Abstract T P179: Atrial Fibrillation Unidentified Prior to Stroke/tia: Background Features, Stroke Severity and Outcome - The Samurai-nvaf Study

Purpose: Atrial fibrillation (AF) is often detected after embolic events occur, and it is an obstacle to effective preventive anticoagulation. We aimed to determine a percentage of patients with nonvalvular AF (NVAF) unidentified prior to stroke in the overall stroke patients with NVAF, as well as background features, stroke severity and outcome of such occult NVAF patients. Methods: A total of 743 acute ischemic stroke/TIA patients with NVAF (344 women, 78±10 years old) were enrolled between Sep 2011 and Jun 2013 from a multicenter prospective registry (the SAMURAI-NVAF study, [NCT01581502][1]). Patients were divided into two groups; those with identified AF before stroke/TIA (Group I) and those with unidentified AF that was documented at emergent visit or later (Group U). Favorable outcome was defined as mRS 0-2 at hospital discharge (for 710 patients, median 23 days) and at 3 months (for 565 patients). Results: 285 patients belonged to the Group U (38%; 138 women, 78±11 years old). Although both median CHADS2 (2 vs. 2, p<0.001) and CHA2DS2-VASc (3 vs. 4, p<0.001) were lower in the Group U than the Group I, patients with the high ischemic risk category (≥2 in each score) accounted for 68% according to CHADS2 and 91% according to CHA2DS2-VASc in the Group U. After multivariate adjustment, paroxysmal AF (OR 1.91, 95% CI 1.24-2.75) was more common and congestive heart failure (0.63, 0.40-0.99) and premorbid use of oral anticoagulants (0.08, 0.04-0.13) were less common in the Group U than the Group I. The median initial NIHSS was higher in the Group U (11 [IQR 4-19]) than in the Group I (6 [2-17], p<0.001). Favorable outcome was less common in the Group U than the Group I both at discharge (41% vs. 51%, p=0.011) and at 3 months (46% vs. 56%, p=0.036). Unidentified AF was independently associated with mRS 3-6 after adjustment for sex and age both at discharge (OR 1.57, 95% CI 1.13-2.18) and at 3 months (1.60, 1.10-2.32), but was no longer associated with mRS 3-6 after further adjustment for the initial NIHSS. Conclusion: Two fifth of the stroke/TIA patients with NVAF were not diagnosed as having AF prior to the attack, though their ischemia-risk indices were generally high. Patients with such occult NVAF had severer stroke and worse outcome than those with identified AF. [1]: /lookup/external-ref?link_type=CLINTRIALGOV&access_num=NCT01581502&atom=%2Fstrokeaha%2F45%2FSuppl_1%2FATP179.atom