Abstract: Objective To identify potential serum metabolic biomarkers in colorectal cancer (CRC) patients using untargeted metabolomics and to evaluate their diagnostic and staging value. Methods Serum samples from 100 healthy controls and 100 CRC patients were analyzed by ultra-performance liquid chromatography-mass spectrometry (UPLC-MS). After data normalization, differential metabolites were screened using multivariate statistical analyses (PCA, OPLS-DA) and subjected to pathway enrichment analysis. Diagnostic performance was assessed via univariate and multivariate regression, while Mfuzz clustering was applied to analyze stage-related metabolites (Ⅰ-Ⅳ). Results A total of 432 metabolites were identified with 59 showing significant alterations. Starch and sucrose metabolism and glycerophospholipid metabolism pathways were significantly enriched. A three-metabolite panel (4,8- dimethylnonanoyl carnitine, 9,13-dihydroxy-4-megastigmen-3-one 9-glucoside and C17 sphingosine-1-phosphate) achieved a diagnostic AUC of 0.907, while L-Carnitine and L-Norleucine showed an AUC of 0.776 in staging analysis. Conclusions Specific serum metabolite panel exhibit high diagnostic accuracy, and dysregulated metabolic pathways are associated with CRC progression, suggesting their potential value as biomarkers.

Key words: colorectal cancer, untargeted metabolomics, UPLC-MS, serum biomarkers