Identification of traumatic intracerebral hemorrhage associated metabolites using untargeted metabolomics

Traumatic intracerebral hemorrhage (TICH) is a severe neurological emergency whose metabolic mechanisms remain largely unresolved. This study used serum metabolomics analysis to compare the metabolic profiles of TICH patients (n = 10) with non-TICH controls (n = 10). LC-MS/MS analysis identified a total of 3183 metabolites. The relative abundances of benzene and its derivatives, organic acids and their derivatives were significantly increased in the serum of ICH patients, while fatty acid, glycerophospholipid, and sphingomyelin metabolites were generally decreased. Multidimensional statistical methods were used to screen for significantly differentially expressed metabolites. Metabolite regulatory networks revealed the potential roles of amino acid metabolism, lipid metabolism (GP/SP/FA), and aromatic compounds (benzenes) in regulating TICH metabolism. Enrichment analysis revealed that these metabolites were highly enriched in lipid metabolism, amino acid metabolism, and neural signaling pathways, and were significantly associated with inflammatory drug pathways. The differentially expressed metabolites were further mapped to the Human Metabolite Database (HMDB) and subjected to metabolic-disease association analysis, suggesting that these metabolites may be closely related to the pathogenesis of TICH. ROC analysis demonstrated good diagnostic performance of five key metabolites in both discovery and validation cohorts (AUC > 0.70), with hypoxanthine and L-histidine showing the highest accuracy (AUC > 0.93). Correlation analysis indicated that key metabolites were associated with hematoma volume, NIHSS and GCS scores. In vitro experiments further confirmed that, under conditions of neuronal injury, the changes in these metabolites exhibit consistency. This study provides a theoretical basis for the development of potential biomarkers and precise intervention strategies.