Multiterritory Brain Infarcts, Anticoagulation, and Recurrence After Cryptogenic Stroke: A Subgroup Analysis of the ARCADIA Trial

BACKGROUND: In patients with cryptogenic stroke, the characteristics of multiterritory brain infarcts, the recurrent stroke risk, and the response to anticoagulation remain unclear. METHODS: The ARCADIA trial (Atrial Cardiopathy and Antithrombotic Drugs in Prevention After Cryptogenic Stroke) screened patients with cryptogenic stroke for atrial cardiopathy at 185 centers in the United States and Canada from 2018 to 2022. Investigators reported baseline acute brain infarction in the left anterior, right anterior, and posterior circulation. Atrial cardiopathy was defined as P-wave terminal force in ECG lead V1s >5000 μV·ms, NT-proBNP (N-terminal pro-B-type natriuretic peptide) >250 pg/mL, or left atrial diameter index ≥3 cm/m 2 . Site echocardiography laboratories determined left atrial diameter index and a central echocardiography laboratory determined LVEF. We used ANCOVA to examine whether atrial cardiopathy biomarkers or LVEF were associated with the number of territories with infarction. Cox regression was used to examine whether the number of infarct territories was associated with recurrent stroke or modified the effect of apixaban compared with aspirin. RESULTS: Among 3464 patients with reported baseline magnetic resonance imaging data, 220 (6.4%) had no visible acute infarct and 2794 (80.7%) had acute infarction in 1, 374 (10.8%) in 2, and 76 (2.2%) in 3 territories. Atrial cardiopathy biomarkers and LVEF were not associated with the number of infarct territories. Among 937 of these 3464 patients who were randomized, we found higher risks of recurrent stroke associated with infarcts in 2 territories (hazard ratio, 2.4 [95% CI, 1.3–4.2]) or 3 territories (hazard ratio, 3.7 [95% CI, 1.5–9.3]) relative to single-territory infarction, whereas the absence of visible infarction was not associated with recurrence (hazard ratio, 1.8 [95% CI, 0.6–4.9]). The number of infarct territories did not modify the effect of apixaban versus aspirin in relation to recurrent stroke ( P for interaction, 0.71). CONCLUSIONS: Multiterritory brain infarction was not associated with atrial cardiopathy biomarkers or LVEF in the ARCADIA trial. Multiterritory infarction was associated with a significantly higher risk of recurrent stroke, but this heightened risk was not reduced by apixaban relative to aspirin.