Abstract WMP64: Global Longitudinal Peak Strain to Predict Atrial Fibrillation in Ischemic Stroke

• Published in: Stroke (1970) Vol. 50; no. Suppl_1
• Main Authors: Ryu, Wi-Sun, Bae, Eun-Kee
• Format: Journal Article
• Language: English
• Published: 01.02.2019
• ISSN: 0039-2499; 1524-4628
• DOI: 10.1161/str.50.suppl_1.WMP64

Abstract only Cerebral infarction due to atrial fibrillation is more severe and has a higher risk of death than other stroke subtypes. Anticoagulation reduces the risk of recurrence in patients with atrial fibrillation, and so it is crucial to find atrial fibrillation in stroke patients. In the present study, we investigated whether left atrial volume index (LAVI) and global longitudinal peak strain (GLPS) on speckle echocardiography could discriminate a high-risk group of atrial fibrillation in stroke patients. We prospectively enrolled 678 ischemic patients. Stroke subtypes were determined by the consensus of experienced neurologists, using a validated MRI-based algorithm. Left ventricular GLPS was measured averaging by speckle tracking imaging echocardiography from 4-chamber, 2-chamber, and the apical longitudinal long-axis view with breathing hold and an echocardiogram recording. To evaluate the predictability of LAVI and GLPS for the presence of atrial fibrillation, we used the optimism-corrected c-statistics and the net reclassification improvement with prespecified atrial fibrillation risk categories of low (10%), intermediate (10% to 30%), and high (≥30%). Patients with atrial fibrillation were likely to be older and to have more vascular risk factors. LAVI (74 ml/m 2 vs. 42 ml/m 2 ) and GLPS (-14% vs. -17%) were higher in patients with atrial fibrillation. Age (adjusted odds ratio 1.04, 95% confidence interval 1.02-1.07), LAVI (1.07, 1.06-1.09), and GLPS (1.08, 1.03-1.12) were independently associated with atrial fibrillation. As a base model, age and LAVI had a C-statistics of 0.89 (95% confidence interval 0.86 - 0.93). When GLPS levels were added to a base-model, net reclassification improvement was 6.1% (p = 0.03), albeit area under curve marginally improved (C-statistics = 0.90, 95% confidence interval 0.87 - 0.93, p = 0.058) compared to a base model. In patients with ischemic stroke, GLPS levels combined with LAVI were associated with the presence of atrial fibrillation. Our data imply that rigorous screening tests to detect atrial fibrillation may be warranted for the patients with high LAVI and GLPS levels.