Abstract P4-07-07: A MAPK microRNA signature significantly associated with poor outcome and predictive of response to tamoxifen therapy in ER+ breast cancer

Abstract

Abstract Background: Hyperactivation of ERK1/2 MAPK (hMAPK) leads to loss of estrogen receptor (ER) expression and poor outcome in breast cancer. Recent evidence suggests that microRNAs (miRNAs) play important regulatory roles and serve as biomarkers of disease. We previously reported a miRNA signature indicative of hyperactivation of MAPK signaling in breast cancer and its associations with pathological features of breast cancer and breast cancer clinical outcomes. This hMAPK-miRNA signature identified ER+ breast cancers with reduced recurrence-free and overall survival. Here we report on a leave-one-out analysis of this hMAPK-miRNA signature to narrow down and identify those miRNAs that may be critically important to poor clinical outcomes associated with hyperactive MAPK signaling among patients with ER+ breast cancer. Methods: We performed a leave-one-out analysis of the full hMAPK-miRNA signature, wherein we determined the impact that removal of individual members of the signature had on the ability of the hMAPK-miRNA signature to predict poor disease survival in patients from the METABRIC breast cancer dataset with ER+ cancers. We identified a subsignature of 21 miRNAs that is very significantly associated with poor clinical outcome in patients with ER+ breast cancer from a large patient cohort, according to multivariate Cox proportional hazard analysis. Summary of Results: Of 57 miRNAs in our original hMAPK-miRNA signature, 21 were retained following leave-one-out analysis. "High-hMAPK" status as described by this signature significantly correlated with adverse pathological characteristics of breast cancer, including ER-negativity, increased tumor grade, and enrichment for Basal and HER2 subtypes. Kaplan-Meier survival analysis of primary breast cancers from the METABRIC dataset indicate that the 21 miRNA subsignature improves upon the prognostic capability of the overall signature in this large patient cohort, and particularly in providing prognostic capability in breast cancers of the luminal-A and luminal-B molecular subtypes. Furthermore, multivariate Cox proportional hazards analysis suggests that classification of breast cancers as "high-hMAPK" based on this leave-one-out miRNA signature is an independent risk factor for poor disease outcome, and is a more significant indicator of poor disease status than hormone receptor status, tumor grade, molecular subtype, and HER2 status. Patients with ER-positive breast cancer who were treated with hormone therapy and classified as "high-hMAPK" by this 21 miRNA signature displayed significantly poorer disease survival compared to patients classified as "low hMAPK". Kaplan-Meier survival analysis and multivariate analysis of independent breast cancer cohorts with miRNA expression data confirm these observations. Conclusions: We report a subset of 21 of the hMAPK-miRNAs that are important prognostic factors in ER-positive breast cancer, and that may have predictive value in estimating whether an ER-positive breast cancer may be resistant to hormone therapy. Additionally, these 21 miRNAs may be potential effectors of MAPK signaling, and could serve as novel biomarkers or therapeutic targets in breast cancer. Citation Format: Philip Miller, Jennifer Clarke, Dorraya El-Ashry. A MAPK microRNA signature significantly associated with poor outcome and predictive of response to tamoxifen therapy in ER+ breast cancer [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P4-07-07.