Identification of three immune subtypes characterized by distinct tumor immune microenvironment and therapeutic response in stomach adenocarcinoma

Abstract

Background: In primary stomach adenocarcinoma (STAD), the tumor immune microenvironment (TIME) is important for cancer occurrence and progression; however, its clinical significance remains unclear. This study investigated the association between patient survival, TIME, and therapeutic response to STAD.

Methods: Gene expression profiles of STAD cases were collected from the Cancer Genome Atlas (TCGA) database and Gene Expression Omnibus. Molecular subtypes were explored with consistent clustering methods according to 119 immune signatures and the infiltrating scores of 22 immune cells using the Multi-Omics Immuno-Oncology Biological Research algorithm. We determined IFNγ scores and immune cytolytic activity (CYT) scores on the basis of corresponding gene signatures via single-sample Gene Set Enrichment Analysis. Comparisons of survival, TIME, 10 immunity-related oncogenic pathways, immune checkpoint expression, and therapeutic response were conducted among the three subtypes. We further applied linear discriminant analysis to construct a characteristic index to classify the subtypes, and the Pearson correlation coefficient for the relationship between the index and immune checkpoint genes. Weighted Correlation Network Analysis (WGCNA) was used to mine the associated modules and specific genes.

Results: We collected gene expression profiles from 352 STAD cases in the TCGA database, 300 in GSE62254, and 344 in GSE84437. Three STAD subtypes (IS1-IS3) were established according to the TIME signatures. The IS3 subtype had the highest immune score and the best prognosis, as well as markedly increased immune T-cell CYT, Th1/IFNγ scores, and immune checkpoint gene expression, compared to the other two subtypes. It was highly similar to the PD-1 response group in the previous study samples of GSE91061. The established TIME classification index performed well in classifying subtypes and was directly proportional to immune checkpoint-related gene expression levels. WGCNA explored 6 modules and 14 genes, namely DYSF, MAN1C1, HTRA3, EMCN, RFLNB, KANK3, MAGEH1, CD93, PCAT19, FUT11, BMP1, FOSB, DCHS1, and TCF3, which were associated with the established TIME classification index and STAD patient prognosis.

Conclusion: TIME phenotypes of STAD patients could be divided into three different molecular subtypes, which displayed different prognoses, immune features, and therapeutic responses. Our results shed new light on predicting patient outcomes and the discovery of new anti-STAD therapeutic strategies according to the TIME.

Keywords: Bioinformatic analysis; Immune microenvironment; Stomach adenocarcinoma.