Anticoagulation-Related Major Bleeding in Patients With Atrial Fibrillation

Looking Behind the CurtainArticle, see p 1792 A ntithrombotic therapy for cardiovascular disease revolves around balancing 2 key outcomes-reducing ischemic events without increasing bleeding.There has been a subtle but progressive shift away from a dominant focus on reducing ischemic events to a more balanced and explicit interest in avoiding bleeding.Bleeding is common and, in part, an expected result of the increased use of anticoagulants in patients at higher bleeding and ischemic risk.The outcomes from bleeding events are at least as bad as those following ischemic events with even minor bleeding associated with increased mortality.1 While the consequences of bleeding are often overt, there is growing recognition of its contribution to subsequent ischemic events including through secondary anemia, creation of a prothrombotic state, and requisite cessation of (and reluctance to restart) antithrombotic therapy.Together these factors suggest that currently available antithrombotic agents may be approaching a point of diminishing returns.Further antithrombotic potency is likely to result in limited incremental anti-ischemic benefit and a price that is too high to pay in bleeding.As the gap between ischemic benefit and bleeding cost narrows, it is vital that the tools used to assess both ischemic and bleeding outcomes are rigorously validated, well-understood by their users, fit for purpose, and able to effectively capture that which is important to patients, providers, investigators, and regulators.In this issue of Circulation, Bergmark et al 2 further inform the assessment of bleeding in anticoagulated patients with atrial fibrillation with an analysis from the ENGAGE AF-TIMI 48 trial (Effective Anticoagulation with Factor Xa Next Generation in Atrial Fibrillation-Thrombolysis in Myocardial Infarction 48) of edoxaban and warfarin.3 In this analysis, the authors compared bleeding rates according to bleeding severity using the TIMI (Thrombolysis in Myocardial Infarction), 4 GUSTO (Global Use of Strategies to Open Occluded Arteries), 5 ISTH (International Society on Thrombosis and Haemostasis), 6 and BARC (Bleeding Academic Research Consortium) 7 bleeding scales and then evaluated the implications of varying bleeding definitions on edoxaban's safety and net clinical benefit compared with warfarin.Many of their findings might be considered "self-evident" and expected based on the bleeding event definitions.There were more total bleeding events classified by GUSTO and BARC than by TIMI or ISTH bleeding scales because bleeding by TIMI or ISTH scales requires the presence of "clinically overt bleeding," which is not required for GUSTO or BARC.The greatest number of "most severe" bleeding events was classified using the ISTH scale (n=1289 [major bleeding]) and the least using the BARC scale (n=276 [fatal or intracranial bleeding]), with intermediate numbers reported using the GUSTO (n=347 [severe or life-threatening bleeding]) and TIMI (n=548 [major bleeding]) scales.Of most interest, however, was a statistically