Open Access

Raffaele De Caterina, Ulrika Andersson, John H. Alexander, Sana M. Al-Khatib, M. Cecilia Bahit, Shinya Goto, Michael Hanna, Claes Held, Stefan H. Hohnloser, Elaine M. Hylek, Fernando Lanas, Renato Delascio Lopes, José López-Sendón, Giulia Renda, John Horowitz, Christopher B. Granger, Lars Wallentin

• Aims: History of bleeding strongly influences decisions for anticoagulation in atrial fibrillation (AF). We analyzed outcomes in relation to history of bleeding and randomization in ARISTOTLE trial patients. Methods and results: The on-treatment safety population included 18,140 patients receiving at least 1 dose of study drug (apixaban) or warfarin. Centrally adjudicated outcomes in relation to bleeding history were analyzed using a Cox proportional hazards model adjusted for randomized treatment and established risk factors. Efficacy end points were analyzed on the randomized (intention to treat) population. A bleeding history was reported at baseline in 3,033 patients (16.7%), who more often were male, with a history of prior stroke/transient ischemic attack/systemic embolism and diabetes; higher CHADS2 scores, age, and body weight; and lower creatinine clearance and mean systolic blood pressure. Major (but not intracranial) bleeding occurred more frequently in patients with versus without a history of bleeding (adjusted hazard ratio 1.35, 95% CI 1.14-1.61). There were no significant interactions between bleeding history and treatment for stroke/systemic embolism, hemorrhagic stroke, death, or major bleeding, with fewer outcomes with apixaban versus warfarin for all of these outcomes independent of the presence/absence of a bleeding history. Conclusion: In patients with AF in a randomized clinical trial of oral anticoagulants, a history of bleeding is associated with several risk factors for stroke and portends a higher risk of major—but not intracranial—bleeding, during anticoagulation. However, the beneficial effects of apixaban over warfarin for stroke, hemorrhagic stroke, death, or major bleeding remains consistent regardless of history of bleeding.