Development of triple-negative breast cancer radiosensitive gene signature and validation based on transcriptome analysis

Triple-negative breast cancer (TNBC) is a heterogeneous disease with highest loco-regional recurrence among breast cancer subtypes. Radiotherapy is indispensable for TNBC loco-regional control. However, intrinsic radiosensitivity differences exist in TNBC patients and RT is still prescribed mainly based on conventional clinicopathologic features of patients without considering the differences. The purpose of the present study is to develop and validate a TNBC radiosensitive gene signature (RSGS) and to guide therapeutic decisions. In this study, we compared transcriptome profiles of 12 locally recurrent TNBCs to 20 non-locally recurrent TNBCs treated with surgery radio-chemotherapy and developed a seven-gene RSGS and a simplified three-gene RSGS by using pathway analysis, univariate Cox proportional hazards regression model and rank-based linear algorithm. They were validated by using transcriptome profiles of 166 TNBC patients. Two gene signatures specifically identified a radiosensitive population that had an improved recurrence-free survival in patients treated with surgery radio-chemotherapy (Radiosensitive patients vs radioresistant patients, for seven-gene RSGS: P = 0.024, HR = 0.35, 95 %CI 0.14–0.87 and for three-gene RSGS: P = 0.035, HR = 0.38, 95 %CI 0.15–0.94). In contrast, there was no significant difference in outcome between predicted radiosensitive and radioresistant patients that treated with other treatment modality. RSGSs provide a useful tool for identification of radiosensitive/radioresistant TNBC patients and they could lead to a better selection of patients for RT protocols.