Prognostic analysis of patients with breast cancer based on tumor mutational burden and DNA damage repair genes

BACKGROUND: Breast cancer has a high tumor-specific death rate and poor prognosis. In this study, we aimed to provide a basis for the prognostic risk in patients with breast cancer using significant gene sets selected by analyzing tumor mutational burden (TMB) and DNA damage repair (DDR). METHODS: Breast cancer genomic and transcriptomic data were obtained from The Cancer Genome Atlas (TCGA). Breast cancer samples were dichotomized into high- and low-TMB groups according to TMB values. Differentially expressed DDR genes between high- and low-TMB groups were incorporated into univariate and multivariate cox regression model to build prognosis model. Performance of the prognosis model was validated in an independently new GEO dataset and evaluated by time-dependent ROC curves. RESULTS: Between high- and low-TMB groups, there were 6,424 differentially expressed genes, including 67 DDR genes. Ten genes associated with prognosis were selected by univariate cox regression analysis, among which seven genes constituted a panel to predict breast cancer prognosis. The seven-gene prognostic model, as well as the gene copy numbers are closely associated with tumor-infiltrating immune cells. CONCLUSION: We established a seven-gene prognostic model comprising MDC1, PARP3, PSMB1, PSMB9, PSMD2, PSMD7, and PSMD14 genes, which provides a basis for further exploration of a population-based prediction of prognosis and immunotherapy response in patients with breast cancer.