A prognostic model based on autophagy-and senescence-related genes for gastric cancer: implications for immunotherapy and personalized treatment.

Background: The process of human aging is accompanied by an increased susceptibility to various cancers, including gastric cancer. This heightened susceptibility is linked to the shared molecular characteristics between aging and tumorigenesis. Autophagy is considered a critical mediator connecting aging and cancer, exerting a dynamic regulatory effect in conjunction with cellular senescence during tumor progression. In this study, a combined analysis of autophagy- and senescence-related genes was employed to comprehensively capture tumor heterogeneity. Methods: The gene expression profiles and clinical data for GC samples were acquired from TCGA and GEO databases. Differentially expressed autophagy- and senescence-related genes (DEASRGs) were identified between tumor and normal tissues. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were carried out to provide insights into biological significance. A prognostic signature was established using univariate Cox and LASSO regression analyses. Furthermore, consensus clustering analyses and nomograms were employed for survival prediction. TME and drug sensitivity analyses were conducted to compare differences between the groups. To predict immunotherapy efficacy, the correlations between risk score and immune checkpoints, MSI, TMB, and TIDE scores were investigated. Results: A fourteen-gene prognostic signature with superior accuracy was constructed. GC patients were stratified into three distinct clusters, each exhibiting significant variations in their prognosis and immune microenvironments. Drug sensitivity analysis revealed that the low-risk group demonstrated greater responsiveness to several commonly used chemotherapeutic agents for gastric cancer, including oxaliplatin. TME analysis further indicated that the high-risk group exhibited increased immune cell infiltration, upregulated expression of ICs, and a higher stromal score, suggesting a greater capacity for immune evasion. In contrast, the low-risk group was characterized by a higher proportion of microsatellite instability-high (MSI-H) cases, an elevated TIDE score, and a greater TMB, indicating a higher likelihood of benefiting from immunotherapy. In addition, Single-cell sequencing demonstrated that TXNIP was expressed in epithelial cells. Cellular experiments preliminarily verified that TXNIP could promote the proliferation and migration of gastric cancer cells. Conclusion: This study presents a robust predictive model for GC prognosis using autophagy- and senescence-related genes, demonstrating its ability to predict immune infiltration, immunotherapy effectiveness, and guide personalized treatment.