Prediction Of Chemosensitivity And Toxicity For Gastric Cancer Based On Metabonomics

Objective:To develop new strategies of individualized therapy for gastric cancer,indicators for chemosensitivity and toxicity prediction based on mtabonomics will be screened and prediction models of chemosensitivity and toxicity for gastric cancer will be built.Methods:Subcutaneous transplanted models of human gastric cancer on nude mice were built and their chemosensitivity and toxicity response to PF regimen was observed.A method of metabonomics based LC/MS techniques was used to analyze the changing patterns of metabolites in the plasma of tumor-bearing nude mice and anti-cancer mechanism of PF regimen for gastric cancer was investigated. Possible reasons why different chemosensitivity and toxicity response occurred in nude mice individuals after the same treatment of PF regimen were also analyzed based on metabonomics.Some metabolites which would be promising indicators for chemosensitivity and toxicity prediction for gastric cancer were screened.Prediction models of chemosensitivity and toxicity for gastric cancer were built based on metabonomics.Results:The anti-cancer mechanism of PF regimen was mainly associated with metabolism of several kinds of glycerophosphocholines and amino acids.Glycerophosphocholines were important indicators of chemosensitivity and toxicity prediction for gastric cancer.A chemosensitivity prediction model with accurate rate 83.3%and a toxicity prediction model with accurate rate 84.3%were built successfully.Conclusion:Several indictors of chemosensitivity and toxicity prediction for gastric cancer were selected out based on metabonomics and two prediction models of chemosensitivity and toxicity with high accuracy were built,which might be helpful to studies of individualized therapy for gastric cancer.