Metabolomic markers of colorectal tumor with different clinicopathological features

BackgroundColorectal cancer (CRC) is the result of complex interactions between the tumors molecular profile and metabolites produced by its microenvironment. Despite recent studies identifying CRC molecular subtypes, a metabolic classification system is still lacking. We aimed to explore the distinct phenotypes and subtypes of CRC at the metabolic level. MethodsWe conducted an untargeted metabolomics analysis of 51 paired tumor tissues and adjacent mucosa using ultra-performance liquid chromatography/quadrupole time-of-flight mass spectrometry. Multivariate analysis including principal component analysis, orthogonal partial least squares discriminant analysis and heat maps, univariate analysis, and pathway analysis were used to identify potential metabolic phenotypes of CRC. Unsupervised consensus clustering were used to identify robust metabolic subtypes, and evaluated their clinical relevance. ResultsA total of 173 metabolites (including nucleotides, carbohydrates, free fatty acids, and choline) were identified between CRC tumor tissue and adjacent mucosa. We found that lipid metabolism was closely related to the occurrence and progression of CRC and CRC tissues could be divided into three subtypes, and statistically significant correlations between different subtypes and clinical prognosis were observed. ConclusionsCRC tumor tissue exhibits distinct metabolic phenotypes. Metabolic differences between subtypes may provide a basis and direction for further clinical individualized treatment planning.