Abstract P2-08-33: A novel seven-gene signature predicts prognosis in early-stage triple-negative breast cancer

Abstract

Abstract Purpose: Chemotherapy remains the only systemic treatment option for patients with triple-negative breast cancer (TNBC). However, due to the heterogeneity of TNBC, not all patients benefit from chemotherapy, especially those with early-stage disease. In order to improve prognostic assessment and reduce unnecessary adjuvant systemic therapy in these patients, we have developed a novel seven-gene signature. Experimental Design: With the ComBat method, we integrated the results from 150 transcriptome microarrays samples and 246 RNA-seq samples of early-stage TNBC patients, and identified mRNAs associated with recurrence-free survival (RFS) using Lasso-Cox model, We further analyzed these TNBC samples and compared them with 60 paired normal breast tissues (40 samples from RNA-seq and 20 samples from microarrays) to identify tumor-specific mRNAs. Twenty-one overlapped mRNAs of the RFS-associated mRNAs and the tumor-specific mRNAs are selected as candidate mRNAs. An additional 371 samples of frozen primary tumors were then collected from early-stage TNBC patients (mean follow-up of 45 months) and randomly divided into two sets: a training set (n = 186) and a validation (n = 185) set. Expression level of candidate mRNAs in these samples were measured using RT-qPCR assays, and a seven-gene signature was built through all subset regression in the training set. The prognostic and predictive accuracy of our signature was tested in the validation set and other public databases (GSE5327, GSE2034 and METABRIC). Results: Twenty-one candidate mRNAs were identified in early-stage TNBC patients, from which we developed a novel seven-gene signature (recurrence risk score [mRNA signature] = 1.108*TMEM101 - 0.213*KRT5 - 0.315*ACAN - 0.464*LCA5 + 0.446*RPP40 - 0.373*LAGE3 - 0.257*CDKL2). Patients in the training set were classified into high- or low-risk group based on our seven-gene signature and an optimum cut-off score derived from x-tile. The patients in high-risk group were more likely to suffer from recurrence (HR, 2.718; 95% confidence interval [CI], 1.928–3.726, P= 0.001), and a time-dependent receiver operating curve showed that the seven-gene mRNA signature had a better prognostic value than the clinicopathologic risk factors in both training set and validation set. The prognostic and predictive accuracy of the signature was also validated in the METABRIC and two other public GEO databases (GSE5327 and GSE2034). The time-dependent receiver operating curve showed that this signature had an area under the curve (AUC) of 0.742 (95% CI, 0.705-0.773) in METABRIC, 0.716 (95% CI, 0.682-0.739) and 0.723 (95% CI, 0.683-0.756) in GSE5327 and GSE2034 respectively. Conclusion: In this study, we developed a novel seven-gene signature which can provide additional prognostic information and may guidance in identifying early-stage TNBC patients eligible for adjuvant therapy or reduction of chemotherapy. To our knowledge, this is the first study investigating the prognostic potential of mRNA signature in early-stage triple-negative breast cancer. Our novel signature may provide an opportunity for de-escalating treatment in early-stage TNBC patients in the future. Citation Format: Ren Y, Jiang Y, Zuo W, Xu X, Jin X, Ma D, Shao Z. A novel seven-gene signature predicts prognosis in early-stage triple-negative breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P2-08-33.