Exploration of Lipid Metabolism in Gastric Cancer: A Novel Prognostic Genes Expression Profile

BACKGROUND: Alterations in lipid metabolism are increasingly being recognized. However, the application of lipid metabolism in the prognosis of gastric cancer (GC) has not yet been explored. METHODS: A total of 204 lipid metabolism relative genes were analyzed in the GC cohort from The Cancer Genome Atlas (TCGA), and four independent cohorts from Gene Expression Omnibus (GEO) and one cohort from Wuhan Union Hospital were applied for external validation. Differential expression and enrichment analyses were performed between GC and normal tissue. The LASSO-Cox proportional hazard regression model was applied to select prognostic genes and to construct a gene expression profile. RESULTS: Our research indicated that higher expression level of AKR1B1, PLD1, and UGT8 were correlated with worse prognosis of GC patients, while AGPAT3 was correlated with better prognosis. Furthermore, we developed a gene profile composed of AGPAT3, AKR1B1, PLD1, and UGT8 suggested three groups with a significant difference in overall survival (OS). The profile was successfully validated in an independent cohort and performed well in the immunohistochemical cohort. Furthermore, we found that ether lipid metabolism, glycerophospholipid metabolism, and glycerolipid metabolism were upregulated, and fatty acid β-oxidation and other lipid peroxidation processes were reduced in GC. CONCLUSION: Collectively, we found lipid metabolism is reliable and clinically applicable in predicting the prognosis of GC based on a novel gene profile.