To detect whether patients with gastric cancer had unique serum metabolomic characteristics by liquid chromatography-mass spectrometry (LC-MS) metabolomic analysis, and to screen potential markers for early gastric cancer and to preliminarily explore the related metabolic pathways.

At the First Affiliated Hospital of Soochow University, 66 patients with gastric cancer and 44 patients with benign gastric disease from July, 2017 to January, 2018 were enrolled, and 50 healthy subjects served as controls. Twenty-five patients with gastric cancer at stage Ⅰ and Ⅱ and 25 patients with gastric cancer at stage Ⅲ and Ⅳ were selected from the 66 patients with gastric cancer, and 25 subjects were also selected from 50 healthy controls. The plasma small molecule metabolites of patients with gastric cancer and benign gastric disease and healthy controls were detected by LC-MS method. Multivariate logistic regression analysis was used to establish and validate the principal component analysis (PCA) model and partial least squares-discriminant analysis (PLS-DA) model and screen the differential metabolites. The receiver operating characteristic curve analysis was used to evaluated the clinical efficacy of differential metabolites.

PCA and PLS-DA models showed that gastric cancer had a obviously specific metabolites profile, the profile of benign gastric disease overlapped with that of gastric cancer and healthy controls. The results of multivariate logistic regression analysis confirmed that four metabolites including isoleucine, benzophenone, sphingosine-1-phosphate and galactopyranose set could be used to establish an optimal diagnostic model. The area under the curve (AUC) (95% confidence interval (CI)) was 0.963 (0.930 to 0.997), and the best cut off value, sensitivity and specificity were 0.871, 93.1% and 94.0%, respectively. Meanwhile, patients with gastric cancer at stage Ⅰ+ Ⅱ and stage Ⅲ+ Ⅳ had a distinct clustering trend compared with the control group. In the serum of patients with gastric cancer at stage Ⅰ+ Ⅱ and stage Ⅲ+ Ⅳ, a total of 24 differential metabolites were identified, the concentration of five of which including lysine, carnitine, benzenesulfonamide, arginine and docosahexaenoic acid ethyl ester, increased along with the progression of gastric cancer. Pipecolic acid and kynurenine might served as biomarkers for early and mid gastric cancer (stage Ⅰ+ Ⅱ) screening.

LC-MS metabolomic effectively confirm the unique changes of serum metabolites in patients with gastric cancer. The screened differential metabolites have potential clinical application value for predicting the risk of gastric cancer.

Stomach neoplasms; Metabolism; Biological markers; Liquid chromatography-mass spectrometer