Abstract
This study employed non-targeted liquid chromatography-mass spectrometry (LC-MS) to investigate the salivary metabolomic profiles of 177 participants, including 147 patients with gastric cancer (GC) and 30 controls with common gastritis. Following rigorous quality control measures, a total of 333 high-confidence metabolites were identified from an initial pool of 368 positive-mode metabolites and 178 negative-mode metabolites. Principal component analysis (PCA) and linear discriminant analysis (LDA) revealed significant differences between the GC and control groups, with stage I GC clearly distinguishable from more advanced stages. Differential metabolite screening was conducted using partial least squares discriminant analysis (PLS-DA) in conjunction with t-tests and fold-change analysis. After controlling for confounding factors such as age, smoking, and alcohol use, 38 salivary metabolites were identified as potential diagnostic markers. Notably, the univariate and multivariate diagnostic models demonstrated excellent discriminative performance in distinguishing GC patients from controls. PLS-DA validated by permutation testing, along with univariate and multivariate ROC analyses, exhibited excellent classification performance based on the 38 salivary metabolites. Metabolic analysis revealed significant downregulation of purines, pyrimidines, amino acids, and carbohydrates in the saliva of GC patients, while sebacic acid and GABA were found to be upregulated. Tyrosine was identified as the most significantly altered metabolite between early and advanced stages of GC. These findings underscore the substantial impact of gastric cancer on the salivary metabolome and suggest the potential of saliva as a promising tool for mass screening of GC.Trial registration DGLES, NCT01420588, Registered 19 August 2011, https://clinicaltrials.gov/ct2/show/NCT01420588 .