Using Extracellular Circulating microRNAs to Classify the Etiological Subtypes of Ischemic Stroke

There is no effective biological method to classify ischemic stroke subtypes. In this study, we first performed a systematical gene array study on serum microRNAs with different ischemic stroke subtypes including 13 normal control subjects (NCs) and 87 ischemic stroke (IS) patients including 23 cardioembolism (CARD), 26 large artery atherosclerosis (LAA), 27 lacunar infarct (LAC), and 11 stroke of undetermined etiology (SUE). Validation was performed by using an independent cohort of 20 NCs and 85 IS patients including 28 CARD, 23 LAA, 18 LAC, and 16 SUE. In the pilot discovery gene array study, we found specific serum microRNA signatures between different ischemic stroke subtypes (CARD, LAA, LAC, and SUE). We further validated 6 microRNAs [miR-125b, miR-125a, let-7b, let-7e, miR-7-2-3p, miR-1908] in a different group of ischemic stroke subtypes by using an independent cohort of 20 NCs, 28 CARD, 23 LAA, 18 LAC, and 16 SUE. Moreover, these circulating miRNAs were further detected to be differentially expressed between pre- vs. post-stroke in different ischemic stroke subtypes. The ROC analysis showed that miR-125b, miR-125a, let-7b, and let-7e could discriminate CARD patients from normal controls and other subtypes. Furthermore, ROC curves shown that miR-7-2-3p and miR-1908 showed significant area-under-the-curve values in both LAA and LAC patients. In conclusion, these results demonstrated that circulating miRNAs in sera could be potentially novel risk factors that involve in the pathogenesis of ischemic stroke subtypes.