A two-gene epigenetic signature for the prediction of response to neoadjuvant chemotherapy in triple-negative breast cancer patients

Abstract

Background: pathological complete response (pCR) after neoadjuvant chemotherapy (NAC) in triple-negative breast cancer (TNBC) varies between 30 and 40% approximately. To provide further insight into the prediction of pCR, we evaluated the role of an epigenetic methylation-based signature. Methods: epigenetic assessment of DNA extracted from biopsy archived samples previous to NAC from TNBC patients was performed. Patients included were categorized according to previous response to NAC in responder (pCR or residual cancer burden, RCB = 0) or non-responder (non-pCR or RCB > 0) patients. A methyloma study was performed in a discovery cohort by the Infinium HumanMethylation450 BeadChip (450K array) from Illumina. The epigenetic silencing of those methylated genes in the discovery cohort were validated by bisulfite pyrosequencing (PyroMark Q96 System version 2.0.6, Qiagen) and qRT-PCR in an independent cohort of TN patients and in TN cell lines. Results: twenty-four and 30 patients were included in the discovery and validation cohorts, respectively. In the discovery cohort, nine genes were differentially methylated: six presented higher methylation in non-responder patients (LOC641519, LEF1, HOXA5, EVC2, TLX3, CDKL2) and three greater methylation in responder patients (FERD3L, CHL1, and TRIP10). After validation, a two-gene (FER3L and TRIP10) epigenetic score predicted RCB = 0 with an area under the ROC curve (AUC) = 0.905 (95% CI = 0.805-1.000). Patients with a positive epigenetic two-gene score showed 78.6% RCB = 0 versus only 10.7% RCB = 0 if signature were negative. Conclusions: these results suggest that pCR in TNBC could be accurately predicted with an epigenetic signature of FERD3L and TRIP10 genes. Further prospective validation of these findings is warranted.