Training and experimental validation a novel anoikis- and epithelial‒mesenchymal transition-related signature for evaluating prognosis and predicting immunotherapy efficacy in gastric cancer.

Anoikis resistance and improper activation of epithelial‒mesenchymal transition (EMT) are critical factors in tumor metastasis and progression. Despite their interaction, the combined impact of anoikis and EMT on prognosis and immunotherapy in gastric cancer remains underexplored. In this study, we identified 354 anoikis- and EMT-related genes (AERGs) through Venn analysis and performed unsupervised clustering to classify gastric cancer patients into two molecular clusters: A and B. Molecular cluster A showed poor prognosis and an immunosuppressive tumor microenvironment, suggesting a "cold tumor" phenotype. Then, a novel AERG-related prognostic model comprising CD24, CRYAB, MMP11, MUC4, PRKAA2, SERPINE1, SKP2, and TP53 was constructed and validated, accurately predicting the 1-, 3-, and 5-year survival rates of gastric cancer patients. Multivariate analysis revealed that the AERG-related risk score was an independent prognostic factor (hazard ratio = 1.651, 95% confidence interval = 1.429-1.907, <0.001). Further studies demonstrated that, compared to the high-risk group, the low-risk group exhibited higher CD8 T cell infiltration, tumor mutational burden, immunophenoscores, and lower tumor immune dysfunction and exclusion scores, indicating potential sensitivity to immunotherapy. RT‒qPCR and immunohistochemical staining validated the expression levels of the model's molecular markers. Overall, our AERG-related model shows promise for predicting outcomes and guiding the selection of tailored and precise therapies for gastric cancer patients.