摘要:Colorectal cancer (CRC) has increasing global prevalence and poor prognostic outcomes, and the development of low- or less invasive screening tests is urgently required. Urine is an ideal biofluid that can be collected non-invasively and contains various metabolite biomarkers. To understand the metabolomic profiles of different stages of CRC, we conducted metabolomic profiling of urinary samples. Capillary electrophoresis-time-of-flight mass spectrometry was used to quantify hydrophilic metabolites in 247 subjects with stage 0 to IV CRC or polyps, and healthy controls. The 154 identified and quantified metabolites included metabolites of glycolysis, TCA cycle, amino acids, urea cycle, and polyamine pathways. The concentrations of these metabolites gradually increased with the stage, and samples of CRC stage IV especially showed a large difference compared to other stages. Polyps and CRC also showed different concentration patterns. We also assessed the differentiation ability of these metabolites. A multiple logistic regression model using three metabolites was developed with a randomly designated training dataset and validated using the remaining data to differentiate CRC and polys from healthy controls based on a panel of urinary metabolites. These data highlight the changes in metabolites from early to late stage of CRC and also the differences between CRC and polyps.
其他摘要:Abstract Colorectal cancer (CRC) has increasing global prevalence and poor prognostic outcomes, and the development of low- or less invasive screening tests is urgently required. Urine is an ideal biofluid that can be collected non-invasively and contains various metabolite biomarkers. To understand the metabolomic profiles of different stages of CRC, we conducted metabolomic profiling of urinary samples. Capillary electrophoresis-time-of-flight mass spectrometry was used to quantify hydrophilic metabolites in 247 subjects with stage 0 to IV CRC or polyps, and healthy controls. The 154 identified and quantified metabolites included metabolites of glycolysis, TCA cycle, amino acids, urea cycle, and polyamine pathways. The concentrations of these metabolites gradually increased with the stage, and samples of CRC stage IV especially showed a large difference compared to other stages. Polyps and CRC also showed different concentration patterns. We also assessed the differentiation ability of these metabolites. A multiple logistic regression model using three metabolites was developed with a randomly designated training dataset and validated using the remaining data to differentiate CRC and polys from healthy controls based on a panel of urinary metabolites. These data highlight the changes in metabolites from early to late stage of CRC and also the differences between CRC and polyps.