Phenotypic Characterization Analysis of Human Hepatocarcinoma by Urine Metabolomics Approach

Abstract

Hepatocarcinoma (HCC) is one of the deadliest cancers in the world and represents a significant disease burden. Better biomarkers are needed for early detection of HCC. Metabolomics was applied to urine samples obtained from HCC patients to discover noninvasive and reliable biomarkers for rapid diagnosis of HCC. Metabolic profiling was performed by LC-Q-TOF-MS in conjunction with multivariate data analysis, machine learning approaches, ingenuity pathway analysis and receiver-operating characteristic curves were used to select the metabolites which were used for the noninvasive diagnosis of HCC. Fifteen differential metabolites contributing to the complete separation of HCC patients from matched healthy controls were identified involving several key metabolic pathways. More importantly, five marker metabolites were effective for the diagnosis of human HCC, achieved a sensitivity of 96.5% and specificity of 83% respectively, could significantly increase the diagnostic performance of the metabolic biomarkers. Overall, these results illustrate the power of the metabolomics technology which has the potential as a non-invasive strategies and promising screening tool to evaluate the potential of the metabolites in the early diagnosis of HCC patients at high risk and provides new insight into pathophysiologic mechanisms.