Triple negative breast cancer (TNBC): BCL11A gene status in prediction of survival in patients treated with anthracycline-based chemotherapy.

Introduction

TNBC is heterogenous, aggressive, the prognosis is poor and patients cannot benefit from targeted treatment. Moreover, validated predictors for chemotherapy sensitivity are not available for TNBC. BCL2 protein was found as an independent predictor of outcome in basal-like TNBC treated with adjuvant anthracycline-based chemotherapy (Bouchalova et al., Tumor Biology 2015). BCL11A transcription factor has been recently described as an oncogene in TNBC with critical functions in stem and progenitor cells and amplified in up to 38% of basal-like breast cancer. BCL11A deletion caused a reduction in the number of mammary epithelial stem and progenitor cells (Khaled et al., Nature Communications 2015). The objective of this study was to determine whether BCL11A status predicts outcome in TNBC patients treated with adjuvant anthracyclinebased regimens.

Materials/methods

Fresh-frozen tumor tissues were collected from 148 patients with TNBC stage I-III (123 out of them were treated by anthracycline-based adjuvant chemotherapy). Genomes of these samples were profiled by Affymetrix SNP6.0 arrays. Arrays were normalized using Aroma and GISTIC 2.0 was used to identify copy number changes. Survival data were analyzed with software R.

Results and conclusions

Segments with BCL11A locus were amplified in 16 (11.6%), normal status was found in 99 (71.7%), and deletion in 23 (16.7%) patiens. Survival analysis revealed that TNBC patients with BCL11A deletion treated with anthracycline-based chemotherapy had worse outcome (breast cancer specific survival, BCSS, logrank p=0.017; relapse free survival, RFS, logrank p=0.021) than those with normal or amplified status. Multivariate analysis found BCL11A as an independent predictor of BCSS (logrank p=0.040, hazard ratio, HR 2.46, 95%CI 1.04-5.83; likelihood ratio test for the whole model p=0.0199), and RFS (logrank p=0.036, hazard ratio, HR 2.50, 95%CI 1.05-5.96, likelihood ratio test for the whole model p=0.058) in TNBC treated with anthracyclinebased adjuvant chemotherapy. The initial model included BCL11A status, age, tumor size, nodal status, and histopathological grade as variables. Conclusion: This is the first study showing BCL11A copy number status as independent predictor of outcome in TNBC treated with anthracycline-based chemotherapy. BCL11A copy number status could facilitate decision making on adjuvant therapy. The data should be validated by prospective study and underlying mechanisms could be revealed by further research. In patients with BCL11A deletion other types of adjuvant therapy should be considered.