Construction of prognostic microRNA signature for human invasive breast cancer by integrated analysis

BACKGROUND: Despite the advances in early detection and treatment methods, breast cancer still has a high mortality rate, even in those patients predicted to have a good prognosis. The purpose of this study is to identify a microRNA signature that could better predict prognosis in breast cancer and add new insights to the current classification criteria. MATERIALS AND METHODS: We downloaded microRNA sequencing data along with corresponding clinicopathological data from The Cancer Genome Atlas (TCGA). Of 1,098 breast cancer patients identified, 253 patients with fully characterized microRNA profiles were selected for analysis. A three-microRNA signature was generated in the training set. Subsequently, the performance of the signature was confirmed in a validation set. After construction of the signature, we conducted additional experiments, including flow cytometry and the Cell Counting Kit-8 assay, to illustrate the correlation of this microRNA signature with breast cancer cell cycle, apoptosis, and proliferation. RESULTS: Three microRNAs (hsa-mir-31, hsa-mir-16-2, and hsa-mir-484) were identified to be significantly and independently correlated with patient prognosis, and performed with good stability. Our results suggest that higher expression of hsa-mir-484 indicated worse prognosis, while higher expression of hsa-mir-31 and hsa-mir-16-2 indicated better prognosis. Moreover, additional experiments confirmed that this microRNA signature was related to breast cancer cell cycle and proliferation. CONCLUSION: Our results indicate a three-microRNA signature that can accurately predict the prognosis of breast cancer, especially in basal-like and hormone receptor-positive breast cancer subtypes. We recommend more aggressive therapy and more frequent follow-up for high-risk groups.