Diagnostic considerations of embolic strokes of undetermined source on admission

Objective: The cause of acute ischemic stroke is unknown in many patients at the time of admission. We aimed to determine the current status of diagnostic test performance and definitive diagnosis at hospital discharge in patients with embolic strokes of unknown source on admission.Methods: Between July 2012 and June 2013, 623 patients with acute ischemic stroke were admitted to our hospital within 7 days from onset. Among them, patients with non-lacunar strokes, excluding those with occlusive atherosclerosis or high-risk sources of cardioembolism, were enrolled. We evaluated the execution rates of examinations and their findings as well as the definitive stroke subtypes of patients at discharge.Results: In total, 147 patients (59 women; mean age: 73 ± 14 years) were enrolled. The execution rate of each examination was: electrocardiogram monitoring in 100%; Holter electrocardiogram in 76%; transthoracic echocardiography in 72%; and transesophageal echocardiography in 68% of patients. Paroxysmal atrial fibrillation and patent foramen ovale were detected in 20% and 46% of patients, respectively, representing the majority of the embolic sources. Stroke subtypes were determined to be: cardioembolism in 54% (High-risk 24%, Medium-risk 30%); Large-artery atherosclerosis in 9%; other determined etiology in 4%; and undetermined etiology in 33% of patients at discharge.Conclusions: Cardioembolic stroke was diagnosed in one-half of patients with embolic stroke, whereas an embolic source could not be detected in one-third of patients. Embolic stroke of undetermined source (ESUS) was found in 13% of total ischemic strokes at discharge. Further studies are needed to investigate the diagnostic characteristics of ESUS.