Pancreatic cancer screening using a multiplatform human serum metabolomics system.

To examine a novel screening method for pancreatic cancer involving gas chromatography/mass spectrometry and liquid chromatography/mass spectrometry-based metabolomics analysis.Sera from pancreatic cancer patients (n = 59) and healthy volunteers (n = 59) were allocated to the training set or validation set. Serum metabolome analysis was carried out using our multiplatform metabolomics system. A diagnostic model was constructed using a two-phase screening method that was newly advocated.When the training set was used, the constructed diagnostic model exhibited high sensitivity (100%) and specificity (80%) for pancreatic cancer. When the validation set was used, the model displayed high sensitivity (84.1%) and specificity (84.1%).We successfully developed a diagnostic model for pancreatic cancer using a multiplatform serum metabolomics system.