Plasma microRNA signature as a potential early detection biomarker in breast cancer.

Abstract

e13780 Background: The 5-year survival rate for women with stage I breast cancer (BC) is 99%, falling to 27% if the disease is diagnosed in stage IV. This demonstrates the importance of early detection for the effectiveness of BC treatment. Among the new diagnostic strategies, the use of liquid biopsies is emerging as a promising alternative for early cancer detection and patient follow-up. miRNAs are small non-coding RNA molecules useful as biomarkers, since they can be found in body fluids. Oncoliq’s approach focuses on the relative quantification of a panel of circulating miRNA biomarkers for the early detection of BC using predictive models built with machine learning techniques. Methods: The aim of this work was to identify a highly sensitive miRNA combination based on liquid biopsies together with predictive models, to generate an assay for BC early detection which we named Oncoliq breast. Human plasma samples (n=239) from BC patients and healthy donors (HD) were divided into two cohorts: exploratory (n=66) and validation (n=173). The exploratory cohort was used for the identification of miRNAs using expression microarrays and miRNA-seq. The validation cohort was used for the analytical validation of potential candidates by RT-qPCR. Results: Using ROC curve analysis, 11 miRNAs were identified as potential biomarkers. By RT-qPCR, we validated the expression of the 11 miRNAs. All of them were significantly increased in BC patient samples (n=100) compared to HD (n=73), in accordance with expression microarrays and miRNA-seq results. Using machine learning we established 7 predictive models that combine these miRNAs. After running all the predictive models using logistic regression, the metrics associated with each one were calculated. From these, model 7 showed a sensitivity of 90%, a specificity of 73% and an accuracy of 83%. Conclusions: MiRNAs show satisfactory advantages as biomarkers as they can circulate in body fluids, such as blood and urine. In the presence of diseases such as BC, the level of some specific miRNAs might be altered and this change can be harnessed for detection, diagnosis and prognosis. Here, we generated a panel of 11 miRNAs that showed increased expression in plasma samples from women diagnosed with BC. Oncoliq breast is a promising low cost and highly sensitive novel liquid biopsy test based on circulating miRNAs for BC screening.