Abstract
BACKGROUND: Gastric cancer (GC) is a leading gastrointestinal malignancy carrying a poor prognosis. Lymphangiogenesis (LYM) refers to the process of forming new lymphatic vessels. This process facilitates tumor metastasis and represents a promising therapeutic target in GC management. However, the exact mechanisms of LYM in GC remain incompletely understood. METHOD: The RNA-sequencing gene expression dataset and clinical characteristics of GC patients were obtained from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) database. The LASSO Cox regression method was utilized to identify feature genes and construct a Lymphangiogenesis Score (LYMS). A nomogram was constructed to assess the predictive efficacy of LYMS in the prognosis of GC patients. The gene set enrichment analysis (GSEA) employed to investigate different molecular functions and pathways. The immune microenvironment analysis, immunotherapy response analysis, and drug sensitivity were conducted to elucidate the association between LYMS and both immune landscape and immunotherapy response. RESULTS: This study selected six LYM-related genes (ADAMTS1, SVEP1, CAV1, NOX4, NPTX1, and SPARC) to construct the LYMS. The results demonstrated that GC patients with a high LYMS exhibited significantly poorer prognosis. Distinct enrichment patterns of molecular functions and pathways were observed between the high and low LYMS groups. Furthermore, marked differences in immune landscape were identified. Immunotherapy response analysis and drug sensitivity analysis further indicated that high-LYMS patients showed reduced benefit to immunotherapy and diminished efficacy of certain chemotherapy agents. CONCLUSION: Overall, this study confirmed that LYMS is an independent prognostic risk factor in GC patients. The LYMS demonstrates significant predictive ability for responses to immunotherapy, suggesting its potential to guide future immunotherapy interventions for GC patients.