ARCADIA-MRI are intriguing, they do not yet justify empiric anticoagulation in patients with ESUS (especially considering the results of the main ARCADIA study). The study does, however, keep the door open to the concept of clinical benefit of anticoagulation for ESUS, and serves to further hone ideal radiographic and clinical endpoints in future studies.