Despite gaps in current knowledge, investigators are already working toward the development of novel treatment strategies to combat the effects of ESR1 mutations.
Although endocrine therapies that block the production or subsequent effects of estrogen have been a bedrock of the treatment for hormonally driven breast cancer for more than 40 years, resistance poses a major threat to their efficacy and has become a focus of ongoing research.
Mutations in the ESR1 gene that encodes estrogen receptor alpha (ERα) have emerged as a central mechanism of resistance thought to be acquired because of the selective pressures of endocrine therapy, particularly with aromatase inhibitors (AIs).
As a result of the technological advances that have yielded the development of liquid biopsies, monitoring the presence of ESR1 mutations in circulating tumor DNA (ctDNA) has the potential to change practice by enabling the early detection of tumor progression and guiding therapeutic decisions.
Despite gaps in current knowledge, investigators are already working toward the development of novel treatment strategies to combat the effects of the mutations and wrestle back control of ER-positive tumor growth.As the cellular orchestrator of the effects of the estrogen hormones, ERs play an important role in a plethora of processes, including its most renowned functions in reproduction and sexual development and in modulating the immune system.1,2Estrogens are steroid hormones that pass freely across the cell membrane and do not need to bind to a membrane-bound receptor to transmit their signal into the cell. Although they can be found in the membranes, ERs predominantly reside in the nucleus, where, after binding to an estrogen ligand, they can act directly as a transcription factor.
An ER molecule is made up of a number of domains; at either end is a transcriptional activation domain, one activated independently of ligand binding (AF-1) and the other ligand-dependent (AF-2). These flank the ligand-binding domain (LBD) and the DNA-binding domain (Figure A).2
Following ligand binding at the LBD, 2 ER molecules pair up and are phosphorylated. This allows the estrogen-ER complex to interact with various corepressors or coactivators, which determine its activation state.
Activated ERs subsequently bind to specific sites within the nuclear DNA, known as estrogen response elements (EREs), found in the promoter region of target genes. In this way, the ER both directly (as a transcription factor) and indirectly (through activation of other transcription factors) activates sets of genes that regulate the proliferation, differentiation, and survival of cells in which it is expressed (Figure B).2-5ER signaling has been shown to be indispensable to mammary gland development,6,7 although the precise mechanisms through which it regulates the proliferation and differentiation of human breast cells is unclear. A relatively small number of cells in the normal mammary gland express ERs.8
There are 2 types of ERs: ERα and ER beta (ERβ). Increased expression of ERα is observed in breast cancers and can be seen from the earliest stages of tumorigenesis,9 suggesting that dysregulated ER signaling plays an important role in breast tumor formation. ERβ mediates many of estrogen’s effects on breast tissue.10
Expression of ERα is observed in around 70% of breast cancers and, as such, ER-positive tumors make up the largest subgroup of breast cancers.11,12 The proportion of patients with ER positivity increases with age; in postmenopausal women, breast cancers are overwhelmingly ER-positive.13
Efforts to treat hormonally driven breast cancers have led to endocrine therapies that are designed to either prevent the production of estrogen or block its effects on the ER. There are 3 classes of endocrine therapy: selective ER modulators (SERMs), selective ER downregulators (SERDs), and AIs.
SERMs are nonsteroidal compounds that bind to the ER and can act as either an agonist or antagonist depending on a number of tissue-specific and other factors. This class of agents includes tamoxifen, which in 1977 became the first FDA-approved antihormonal therapy for breast cancer. As a result of its agonist effects in certain tissues, tamoxifen is associated with adverse events (AEs) with longterm use, such as an increased risk of endometrial cancer and thrombotic events.14 Several novel SERMs have been developed with reduced agonist activity, including raloxifene, which has significantly less of an impact on endometrial cells in preclinical models.14
SERDs are pure antiestrogens, with a high affinity for the ER and none of the agonist activities of SERMs. Fulvestrant (Faslodex) is the only SERD currently approved by the FDA for breast cancer settings.
Although both SERMs and SERDs bind to the ER, their downstream effects are different. SERDs trigger rapid degradation of the ER and interfere with dimerization and nuclear localization, among other effects.
The AIs have a different mechanism of action. These agents block the production of estrogen via the aromatase enzyme, thereby reducing circulating levels of estrogen and preventing activation of the ER.14,15 The FDA has approved 3 drugs in this class: anastrozole, letrozole, and exemestane.
Although endocrine therapy has become the backbone of treatment for ER-positive metastatic breast cancer, approximately 40% of patients do not respond and those who do invariably develop disease progression after a period of response. Intrinsic and acquired resistance poses a major challenge to the control of hormonally driven cancers.16Research studies into the mechanisms of resistance to endocrine therapy have uncovered a number of ways in which tumors essentially can become estrogen independent. Recently, significant buzz has surrounded the discovery of mutations in the gene that encodes the ER.ESR1 mutations were first noted in patients with breast cancer in the 1990s.17-19 In many subsequent studies, however, the aberrations were barely detectable. In the genome-wide breast cancer sequencing study performed by The Cancer Genome Atlas (TCGA), the prevalence was low,20 which led researchers to discount the mutations as having little or no relevance to breast cancer.
In 2013, findings from a series of next-generation sequencing (NGS) studies revealed that ESR1 mutations were significantly more prevalent in metastatic disease, particularly in patients who had received treatment with endocrine therapy.21-26 In the past several years, the use of circulating tumor DNA (ctDNA) testing in plasma samples from patients with metastatic disease has revealed a prevalence of up to 40% (Table 1).21-25,27-32 The MammaSeq assay, a breast cancer—specific NGS panel developed at the University of Pittsburgh Medical Center in Pennsylvania, detected ESR1 mutations in 21% of solid tumor and cfDNA samples from patients with advanced breast cancer.32
Although technological advances in sequencing have determined that ESR1 mutations are likely present in primary tumors at a slightly higher rate than previously thought (at an estimated range of 2.5%-7%), it is clear that the mutation rate in this gene is dramatically higher in metastatic tumors.25,33,34
Approximately a dozen different ESR1 mutations have been described to date, most located within the LBD. They are gain-of-function mutations that lead to ligand-independent, constitutive activation of the ER. Five mutations—D538G, E380Q, D537S, D537N, and D537C—represent the vast majority of cases. Up to 40% of patients have more than 1 ESR1 mutation.29,31,34
The precise mechanism by which ESR1 mutations are enriched in patients with metastatic breast cancer is unknown; however, studies suggest that they could be present subclonally in primary tumors at low levels and become enriched by the use of ER-targeted endocrine therapies.
Preclinical functional models have suggested that ESR1 mutations may be a specific mechanism of resistance to AIs, rather than a broad mechanism of resistance to all endocrine therapies, but the details are still being teased out. Cell lines transfected with ESR1 mutants were less sensitive to AIs but still responsive to treatment with tamoxifen and fulvestrant, although this sensitivity was dose-dependent and still impaired relative to wild-type ESR1 cell lines. This suggests that these mutations confer relative resistance to SERMs and SERDs, although the underlying mechanism has not yet been elucidated.23,24,25
Recently, investigators uncovered another possible effect of ESR1 mutations in cancer cells. In a study published in Cancer Cell, RNA sequencing was performed on ESR1 mutant and wild-type—expressing breast cancer cell lines. The investigators aimed to assess differences in the global transcriptome and in the cistrome, the genome-wide ERα binding sites. They found that ESR1-mutant cells had a unique transcriptional network and a very different cistrome.
Even when comparing ESR1 mutant alleles (Y537S and D538G), distinct differences were observed. Interestingly, the ESR1 mutation-driven cistrome was found to activate genes that drive breast cancer cells to metastasize, suggesting that ESR1 mutations play a role in metastasis.35Retrospective analyses of various clinical trials of endocrine therapy have indicated the potential role of ESR1 mutations as a predictive biomarker of response. In 144 patients who experienced disease progression following AI treatment, median OS and progression-free survival (PFS) were significantly shorter in patients with ESR1 mutations at 15.5 months and 5.9 months, respectively, compared with 23.8 months and 7 months for those without mutations.28 A long-term analysis also suggested that earlier emergence of ESR1 mutations was associated with a worse PFS.36
The presence of ESR1 mutations eventually could guide therapeutic decisions following progression, with several studies indicating poorer outcomes in patients who are subsequently treated with AIs compared with those who received SERDs. Interestingly, ESR1 mutations do not seem to be as common in patients treated with AIs in the adjuvant setting or in both the adjuvant and metastatic settings, compared with those who receive AIs just in the metastatic setting.29
Data from the FERGI study, in which patients failing AI treatment were randomized to either fulvestrant in combination with pictilisib, a PI3K inhibitor, or fulvestrant plus placebo suggested that fulvestrant does not have reduced activity in patients with ESR1 mutations.31
The SoFEA trial compared fulvestrant in combination with anastrozole or placebo with exemestane alone in patients with metastatic breast cancer following progression on AIs. Patients with an ESR1 mutation treated with fulvestrant had an improved PFS compared with those who received exemestane.29
Meanwhile, the PALOMA-3 study evaluated fulvestrant in combination with palbociclib (Ibrance), a cyclin-dependent kinase (CDK) inhibitor, or placebo. The median PFS for patients who received the combination was similar for those with and without ESR1 mutations.29 However, among participants who received fulvestrant alone, patients with baseline ESR1 mutations had statistically significantly worse PFS outcomes compared with those without baseline ESR1 mutations.37
In the BOLERO-2 trial, which evaluated the addition of everolimus (Afinitor), a mammalian target of rapamycin (mTOR) inhibitor, to exemestane, the median OS for patients with wild-type ESR1 was 32 months, compared with 26 months for those with ESR1 D538G mutations.27
These studies also suggest that the benefits of adding mTOR or CDK inhibitors to endocrine therapy are independent of ESR1 mutation status. However, only palbociclib has been evaluated in this context; there are no data on the impact of ESR1 mutations on patient outcomes following treatment with the 2 other approved CDK inhibitors, ribociclib (Kisqali) and abemaciclib (Verzenio). Nor is there any information on the potential impact on other drugs commonly used to treat ER-positive breast cancer.
A recent analysis of ESR1 mutations from more than 900 patients suggested that select mutation alleles may be associated with resistance to fulvestrant. The study demonstrated that the Y537S mutation was the least effectively inhibited by fulvestrant.38ESR1 mutations could undoubtedly guide therapeutic decision making in ER-positive breast cancer cases. Helping these patients achieve their clinical potential will require continuous monitoring of ESR1 mutation status over the course of the disease.Although this can be achieved through serial tumor biopsies, this method has several drawbacks, and researchers are turning to liquid biopsies for analyses. These assays isolate circulating tumor cells or ctDNA released into the circulation from necrotic tumor cells and can help overcome the potential for sampling bias due to tumor heterogeneity, the difficulty in obtaining biopsy samples from metastatic tumors, and other issues.39-42
They offer a faster, more convenient, less invasive alternative with the potential to analyze tumor dynamics in real time. In addition, using ctDNA-based liquid biopsies has been shown to detect tumor progression significantly earlier than radiographic progression.43
The use of ctDNA to detect ESR1 mutations in patients with metastatic ER-positive breast cancer has been tested in a plethora of clinical trials. This technique has demonstrated excellent sensitivity and concordance with mutation status per tumor tissue biopsies.30,44,45
Several techniques can be used to interrogate mutation status in plasma ctDNA. Digital droplet polymerase chain reaction (ddPCR) is the most commonly used in clinical trials and has shown better sensitivity than conventional NGS.46 ddPCR has been shown to be capable of detecting ESR1 mutations in patients without radiographic evidence of disease, suggesting the mutations may be indicative of micrometastases.44 In 1 study, 75% of patients had detectable mutations at least 3 months prior to clinical progression.28
Despite this significant potential for clinical utility, ESR1 mutations are not yet integrated into the treatment landscape: many details must be worked out. Larger, prospective studies, in which patients are randomized and treated according to their mutation status, have been lacking but are starting to be conducted. There also is a need for standardized testing protocols.
We also have a limited understanding of the roles of different ESR1 mutations and have queried just the most common ones. The potential clinical implications of other, less common mutations remain unknown. It also has been suggested that the ratio of wild-type to mutant ESR1 may be a more precise biomarker and this warrants further investigation.In addition to these challenges, the most effective treatment strategies for ESR1-mutant breast cancer have yet to be determined. Several strategies have been proposed, including using higher doses of fulvestrant or tamoxifen, which could help overcome the relative resistance that has been observed. The CONFIRM trial demonstrated that a 500-mg dose of fulvestrant significantly prolonged OS compared with a 250-mg dose, with no increased toxicity.47
Newer, more potent antiestrogens may be more effective against ESR1 mutant tumors (Table 2). Several agents are being developed, including AZD9496, a novel nonsteroidal SERD. AZD9496 has been shown to potently bind and downregulate D538G and ESR1 Y537C/N/S mutants in preclinical studies.48
Meanwhile, there have been disappointments in the field. In 2017, Roche dropped 2 SERD agents, GDC-0810 and GDC-0927, which had shown early promise, from its development program.49 Genentech, a member of the Roche Group, continues work on a third drug in this class, GDC-9545, in a multiarm phase I trial (NCT03332797).
Other potential strategies include drugs targeting the coactivators, chaperones, and other associated proteins upon which ER activity depends. For example, SRC-3 is 1 of the coactivators of the ER, but small molecule inhibitors are still very much in the discovery phase. ER activity is also highly dependent on the heatshock protein 90 (HSP90). HSP90 inhibitors are already being developed in cancer but have yet to be specifically explored in the treatment of ER-positive breast cancer.
The bromodomain and extra terminal (BET) family of proteins includes epigenetic “readers” that play important roles in regulating transcription by binding to acetylated lysines in histone tails. BET proteins are highly involved in cancer and directly regulate the expression of certain cancer-related genes.
In breast cancer, the BRD3/4 proteins, which are members of the BET family, have been shown to promote ESR1 transcription and could contribute to endocrine therapy resistance. BET inhibitors are already in development for the treatment of cancer; GSK525762 is being evaluated in a phase II study in combination with fulvestrant in advanced/ metastatic ER-positive breast cancer that has progressed after prior treatment (NCT02964507).
Among 46 patients treated with at least 1 dose of GSK525762 during a phase I/II study, there were 5 responses: 3 partial; 1 complete with incomplete hematologic recovery; and 1 complete with incomplete platelet recovery. There were 2 dose-limiting toxicities: 1 grade 3 diarrhea and 1 grade 3 ejection fraction decrease. The most common AEs were dysgeusia, diarrhea, nausea, and elevated bilirubin levels.50
This study did not separatepatients according to ESR1 mutation status; however, a preclinical study of the BET inhibitor OTX015 found it to be highly effective in inhibiting the growth of breast cancer cells expressing ESR1 mutations.51