Menopausal Medicine: For clinicians who provide care for women

Selective Estrogen Receptor Modulators: From bench, to bedside, and back again

Department of Pharmacology and Cancer Biology, Duke University Medical Center, Durham, North Carolina

When the search began for estrogen receptor (ER) ligands that could treat and prevent osteoporosis while reducing the risk and incidence of gynecologic cancers, it seemed unlikely that a single drug would exhibit clinically useful selectivity. In the meantime, it has become clear that the ER does not function in the same manner in all cells. This observation led to the development of a new class of molecules, selective estrogen receptor modulators (SERMs), which manifest estrogenic activities in a cell-selective manner.

As effective as the current SERMs are for conditions associated with long-term estrogen deprivation, new SERMs with more clinically useful activities are likely to emerge in the near future. The science that is driving this quest for the perfect SERM will be discussed in detail below.

Why target estrogen action in a tissue-specific manner?

Estrogen is a key regulator of the processes involved in the growth, differentiation, and function of a wide variety of tissues. Estrogen is important in sustaining overall health: as evidence, consider that estrogen deficiency in postmenopausal women has adverse effects that can be alleviated with estrogen therapy (ET).

The beneficial effects of ET on menopausal symptoms have been well established. However, estrogen can facilitate the growth of breast tumors in some circumstances, and unopposed estrogen is an established risk factor for uterine cancer. Thus there is a clear unmet medical need for drugs which, acting through the ER, manifest estrogenic activities in a tissue- or process-selective manner.

Key molecular mechanisms in ER pharmacology

The molecular pharmacology of estrogens and SERMs is complex. It is clear, however, that all of these compounds mediate their biological activities through 2 intracellular ERs (ERα and ERβ), which function as ligand-inducible transcription factors in target cell nuclei. These receptors are ligand-activated transcription factors that, when activated, facilitate an up-regulation or down-regulation of specific genes at the messenger RNA (mRNA) level. The mRNAs are then translated into proteins that function within the cells of hormone-responsive tissues to regulate proliferation, differentiation, and homeostasis.¹

Classic receptor theory: On and off

Ligand binding initiates ER signaling. According to classic receptor theory, agonists (such as endogenous estrogens) act as molecular switches, converting ERs from an inactive to an active form. Anti-estrogens—synthetic compounds developed to oppose the action of the natural hormone—were thought to competitively inhibit agonist binding and, in doing so, were able to lock the receptor in an inactive state. Thus, it was considered that when corrected for affinity, all agonists were functionally indistinguishable and, likewise, antagonists were qualitatively similar.

The “tamoxifen paradox”

The clinical profiles of SERMs, such as tamoxifen and raloxifene, indicate that the classic model is oversimplified.² Studies of patients who received tamoxifen as adjuvant therapy for estrogen-dependent breast tumors provided some of the first evidence that these “anti-estrogens” do more than freeze ERs in a latent state. Strikingly, while tamoxifen blocked the actions of estrogen in breast cancer cells, it was shown to function as an agonist in the bone and uterus, mimicking the actions of estrogen.³ The observation that tamoxifen displays tissue-specific agonist/antagonist activities was inconsistent with the classic definition of antagonist action and, furthermore, suggested that this compound may alter ERs in such a way that the liganded receptors could be recognized differently in distinct cell types.

These early findings led to reclassification of these agents, additional research, and new indications. Tamoxifen, originally considered an “anti-estrogen,” was reclassified as a SERM to reflect its complex pharmacologic activity. Raloxifene, another SERM, has been approved for the prevention and treatment of osteoporosis. And both have been approved for use as preventatives (early treatment) in patients at high risk of breast cancer.

Mediation of diverse responses

Two major discoveries in the past 10 years have helped us understand the molecular basis of SERM action. The first was that ERs do not merely exist in an on or off form, but rather, the overall conformation of the receptors in the presence of different ligands is not identical. The second discovery was that of cofactors, proteins that are required for ER signaling but whose interactions with the receptor are influenced by receptor conformation. These proteins—termed coactivators and corepressors—bind the liganded ER and enhance or decrease ER-mediated transcription of target genes, respectively.¹

The ability of different ER ligands to induce different conformations of the receptors influences their ability to interact with coactivators and corepressors. We now know that different ligands induce unique structural changes in ERs and that this results in differential recruitment of coactivators and corepressors, leading to diversity in biological response. This provides a mechanism by which different ligands, acting through the same receptor, can mediate unique biological effects. For example, estrogen induces coactivator recruitment to ERs, whereas when bound to the pure antiestrogen ICI 182780 (fulvestrant), the conformation of the ERs is compatible with corepressor (but not coactivator) binding. Correspondingly, estrogen is a full agonist of ER, whereas ICI functions as a pure antagonist on the receptor. In contrast, when bound to the SERM tamoxifen (which displays both agonist and antagonist activities), the ER is capable of interacting with either coactivators or corepressors. Therefore, it is likely that ligand-induced changes in receptor shape and the relative expression of functionally distinct coactivators or corepressors in different target cells are sufficient to explain the tissue-specific agonist/antagonist activities of tamoxifen and other SERMs. Thus, it will be important to identify the cofactor proteins present in different ER target cells in order to allow mechanism-based screening for new tissue-targeted SERMs.^1,2,4

Clinical implications of mechanistic differences

ER “antagonists” comprise 2 broad categories: pure anti-estrogens and SERMs, which are further divided by generation. TABLE 1 presents a summary of the different biological functions associated with some of the known ER-targeted pharmaceutical agents.

Pure anti-estrogens. Pure anti-estrogens (type I), represented by ICI 182780, oppose estrogen and ER (α and β) activity in all tissues. These compounds differ considerably from SERMs in their mechanism of action and effect a more complete blockade of estrogen signaling in target cells. Specifically, these compounds induce conformations in the ERs that are permissive to corepressor (and not coactivator) association and that also lead to a proteasome-mediated decrease in cellular ERα protein. The latter observation resulted in the renaming of some members of this drug class as selective estrogen receptor downregulators (SERDs).

Biological Activities of ER Ligands in Selected Target Tissues

Type I anti-estrogens are very useful in cancer therapy because they block the mitogenic activities of estrogen in both the breast and reproductive system. However, since ICI 182780 lacks the beneficial agonist effects of estrogen in the bone, cardiovascular system, and central nervous system, the clinical use of this compound is currently limited to patients with advanced breast cancer that has become resistant to tamoxifen.

First-generation SERMs. Tamoxifen (Nolvadex) belongs to the first generation of SERMs, having been used clinically since 1971. In addition to maintaining bone mineral density (BMD) in postmenopausal women, tamoxifen decreases low-density lipoprotein cholesterol (LDL-C) levels and antagonizes estradiol-stimulated growth in the breast. Until recently, this latter effect of tamoxifen made it a standard endocrine therapy for ER-positive breast cancers in both the neoadjuvant and postoperative adjuvant settings and a first-line therapy for the treatment of pre- and postmenopausal women with estrogen-responsive advanced (stage IV) breast cancer.

A series of clinical studies has also demonstrated the efficacy of tamoxifen as a chemopreventative agent. In the Breast Cancer Prevention Trial (BCPT), which involved 13,388 pre- and postmenopausal women, tamoxifen dramatically reduced the risk of both in situ and locally invasive breast cancers by approximately 50%.^5,6

A significant proportion of women develop “tamoxifen resistance” after 5 years of anti-estrogen therapy. This term refers to the phenomenon by which certain breast cancers alter their biology and recognize the compound as an agonist for growth. Further, it appears that most tamoxifen-resistant tumors possess or are capable of developing cross-resistance to structurally related compounds such as toremifene, droloxifene, and idoxifene. Thus, mechanistically distinct pharmaceuticals such as pure anti-estrogens and aromatase inhibitors are commonly used as second-line therapies for breast cancer.^6,7

Although tamoxifen is safe and well tolerated, it is not suitable for the treatment of the climacteric patient. Most notably, it induces hot flashes in about 40% of women.⁸ Of greatest concern, perhaps, are its uterotrophic effects—clinical trials have found a clear association between tamoxifen and endometrial cancer. In the BCPT, the risk of invasive endometrial cancer was 2.5 times greater with tamoxifen. These findings of a significant benefit of tamoxifen in breast cancer tempered by serious risk in the uterus have been verified in other trials. Regardless, tamoxifen has been a prototype for other SERMs with improved therapeutic profiles, as will be discussed below.⁹

Second-generation SERMs. Second-generation SERMs were developed with the objective of obtaining ER-targeted pharmaceuticals that do not have the uterotrophic effects of tamoxifen. The most well-characterized second-generation SERM is raloxifene (Evista). This compound functions as an agonist in bone and the cardiovascular system but acts as a pure antagonist in the breast and uterus (TABLE 1). Mechanistic studies showed that raloxifene induces a unique conformational state of the ER that is thought to utilize distinct cofactor complexes, compared with tamoxifen and other ER ligands.^2,10

Raloxifene has higher efficacy in maintaining BMD compared with first-generation SERMs, which is likely due to the unique manner in which it alters the ER structure. Indeed, raloxifene is now approved for both the treatment and prevention of osteoporosis.

More recently, raloxifene has emerged as a promising breast cancer chemopreventative. The Multiple Outcomes of Raloxifene Evaluation (MORE) trial of 7705 postmenopausal women from 25 countries evaluated whether women with osteoporosis taking raloxifene have a lower risk of invasive breast cancer. Raloxifene was found to decrease the risk of all ER-positive breast cancers by 90% and invasive breast cancers by 72%.¹¹

As a result of these findings, the National Cancer Institute sponsored the Study of Tamoxifen and Raloxifene (STAR) trial, which directly compared the effects of tamoxifen and raloxifene in postmenopausal women at higher-than-average risk of breast cancer. The positive results of this trial led a recent US Food and Drug Administration (FDA) advisory panel to recommend approval of this drug as a breast cancer preventative in high-risk patients. Interestingly, this trial indicated that tamoxifen and raloxifene did not behave in the same manner with respect to breast cancer prevention. Specifically, whereas raloxifene was found to work as well as tamoxifen in reducing risk by about 50% for invasive breast cancer in postmenopausal women, tamoxifen was also found to reduce the incidence of both lobular carcinoma in situ (LCIS) and ductal carcinoma in situ (DCIS) by 50%. On the upside, participants in STAR who were assigned to take raloxifene had 36% fewer uterine cancers than did women assigned to take tamoxifen, reflecting differences in the ability of the 2 compounds to stimulate growth in the endometrium. In fact, the generally positive effects of raloxifene in the uterus are likely to make it the agent of choice for use as a chemopreventative in postmenopausal women at high risk of breast cancer.^5,6

Third-generation SERMs. The goal of developing a third-generation of SERMs was to create pharmaceuticals that are structurally distinct from first- and second-generation agents of this class. These SERMS should possess the favorable qualities of raloxifene but should be more effective in enhancing BMD and in decreasing serum LDL-C levels, and should have a reduced propensity to induce hot flashes.

Two of the most advanced third-generation SERMs, bazedoxifene and lasofoxifene, have completed registration trials. The initial New Drug Application (NDA) for lasofoxifene received an unapprovable letter from the FDA. The ongoing Postmenopausal Evaluation and Risk Reduction with Lasofoxifene (PEARL) trial should resolve this issue. The second of this generation of SERMs, bazedoxifene, has recently received an approvable letter from the FDA and should be commercially available soon.

Both of these drugs have improved pharmaceutical properties and differ mechanistically from each other and from existing SERMs. How this translates into clinical benefit remains to be seen. Of note, however, are the observations from a 2-year randomized, double-blind treatment study in postmenopausal women. Lasofoxifene was found to be similar to raloxifene with regard to maintaining BMD in the hip; yet in the lumbar spine, lasofoxifene increased BMD by 2%. Comparatively, there was no mean improvement in spine BMD in patients assigned to raloxifene and a 2% decrease in density in patients assigned to placebo.

Bazedoxifene exhibited an improved vasomotor profile in phase 3 clinical trials compared with tamoxifen or raloxifene.^12,13 However, the clinical significance of any of the differences between the second- and third-generation SERMs remains to be determined.

Fourth-generation SERMs. A list of desired properties for a fourth generation of SERMs was recently presented (TABLE 2). In general, future SERMs must oppose endogenous hormone action in the breast and reproductive system while displaying salutary estrogenic effects in the cardiovasculature, bone, and central nervous system. Specifically, these agents must relieve hot flashes and both lower LDL-C and raise high-density lipoprotein cholesterol levels. Fourth-generation SERMs should possess superior bioavailability compared with existing ER modulators and should have additional indications in men for bone protection and cardioprotection. The science that will drive the search for drugs of this class is discussed below.

Criteria for Fourth-Generation SERMs

A new paradigm: Tissue selective estrogen complex

The tissue selective estrogen complex (TSEC) is a new concept that has recently received some attention and is likely to be available in the near future. The most advanced drug in this class, Aprela, is a combined therapy that partners the third-generation SERM bazedoxifene with conjugated estrogens. A 2-year study has demonstrated efficacy of TSEC medications in offering the benefits of estrogen while minimizing the most concerning side effects and risks of traditional ET. The TSEC provides a blend of both tissue selectivity and favorable physiologic profiles, including increases in BMD superior to those seen with SERMs, and substantial improvement in hot flashes and genitourinary atrophy, without the common side effects of hormonal therapy such as endometrial hyperplasia, bleeding, and breast pain. Interestingly, the outcome is not just that of one medication added to the other but, rather, a brand new profile, as for example, no adverse cardiovascular events have been reported.¹⁴ The mechanism(s) underlying the complex pharmacologic effect of the TSECs remain to be determined. The NDA submission for this new drug is expected in the near future, and it will be interesting to see if it is perceived in the marketplace as the “perfect” menopausal therapy.

The future of SERMs: Mechanism-based drug discovery

In the future, classic drug discovery approaches that rely primarily on ligand-receptor binding assays are unlikely to identify agents that provide favorable biology in most or all estrogen target tissues. Rather, efforts focused toward understanding the molecular determinants of ER pharmacology may lead to incorporation of functional assays as primary drug screens. This “mechanism-based” approach to drug discovery should, in theory, permit identification of novel classes of modulators.^4,15 Three mechanism-based approaches to new SERM discovery that are currently being put into practice are described in brief below: Cofactoromics, ER subtype-selective modulators, and pathway-targeted SERMs.

Cofactoromics. More than 50 coactivators and corepressors that bind the ER and alter its activities have been identified. The relative and absolute expression of these factors, however, can differ significantly among different cell and tissue types. Thus, a new type of mechanism-based approach to SERM discovery involves screening for compounds that facilitate particular ER-cofactor interactions in specific tissues.⁴ Clearly, this approach will require an understanding of the expression levels and biological roles of relevant ER cofactors in estrogen target tissues, a tremendous task in which our laboratory and others are currently engaged.⁴

ER subtype-selective modulators. Estrogens manifest their biological effects through 2 distinct receptors, ERα and ERβ. Both ERs possess similar affinities for estrogens and most of the currently available SERMs; however, they display significant sequence and structural differences in their ligand-binding domains and have dissimilar biological activities. This observation, together with the apparent selective tissue distribution of the 2 ERs, suggest that it might be possible to pharmacologically separate estrogen biology by targeting a single ER.

To date, in vivo studies have been limited to animals, but these agents have sufficient efficacy to prompt exploration of the clinical utility of ERβ-selective agonists in inflammatory diseases, including arthritis, endometriosis, inflammatory bowel disease, and sepsis.¹⁶

Pathway-targeted SERMs. Not only do estrogen-activated ERs bind to their specific target gene promoters and subsequently activate gene transcription, but ERs can manifest cellular actions independent of DNA binding by associating with other transcription factors (such as AP-1, Sp1, progesterone receptor, and others) and interfering with their activities.¹⁵ If specific biological outcomes of estrogens could be ascribed to a particular ER-transcription factor pair, it might be possible to develop agents that target the relevant interaction in order to achieve process selectivity and “pathway-targeted SERMs.”

This mechanism has been explored with regard to one alternate pathway of ER action that provides for the anti-inflammatory actions of estrogens. Chadwick et al recently identified a pathway-selective ER ligand, WAY-169916. This agent, devoid of classic estrogenic activity, can inhibit NF-κB activity and function as a potent anti-inflammatory agent in animal models of inflammatory bowel disease.¹⁷ The clinical utility of this class of compound remains to be determined.

Funding Acknowledgment

This work was supported by NIH Grant DK48807 (to D.P.M.).

Disclosure

Dr Hall has nothing to disclose. Dr McDonnell reported the following potential conflicts of interest: he receives research grants from Organon Pharmaceuticals and Wyeth Pharmaceuticals; serves as a consultant to Wyeth Pharmaceuticals; is a member of the speakers’ bureau of Wyeth Pharmaceuticals; and is council member of the Endocrine Society.

Sexuality, Reproduction & Menopause ©2008 Quadrant HealthCom Inc.