Abstract
BACKGROUND: Gastric cancer remains one of the major causes for tumor-related deaths worldwide. Our study aimed to provide an understanding of primary gastric cancer and prompt its clinical diagnosis and treatment. METHODS: We integrated the expression profiles and overall survival information of primary gastric cancer in TCGA and GEO database and estimated the stromal score of each sample by the estimate R package. Stromal score and clinicopathologic characteristics associated with overall survival were analyzed by using Cox regression and the Kaplan-Meier method. Gene set enrichment analysis (GSEA) and KEGG analysis were performed to explore the potential molecular mechanism in TCGA dataset. The relationship between immunotherapy-associated markers or immune cell types and stromal score was explored by using Pearson correlation analysis. RESULTS: A total of 796 samples were collected for the analysis. Patients with stromal score-high showed poor overall survival (P < .01, HR: 1.407, 95% CI: 1.144-1.731) and identified as an independent prognostic factor. KEGG analysis revealed that stromal score actively involved in diverse tumor-associated pathways. GSEA analysis also revealed stromal score associated with diverse immune-related biological processes. Furthermore, stromal score was related with immunotherapy-associated markers and multiple immune cells. CONCLUSION: Our results showed that stromal score could serve as a potential prognostic biomarker in primary gastric cancer and play an important role in the recognition, surveillance, and prognosis of gastric cancer.