Embolic Stroke, Atrial Fibrillation, and Microbleeds

Patients with AF have a high risk of ischemic stroke, which can be dramatically reduced by oral anticoagulation with vitamin K antagonists (VKAs). I will look at our patient from different positions. First, the patient has paroxysmal AF. The type of AF (paroxysmal, persistent, or permanent) has only a minor impact on the stroke risk. All types of AF respond to oral anticoagulation with VKAs or non–vitamin K oral anticoagulants (NOACs) in a similar way. Therefore, our patient needs to be anticoagulated. Second, the patient’s MRI shows a cortical embolic looking stroke. Therefore, it is likely that the stroke was due to AF. Lacunar strokes, which are treated with antiplatelet therapy, are located in the white matter, basal ganglia or in the brainstem and midbrain. The location and type of infarction makes a hemodynamic stroke due to large vessel disease unlikely. A patient with a severe carotid stenosis and AF would need to undergo carotid endarterectomy. Third, the patient has a small number of cerebral microbleeds. These do not constitute a contraindication to oral anticoagulation. Microbleeds are key markers of cerebral small vessel disease. In a retrospective analysis, Horstmann et al analyzed data from a prospectively recruiting stroke registry in patients with AF; 785 patients with ischemic stroke or transient ischemic attack with brain MRI including susceptibility weighted imaging were consecutively enrolled during a 3-year period. At least 1 cerebral microbleed was detected in 186 (23.7%) patients. Microbleeds were significantly more frequent in patients with AF (30.5% versus 22.4%). Patients with previous OAC (VKAs) treatment were more likely to have cerebral microbleeds (36.7% versus 22.8%; P=0.03), and abundant cerebral microbleeds (n>10) were more frequent in anticoagulated patients even after adjustment for age. NOACs do not seem to increase the prevalence of microbleeds. There is uncertainty in which load of microbleeds and their location would constitute an increased risk of bleeding, in particular, in patients on oral anticoagulation. Patients with a high number of cortical microbleeds in addition to severe white matter disease on MRI might have amyloid angiopathy and should not be anticoagulated. Fourth, the 4 large trials, which compared VKAs with NOACs in patients with AF, had subgroups of patients with transient ischemic attack or ischemic stroke. The results for these subgroups for apixaban, dabigatran, and rivaroxaban have been published. There was a trend toward higher efficacy in preventing stroke or systemic embolism in favor of the NOACs. The subgroups, however, were not powered to show a significant benefit. A meta-analysis of 3 trials in 14 527 patients showed that NOACs were associated with a significant reduction of stroke/systemic embolism (odds ratio, 0.85 [95% confidence interval, 074–0.99]) compared with warfarin. Non-VKAs were also associated with a significant reduction of major bleeding compared with warfarin (odds ratio, 0.86 [95% confidence interval, 075–0.99], mainly driven by the significant reduction of hemorrhagic stroke (odds ratio, 0.44 [95% confidence interval, 032–0.62]). Data for edoxaban have not yet been published. Embolic Stroke, Atrial Fibrillation, and Microbleeds Is There a Role for Anticoagulation?