Efficacy and safety of apixaban compared with warfarin according to patient risk of stroke and of bleeding in atrial fibrillation: A secondary analysis of a randomised controlled trial

Background: The Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation ( ARISTOTLE ) trial showed that apixaban is better than warfarin at prevention of stroke or systemic embolism, causes less bleeding, and results in lower mortality. We assessed in this trial's participants how results differed according to patients' CHADS2, CHA2DS2VASc, and HAS-BLED scores, used to predict the risk of stroke and bleeding. Methods: ARISTOTLE was a double-blind, randomised trial that enrolled 18 201 patients with atrial fibrillation in 39 countries. Patients were randomly assigned apixaban 5 mg twice daily ( n=9120 ) or warfarin ( target international normalised ratio 2·0–3·0; n=9081 ). The primary endpoint was stroke or systemic embolism. The primary safety outcome was major bleeding. We calculated CHADS2, CHA2DS2VASc, and HAS-BLED scores of patients at randomisation. Efficacy analyses were by intention to treat, and safety analyses were of the population who received the study drug. ARISTOTLE is registered with ClinicalTrials.gov, number NCT00412984. Findings: Apixaban significantly reduced stroke or systemic embolism with no evidence of a differential effect by risk of stroke ( CHADS2 1, 2, or ≥3, p for interaction=0·4457; or CHA2DS2VASc 1, 2, or ≥3, p for interaction=0·1210 ) or bleeding ( HAS-BLED 0–1, 2, or ≥3, p for interaction=0·9422 ). Patients who received apixaban had lower rates of major bleeding than did those who received warfarin, with no difference across all score categories ( CHADS2, p for interaction=0·4018; CHA2DS2VASc, p for interaction=0·2059; HAS-BLED, p for interaction=0·7127 ). The relative risk reduction in intracranial bleeding tended to be greater in patients with HAS-BLED scores of 3 or higher ( hazard ratio [HR] 0·22, 95% CI 0·10–0·48 ) than in those with HAS-BLED scores of 0–1 ( HR 0·66, 0·39–1·12; p for interaction=0·0604 ). Interpretation: Because apixaban has benefits over warfarin that are consistent across patient risk of stroke and bleeding as assessed by the CHADS2, CHA2DS2VASc, and HAS-BLED scores, these scores might be less relevant when used to tailor apixaban treatment to individual patients than they are for warfarin. Further improvement in risk stratification for both stroke and bleeding is needed, particularly for patients with atrial fibrillation at low risk for these events. Funding: Bristol-Myers Squibb and Pfizer.