Emerging Management Challenges Posed by Atrial Fibrillation Detected After Stroke

Atrial fibrillation (AF) is an important risk factor for acute ischemic stroke, and among patients with overt or clinically symptomatic AF, studies have shown that anticoagulation reduces that risk.1 In patients with acute ischemic stroke and AF, nearly 25% of AF is detected after the stroke2 and patients with AF detected after acute ischemic stroke of presumed embolic etiology are typically treated with anticoagulation. Although some studies suggest that patients with AF detected after stroke (AFDAS) have a lower stroke recurrence risk than patients with AF known prior to the stroke (Known AF),3 reasons for these associations remain uncertain. Nevertheless, AF is a potent risk factor for future stroke, and the evaluation and treatment for AF as well as the predictive models in current use such as CHA2DS2-VASc do not distinguish these two categories of AF as distinct pathophysiologic entities. When identified, AF should be factored into the stroke secondary prevention strategy in a manner similar to that of other vascular risk factors such as hypertension and diabetes. Insertable cardiac monitors detecting AFDAS at rates up to 8 times greater than conventional care4, 5 does not warrant dismissing it as a fleeting and clinically insignificant poststroke artifact. Clinicians who are comfortable with readily setting aside treatment for AFDAS in patients with stroke of presumed known origin should likewise be less comfortable embracing it as a treatable cause of cryptogenic embolic stroke of uncertain source, because the pattern and burden of AFDAS in these two populations has not been shown to differ. It may be framing bias of suspected occult AF in patients with cryptogenic stroke, rather than the actual arrhythmic burden, that sways opinion in these matters. In a study published in this issue of Annals of Neurology, Lyrer et al6 analyzed data from a multicenter study of patients with acute ischemic stroke in the setting of AF. They evaluated the risk of recurrent stroke in patients with AF, comparing those with two different categories of AF detection: AFDAS versus those with Known AF. Like others, they found that in a simple Cox model including only an AF category, Known AF was associated with a higher hazard for ischemic stroke recurrence than AFDAS, but in combined Cox models with both an AF category and oral anticoagulation (OAC) before stroke, Known AF was no longer associated with ischemic stroke recurrence, whereas pre-existing anticoagulation retained its strong association with a higher hazard for the primary outcome of recurrent stroke. Reassuringly, in the subgroup of OAC-naïve patients, there was no association of AF category with the hazard for the primary outcome of ischemic stroke recurrence in simple and adjusted Cox models. This finding is important, because current skepticism about the relative risk of AFDAS versus Known AF fueled by prior studies may lead many providers to withhold OAC in the setting of AFDAS. In the Stroke AF study, where patients had a median CHA2DS2-VASc of 5, ~33% of patients were not treated with OAC despite AF detection by an insertable cardiac monitor.5 This study challenges the conventional wisdom that Known AF and AFDAS are clinically relevant and distinct prognostic entities. Instead of an independently high stroke recurrence risk being attributable to Known AF, it is likely that this is due to existing and residual confounding by the presence of prestroke OAC and higher prevalence of vascular risk factors in the Known AF group. Although this study6 is important, it has several limitations worth noting. Authors lack data on AF burden, which is known to be associated with ischemic stroke risk in AF. Furthermore, the diagnostic evaluation performed was not reported, and thus there also may be residual confounding in the Known AF group from underlying atherosclerosis or malignancy as causes of the ischemic stroke rather than pure AF-related cardioembolism, and these also have a higher recurrence risk. The question remains as to the optimal antithrombotic choice for patients with AFDAS. If it is merely the earlier phase of detection of AF due to more sensitive monitoring, then we can expect stroke rates to increase as AF burden increases over time, and it becomes a question of when, not if, OAC is indicated. The increased risk of bleeding, which is already quite low with direct oral anticoagulants, should be considered only for the first few years of increasing AF burden rather than the many years of established AF for which OAC is already clearly beneficial and incorporates the risk of harm due to bleeding. Failure of timely initiation of OAC in patients with AFDAS in favor of continued antiplatelet therapy may result in more cardioembolic strokes with greater disability, even if the overall rate of atherothrombotic stroke is lower. Although anticoagulation is the current standard of care for most patients with AF and ischemic stroke, anticoagulation has several known challenges, including compliance, cost, hemorrhage risk, need for interruption for surgical procedures, and the residual stroke risk despite therapeutic anticoagulation. Patients with AF who develop an ischemic stroke despite anticoagulation therapy have been shown to carry a high risk of recurrence in several patient cohorts.7, 8 Resuming anticoagulation or changing antithrombotic strategy has not been shown to reduce the risk of ischemic stroke recurrence in such patients and may increase risk of bleeding.7, 8 It may be beneficial in some of these patients to screen pharmacy refill data for compliance or measure OAC drug levels to ensure therapeutic anticoagulation has been achieved. Left atrial appendage closure (LAAC) is an emerging treatment for stroke prevention in patients with AF and contraindications to OAC, with initial studies showing comparable ischemic stroke risk but lower hemorrhagic stroke risk than vitamin K antagonists.9, 10 Furthermore, limited studies have suggested that LAAC may result in rates of ischemic stroke recurrence comparable to those of direct-acting OAC. This treatment option raises interesting possibilities for combined mechanical and pharmacological therapy, especially in patients with AF and multiple stroke risk factors, including advanced atherosclerosis. The data to date emphasize the importance of conducting a thorough diagnostic evaluation in patients with AFDAS or Known AF to identify coexisting high-risk conditions such as atherosclerosis with vulnerable plaque or active malignancy with its associated hypercoagulability. Furthermore, it is essential that all risk factors for stroke are controlled and lifestyle modifications are implemented effectively, because uncontrolled smoking, hypertension, and diabetes all can indirectly increase the risk of thrombosis and thereby stroke risk. Furthermore, in patients with AF suffering an ischemic stroke despite anticoagulation with no evidence of a competing mechanism, it is possible that this subgroup benefits from the combined approach of LAAC plus anticoagulation therapy. The left atrial appendage occlusion study III trial randomized patients with AF undergoing cardiac surgery to LAAC versus no LAAC, but nearly 75% of patients remained on pre-enrollment OAC. In this trial, patients with LAAC had a significantly lower risk of stroke and systemic embolism during follow-up.11 As we continue to improve and refine secondary stroke prevention, new challenges and new questions emerge. As shown by Lyrer et al,6 there is no substitute for randomized trial data to help avoid the confounding inherent in observational data that has led many to treat AFDAS and Known AF as distinct clinical entities. Well-designed and adequately powered studies are urgently needed to determine optimal stroke prevention therapy for patients with AFDAS.