Hypertension Subtype Predicts Who Benefits From Blood Thinners After Cryptogenic Stroke

Cryptogenic stroke — ischemic stroke with no identifiable cause after thorough workup — accounts for roughly 25-30% of all ischemic strokes and represents a major clinical challenge for secondary prevention. Because many cryptogenic strokes are thought to be embolic in origin, anticoagulation has long been hypothesized to outperform antiplatelet therapy. Yet multiple large randomized trials, including NAVIGATE ESUS, RESPECT ESUS, and ARCADIA itself, have failed to demonstrate superiority of anticoagulation. This exploratory analysis proposes a compelling explanation: these trials may have inadvertently enrolled patients whose strokes were driven not by embolism but by hypertensive arteriopathy — a noncardioembolic mechanism that anticoagulation would not be expected to prevent.

The ARCADIA trial enrolled 1,015 patients with recent cryptogenic stroke and evidence of atrial cardiopathy across 185 North American sites from 2018 to 2023, randomizing them to apixaban (5 mg or 2.5 mg twice daily) versus aspirin (81 mg daily). This post-hoc exploratory analysis included 945 participants with complete blood pressure and echocardiography data (mean age 68.0 years, 54.3% female). Hypertension with high-risk features (HHF) was defined as systolic blood pressure ≥160 mmHg at enrollment, left ventricular hypertrophy on echocardiography (sex-specific LV mass index thresholds), or both — mirroring the 2023 European Society of Hypertension risk stratification framework. Of the 945 participants, 351 (37.1%) met criteria for HHF.

Over a median follow-up of 1.6 years, 67 patients experienced the primary composite outcome of recurrent ischemic stroke or systemic embolism. A statistically significant interaction was detected between HHF status and antithrombotic treatment assignment. Among the 594 patients without HHF, apixaban was associated with a markedly lower risk of recurrent events compared to aspirin (HR 0.43; 95% CI 0.22-0.85; annualized rate difference: -3.4%). In stark contrast, among the 351 patients with HHF, apixaban showed no benefit and trended toward harm (HR 1.68; 95% CI 0.78-3.62; annualized rate difference: +2.4%). Secondary outcomes including recurrent ischemic stroke alone and any recurrent stroke showed consistent directional patterns.

The biological rationale is straightforward: patients with severe hypertension or hypertension-mediated organ damage (LV hypertrophy, impaired kidney function) are more likely to have strokes driven by small vessel disease and hypertensive arteriopathy rather than cardioembolism. Anticoagulation does not address this mechanism and may even increase hemorrhagic risk in the setting of damaged cerebral vasculature. The authors tested secondary HHF definitions incorporating eGFR <60 mL/min/1.73m² and found consistent results, strengthening the robustness of the finding. Fully adjusted models controlling for CHA₂DS₂-VASc score and Black race did not materially change the interaction estimates.

This analysis carries important implications for trial design and clinical practice. If roughly 37% of cryptogenic stroke patients have HHF and derive no benefit — or possible harm — from anticoagulation, their inclusion in trials would substantially dilute any treatment signal. Future trials of anticoagulation in embolic stroke of undetermined source should consider prospectively stratifying by hypertension risk features. Clinically, a simple assessment of blood pressure and echocardiographic LV mass at enrollment could help neurologists and cardiologists personalize antithrombotic selection. The authors appropriately caution that this is an exploratory, post-hoc analysis not powered for subgroup testing, and prospective validation is essential before changing practice.