Stroke remains a major public health problem worldwide; it causes severe disability and is associated with high mortality rates. However, early diagnosis of stroke is difficult, and no reliable biomarkers are currently established. In this study, mass-spectrometry-based metabolomics was utilized to characterize the metabolic features of the serum of patients with acute ischemic stroke (AIS) to identify novel sensitive biomarkers for diagnosis and progression. First, global metabolic profiling was performed on a training set of 80 human serum samples (40 cases and 40 controls). The metabolic profiling identified significant alterations in a series of 26 metabolites with related metabolic pathways involving amino acid, fatty acid, phospholipid, and choline metabolism. Subsequently, multiple algorithms were run on a test set consisting of 49 serum samples (26 cases and 23 controls) to develop different classifiers for verifying and evaluating potential biomarkers. Finally, a panel of five differential metabolites, including serine, isoleucine, betaine, PC(5:0/5:0), and LysoPE(18:2), exhibited potential to differentiate AIS samples from healthy control samples, with area under the receiver operating characteristic curve values of 0.988 and 0.971 in the training and test sets, respectively. These findings provided insights for the development of new diagnostic tests and therapeutic approaches for AIS.
Keywords: acute ischemic stroke; classify; human serum biomarkers; logistic regression; mass spectrometry; metabolomics.