Analysis Of The Difference Of Plasma Amino Acids In Patients With Early Gastric Cancer And Benign Gastric Disease By Liquid Chromatography-Tandem Mass Spectrometry

ObjectiveGastric cancer(GC)is a malignant tumor with high incidence in the world.In benign gastric diseases,the risk of GC in patients with gastric ulcer and chronic atrophic gastritis is significantly higher than that of the healthy population.The effective way to improve the prognosis of GC is early screening,but there is no noninvasive screening tool for accurate diagnosis of GC.The purpose of this study was to find potential molecular markers for the discrimination of early GC and benign gastric diseases,in order to improve the diagnosis and screening techniques of early GC.MethodsThis study used metabonomics method based on liquid chromatography-tandem mass spectrometry(LC-MS/MS)to analyze the plasma amino acids of 50 patients with early GC and 50 patients with gastric benign diseases(gastric ulcer in 27 cases and chronic atrophic gastritis in 23 cases),and 23 amino acids were detected in each patient.Partial least squares-discriminant analysis(PLS-DA)was used to analyze amino acid data and observe the differences in metabolism of amino acid profiles between early GC and benign gastric diseases groups.The differences of amino acid levels between the two groups were further verified by the parameter test.The Logistic regression model was established for the differential amino acids between the two groups,finally screening the amino acids that were valuable for the differential diagnosis of early GC and gastric benign diseases.The diagnostic ability of the regression model was evaluated by the receiver operating characteristic(ROC)curve.ResultsPLS-DA model showed metabolic differences in the amino acid profiles between early GC and gastric benign disease groups.According to results from parameter test and PLS-DA model,7 amino acids showed significant differences between the two groups in the metabolism.Logistic regression model was constructed using these 7 differential amino acids.Three amino acids,including glutamine(Gln),arginine(Arg)and ornithine(Orn),were screened in the final regression model for discriminating early GC and gastric benign diseases with good specificity and sensitivity.The ROC curve was used to evaluate the diagnostic ability of the regression model.The area under the curve(AUC)of the model was 0.937,and the diagnostic sensitivity and specificity was 90.7% and 90.2%,respectively.ConclusionsIn this study,a reasonable regression diagnosis model was established by Gln,Arg and Orn,and this model could be used to differentiate early GC and gastric benign diseases.Metabolomics based on LC-MS/MS could facilitate to diagnose and screen early GC and provide a scientific basis for studying the metabolic mechanisms of tumors.