Identification and evaluation of a risk model predicting the prognosis of breast cancer based on characteristic signatures

Background: Breast cancer (BC) is one of the most common fatal cancers in women. Identifying new biomarkers is thus of great significance for the diagnosis and prognosis of BC.

Methods: In this study, 1,030 BC cases from The Cancer Genome Atlas (TCGA) were obtained for differential expression analysis and Short Time-series Expression Miner (STEM) analysis to identify characteristic BC development genes, which were further divided into upregulated and downregulated genes. Two predictive prognosis models were both defined by Least Absolute Shrinkage and Selection Operator (LASSO). Survival analysis and receiver operating characteristic (ROC) curve analysis were used to determine the diagnostic and prognostic capabilities of the two gene set model scores, respectively.

Results: Our findings from this study suggested that both the unfavorable (BC1) and favorable (BC2) gene sets are reliable biomarkers for the diagnosis and prognosis of BC, although the BC1 model presents better diagnostic and prognostic value. Associations between the models and M2 macrophages and sensitivity to Bortezomib were also found, indicating that unfavorable BC genes are significantly involved in the tumor immune microenvironment.

Conclusions: We successfully established one predictive prognosis model (BC1) based on characteristic gene sets of BC to diagnose and predict the survival time of BC patients using a cluster of 12 differentially expressed genes (DEGs).