Identification of a DNA Methylation-Based Prognostic Signature for Patients with Triple-Negative Breast Cancer

Background: Aberrant DNA methylation is an important biological regulatory mechanism in malignant tumors. However, it remains underutilized for establishing prognostic models for triple-negative breast cancer (TNBC). Material/Methods: Methylation data and expression data downloaded from The Cancer Genome Atlas (TCGA) were used to identify differentially methylated sites (DMSs). The prognosis-related DMSs were selected by univariate Cox regression analysis. Functional enrichment was analyzed using DAVID. A protein-protein interaction (PPI) network was constructed using STRING. Finally, a methylation-based prognostic signature was constructed using LASSO method and further validated in 2 validation cohorts. Results: Firstly, we identified 743 DMSs corresponding to 332 genes, including 357 hypermethylated sites and 386 hypomethylated sites. Furthermore, we selected 103 prognosis-related DMSs by univariate Cox regression. Using a LASSO algorithm, we established a 5-DMSs prognostic signature in TCGA-TNBC cohort, which could classify TNBC patients with significant survival difference (log-rank p=4.97E-03). Patients in the high-risk group had shorter overall survival than patients in the low-risk group. The excellent performance was validated in GSE78754 (HR=2.42, 95%CI: 1.27-4.59, log-rank P=0.0055). Moreover, for disease-free survival, the prognostic performance was verified in GSE141441 (HR=2.09, 95%CI: 1.28-3.44, log-rank P=0.0027). Multivariate Cox regression analysis indicated that the 5-DMSs signature could serve as an independent risk factor. Conclusions: We constructed a 5-DMSs signature with excellent performance for the prediction of disease-free survival and overall survival, providing a guide for clinicians in directing personalized therapeutic regimen selection of TNBC patients.