Metabolomics on vascular events and death after acute ischemic stroke: A prospective matched nested case-control study

Abstract

Background and aimsIschemic stroke is a major contributor to global mortality and disability. Metabolomics represents a powerful tool for discovering biomarkers of ischemic stroke due to its ability to detect metabolites small enough to cross the blood–brain barrier. The aim of this study is to identify potential metabolic biomarkers for predicting 1-year vascular events and death after acute ischemic stroke (AIS). MethodsWe adopted a nested case-control design from a multicenter prospective cohort study. A total of 143 AIS patients with 1-year vascular events/death and 143 sex-, age- and center-matched patients without 1-year vascular events/death were selected, and further divided into the discovery set (n = 140) and validation set (n = 146). We then performed untargeted metabolomics analysis by liquid chromatography coupled to mass spectrometry. ResultsMetabolomics analyses could predict 1-year vascular events and death after AIS using pattern recognition methods. Eight metabolites (e.g., LysoPC(18:1)) were identified as potential biomarkers of AIS prognosis. Four of them (e.g., PS(O-18:0/0:0)) were found to be significantly decreased in patients with early vascular events/death. Adding these metabolic biomarkers to traditional factors resulted in a great improvement of the predictive utility for 1-year vascular events/death, with net reclassification index and integrated discrimination improvement being 0.9143 (p < 0.0001) and 0.1906 (p < 0.0001) in the discovery cohort, and 0.9041 (p < 0.0001) and 0.1896 (p < 0.0001) in the validation cohort. ConclusionsThis study found several metabolic biomarkers for 1-year vascular events and death after AIS, providing opportunities for the construction of prognostic models and the discovery of novel therapeutic targets.