Non-invasively predicting differentiation of pancreatic cancer through comparative serum metabonomic profiling

Research paper by Shi Wen, Bohan Zhan, Jianghua Feng, Weize Hu, Xianchao Lin, Jianxi Bai, Heguang Huang

Indexed on: 03 Nov '17Published on: 02 Nov '17Published in: BMC Cancer


The differentiation of pancreatic ductal adenocarcinoma (PDAC) could be associated with prognosis and may influence the choices of clinical management. No applicable methods could reliably predict the tumor differentiation preoperatively. Thus, the aim of this study was to compare the metabonomic profiling of pancreatic ductal adenocarcinoma with different differentiations and assess the feasibility of predicting tumor differentiations through metabonomic strategy based on nuclear magnetic resonance spectroscopy.By implanting pancreatic cancer cell strains Panc-1, Bxpc-3 and SW1990 in nude mice in situ, we successfully established the orthotopic xenograft models of PDAC with different differentiations. The metabonomic profiling of serum from different PDAC was achieved and analyzed by using 1H nuclear magnetic resonance (NMR) spectroscopy combined with the multivariate statistical analysis. Then, the differential metabolites acquired were used for enrichment analysis of metabolic pathways to get a deep insight.An obvious metabonomic difference was demonstrated between all groups and the pattern recognition models were established successfully. The higher concentrations of amino acids, glycolytic and glutaminolytic participators in SW1990 and choline-contain metabolites in Panc-1 relative to other PDAC cells were demonstrated, which may be served as potential indicators for tumor differentiation. The metabolic pathways and differential metabolites identified in current study may be associated with specific pathways such as serine-glycine-one-carbon and glutaminolytic pathways, which can regulate tumorous proliferation and epigenetic regulation.The NMR-based metabonomic strategy may be served as a non-invasive detection method for predicting tumor differentiation preoperatively.