Abstract P2-11-25: A 10-gene signature to predict the prognosis of early-stage triple-negative breast cancer

Background: Triple-negative breast cancer (TNBC) is highly heterogeneous cancer and the most challenging subtype of breast cancer. And its prognosis is poor compared to the other subtypes. Unlike luminal type cancers, there is no valid biomarker to predict the prognosis of patients with early TNBC. To establish an elaborate therapeutic strategy for TNBC, biomarkers that accurately predict the prognosis and response to treatment are needed. Method: 184 patients with early stage TNBC (training cohort, n = 76; validation cohort, n = 108) were enrolled. Median Follow-up period was 51.5 months (range: 4.6-230.8) for training cohort and 58.3 months (range: 6.6-99.8) for validation cohort. Of the patients in training cohort, 13 patients had recurrence or metastasis. Of the patients in validation cohort, 23 patients had recurrence or metastasis. Using a HiSeq sequencer, RNA sequencing was conducted to analyze the gene expression profiles of tumor samples from TNBC patients. Gene signature was analyzed by combination of DEGs which found in gene expression profiles. Cross validation and meta-analysis were conducted as pre-validation. Meta-analysis was conducted using CBS probePINGS. To compare gene signature and other methods, PAM 50 call and TCR diversity analysis were investigated. Statistical analyses were conducted using R language (v.3.4.3). Result: To predict prognosis of recurrence or metastasis for TNBC patients, we identified the 10-gene signature (DGKH, GADD45B, KLF7, LYST, NR6A1, PYCARD, ROBO1, SLC22A20P, SLC24A3 and SLC45A4) that stratified patients with TNBC by risk score (sensitivity = 92.31%; specificity = 92.06%; accuracy = 92.11%) and validated in a cohort of separate institutions. Meta-analysis supported the biological relevance of the 10-gene signature to well-known driving pathways in TNBC. When compared with other potential biomarkers like PAM 50 call and T-cell receptor β diversity, the 10-gene signature was the only independent factor that can predict prognosis for invasive disease-free survival in multivariate analysis. Conclusion: Our novel findings can contribute to solving the diagnostic challenges in TNBC and the 10-gene signature may serve as a novel biomarker for risk-based patient care. Citation Format: Chang Min Kim, Kyong Hwa Park, Yun Suk Yu, Ju Won Kim, Jin Young Park, Jeong Eon Lee, Sung Hoon Sim, Bo Kyoung Seo, Jin Kyeoung Kim, Eun Sook Lee, Yeon Hee Park, Sun-Young Kong. A 10-gene signature to predict the prognosis of early-stage triple-negative breast cancer [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P2-11-25.