Candidate methylation sites associated with endocrine therapy resistance in ER+/HER2- breast cancer

Abstract

Background Estrogen receptor (ER) positive breast cancer is often effectively treated with drugs that inhibit ER signaling, i.e., tamoxifen (TAM) and aromatase inhibitors (AIs). However, 30% of ER+ breast cancer patients develop resistance to therapy leading to tumour recurrence. Changes in the methylation profile have been implicated as one of the mechanisms through which therapy resistance develops. Therefore, we aimed to identify methylation loci associated with endocrine therapy resistance. Methods We used genome-wide DNA methylation profiles of primary ER+/HER2- tumours from The Cancer Genome Atlas in combination with curated data on survival and treatment to predict development of endocrine resistance. Association of individual DNA methylation markers with survival was assessed using Cox proportional hazards models in a cohort of ER+/HER2- tumours ( N = 552) and two sub-cohorts corresponding to the endocrine treatment (AI or TAM) that patients received ( N = 210 and N = 172, respectively). We also identified multivariable methylation signatures associated with survival using Cox proportional hazards models with elastic net regularization. Individual markers and multivariable signatures were compared with DNA methylation profiles generated in a time course experiment using the T47D ER+ breast cancer cell line treated with tamoxifen or deprived from estrogen. Results We identified 134, 5 and 1 CpGs for which DNA methylation is significantly associated with survival in the ER+/HER2-, TAM and AI cohorts respectively. Multi-locus signatures consisted of 203, 36 and 178 CpGs and showed a large overlap with the corresponding single-locus signatures. The methylation signatures were associated with survival independently of tumour stage, age, AI treatment, and luminal status. The single-locus signature for the TAM cohort was conserved among the loci that were differentially methylated in endocrine-resistant T47D cells. Similarly, multi-locus signatures for the ER+/HER2- and AI cohorts were conserved in endocrine-resistant T47D cells. Also at the gene set level, several sets related to endocrine therapy and resistance were enriched in both survival and T47D signatures. Conclusions We identified individual and multivariable DNA methylation markers associated with therapy resistance independently of luminal status. Our results suggest that these markers identified from primary tumours prior to endocrine treatment are associated with development of endocrine resistance.

SAM-DNMT3A, a strategy for induction of genome-wide DNA methylation, identifies DNA methylation as a vulnerability in ER-positive breast cancers

“…These findings could also have implications to ER-positive cancers that are resistant to ER inhibitors. Previous reports suggest that ER-positive breast cancers that are resistant to endocrine therapy have higher levels of DNA methylation 31, 32 . Based on our findings showing ER-positive breast cancers require low levels of DNA methylation suggests that ER dependency and DNA methylation are directly connected and that it is possible that escapers of this process are no longer sensitive to endocrine therapy or to DNA methylation induction.…”

Novel Endocrine Therapeutic Opportunities for Estrogen Receptor-Positive Ovarian Cancer—What Can We Learn from Breast Cancer?

“…Endocrine resistance has previously been shown to be associated with epigenetic alterations in breast cancer, including DNA methylation, chromatin accessibility, histone modifications, and the binding of different transcription factors, due to their effect on gene expression [47,48]. Differential DNA methylation has been implicated in endocrineresistant tumors [49]. For example, in breast cancer, different methylation profiles were found between tamoxifen-sensitive and tamoxifen-resistant cell lines [50].…”

Epigenetic Factors as Etiological Agents, Diagnostic Markers, and Therapeutic Targets for Luminal Breast Cancer

“…In particular, circulating tumor DNA can be used to quantify DNA methylation and detect ERα-encoding gene mutations in patients with breast cancer [87,88]. Genome-wide DNA methylation profiling is a powerful approach that can comprehensively characterize the function of CpG islands in cancer recurrence and poor survival outcomes of patients with ER+ breast cancer [89]. DNA hypermethylation at the enhancer sites of ERE-containing genes regulates ER signaling.…”