Abstract P302: Neuroimaging Embolic Infarct Pattern in Work-Up For Embolic Source of Undetermined Etiology in Acute Ischemic Stroke

Abstract

Introduction: Embolic infarct pattern on Neuroimaging as a marker for positive cardioembolism results in the work-up for Embolic Source of Undetermined Etiology (ESUS) in acute ischemic stroke (AIS) is unclear. Methods: Among 1625 acute ischemic stroke (AIS) hospitalized at our academic stroke program from July 2017 to April 2019, 350 (21.5%) had complete ESUS work-up performed (defined as including TEE and cardiac monitoring). Data was abstracted on demographics, medical history, initial National Institutes of Health Stroke Scale (NIHSS), systolic blood pressure (SBP), troponin level, neuroimaging findings, and cardiac test results. Positive ESUS work-up for cardioembolism was defined as results that prompted management change. Embolic infarct pattern was defined as infarct confirmed on neuroimaging to affect 2 or more vessel territories. Results: Among 350 with complete ESUS work-up performed, the mean age was 59.0 (range, 23 to 88) and 170 (48.6%) were female. The mean initial NIHSS was 5.4 (range, 0 to 32). Neuroimaging infarct pattern suggested embolic pattern in 75 (21.4%), single vessel pattern in 274 (78.3%), and not visible in 1 (0.3%). Among 274 with single vessel infarct pattern on neuroimaging, affected territory was MCA in 184, Vertebrobasilar in 46, PCA in 34, and ACA in 10. Cardioembolism source was identified in 96 (27.4%) of patients with completed ESUS work-up (7.1% had positive findings on TEE and 20.3% had new-atrial fibrillation diagnosed on cardiac monitoring). Compared to patients with single territory infarct pattern, patients with embolic pattern were more likely to have a history of diabetes mellitus (44.0% vs 31.3%, p 0.0392), higher SBP (159.0 vs 147.9mm Hg, p 0.002) and higher troponin (0.87 vs 0.07 ng/mL, p 0.0124), but there was no significant difference based upon age, sex, history of hypertension, initial NIHSS, HgbA1c, LDL, or rate of positive ESUS work up for cardioembolism (22.7% vs 28.5%, p 0.0088). Conclusion: Among AIS patients with complete ESUS work-up, the most common neuroimaging pattern was single territory MCA infarct (52.6%) followed by embolic infarct pattern (21.4%). Embolic pattern did not increase the positive yield of ESUS work-up for cardioembolism.