Background
Gastric cancer (GC) is one of the most common malignancies in the human digestive tract. CD4+T cells can eliminate tumor cells directly through the mechanism of cytolysis, they can also indirectly attack tumor cells by regulating the tumor TME. A prognostic model of CD4+T cells is urgently needed to improve treatment strategies and explore the specifics of this interaction between CD4+T cells and gastric cancer cells.
Conclusion
The risk model not only provides a basis for better prognosis in GC patients, but also is a potential prognostic indicator to distinguish the molecular and immune characteristics of the tumor, and its response to immune checkpoint inhibitor (ICI) therapy and chemotherapy.
Methods
The detailed data of GC samples were downloaded from the Cancer Genome Atlas (TCGA), GSE66229, and GSE84437 datasets. CD4+ T cell-related genes were identified to construct a risk-score model by using the Cox regression method and validated with the Gene Expression Omnibus (GEO) dataset. In addition, postoperative pathological tissues of 139 gastric cancer patients were randomly selected for immunohistochemical staining, and their prognostic information were collected for external verification. Immune and molecular characteristics of these samples and their predictive efficacy in immunotherapy and chemotherapy were analysed.
Results
The training set and validation set had consistent results, with GC patients of high PROC and SERPINE1 expression having poorer prognosis. In order to improve their clinical application value, we constructed a risk scoring model and established a high-precision nomogram. Low-risk patients had a better overall survival (OS) than high-risk patients, consistent with the results from the GEO cohort. Furthermore, the risk-score model can predict infiltration of immune cells in the tumor microenvironment of GC, as well as the response of immunotherapy. Correlations between the abundance of immune cells with PROC and SERPINE1 genes were shown in the prognostic model according to the training cohort. Finally, sensitive drugs were identified for patients in different risk subgroup.