Selecting the optimal stroke prevention therapy in atrial fibrillation.

Stroke is a feared complication of atrial fibrillation and leads to substantial morbidity and mortality (1). Atrial fibrillation not only increases the risk for ischemic stroke by approximately 5-fold, but atrial fibrillation–related strokes result in more deaths and disability than strokes of other causes (2). Anticoagulant therapy with vitamin K antagonists, such as warfarin, effectively reduces the risk for atrial fibrillation–related ischemic stroke but is associated with potentially serious bleeding complications, as well as challenges in maintaining a therapeutic level (3). Antiplatelet agents, although less effective than oral anticoagulants, have better safety profiles and are easier to administer (3). Selecting the most appropriate antithrombotic therapy for a patient is one of the most important management decisions in atrial fibrillation, and the net clinical benefit associated with a given therapeutic choice should guide this decision. Net clinical benefit can be conceptualized as the balance between the benefits of a given therapy and its associated risks. For atrial fibrillation, the benefit is largely the reduction in thromboembolic events; the risks include increased bleeding events—in particular, intracranial hemorrhage (4). However, because various outcome events are associated with different rates of morbidity and mortality, a simple summation of outcome rates does not capture the true consequences of a given therapeutic choice (5). A more sophisticated approach is to establish a standard reference for the severity of each outcome to allow a more direct comparison. As the baseline risk for ischemic stroke increases, so does the net clinical benefit observed with warfarin (4). However, warfarin has many well-known disadvantages, including a narrow therapeutic window, numerous drug– drug interactions, and the need for frequent monitoring and dosage adjustment (6). The challenges of safely maintaining a patient in a therapeutic range while receiving warfarin have spurred investigation into alternative therapies that are easier to administer, including antiplatelet agents and newer, fixed-dose oral anticoagulants. The 2 ACTIVE (Atrial Fibrillation Clopidogrel Trial With Irbesartan for Prevention of Vascular Events) studies investigated the potential role of dual antiplatelet therapy in atrial fibrillation (7, 8). ACTIVE-W compared warfarin monotherapy with combination aspirin and clopidogrel therapy, whereas ACTIVE-A compared combination aspirin and clopidogrel therapy with aspirin monotherapy in patients who were not candidates for warfarin. These trials demonstrated that combination aspirin and clopidogrel therapy was more effective than aspirin alone but was inferior to warfarin anticoagulation. In this issue, Connolly and colleagues (9) extend the ACTIVE studies in an important direction by attempting to account for the consequences of various clinical outcomes. By weighting the effect of each outcome relative to the death and disability resulting from ischemic stroke, the investigators provide a standardized metric to calculate net clinical benefit. Their primary findings show that dualantiplatelet therapy might have a net clinical benefit over aspirin monotherapy, but this benefit is small at best and does not reach statistical significance. This analysis provides empirical evidence about the relative consequences of various outcomes experienced by patients with atrial fibrillation. It confirms that intracerebral hemorrhage is considerably more serious than ischemic stroke, which is in turn more severe than extracranial hemorrhage (5). Intracerebral hemorrhage was particularly fatal in this study; major extracranial hemorrhage was also associated with surprisingly poor outcomes. Subgroup analyses indicated no subcategory of patients that derived greater net benefit from dual-antiplatelet therapy, although the point estimate trended toward greater benefit in patients with previous stroke and a higher baseline risk for stroke. The role of dual-antiplatelet therapy in atrial fibrillation has been overshadowed by the introduction of newer oral anticoagulants, including direct thrombin inhibitors and factor Xa inhibitors, that not only have the convenient fixed-dose administration of antiplatelet agents but also seem to be at least as effective and safe as warfarin (10–12). Dabigatran, which is available in the United States for use in atrial fibrillation, has been associated with lower rates of ischemic stroke and intracranial hemorrhage than warfarin (10). Rivaroxaban has been shown to be noninferior to warfarin and has resulted in fewer intracranial hemorrhages (11). Apixaban has resulted in both lower stroke rates and lower all-cause mortality than warfarin (12). As newer anticoagulants become available, the role of dual-antiplatelet therapy in atrial fibrillation may become even more limited. Apixaban has been shown to be significantly more effective than aspirin in reducing stroke, with no significant increase in bleeding complications (13). Although dual-antiplatelet therapy has not yet been directly compared with newer anticoagulants, it is unlikely (extrapolating from the comparisons with warfarin) to be as effective as these agents. Combination aspirin and clopidogrel therapy might eventually be considered only for patients with an elevated risk for stroke who are intolerant of or decline to receive anticoagulants. The optimal antithrombotic therapy should be selected in the context of the evolving field of stroke prediction in atrial fibrillation. Although considerable efforts have been directed at attempting to predict which patients with atrial fibrillation will develop stroke, current prediction methods are limited at best (14). Imperfect risk stratification may result in overtreating patients who would not otherwise develop stroke and undertreating those who Annals of Internal Medicine Editorial