Identification and verification of immune-related gene prognostic signature based on ssGSEA for breast cancer

INTRODUCTION: Breast cancer (BC) is the most common cancer in women worldwide and has a high mortality rate. The fact that the tumor microenvironment affects clinical outcomes of all types of cancers underlines the involvement of various immune-related genes (IRGs). Therefore, this study aimed to establish an IRGs-based signature for the prognosis of BC patients. MATERIAL AND METHODS: In this study, 12 immune cell infiltrating degrees in 1,102 BC cases from The Cancer Genome Atlas (TCGA) database were assessed, and RNA-sequencing (RNA-seq) data of these samples were analyzed by single-sample gene set enrichment analysis (ssGSEA). Based on the results, high, low, and middle immune infiltrating clusters were constructed. A total of 138 overlapped differentially expressed genes (DEGs) were identified in the high and low infiltrating clusters, as well as in normal and BC samples. Univariate Cox regression and LASSO analyses were also performed. Furthermore, GSEA suggested some highly enriched pathways in the different immune infiltrating clusters, leading to a better understanding of potential mechanisms of immune infiltration in BC. RESULTS: Finally, 19 immune-related genes were identified that could be utilized as a potential prognostic biomarker for BC. Kaplan-Meier plot and ROC curve, univariate as well as multivariate Cox analyses were carried out, which suggested that the 19-IRG-based signature is a significant prognosis factor independent of clinical features. Based on the analysis of protein-protein interactions (PPI), the three hub genes were identified. CONCLUSIONS: These results provide a new method to predict the prognosis and survival of BC based on the three genes’ features.