A novel immune-related prognostic index for predicting breast cancer overall survival

To find immune-related genes with prognostic value in breast cancer, and construct a prognostic risk assessment model to make a more accurate assessment. Moreover, looking for potential immune markers for breast cancer immunotherapy.

The breast cancer (BC) data were retrieved from The Cancer Genome Atlas (TCGA) database as a training set. Through the Weighted gene co-expression network analysis (WGCNA), Kaplan-Meier (KM) analysis, lasso regression analysis and stepwise backward Cox regression analysis, screening for prognosis-related immune genes, a prognostic index was built, and external validation with two data sets of Gene Expression Omnibus (GEO) database was performed. Transcription factor (TF) regulatory network was constructed to identify key transcription factors that regulate prognostic immune genes. Gene set enrichment analysis (GSEA) was used to explore the signal pathways differences between high and low-risk groups, estimate package and TIMER database were used to evaluate the relationship between risk score and tumor immune microenvironment.

We obtained 10 prognosis-related immune genes, and the index showed accurate prognostic value. We also identified 7 prognostic transcription factors. Multiple signaling pathways that inhibit tumor progression were enriched in the low-risk group, and risk score was significantly negatively related to the degree of immune infiltration and the expression level of immune checkpoint genes.

We successfully constructed an independent prognostic index, which not only has a stronger predictive ability than the tumor pathological stage, but also can reflect the immune infiltration of breast cancer patients.