Comprehensive analysis of a prognostic immune-related gene signature in esophageal carcinoma predicts disease prognosis and treatment response

Abstract Background: Esophageal cancer (EC) is a lethal disease with dismal survival rates. The tumor immune microenvironment plays a key contributor to the poor prognosis of EC. Here, we aimed to identify an immune-associated gene biomarker signature in EC and explore potential therapeutic strategies. Methods: We combined the publicly available TCGA dataset with qRT-PCR results and constructed an immune-related 3-genemodel related to the survival of patients with EC. Kaplan-Meier survival analysis and time-dependent receiver operating curve (ROC) were then conducted to determine the predictive performance of the model, followed by clinical validation and nomogram establishment using Cox regression analysis. The underlying molecular mechanisms were explored with gene set enrichment analysis and tumor mutational burden (TMB). Finally, the immune microenvironment landscape was described with ESTIMATE and CIBERSORT, and the TIDE algorithm and the pRRophetic package were used to evaluated the model’s ability to predict immunotherapeutic and chemotherapeutic efficacy for different EC patient sub-groups. Results: We identified a 3-gene signature that significantly associated with poor overall survival in the TCGA and clinical data sets. And the EC samples were divided into the high- and low-risk groups depending on their risk scores. Then, a nomogram was established which included the gene model, clinical stage and positive lymph node metastasis for potential clinical translation. There were significant differences in the tumor immune microenvironment, TMB and metabolic pathways between the two groups. Cell type composition analysis found that M0 Macrophages, CD8 T cell and other immune cells not only have increased abundance in the tumor, but also can be reflected by the model. Additionally, while low-risk patients were found to benefit more from immunotherapy, high-risk patients with a worse prognosis could benefit more from the application of agents such as Bryostatin.1and FH535. Conclusion: We report a novel gene model which reveals the driving immune-related factors in EC. These factors may act as immunotherapeutic biomarkers and potential therapeutic targets for novel EC treatments.