Background and Purpose:
Radiotherapy following breast cancer (BrC) surgery is highly effective, with an improvement in local control translating into a survival benefit1. However, local recurrence and distant failure remain significant clinical problems, poorly predicted by conventional clinico-pathological markers. Previous work carried out by our group has identified specific microRNAs (miRNAs), 139-5p and 222, as key regulators of the response to radiation in BrC cells in vitro. We therefore hypothesized that these miRNAs and their target genes were promising biomarkers in four large cohorts of BrC patients.
Methods and Materials:
Four publically-available microarray mRNA/miRNA expression datasets (METABRIC2, Oxford-GSE222203, Lund-GSE318634 and NKI-2955) totalling > 1,700 early-stage BrC patients, treated with adjuvant radiotherapy, each with >10 years of follow-up, were included in the analyses. More detailed clinic-pathological and radiotherapy information was obtained for the Oxford cohort. The expression of 4 putative miRNA target biomarkers – TOP2A, POLQ, RAD54L, and NEK1 – were correlated to outcome and to standard clinico-pathological variables using non-parametric tests and Kaplan-Meier/Cox Regression as appropriate.
Results:
Low expression of miR-139-5p and miR-222 and/or high expression of their putative targets TOP2A, POLQ, RAD54L, and NEK1 were correlated with poor prognosis in radiotherapy-treated patients in multiple datasets. Target genes ‘outperformed’ their miRNAs as biomarkers. Importantly, in the dataset derived from a randomised radiotherapy clinical trial, RAD54L was significantly prognostic specifically in patients that received radiotherapy but not in matched patients who did not receive radiotherapy, suggesting a true predictive role. Target gene expression was also significantly correlated with clinico-pathological variables associated with poor outcome, including grade, size, stage, and intrinsic subtype.
Conclusion:
Targets of miRNAs involved in response to radiotherapy are promising novel biomarkers of response to adjuvant radiotherapy for breast cancer. We next aim to validate these by immunohistochemistry in independent clinical cohorts for which tissue microarrays are available.