Abstract A033: Potential Benefit of Apixaban Among Patients Without High-Risk Hypertension In Cryptogenic Stroke: A Secondary Analysis of the ARCADIA Trial

Background: Multiple trials have found no difference in stroke prevention comparing anticoagulation to antiplatelet therapy in patients with stroke from a potential occult cardioembolic source. One explanation is the failure to exclude patients with index or subsequent strokes due to hypertensive arteriopathy. We sought to determine whether systemic evidence of high-risk hypertension (HRH) modifies the treatment effect of anticoagulation versus antiplatelet therapy after cryptogenic stroke. Methods: This secondary analysis of the Apixaban to Prevent Recurrence After Cryptogenic Stroke in Patients With Atrial Cardiopathy (ARCADIA) trial assessed whether HRH modified the effect of apixaban versus aspirin in patients with cryptogenic stroke and atrial cardiopathy. HRH was defined as systolic blood pressure ≥160 mmHg at enrollment, left ventricular hypertrophy on echocardiography using left ventricular mass index, or both. The primary outcome was recurrent ischemic stroke or systemic embolism. Cox proportional hazards models evaluated treatment effects and interaction with HRH adjusted for 1) CHA 2 DS 2 VASc score and race; 2) individual variables associated with the primary outcome. Unadjusted stratified analyses and cumulative event rate curves were performed. Results: In 945 randomized patients followed over a median of 1.8 years, 351 (37%) had evidence of HRH, and the primary outcome occurred in 67 patients. Among 594 patients without HRH, a lower incidence rate of the primary outcome was observed in patients randomized to apixaban compared to aspirin (21.5 vs 55.1 per 1000 person-years), whereas patients with HRH had higher incidence with apixaban (55.8 vs 31.8 per 1000 person-years). In the crude and fully adjusted models, a significant interaction effect between HRH and antithrombotic treatment arm was observed (p<0.05) (Table 1). Stratified analysis in patients without HRH demonstrated a lower risk for recurrent ischemic stroke or systemic embolism with apixaban (HR 0.40, CI 0.21-0.79, p=0.008). In patients with HRH, there was no evidence for a beneficial effect of apixaban (HR 1.73, CI 0.80-3.72, p=0.164) (Table 2, Figure 1). Conclusion: Evidence of HRH modified the effect of anticoagulation treatment in patients with cryptogenic stroke and atrial cardiopathy. An underappreciated inclusion of strokes due to hypertensive arteriopathy may account for the absent benefit of anticoagulation in prior trials of embolic stroke of unknown source.