Arzoxifene: A Promising New Selective Estrogen Receptor Modulator for Clinical Chemoprevention of Breast Cancer: Commentary re C. J. Fabian et al., Breast Cancer Chemoprevention Phase I Evaluation of Biomarker Modulation by Arzoxifene, a Third Generation Selective Estrogen Receptor Modulator. Clin Cancer Res 2004;10:5403–17.

Although the original discovery of the mixed functionality of an agent such as tamoxifen was extremely puzzling (at one time, it was thought that differential metabolism of tamoxifen accounted for its different actions in different organs), there is now a much better understanding of the molecular basis for the selective estrogen receptor modulator concept. Selective estrogen receptor modulators are ligands for the two estrogen receptors, estrogen receptor α and estrogen receptor β, which, in turn, are transcription factors controlling the expression of numerous genes. The main advance in understanding the complexity of the actions of selective estrogen receptor modulators has been the discovery that the functional transcriptional unit that enables an individual selective estrogen receptor modulator to regulate the activity of any specific gene is not merely the selective estrogen receptor modulator and its cognate estrogen receptor but rather a multimolecular transcriptional complex, which includes not only the selective estrogen receptor modulator and the estrogen receptor but also a series of coregulatory proteins known as coactivators or corepressors. These critical coregulators interact with the receptor and, in turn, with the response machinery that ultimately controls the expression of any specific gene (4, 5, 6). These coactivators and corepressors thus control the ultimate agonistic or antagonistic action of the selective estrogen receptor modulator in any particular organ. As an example, in an elegant series of experiments, Shang and Brown (7) have shown that both tamoxifen and raloxifene induce the recruitment of corepressors to the promoters of target genes in mammary cells, and thus both agents can act as estrogen antagonists in the breast. Conversely, they found that in endometrial cells, tamoxifen, but not raloxifene, is proestrogenic by stimulating the recruitment of coactivators to the transcriptional complex; this is presumed to be the basis of the undesirable proliferative and carcinogenic actions of tamoxifen in the uterus (8). However, raloxifene, which does not recruit these coactivators in uterine cells, does not stimulate proliferation in the uterus and does not cause uterine cancer (3). Coregulator function is now believed to be a critical modulator of the agonist or antagonist activity of an entire series of related transcription factors, all of which are members of the nuclear receptor superfamily (6). Thus, in recent years, the selective estrogen receptor modulator concept has been extended to selective ligands for the androgen receptor, the progesterone receptor, the glucocorticoid receptor, and the peroxisome proliferator-activated receptor, and an entire new field of pharmacology has been developed in which the desirable activities of a hormone-like agent can be maximized and the undesirable activities minimized (6).

Therefore, it is not surprising that there has been intense activity in the pharmaceutical industry directed toward synthesis and testing of new selective estrogen receptor modulators. Arzoxifene is an excellent example of this effort, and the article by Fabian et al. (9) in this issue represents the first report of the use of this agent in a clinical chemoprevention study, although promising results have already been reported with arzoxifene in Phase II studies of treatment of breast cancer (10, 11). The development of arzoxifene as a preventive agent has been propelled by the clinical utility of raloxifene, which is effective for both treatment of osteoporosis and prevention of breast cancer in postmenopausal women (2, 3). As can be seen in Fig. 1, arzoxifene and raloxifene are closely related chemical structures. The replacement of the carbonyl function in raloxifene with an ether linkage in arzoxifene has created a new molecule with much higher affinity for the estrogen receptor, and in arzoxifene, the phenolic hydroxyl group of raloxifene has been methylated to improve the pharmacokinetic properties of the molecule (12). Thus, arzoxifene is clearly a much better selective estrogen receptor modulator than raloxifene in preclinical animal studies that attempt to measure potential clinical usefulness either for treatment of osteoporosis (13) or prevention of breast cancer (12). Arzoxifene is exceptionally potent in these animal studies; dosages as low as 0.1 mg/kg by mouth have marked pharmacodynamic activity. Additionally, and most notably with respect to the safety issue, arzoxifene appears to be devoid of the proestrogenic activity of tamoxifen when these agents are compared for stimulation of growth of MCF-7 cells in the absence of 17β-estradiol (12). Moreover, arzoxifene appears to be totally devoid of any proestrogenic activity when tested on the uterus of either intact or ovariectomized rats (12). This absence of uterotrophic action has been confirmed in Phase II studies evaluating efficacy of arzoxifene in women with advanced breast cancer. Mechanistic studies indicate that arzoxifene is a potent inducer of the expression of transforming growth factor β (14); this action may contribute to the suppression of early lesions in the breast (15).

Arzoxifene thus is an ideal candidate for a clinical study of chemoprevention of breast cancer, and in this context, this first report from Fabian et al. (9) is most welcome. Two trials are reported, in which women with newly diagnosed ductal carcinoma in situ or T1/T2 invasive breast cancer were studied. In both trials, women were biopsied then treated with arzoxifene for between 2 and 6 weeks before re-excision of their lesion. A very large battery of both serum and tissue biomarkers have been evaluated, and with this short a treatment period, many of the results were negative. However, there were some strikingly positive findings, most notably decreases in serum insulin-like growth factor I (IGF-I) and serum IGF-1:IGF binding protein-3 ratio (IGF binding protein-3 is one of the specific binding proteins that maintain IGF-I in an inactive state), as well as a marked decrease in proliferation indices, especially as measured by proliferating cell nuclear antigen expression. The effects of arzoxifene on IGF-I are particularly notable, because there are extensive data to indicate that high levels of circulating IGF-I and low levels of IGF binding protein-3 are clinically associated with increased risk of breast cancer (16, 17). Moreover, in the second trial, a statistically significant decrease in estrogen receptor expression was found in the women treated with arzoxifene. In short trials such as those reported here, it is clearly impossible to evaluate long-term effects of a preventive agent on cancer incidence, but the authors conclude that, given the favorable side effect profile and the biomarker modulations, arzoxifene remains a reasonable candidate for additional study as a breast cancer chemopreventive agent.

Clearly this entire approach to prevention of breast cancer is now ready for additional clinical study. Synthetic chemistry continues to generate a large number of new selective estrogen receptor modulators that have yet to be explored in the clinic as preventive agents. Just as the second and third generation selective estrogen receptor modulators, raloxifene and arzoxifene represent a totally new class of chemical structures (benzothiophenes) that are chemically distinct from tamoxifen (a triphenylethylene), so there is now an even greater choice of new selective estrogen receptor modulators to consider for prevention studies. These include agents of very diverse structure, such as acolbifene (a benzopyran) and lasofoxifene (a tetrohydronaphthalene), both of which share the common property of being antiestrogenic in the breast and proestrogenic on bone in experimental animals; neither one has the undesirable uterotrophic effects of tamoxifen (18, 19, 20, 21). The ultimate practical usefulness of any of these new agents no doubt will depend on their total selective estrogen receptor modulator profile, which has yet to be explored in a clinical context. Studies of the interaction of these drugs with coactivators and corepressors will be of major importance for the understanding of the selectivity, which is desired.

As can be seen from the present report by Fabian et al. (9), there is still a major need to develop new and better biomarkers to evaluate the efficacy of any selective estrogen receptor modulator in a clinical trial. Many of the biomarkers in present use are more than a decade old, and their sensitivity is hardly optimal. The recent emphasis on proteomics should offer better hope for the future (22, 23). Additional cooperation among academia, the pharmaceutical industry, and the Food and Drug Administration will be essential in achieving this desired goal of clinical chemoprevention of breast cancer. It is clear by now that we have the means to prevent breast cancer in numerous animal models, whether with single agents or with combinations of drugs (24, 25). The current challenge is to apply this knowledge in the clinic. The study by Fabian et al. (9) clearly shows that clinical evaluation of biomarkers can be an important component in this complex process.