Plasma Lipidomic Profiling Reveals Predictive Biomarkers and Pathogenic Pathways in Recurrent Ischemic Stroke

Objective Recurrent ischemic stroke (RIS) poses a major challenge to clinical cerebrovascular care, with high disability and mortality rates. This study aimed to identify plasma lipid biomarkers specific to RIS, characterize dynamic lipid metabolic alterations across the healthy, acute ischemic stroke (AIS), recovery-phase non-recurrent IS (RIS-free IS), and RIS stages, and provide molecular evidence for early RIS prediction and precision intervention. Methods A total of 40 participants were enrolled from the Department of Cerebrovascular Diseases, The Second Affiliated Hospital of Anhui University of Traditional Chinese Medicine, and equally divided into four groups (n = 10 each): healthy controls (HC), AIS, RIS-free IS, and RIS. Plasma samples were analyzed using non-targeted lipidomics based on liquid chromatography–tandem mass spectrometry (LC-MS/MS). Data preprocessing included peak detection, alignment, and normalization. Multivariate statistical analyses (principal component analysis [PCA], orthogonal partial least squares discriminant analysis [OPLS-DA]) and univariate analyses (Student’s t-test with false discovery rate [FDR] correction) were applied to screen differential lipid species. Candidate biomarkers were further validated using variable importance in projection (VIP) scores (VIP > 1.0), receiver operating characteristic (ROC) curves, and least absolute shrinkage and selection operator (LASSO) regression. Kyoto Encyclopedia of Genes and Genomes (KEGG) and MetaboAnalyst were used for metabolic pathway enrichment analysis. Results Distinct plasma lipid profiles were observed among the four groups. Compared with non-recurrent groups (AIS + RIS-free IS), RIS patients showed significantly elevated levels of lysophosphatidylcholine (LysoPC)(16:0), LysoPC(20:4), phosphatidylcholine (PC)(34:1), and free fatty acid (FA)(18:2) (all P < 0.05). ROC analysis demonstrated that LysoPC(16:0) alone achieved an area under the curve (AUC) of 0.81 (95% confidence interval [CI]: 0.73–0.89) for RIS prediction, while the combined panel of the four differential lipids improved the AUC to 0.91 (95% CI: 0.85–0.96). Pathway enrichment analysis revealed significant perturbations in glycerophospholipid metabolism, fatty acid metabolism, and arachidonic acid metabolism in RIS patients. Comparative analysis across stages indicated dynamic lipid metabolic shifts: AIS was associated with increased phospholipids and free fatty acids, RIS-free IS showed heterogeneous remodeling of glycerophospholipids and sphingolipids, and RIS was characterized by altered triglycerides and ceramides. Conclusion Plasma lipidomics can effectively identify characteristic metabolic signatures of RIS. LysoPC(16:0) and the combined lipid panel exhibit high potential for early RIS risk stratification, and the dysregulated lipid metabolic pathways provide novel insights into RIS pathogenesis, laying a foundation for precision prevention and intervention strategies.