A 3-miRNA signature for noninvasive breast cancer detection

As a key component of epigenetics, microRNAs (miRNAs) have provided promising insights into several aspects of Breast Cancer (BC). We have analyzed 2 BC tissue microarray datasets (GSE26659 and GSE40525), as well as 2 serum datasets (GSE106817 and GSE113486). The results were then intersected to identify commonly dysregulated miRNAs in the tissue and serum of BC patients. RNA-seq analysis was then applied to The Cancer Genome Atlas (TCGA) data. Briefly, 79 dysregulated miRNAs were identified in the tissue and serum of patients with BC of which 3 significantly dysregulated and previously unstudied miRNAs, let-7e-5p, miR-151a-5p and miR-887-3p, were chosen for quantification in the serum of cancer patients by RT-PCR followed by evaluation of their diagnostic and prognostic features. RT-PCR analysis revealed overexpression of let-7e-5p (logFC = 2.01, P < .05) and miR-151a-5p (logFC = 1.48, P < .05) whereas miR-887-3p was downregulated (logFC = 0.62, P < .05) similar to microarray and RNA-seq data analysis. Based on regression analysis, a 3 miRNA-signature biomarker was proposed which had better diagnostic ability (AUC = 84.17%) compared to the ability of these miRNAs when assessed individually. Moreover, enrichment analysis revealed these miRNAs mediate vital cellular processes and biological functions that influence the development of cancer. Similarly, significant prognostic clinical characteristics were observed for these miRNAs. Overall, we have identified and validated a novel and proficient signature biomarker in serum of BC patients consisting of 3 miRNAs.