Lipidomic profiling identifies key pathways and a 5-lipid panel with high diagnostic efficacy for ischemic stroke

Abstract

Abstract Ischemic stroke (IS) accounts for over 80% of all stroke cases, presenting as a prevalent, debilitating cerebrovascular disorder with limited therapeutic options. The urgent need for early diagnostic biomarkers and insights into pathogenesis has highlighted dysregulated lipid metabolism as a key contributor, while metabolomics advances enable novel biomarker exploration. This study integrated bioinformatics and a case-control design to investigate IS-related lipid metabolism pathways and blood lipid biomarkers. Gene Expression Omnibus (GEO) gene expression datasets were analyzed via Gene Set Enrichment Analysis (GSEA) to identify lipid pathways, and case-control analyses employed Chi-square/Z tests for conventional blood lipids, Liquid Chromatography-Mass Spectrometry (LC-MS) for plasma small-molecule lipids, and orthogonal partial least squares discriminant analysis, t-tests, and Receiver Operating Characteristic (ROC) curves for validation. Results revealed five significantly downregulated lipid pathways (α-linolenic acid, linolenic acid, ether lipid, glycerophospholipid, and sphingolipid metabolism). IS patients exhibited dyslipidemia (elevated TC/TG/LDL-C, reduced HDL-C). Additionally, 15 differentially expressed lipid molecules were identified in a validation cohort after excluding the influence of comorbidities. Among these, five representative lipids (e.g., PE(P-18:1/22:4)) demonstrated potential diagnostic performance, with an area under the receiver operating characteristic curve (AUC) of 0.917, sensitivity of 60.0%, and specificity of 96.7%, indicating their potential utility as biomarkers for the early detection of IS.