A New Selective Estrogen Receptor Modulator with Potent Uterine Antagonist Activity, Agonist Activity in Bone, and Minimal Ovarian Stimulati
http://www.100md.com
《内分泌学杂志》
Lilly Research Laboratories (A.G.G., M.W.D., J.W.H., I.R.C., D.G.R., K.B.D., M.D.A., T.A.S., O.W.B., D.J.M., S.W.O., H.U.B., M.S., J.A.D.), Eli Lilly & Co., Indianapolis, Indiana 46285
Array Biopharma (C.W.H.), Boulder, Colorado 80301
Abstract
The use of selective estrogen receptor modulators for the treatment of estrogen-dependent diseases in premenopausal women has been hindered by undesirable ovarian stimulation and associated risks of ovarian cysts. We have identified a selective estrogen receptor modulator compound (LY2066948) that is a strong estrogen antagonist in the uterus yet has minimal effects on the ovaries of rats. LY2066948 binds with high affinity to both estrogen receptors and has potent estrogen antagonist activity in human uterine and breast cancer cells. Oral administration of LY2066948 to immature rats blocked uterine weight gain induced by ethynyl estradiol with an ED50 of 0.07 mg/kg. Studies in mature rats demonstrated that LY2066948 decreases uterine weight by 51% after 35 d treatment, confirming potent uterine antagonist activity over several estrous cycles. This strong uterine response contrasted with the minimal effects on the ovaries: serum estradiol levels remained within the normal range, whereas histologic evaluation showed granulosa cell hyperplasia in few of the rats. Bone studies demonstrated that LY2066948 prevented ovariectomy-induced bone loss and treatment of ovary-intact rats caused no bone loss, confirming estrogen receptor agonist skeletal effects. Collectively, these data show that LY2066948 exhibits a tissue-specific profile consistent with strong antagonist activity in the uterus, agonist activity in bone, and minimal effects in the ovaries.
Introduction
UTERINE FIBROIDS (LEIOMYOMAS) are benign tumors arising from smooth muscle cells of the myometrium (1, 2, 3). They are the most common type of solid tumor in adult women, clinically apparent in approximately 25% of those of reproductive age. In women who experience symptoms, abnormal menstrual bleeding, pelvic pain, pressure symptoms, and infertility are the most common. Uterine fibroids and their associated complications often necessitate hysterectomy, accounting for more than 200,000 of these procedures each year in the United States. Less drastic surgical procedures such as myomectomy (fibroid removal with uterine retention), laser ablation, or uterine artery embolization are available but have been associated with complications and fibroid recurrence (3).
Clinical data suggest that estrogens are important in the regulation of fibroid growth. These tumors increase rapidly in size during the hyperestrogenic state brought on by pregnancy and regress in hypoestrogenic states brought on by menopause or ovariectomy. Thus, pharmacologic approaches have centered on the hormone dependency of fibroids, especially on blockade of estrogenic stimulation. One such therapeutic approach is treatment with GnRH agonists. GnRH efficacy is the result of down-regulation of the hypothalamo-pituitary-ovarian (HPO) axis. GnRH agonists desensitize the pituitary gland to GnRH, decreasing release of the gonadotropins FSH and LH and in turn reduce the amount of estrogens produced in the ovaries to castrate levels. These agents show efficacy but carry a high side effect burden, limiting chronic treatment and relegating use to preoperative applications (4). Side effects associated with lowered levels of ovarian estrogens, such as hot flashes and bone loss, are common with GnRH use.
Clearly there exists a need for improved therapies to treat uterine fibroids but also to avoid the drawbacks without the risks and side effects associated with current surgical and pharmacological interventions. One particularly promising class of compounds that could fill this need is selective estrogen receptor modulators (SERMs) (5). SERMs bind to estrogen receptors (ER) and impart tissue-dependent (selective) antagonistic or agonistic effects. SERMs currently on the market include tamoxifen (Nolvadex; AstraZeneca PLC, London, UK), primarily indicated for adjuvant treatment of breast cancer, and raloxifene (Evista; Eli Lilly and Co., Indianapolis, IN), indicated for prevention and treatment of postmenopausal osteoporosis. Both of these SERMs have tissue-specific ER-mediated pharmacology. They share antagonist properties in breast and agonist properties in bone but differ at the uterus in that tamoxifen acts as a partial agonist, whereas raloxifene is an antagonist (6). Another SERM, clomiphene (Clomid; Sanofi-Aventis SA, Paris, France), is used in premenopausal women to induce ovulation by blocking ERs in the hypothalamus (7). This antiestrogenic action results in stimulation of the HPO axis through antagonism of the steroidal negative feedback loop and stimulation of GnRH release, resulting in increased production of LH and FSH and ultimately ovarian stimulation. Evidence has also pointed to a direct estrogenic effect of clomiphene on the ovaries, leading to ovarian stimulation (8). Clinical results in premenopausal women have suggested that tamoxifen and raloxifene affect the ovaries by stimulating the HPO axis and, possibly, the ovary directly (9, 10). Ovarian stimulation is not desired in many premenopausal applications because of the risk of follicular cyst formation. Tamoxifen treatment of premenopausal women does in fact increase circulating estrogens and the incidence of follicular cysts (11).
To address the long-standing issue of ovarian hyperstimulation observed with currently available SERMs, we sought to identify a more tissue-selective agent for potential use in the treatment of uterine fibroids. Structure-activity studies were directed at identifying compounds that displayed a high degree of estrogen antagonist activity in the uterus yet were relatively neutral for the ovaries and displayed estrogen agonist properties in bone. In this paper we describe the discovery of LY2066948, a novel SERM that possesses potent uterine antagonist activity and minimal ovarian stimulation, making this a potential drug candidate suitable for treatment of premenopausal women suffering from uterine fibroids and other estrogen-dependent gynecologic applications.
Materials and Methods
SERM compound sources
Tamoxifen and 4-hydroxytamoxifen were purchased from Sigma (St. Louis, MO). Note that tamoxifen was used in the in vivo studies (structure shown, Table 1) whereas 4-hydroxytamoxifen was used in the in vitro studies. ICI 182780 was purchased from Tocris (Ellisville, MO). All other SERMs used in this study were synthesized in the Lilly Research Labs (Indianapolis, IN). All studies were done with the hydrochloride salt form of 2066948 (2066948-HCl).
ER binding
The competition binding assay was run in a buffer containing 50 mM HEPES (pH 7.5), 1.5 mM EDTA, 150 mM NaCl, 10% glycerol, 1 mg/ml ovalbumin, and 5 mM dithiothreitol, using 0.025 μCi/well 3H-estradiol [no. NET517 at 118 Ci/mmol, 1.5 nM 17-estradiol (E2); NEN Life Science Products/PerkinElmer, Boston, MA] 10 ng/well ER or ER receptor (PanVera/Invitrogen, Carlsbad, CA). Nonspecific binding was determined in the presence of 1 μM E2. The binding reaction was incubated for 4 h at room temperature, and then cold dextran-coated charcoal (DCC) buffer was added to each reaction [DCC buffer contains, per 50 ml of assay buffer, 0.75 g charcoal (Sigma) and 0.25 g dextran (Pharmacia, Uppsala, Sweden)]. Plates were mixed 8 min on an orbital shaker at 4 C and then centrifuged at 3000 rpm at 4 C for 10 min. An aliquot of the mix was added to Optiphase Hisafe 3 scintillation fluid [Wallac (PerkinElmer)], incubated for 2.5 h, and read in a Wallac Microbeta counter. The equilibrium dissociation (Kd) constant for 3H-estradiol was determined by saturation binding to ER and ER receptors. The IC50 values for compounds were converted to inhibitory equilibrium dissociation constant values (Ki) using the Cheng-Prusoff equation.
Ishikawa assay
Estrogenic stimulation and antagonism were measured in Ishikawa human endometrial tumor cells by alkaline phosphatase quantitation. The cells were incubated in DMEM/F-12 (3:1) supplemented with 5% DCC-stripped fetal bovine serum (Hyclone, Logan, UT), L-glutamine (2 mM), MEM sodium pyruvate (1 mM), HEPES (2 mM) (all from Life Technologies, Inc., Gaithersburg, MD). For the agonist mode, plates received diluted compounds only, whereas antagonist mode plates additionally received 1 nM estradiol (E2, Sigma). Cells were incubated for 48 h, and then fresh compounds were added for an additional 72 h. The assay was quenched by removing medium and rinsing plates twice in PBS, and the plates were dried for 5 min and frozen at –70 C. After thawing, 100 μl of 1-Step p-nitrophenyl phosphate (PNPP) (Pierce Chemical Co., Rockford, IL) were added for 20 min, and plates were read on a spectrophotometer at 405 nm. For the agonist mode, a percentage increase over control was calculated. The data were fitted to a linear interpolation to derive IC50 values for the antagonist mode, and a percentage efficacy was calculated that blocks the E2 (1 nM) stimulus.
MCF-7 assay
MCF-7 breast adenocarcinoma cells (HTB 22; American Type Culture Collection, Manassas, VA) were grown in assay medium MEM (phenol red-free; Life Technologies) supplemented with 10% DCC-stripped fetal bovine serum. Cells were assayed by plating 8000 cells in each well of 96-well Cytostar T scintillation plates (Amersham, Piscataway, NJ) and incubated at 37 C for 24 h before adding compounds. For antagonist mode, 10 pM E2 were added along with dilutions of compound for 48 h, and then medium containing 0.01 μCi of 14C-thymidine (52 mCi/mmol, 50 μCi/μl; Amersham) was added to each well. The plates were incubated overnight and then quantitated on a Wallac Microbeta counter. The data were averaged to calculate an IC50 and percent inhibition at 1 μM compound.
Three-day rat uterine antagonist
All rat experiments were conducted in accord with accepted standards of humane animal care and in accordance of the Animal Care and Use Committee at Eli Lilly & Co.
Female Sprague Dawley (SD) rats, six per group and 19–21 d of age, were orally treated with ethynyl estradiol (EE; 0.1 mg/kg) and 10, 1.0, 0.1, or 0.01 mg/kg SERM for 3 d. Test compounds were dissolved in 20% (wt/vol) -hydroxycyclodextrin in water and administered by oral gavage in a volume of 0.2 ml daily (15 min before the EE gavage). Groups of six rats were also given vehicle as a negative control, EE alone as a positive control, and EE plus LY117018 as a standard antagonist control. The animals were fasted overnight after the final dose. On the following morning, the animals were weighed and then killed (by carbon dioxide asphyxiation), and the uteri were rapidly collected (via a midline ventral incision), stripped of adipose tissue, luminal fluid removed by blotting onto absorbant paper, and weighed.
Uterine weight to body weight ratios (UWR) were calculated for each animal. The percentage inhibition of the estrogen-induced response was then calculated by the following formula: percent inhibition = 100 x (UWRE2 – UWRtest compound/UWRE2 – UWRcontrol). ED50 values were derived from a semilog regression analysis of the linear aspect of the dose-response curve.
Ovariectomized (OVX) rat assays
For the 42-d assay (see Figs. 2 and 6), virgin 6-month-old, SD female rats (Harlan, Indianapolis, IN) weighing about 270 g were bilaterally ovariectomized under isoflurane anesthesia, randomized to treatment groups, and dosed, beginning on the third day after surgery. For the 7-d assay (see Fig. 7), virgin 75-d-old, SD female rats (Harlan) weighing about 240 g were treated as above with dosing beginning 14 d after surgery (rats now 3 months of age). Groups from both experiments were orally dosed each day in a vehicle of 20% hydroxypropyl--cyclodextrin (Aldrich Chemical Co., Milwaukee, WI). Uteri were removed and analyzed as in the 3-d rat assay above. Quantitative determination of cholesterol levels was achieved by measurement of cholesterol esterase/cholesterol oxidase activity using a Hitachi-917 automated chemistry analyzer manufactured by Roche Diagnostics (Indianapolis, IN). LH was quantitated from sera using a rodent LH ELISA kit from Endocrine Technologies, Inc. (Newark, CA).
Ten-day hormone assay
Female Fischer 344 rats (9–10 wk old) were dosed with 30 times the ED50 dose value derived from the 3-d rat uterine antagonist model. Trunk blood was collected 2 h after the final dose. Serum levels of estradiol were measured by RIA, using an assay kit from DiaSorin (Saluggia, Italy).
Fourteen- and 35-d intact rat uterine efficacy
Mature, virgin, 3- (14-d assay) or 6-month-old (35-d assay) SD rats were administered compound by oral gavage daily. Animals were killed and the uteri collected and weighed as described for the 3-d rat uterine antagonist assay. For histology and morphometry in the 35-d study, uteri and ovaries were removed, weighed, and fixed in 10% neutral buffered formalin. Tissues were trimmed, processed, and embedded in paraffin blocks for histology. For trimming, transverse sections of the ovary and the midpoint of the left uterine horn were taken. Five-micrometer sections were stained with hematoxylin and eosin and evaluated by a board-certified veterinary pathologist for histologic changes. The tissue sections were examined by light microscopy by an American College of Veterinary Pathologists board-certified veterinary pathologist. Histologic changes were described, when applicable, according to their distribution, severity, and morphologic character (12) using standardized nomenclature for the rat and lesion definitions for SERM ovarian effects published elsewhere (13). The diagnosis of ovarian follicular cyst was reserved for prominent ovarian follicles that had a diameter of at least 2 mm, which was not observed in this study (14). Qualitative severity scores for the ovarian changes were assigned as follows: 1, minimal; 2, slight; 3, moderate; and 4, marked. Mean group lesion scores for individual groups were calculated by adding the individual animal severity scores and dividing by the total number of animals.
Morphometry was done on the uteri from all control (n = 6) and 0.5 mg/kg LY2066948-treated (n = 7) rats. Uteri were imaged at x25 magnification using a digital camera (Spot Diagnostic Instruments, Sterling Heights, MI) attached to a DMLB microscope (Leica Microsystems AG, Wetzlar, Germany). Cross-sectional area (square micrometers) of the uterine wall, endometrium, and myometrium were measured using Spot morphometry software. LH was measured by RIA as described in (13), progesterone (P4) was measured by active progesterone RIA (Diagnostic Systems Laboratories Inc., Webster TX). For the progesterone assay, the upper limit of detection was 60 ng/ml and the lower limit of detection was 0.3 ng/ml for a 25-μl sample volume.
Bone analysis
Vertebrae and femora were excised at necropsy and the midtransverse section of the lumbar vertebra L-4 (see Table 3) and distal femur metaphysis (see Fig. 6) was scanned in 50% ethanol/saline, using quantitative computed tomography (QCT) (Research M, Norland/Stratec, Ft. Atkinson, WI). Cross-sectional area, bone mineral content (BMC, milligrams), and volumetric bone mineral density (milligrams per cubic centimeter) were quantitated, using voxel dimensions of 148 x 148 x 500 μm as previously described (15).
Statistical analyses
One-way ANOVA was followed by Dunnett’s post hoc test when the one-way ANOVA was statistically significant. Dunnett’s test (using JMP 5.1 software, SAS Institute Inc., Cary, NC) was used to assess the statistical significance (P < 0.05) of treatment groups with respect to a nontreatment vehicle control group for all analyses.
Results
In vitro characterization
Compounds were first selected for high affinity binding to both estrogen receptor types (ER and ER). A cell-free competition-binding assay was used to measure the propensity to displace 3H-estradiol and generate Ki values for a library of SERM compounds. Our eventual lead compound, LY2066948-HCl (herein referred to as LY2066948), demonstrated high affinity binding to both human ERs, with a Ki- of 0.51 nM and Ki- of 1.36 nM. Table 1 lists ER binding affinities and other activities described below for LY2066948 and other SERMs including raloxifene, the mixed uterine agonist/antagonist tamoxifen (6), and the pure ER antagonist ICI 182780 (16).
The Ishikawa human uterine cell line was used to investigate potential uterine antagonist activity in vitro. Alkaline phosphatase activity was measured as an end point for relative estrogenic stimulation in both an agonist mode (in the presence of the SERM alone) and an antagonist mode (dose response for the SERM to block the stimulatory effect of a fixed dose of E2) (17, 18). In the antagonist mode, LY2066948 blocked the 1 nM E2 stimulatory response by 87.5% with an IC50 of 10.7 nM, compared with 53.4% inhibition and an IC50 of 421 nM for 4-hydroxytamoxifen, a first-generation SERM with known uterotropic activity (see Table 1). The agonist activity of LY2066948 alone was only 29% over control, compared with 123% with 4-hydroxytamoxifen. These data indicate that LY2066948 is a potent antagonist of uterine cell stimulation with significantly less agonist characteristics than 4-hydroxytamoxifen.
We also examined growth-inhibitory effects of LY2066948 in MCF-7 cells that were originally derived from a human breast adenocarcinoma and are frequently used in estrogenicity studies (19). These mammary cells express high numbers of ERs and demonstrate a dose-dependent increase in proliferation with increasing concentrations of estradiol. The growth effects of SERMs in this assay were examined by conducting the assay in both the agonist mode (absence of E2) and antagonist mode (presence of 10 pM E2). LY2066948 demonstrated no agonist activity up to 1000 nM (data not shown) but was able to antagonize E2 stimulation with an IC50 of 0.86 nM, confirming potent antiestrogenic activity in this cell line and suggesting estrogen antagonist potential in the mammary gland (Table 1).
In vivo uterine efficacy
To assess the potential of compounds to block uterine fibroid growth, we relied on the ability of compounds to oppose either exogenous or endogenous estradiol stimulation of the rat uterus as measured by changes in uterine wet weight. All of the following in vivo assays were carried out with various doses of test compound administered orally on a daily basis.
Uterine antagonism was measured in immature (3 wk old) female rats, which are highly sensitive to estrogenic stimulation of the uterus, given that their circulating estrogen levels are prepubertal (20). Administration of EE to immature rats produces a reliable increase (typically 3- to 4-fold) in uterine weight. In this assay, LY2066948 was found to be a potent uterine antagonist, completely inhibiting the estrogenic stimulus at 10 mg/kg and giving an ED50 value of 0.07 (± 0.02) mg/kg (Fig. 1 and Table 1). This potency compares favorably with raloxifene (21, 22), tamoxifen (partial agonist, partial antagonist), and the pure ER antagonist ICI 182780 (Table 1).
To verify that LY2066948 does not act as a mixed agonist/antagonist in the uterus after an extended period of treatment, the compound was administered to mature, OVX rats for 42 d and then uteri were analyzed. Castration dramatically decreases uterine weight and allows for sensitive measurements of potential estrogenic agonist activity. Slight uterine weight gains over OVX control were noted at all doses (no dose response seen), amounting to an 18% increase at the highest (10 mg/kg) dose administered and comparable with that seen with the raloxifene analog, LY117018 (23) (Fig. 2A). In this same experiment, we measured serum cholesterol levels because many SERMs have demonstrated a cholesterol-lowering effect (6, 22). LY2066948 lowered total serum cholesterol by 62–74% at tested doses, confirming this desirable SERM property (Fig. 2B).
Estrogen antagonism in the uterus of ovary-intact rats
LY2066948 was examined for the ability to antagonize estrogen effects on uterine wet weight in ovary-intact rats with circulating endogenous estrogen. Two different experiments were performed, differing in the duration of dosing. In the first experiment, 75-d-old rats were given LY2066948 for 14 d, resulting in a reduction in uterine wet weight of 37% at a dose of 1 mg/kg (data not shown). In the second study, 6-month-old rats were administered LY2066948 for 35 d at doses up to 3 mg/kg, greater than 40 times the ED50 of 0.07 mg/kg for uterine weight reduction in the immature rat study. Uterine antagonism was maximal at 1 mg/kg, with a 51% reduction in uterine wet weight (Fig. 3A), and the ED50 was estimated to be approximately 0.15 mg/kg. Morphometric measurements were carried out on all animals of the 0.5 mg/kg dose group and compared with the control group. At this dose of LY2066948, a uterine weight reduction of 43% was noted, compared with control. The total area of the uterine wall (endometrium and myometrium combined) was significantly reduced by 53% with nearly equal contributions in reduction of endometrial area (54%) and myometrial area (51%) (Fig. 3, B and C). Thus, these data confirm substantial antagonist activity in both the endometrial and myometrial compartments of the uterus in the presence of circulating levels of estradiol over several estrous cycles (one cycle is 4–5 d) in the rat.
Ovarian evaluation
To assess potential stimulatory effects of SERMs on the ovaries, we measured serum estradiol levels after 10 d of oral administration at doses exceeding the uterine ED50 dose by 30-fold. Compounds chosen for further analysis were those that produced less than 2-fold increases in serum estradiol concentrations. Nearly all compounds failed this ovarian safety hurdle with the exception of LY2066948, which had no significant effects on serum estradiol concentrations. Given this unique profile among the more than 120 compounds evaluated in this assay, LY2066948 became the focus for longer-term studies to further evaluate ovarian effects.
The effect of LY2066948 on the ovaries was assessed by evaluating changes in serum estrogen and LH, ovarian weight, and morphology after 35 d of oral administration. There was a dose-dependent increase in serum estradiol levels with doses greater than 0.5 mg/kg, and at 3.0 mg/kg, this increase was significant (Fig. 4). However, these levels did not exceed the normal mean proestrus level for SD rats of approximately 52 pg/ml (Cohen, I. R., unpublished observation). Serum LH was comparable with controls in rats administered LY2066948, although there was a trend toward LH elevation at higher doses. Mean ovarian weights were decreased in LY2066948-treated rats given 0.5 mg/kg or more, compared with controls (Table 2). The decreased ovarian weights corresponded to the moderate to marked depletion of corpora lutea (CL) that was observed in rats given LY2066948. Despite the general finding of corpora lutea depletion in LY2066948-treated rats, a few rats given 1 mg/kg or less had basophilic CL consistent with recent ovulation. Serum P4 levels were generally less than control, consistent with CL depletion in LY2066948-treated animals (see Table 2). However, P4 levels of some treated rats were of sufficient magnitude to indicate that ovulation had occurred.
The follicular prominence scores were generally comparable at doses between 0.05 and 1 mg/kg. Follicular prominence was characterized by dilated antral follicles that were lined by an orderly layer of attenuated granulosa cells. Oocytes within prominent follicles were uncommon and if present were often degenerate. The follicular prominence was consistent with retention of anovulatory follicles and in combination with the CL depletion indicated a prolongation of the ovarian cycle at the preovulatory phase.
Granulosa cell hyperplasia was observed in only a few animals. No rats had hyperplasia at a dose comparable with the ED50 for uterine antagonism in the immature rat assay, and only a single rat had minimal hyperplasia at the 0.5 mg/kg dose. When present, hyperplasia was usually graded minimal. The minimal grade was based on the small number of follicles affected (usually one or two), the focal nature of the hyperplasia, and the morphologic characteristics of the hyperplasia change; the hyperplasia consisted of papillary infoldings of the granulosa cells into the follicular lumen with subadjacent infoldings of thecal cells supported by a vascular bud. Ovarian cysts, a common manifestation of most SERMs, were not observed in rats given LY2066948 at doses up to 3 mg/kg.
We compared the ovarian effects of LY2066948 with those of rats treated with LY560212, a close analog of LY2066948 in which the methyl sulfone moiety was replaced with a fluorine atom. Whereas LY560212 is a very potent uterine antagonist (3-d uterine ED50 of 0.02 mg/kg), it also displays the more typical SERM profile of stimulating estradiol levels (4.6-fold, data not shown) in the 10-d hormone studies. In contrast to the modest ovarian effects seen with LY2066948, dosing with 0.45 mg/kg of LY560212 resulted in marked ovarian stimulation characterized by a 43% increase in ovarian weight and nearly a 1.5- and 3-fold higher histologic score for follicular prominence and granulosa cell hyperplasia, respectively (Table 2 and Fig. 5). The accentuated ovarian stimulation may have been due in part to the effects of increased levels of the trophic hormone, LH. LH was approximately 2-fold greater in LY560212 rats, compared with controls, but did not reach statistical significance (P = 0.13, Table 2).
Skeletal effects
Another end point of interest in the 35-d assay (uterine and ovarian effects shown in Figs. 3–5) was the analysis of bone to assure that a relevant dose of LY2066948 would not act as an estrogen antagonist in the skeleton and lead to bone loss in ovary intact rats. This was a concern because there have been reports of bone loss associated with another SERM, tamoxifen, in ovary-intact rats and also premenopausal women (24, 25). Cancellous bone from the fourth lumbar vertebra was evaluated by QCT analysis. QCT parameters included bone mineral density (BMD), BMC, and cross-sectional area. Vertebral BMD was significantly reduced in OVX animals relative to intact controls, but LY2066948 had no effect, compared with intact controls (Table 3). BMC and cross-sectional vertebral area were not significantly different between groups. These data showed no significant effects of LY2066948 at any dose on the bones of ovary-intact rats after 35 d of treatment. Whereas definitive effects of LY2066948 on the skeleton may require longer periods of treatment, these findings contrast with the significant loss of bone (femoral BMD and BMC) seen after 35 d of treatment with a GnRH agonist (Alzet pump-infused D-Trp6-LHRH), which was comparable with the loss seen in OVX rats (data not shown).
Skeletal efficacy of 0.01–10 mg/kg LY2066948 was also evaluated for 42 d in OVX animals and compared with 1 mg/kg of LY117018 (uterine data shown in Fig. 2). Treatment was initiated 3 d after ovariectomy to measure the ability of LY2066948 to prevent bone loss. Measurements were carried out on the distal femur metaphysis (cancellous bone measurement) and the femur midshaft (cortical bone measurement) (15), but the latter was not significantly different among any of the groups, including intact vs. OVX (data not shown). LY2066948 prevented ovariectomy-induced bone loss in the distal femur measured as preservation of both BMC and BMD, with effects comparable with those seen with LY117018 or raloxifene (26) (Fig. 6). Thus, LY2066948 acts as an ER agonist on bone in OVX rats and does not antagonize endogenous estrogen effects in the skeleton of ovary-intact rats.
Hypothalamic-pituitary effects
We were interested in determining whether LY2066948 had any effect on the hypothalamic-pituitary axis, given that estradiol normally feeds back on the hypothalamus to regulate pituitary gonadotrophins. We therefore treated OVX rats for 7 d with LY2066948 to see whether there was any effect on lowering the high levels of LH seen in the absence of estradiol. As shown in Fig. 7A, the high levels of serum LH were decreased by an average of 84% with estradiol treatment, whereas LY2066948 showed a significant but non-dose-dependent reduction (over the dose range tested) of about 40%, suggestive of a slight agonist activity for the hypothalamo-pituitary axis. In this same experiment, we measured uterine wet weight changes (Fig. 7B), which were consistent with the longer-term dosing of OVX rats shown in Fig. 2A, emphasizing the minimal uterine agonist activity of LY2066948.
Discussion
Structure-activity studies in a series of estrogen-dependent in vitro and in vivo assays have led to the discovery of a novel SERM that displays potent uterine antagonist activities, ER agonist effects on bone, and minimal ovarian stimulation in rats. A compound with such a pharmacological profile may be ideal for the treatment of uterine fibroids. In premenopausal women, SERMs such as clomiphene, tamoxifen, and raloxifene have shown ovarian stimulatory activity with the potential for ovarian cyst formation presumably through stimulation of the HPO axis (7, 8, 9, 10, 11). Currently, the main pharmacologic approach for treatment of uterine fibroids, GnRH analogs, is contraindicated for chronic use due to concerns over excessive bone loss with prolonged administration. A SERM with potent uterine antagonist activity that is safe with respect to the ovary and does not cause bone loss would represent an important therapeutic advance for premenopausal women with uterine fibroids.
For measurement of uterine antagonist efficacy and potency, we relied on the ability of the SERM to oppose estradiol-dependent uterine weight gains in rats. We used this model as a surrogate for changes in fibroid size, given the absence of animal models for uterine fibroids that truly represent human disease. Although there are two noteworthy models that have approximated human fibroids, it is unclear how relevant they are to human disease. For example, the Eker rat develops spontaneous uterine fibroids, but these are associated with a genetic mutation in the tuberous sclerosis-2 tumor suppressor gene that also predisposes the animals to renal cell carcinoma (27). This is not thought to be a mechanism involved in human predisposition to uterine fibroids. Moreover, the efficacy of fibroid reduction by a raloxifene analog was found to be equal to that of tamoxifen in the Eker rat (28), whereas raloxifene has been shown to be nonstimulatory in the human uterus and shows efficacy against human fibroids (29, 30) and tamoxifen is stimulatory and increases fibroid size (31, 32). Our rat uterine weight models clearly show that raloxifene is a much stronger antagonist in the uterus, compared with the partial agonist/antagonist character of tamoxifen. Another fibroid model of consideration is the hormone-dependent guinea pig model in which the animals are first ovariectomized and then stimulated with exogenous estradiol (33). Whereas this model has the advantage of estrogen dependence (like that seen with human fibroids) and raloxifene was found to be effective in decreasing fibroid size (34), the guinea pig fibroids occur mainly abdominally instead of in the uterus, a situation not normally seen in humans. Of note is the recent description of a guinea pig model in which a cocktail of nuclear receptor ligands was administered and in which fibroids were induced quickly in the correct uterine compartment (35). This model shows promise for assessing antileiomyoma efficacy; however, additional validation is necessary to determine whether this model predicts clinical efficacy.
Clinical data have clearly demonstrated that raloxifene decreases both uterine and leiomyoma size in postmenopausal patients (30). There are, however, conflicting data on the efficacy of SERMs in premenopausal women with uterine fibroids. In one study, raloxifene alone did not demonstrate efficacy, suggesting that the doses (60 and 180 mg/d for 24 wk) were not sufficient to antagonize the plasma estrogen levels in ovulatory women or that some other factor was responsible for differences in fibroid response in post and premenopausal women (36). Interestingly, when these authors administered GnRH agonist along with raloxifene to premenopausal women with fibroids, raloxifene had an additional effect of decreasing fibroid size, suggesting that depletion of estradiol via the GnRH agonist allowed raloxifene to antagonize fibroid growth (37). More recently, studies in premenopausal women treated with raloxifene concluded that raloxifene prevented the progression of fibroids and resulted in an average 22% reduction in fibroid volume after 3 months treatment with a dose of 180 mg/d (38). The difference in efficacy between this study and the study by Palomba et al. (36) was speculated to be due to differences in the average age of study subjects (36 vs. 40 yr), and therefore, potential differences in circulating estradiol levels, though other protocol dissimilarities bring this theory into question (39, 40).
The effects of tamoxifen on fibroid size have also been investigated, but no effect was observed in premenopausal women (41, 42). Furthermore, these studies demonstrated increased estradiol levels in treated patients and seven of 10 women developed ovarian cysts during the study. Both groups suggested that the increased hormonal levels and ovarian hyperstimulation were a result of indirect as well as direct effects of tamoxifen on the ovaries. There have also been studies or case reports in which tamoxifen was shown to increase fibroid growth as well as endometrial proliferation in postmenopausal women (32, 43, 44).
Ovarian stimulation and subsequent risk for ovarian cyst formation is an undesired effect of SERM treatment in premenopausal women. Preclinical rat and human clinical studies in premenopausal women have demonstrated that SERMs like raloxifene and tamoxifen stimulate the HPO axis and possibly the ovary directly (9, 13, 42). We found that doses of LY2066948 that produced potent uterine efficacy had limited ovarian effects. The effects on ovarian antral follicles and CL number were consistent with LY2066948-mediated prolongation of the follicular phase of the ovarian cycle. This pertubation of the follicular phase is likely due to loss of estrogen feedback on the hypothalmic-pituitary axis resulting in an absence of a LH surge (13). Without the LH surge, follicles will fail to ovulate and CL will decrease in number. The relative decrease in CL and increase in anovulatory follicles results in the ovarian morphology observed. Interestingly, there was clear histologic evidence that a few animals had ovulated during the 35-d in vivo study. Basophilic CL were diagnostic for recent (few days) ovulation. These data along with decreased progesterone levels suggest that this SERM may have a contraceptive effect, but the consistency of the effect is unknown and suggests the need for further studies.
There was minimal ovarian stimulation in LY2066948-treated rats. There were no ovarian cysts, and granulosa cell hyperplasia observed in a few animals was usually minimal. The ovarian effects seen at higher dose levels would likely be fully reversible, given previous findings with raloxifene (13). SERM-mediated ovarian stimulation in rats was proposed to be primarily the result of sustained, elevated LH levels due to loss of estrogen feedback mechanisms in the hypothalamic-pituitary axis (13). Therefore, it was not unexpected to find that LH levels were not significantly elevated in rats given LY2066948. Treatment of rats with LY560212 resulted in clear ovarian stimulation with an approximately 3-fold higher granulosa cell hyperplasia score, compared with rats treated with the highest dose of LY2066948. However, LH levels were only modestly elevated by LY560212 relative to LY2066948, suggesting that LY560212 (and other SERMs) exert direct effects on the ovaries as well as through the HPO axis (see below).
The reason(s) for the improved ovarian profile of LY2066948 are not clear. One contributing possibility may be low brain exposure of the compound, decreasing the chance that the compound can bind to hypothalamic estrogen receptors and stimulate the HPO axis. This hypothesis is consistent with the relative whole brain concentrations observed for LY2066948 in comparison with LY560212. The brain level of LY2066948 after oral administration to rats (10 mg/kg, n = 3/group) was found to be 103 ± 28 ng/g, and that of LY560212 was 2692 ± 211 ng/g after 6 h. This represents a more than 25 times increase in brain exposure of LY560212 relative to LY2066948. Many SERMs cause elevated levels of LH (and the highest stimulation of estrogen production), suggesting a hypothalamic GnRH mechanism or direct pituitary effects in animals and women (13, 45). Decreased LY2066948 exposure in the hypothalamic loci responsible for GnRH production may be responsible for limiting HPO stimulation relative to other SERMs. However, because it appears that at least part of the hypothalamus is outside the blood-brain barrier and the pituitary is clearly outside the blood-brain barrier (46), a relatively low brain exposure may not necessarily lead to neutral HPO effects. In fact, we did see evidence of an estrogen-like effect of LY2066948 on the hypothalamo-pituitary axis (Fig 7) in OVX rats, suggesting that even if the compound reaches the hypothalamus or pituitary in pharmacologic concentrations, its activity there might not lead to a block of hormonal feedback mechanisms.
Cells in the ovary express ERs (47), and estrogen stimulatory effects on follicular granulosa cell growth and differentiation have been widely reported (48). Therefore, it is possible that some SERMs have direct stimulatory effects on the ovaries, whereas LY2066948 does not or the effects are reduced. Reduction of direct ovarian effects could be a reflection of unique pharmacokinetic properties of the compound that limit exposure to the ovaries or unique conformation of the estrogen receptors induced by binding LY2066948. This conformational change may lead to unique cofactor/ER interactions. It is interesting to note that the ovary is a tissue in which ER expression appears to be high relative to ER (49). After examining the data from all of the SERMs used in our studies, we assessed whether there were differences in relative ER/ binding affinity that might correlate with the degree of ovarian stimulation, but no such correlation was found.
In summary, we have discovered a SERM, LY2066948, that shows good separation between desired pharmacology in the uterus and undesired pharmacology in the ovary. These studies have provided compelling data to support clinical evaluation of uterine antagonist efficacy as a possible treatment for gynecologic diseases. In addition to the desirable aspects of a safe pharmacologic treatment for uterine fibroids, other benefits over current therapies may be realized through reduced side effects mediated by estrogen agonism in selected tissues, allowing preservation of bone mass and normal vasomotor and cardiovascular homeostasis.
Acknowledgments
We thank Chahrzad Montrose-Rafizadeh, Sandy Cockerham, Ed Osbourne, Robert Amos, Pam Shetler, Ellen Rowley, and Harlan Cole for excellent help they have provided in the lead optimization biology labs and the animal rooms to make this study possible. We also thank Steve Iturria for statistical interpretation.
Footnotes
Abbreviations: BMC, Bone mineral content; BMD, bone mineral density; CL, corpora lutea; DCC, dextran-coated charcoal; E2, 17-estradiol; EE, ethynyl estradiol; ER, estrogen receptor; HPO, hypothalamo-pituitary-ovarian; Ki, inhibitory constant; OVX, ovariectomized; P4, progesterone; QCT, quantitative computed tomography; SD, Sprague Dawley; SERM, selective estrogen receptor modulator; UWR, uterine weight to body weight ratio.
References
Buttram Jr VC, Reiter RC 1981 Uterine leiomyomata: etiology, symptomatology, and management. Fertil Steril. 36:433–445
Cramer SF, Patel A 1990 The frequency of uterine leiomyomas. Am J Clin Pathol. 94:435–438
Stewart EA 2001 Uterine fibroids. Lancet 357:293–298
Friedman AJ, Hoffman DI, Comite F, Browneller RW, Miller JD 1991 Treatment of leiomyomata uteri with leuprolide acetate depot: a double-blind, placebo-controlled, multicenter study. The Leuprolide Study Group. Obstet Gynecol. 77:720–725
Draper MW, Chin WW 2003 Molecular and clinical evidence for the unique nature of individual selective estrogen receptor modulators. Clin Obstet Gynecol. 46:265–297
Sato M, Rippy MK, Bryant HU 1996 Raloxifene, tamoxifen, nafoxidine, or estrogen effects on reproductive and nonreproductive tissues in ovariectomized rats. FASEB J. 10:905–912
Huppert LC 1979 Induction of ovulation with clomiphene citrate. Fertil Steril. 31:1–8
Adashi EY 1984 Clomiphene citrate: mechanism(s) and site(s) of action—a hypothesis revisited. Fertil Steril. 42:331–344
Baker VL, Draper M, Paul S, Allerheiligen S, Glant M, Shifren J, Jaffe RB 1998 Reproductive endocrine and endometrial effects of raloxifene hydrochloride, a selective estrogen receptor modulator, in women with regular menstrual cycles. J Clin Endocrinol Metab. 83:6–13
Plouffe Jr L, Siddhanti S 2001 The effect of selective estrogen receptor modulators on parameters of the hypothalamic-pituitary-gonadal axis. Ann NY Acad Sci. 949:251–258
Mourits MJ, de Vries EG, Willemse PH, ten Hoor KA, Hollema H, Sluiter WJ, de Bruijn HW, van der Zee AG 1999 Ovarian cysts in women receiving tamoxifen for breast cancer. Br J Cancer. 79:1761–1764
Bucci TJ 1991 Evaluation of altered morphology. In: Haschek WM, Rousseaux CG, eds. Handbook of toxicologic pathology. San Diego: Academic Press; 23–35
Long GG, Cohen IR, Gries CL, Young JK, Francis PC, Capen CC 2001 Proliferative lesions of ovarian granulosa cells and reversible hormonal changes induced in rats by a selective estrogen receptor modulator. Toxicol Pathol. 29:403–410
Peluso J 1992 Morphologic and physiologic features of the ovary. In: Mohr U, Dungwort DL, Capen CC, eds. Pathobiology of the aging rat. Vol 1. Washington, DC: ILSI Press; 342–344
Sato M 1995 Comparative x-ray densitometry of bones from ovariectomized rats. Bone 17:157S–162S
DeFriend DJ, Howell A, Nicholson RI, Anderson E, Dowsett M, Mansel RE, Blamey RW, Bundred NJ, Robertson JF, Saunders C, Baum M, Walton P, Sutcliffe F, Wakeling AE 1994 Investigation of a new pure antiestrogen (ICI 182780) in women with primary breast cancer. Cancer Res. 15:408–414
Littlefield BA, Gurpide E, Markiewicz L, McKinley B, Hochberg RB 1990 A simple and sensitive microtiter plate estrogen bioassay based on stimulation of alkaline phosphatase in Ishikawa cells: estrogenic action of delta 5 adrenal steroids. Endocrinology 127:2757–2762
Bramlett KS, Burris TP 2003 Target specificity of selective estrogen receptor modulators within human endometrial cancer cells. J Steroid Biochem Mol Biol. 86:27–34
Katzenellenbogen B, Kendra K, Norman M, Berthois Y 1987 Proliferation, hormonal responsiveness, and estrogen receptor content of MCF-7 human breast cancer cells grown in the short-term and long-term absence of estrogens. Cancer Res. 47:4355–4360
Black L, Goode R 1980 Uterine bioassay of tamoxifen, trioxifene, and a new estrogen antagonist (LY117018) in rats and mice. Life Sci. 26:1453–1458
Dodge JA, Lugar CW, Fahey KJ, Cullinan GJ, Phillips DL, Short LL, Glasebrook AL, Sato M, Muehl BS, Kim JK, Yang NN, Magee DE, Cole HW, Bryant HU 1996 A novel estrogen receptor modulator with selectivity for bone over uterus in rats. J Bone Mineral Res. 8:T589
Palkowitz AD, Glasebrook AL, Thrasher KJ, Hauser KL, Short LL, Phillips DL, Muehl BS, Sato M, Shetler PK, Cullinan GJ, Pell TR, Bryant HU 1997 Discovery and synthesis of [6-Hydroxy-3-[4-[2-(1-piperidinyl)ethoxy]phenoxy]-2-(4-hydroxyphenyl)]bezo[b]thiophene: a novel, highly potent, selective estrogen receptor modulator. J Med Chem. 40:1407–1416
Kym P, Anstead G, Pinney K, Wilson SR, Katzenellenbogen J 1993 Molecular structures, conformational analysis, and preferential modes of binding of 3-aroyl-2-arylbenzo[b]thiophene estrogen receptor ligands: LY117018 and aryl azide photoaffinity labeling analogs. J Med Chem. 36:3910–3922
Sibonga JD, Evans GL, Hauck ER, Bell NH, Turner RT 1996 Ovarian status influences the skeletal effects of tamoxifen in adult rats. Breast Cancer Res Treat. 41:71–79
Powles T, Hickish T, Kanis J, Tidy A, Ashley S 1996 Effect of tamoxifen on bone mineral density measured by dual-energy x-ray absorptiometry in healthy premenopausal and postmenopausal women. J Clin Oncol. 4:78–84
Black L, Sato M, Rowley E, Magee D, Bekele A, Williams D, Cullinan G, Bendele R, Kauffman R, Bensch W, Frolik C, Termine J, Bryant HU 1994 Raloxifene (LY139481 HCl) prevents bone loss and reduces serum cholesterol without causing uterine hypertrophy in ovariectomized rats. J Clin Invest. 93:63–69
Everitt JI, Wolf DC, Howe SR, Goldsworthy TL, Walker C 1995 Rodent model of reproductive tract leiomyomata. Clinical and pathological features. Am J Pathol. 146:1556–1567
Walker CL, Burroughs KD, Davis B, Sowell K, Everitt JI, Fuchs-Young R 2000 Preclinical evidence for therapeutic efficacy of selective estrogen receptor modulators for uterine leiomyoma. J Soc Gynecol Investig. 7:249–256
Delmas PD, Bjarnason NH, Mitlak BH, Ravoux AC, Shah AS, Huster WJ, Draper M, Christiansen C 1997 Effects of raloxifene on bone mineral density, serum cholesterol concentrations, and uterine endometrium in postmenopausal women. N Engl J Med. 337:1641–1647
Palomba S, Sammartino A, Di Carlo C, Affinito P, Zullo F, Nappi C 2001 Effects of raloxifene treatment on uterine leiomyomas in postmenopausal women. Fertil Steril. 76:38–43
Fisher B, Costantino JP, Redmond CK, Fisher ER, Wickerham DL, Cronin WM 1994 Endometrial cancer in tamoxifen-treated breast cancer patients: findings from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14. J Natl Cancer Inst. 86:527–537
Schwartz LB, Rutkowski N, Horan C, Nachtigall LE, Snyder J, Goldstein SR 1998 Use of transvaginal ultrasonography to monitor the effects of tamoxifen on uterine leiomyoma size and ovarian cyst formation. J Ultrasound Med. 17:699–703
Porter K, Tsibris J, Nocosia S, Murphy J, O’Brien W, Rao P, Spellacy W 1995 Estrogen-induced guinea pig model for uterine leiomyomas: do the ovaries protect Biol Reprod. 52:824–832
Porter KB, Tsibris JC, Porter GW, Fuchs-Young R, Nicosia SV, O’Brien WF, Spellacy WN 1998 Effects of raloxifene in a guinea pig model for leiomyomas. Am J Obstet Gynecol. 179:1283–1287
Tsibris JC, Porter KB, Jazayeri A, Tzimas G, Nau H, Huang H, Kuparadze K, Porter GW, O’Brien WF, Spellacy WN 1999 Human uterine leiomyomata express higher levels of peroxisome proliferator-activated receptor , retinoid X receptor , and all-trans retinoic acid than myometrium. Cancer Res. 59:5737–5744
Palomba S, Orio Jr F, Morelli M, Russo T, Pellicano M, Zupi E, Lombardi G, Nappi C, Panici PL, Zullo F 2002 Raloxifene administration in premenopausal women with uterine leiomyomas: a pilot study. J Clin Endocrinol Metab. 87:3603–3608
Palomba S, Russo T, Orio Jr F, Tauchmanova L, Zupi E, Panici PL, Nappi C, Colao A, Lombardi G, Zullo F 2002 Effectiveness of combined GnRH analogue plus raloxifene administration in the treatment of uterine leiomyomas: a prospective, randomized, single-blind, placebo-controlled clinical trial. Hum Reprod. 17:3213–3219
Jirecek S, Lee A, Pavo I, Crans G, Eppel W, Wenzl R 2004 Raloxifene prevents the growth of uterine leiomyomas in premenopausal women. Fertil Steril. 81:132–136
Palomba S, Zullo F, Orio Jr F, Lombardi G 2004 Does raloxifene inhibit the growth of uterine fibroids Fertil Steril. 81:1719–1720
Jirecek S, Lee A, Eppel W, Wenzl R, Pavo I, Gerald C 2004 Reply: does raloxifene inhibit the growth of uterine fibroids Fertil Steril. 81:1720–1721
Lumsden MA, West CP, Baird DT 1989 Tamoxifen prolongs luteal phase in premenopausal women but has no effect on the size of uterine fibroids. Clin Endocrinol. 31:335–343
Sadan O, Ginath S, Sofer D, Rotmensch S, Debby A, Glezerman M, Zakut H 2001 The role of tamoxifen in the treatment of symptomatic uterine leiomyomata—a pilot study. Eur J Obstet Gynecol Reprod Biol. 96:183–186
Ugwumadu AH, Harding K 1994 Uterine leiomyomata and endometrial proliferation in postmenopausal women treated with the anti-oestrogen tamoxifen. Eur J Obstet Gynecol Reprod Biol. 54:153–156
Kang J, Baxi L, Heller D 1996 Tamoxifen-induced growth of leiomyomas. A case report. J Reprod Med. 41:119–120
Dukes M, Chester R, Yarwood L, Wakeling AE 1994 Effects of a nonsteroidal pure antioestrogen, ZM 189,154, on oestrogen target organs of the rat including bones. J Endocrinol. 141:335–341
Landas S, Fischer J, Wilkin LD, Mitchell LD, Johnson AK, Turner JW, Theriac M, Moore KC 1985 Demonstration of regional blood-brain barrier permeability in human brain. Neurosci Lett. 57:251–256
Brandenberger AW, Tee MK, Jaffe RB 1998 Estrogen receptor (ER-) and (ER-) mRNAs in normal ovary, ovarian serous cystadenocarcinoma and ovarian cancer cell lines: down-regulation of ER- in neoplastic tissues. J Clin Endocrinol Metab. 83:1025–1028
Palter SF, Tavares AB, Hourvitz A, Veldhuis JD, Adashi EY 2001 Are estrogens of import to primate/human ovarian folliculogenesis Endocr Rev. 22:389–424
Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson JA 1996 Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA. 93:5925–5930(Andrew G. Geiser, Conrad )
Array Biopharma (C.W.H.), Boulder, Colorado 80301
Abstract
The use of selective estrogen receptor modulators for the treatment of estrogen-dependent diseases in premenopausal women has been hindered by undesirable ovarian stimulation and associated risks of ovarian cysts. We have identified a selective estrogen receptor modulator compound (LY2066948) that is a strong estrogen antagonist in the uterus yet has minimal effects on the ovaries of rats. LY2066948 binds with high affinity to both estrogen receptors and has potent estrogen antagonist activity in human uterine and breast cancer cells. Oral administration of LY2066948 to immature rats blocked uterine weight gain induced by ethynyl estradiol with an ED50 of 0.07 mg/kg. Studies in mature rats demonstrated that LY2066948 decreases uterine weight by 51% after 35 d treatment, confirming potent uterine antagonist activity over several estrous cycles. This strong uterine response contrasted with the minimal effects on the ovaries: serum estradiol levels remained within the normal range, whereas histologic evaluation showed granulosa cell hyperplasia in few of the rats. Bone studies demonstrated that LY2066948 prevented ovariectomy-induced bone loss and treatment of ovary-intact rats caused no bone loss, confirming estrogen receptor agonist skeletal effects. Collectively, these data show that LY2066948 exhibits a tissue-specific profile consistent with strong antagonist activity in the uterus, agonist activity in bone, and minimal effects in the ovaries.
Introduction
UTERINE FIBROIDS (LEIOMYOMAS) are benign tumors arising from smooth muscle cells of the myometrium (1, 2, 3). They are the most common type of solid tumor in adult women, clinically apparent in approximately 25% of those of reproductive age. In women who experience symptoms, abnormal menstrual bleeding, pelvic pain, pressure symptoms, and infertility are the most common. Uterine fibroids and their associated complications often necessitate hysterectomy, accounting for more than 200,000 of these procedures each year in the United States. Less drastic surgical procedures such as myomectomy (fibroid removal with uterine retention), laser ablation, or uterine artery embolization are available but have been associated with complications and fibroid recurrence (3).
Clinical data suggest that estrogens are important in the regulation of fibroid growth. These tumors increase rapidly in size during the hyperestrogenic state brought on by pregnancy and regress in hypoestrogenic states brought on by menopause or ovariectomy. Thus, pharmacologic approaches have centered on the hormone dependency of fibroids, especially on blockade of estrogenic stimulation. One such therapeutic approach is treatment with GnRH agonists. GnRH efficacy is the result of down-regulation of the hypothalamo-pituitary-ovarian (HPO) axis. GnRH agonists desensitize the pituitary gland to GnRH, decreasing release of the gonadotropins FSH and LH and in turn reduce the amount of estrogens produced in the ovaries to castrate levels. These agents show efficacy but carry a high side effect burden, limiting chronic treatment and relegating use to preoperative applications (4). Side effects associated with lowered levels of ovarian estrogens, such as hot flashes and bone loss, are common with GnRH use.
Clearly there exists a need for improved therapies to treat uterine fibroids but also to avoid the drawbacks without the risks and side effects associated with current surgical and pharmacological interventions. One particularly promising class of compounds that could fill this need is selective estrogen receptor modulators (SERMs) (5). SERMs bind to estrogen receptors (ER) and impart tissue-dependent (selective) antagonistic or agonistic effects. SERMs currently on the market include tamoxifen (Nolvadex; AstraZeneca PLC, London, UK), primarily indicated for adjuvant treatment of breast cancer, and raloxifene (Evista; Eli Lilly and Co., Indianapolis, IN), indicated for prevention and treatment of postmenopausal osteoporosis. Both of these SERMs have tissue-specific ER-mediated pharmacology. They share antagonist properties in breast and agonist properties in bone but differ at the uterus in that tamoxifen acts as a partial agonist, whereas raloxifene is an antagonist (6). Another SERM, clomiphene (Clomid; Sanofi-Aventis SA, Paris, France), is used in premenopausal women to induce ovulation by blocking ERs in the hypothalamus (7). This antiestrogenic action results in stimulation of the HPO axis through antagonism of the steroidal negative feedback loop and stimulation of GnRH release, resulting in increased production of LH and FSH and ultimately ovarian stimulation. Evidence has also pointed to a direct estrogenic effect of clomiphene on the ovaries, leading to ovarian stimulation (8). Clinical results in premenopausal women have suggested that tamoxifen and raloxifene affect the ovaries by stimulating the HPO axis and, possibly, the ovary directly (9, 10). Ovarian stimulation is not desired in many premenopausal applications because of the risk of follicular cyst formation. Tamoxifen treatment of premenopausal women does in fact increase circulating estrogens and the incidence of follicular cysts (11).
To address the long-standing issue of ovarian hyperstimulation observed with currently available SERMs, we sought to identify a more tissue-selective agent for potential use in the treatment of uterine fibroids. Structure-activity studies were directed at identifying compounds that displayed a high degree of estrogen antagonist activity in the uterus yet were relatively neutral for the ovaries and displayed estrogen agonist properties in bone. In this paper we describe the discovery of LY2066948, a novel SERM that possesses potent uterine antagonist activity and minimal ovarian stimulation, making this a potential drug candidate suitable for treatment of premenopausal women suffering from uterine fibroids and other estrogen-dependent gynecologic applications.
Materials and Methods
SERM compound sources
Tamoxifen and 4-hydroxytamoxifen were purchased from Sigma (St. Louis, MO). Note that tamoxifen was used in the in vivo studies (structure shown, Table 1) whereas 4-hydroxytamoxifen was used in the in vitro studies. ICI 182780 was purchased from Tocris (Ellisville, MO). All other SERMs used in this study were synthesized in the Lilly Research Labs (Indianapolis, IN). All studies were done with the hydrochloride salt form of 2066948 (2066948-HCl).
ER binding
The competition binding assay was run in a buffer containing 50 mM HEPES (pH 7.5), 1.5 mM EDTA, 150 mM NaCl, 10% glycerol, 1 mg/ml ovalbumin, and 5 mM dithiothreitol, using 0.025 μCi/well 3H-estradiol [no. NET517 at 118 Ci/mmol, 1.5 nM 17-estradiol (E2); NEN Life Science Products/PerkinElmer, Boston, MA] 10 ng/well ER or ER receptor (PanVera/Invitrogen, Carlsbad, CA). Nonspecific binding was determined in the presence of 1 μM E2. The binding reaction was incubated for 4 h at room temperature, and then cold dextran-coated charcoal (DCC) buffer was added to each reaction [DCC buffer contains, per 50 ml of assay buffer, 0.75 g charcoal (Sigma) and 0.25 g dextran (Pharmacia, Uppsala, Sweden)]. Plates were mixed 8 min on an orbital shaker at 4 C and then centrifuged at 3000 rpm at 4 C for 10 min. An aliquot of the mix was added to Optiphase Hisafe 3 scintillation fluid [Wallac (PerkinElmer)], incubated for 2.5 h, and read in a Wallac Microbeta counter. The equilibrium dissociation (Kd) constant for 3H-estradiol was determined by saturation binding to ER and ER receptors. The IC50 values for compounds were converted to inhibitory equilibrium dissociation constant values (Ki) using the Cheng-Prusoff equation.
Ishikawa assay
Estrogenic stimulation and antagonism were measured in Ishikawa human endometrial tumor cells by alkaline phosphatase quantitation. The cells were incubated in DMEM/F-12 (3:1) supplemented with 5% DCC-stripped fetal bovine serum (Hyclone, Logan, UT), L-glutamine (2 mM), MEM sodium pyruvate (1 mM), HEPES (2 mM) (all from Life Technologies, Inc., Gaithersburg, MD). For the agonist mode, plates received diluted compounds only, whereas antagonist mode plates additionally received 1 nM estradiol (E2, Sigma). Cells were incubated for 48 h, and then fresh compounds were added for an additional 72 h. The assay was quenched by removing medium and rinsing plates twice in PBS, and the plates were dried for 5 min and frozen at –70 C. After thawing, 100 μl of 1-Step p-nitrophenyl phosphate (PNPP) (Pierce Chemical Co., Rockford, IL) were added for 20 min, and plates were read on a spectrophotometer at 405 nm. For the agonist mode, a percentage increase over control was calculated. The data were fitted to a linear interpolation to derive IC50 values for the antagonist mode, and a percentage efficacy was calculated that blocks the E2 (1 nM) stimulus.
MCF-7 assay
MCF-7 breast adenocarcinoma cells (HTB 22; American Type Culture Collection, Manassas, VA) were grown in assay medium MEM (phenol red-free; Life Technologies) supplemented with 10% DCC-stripped fetal bovine serum. Cells were assayed by plating 8000 cells in each well of 96-well Cytostar T scintillation plates (Amersham, Piscataway, NJ) and incubated at 37 C for 24 h before adding compounds. For antagonist mode, 10 pM E2 were added along with dilutions of compound for 48 h, and then medium containing 0.01 μCi of 14C-thymidine (52 mCi/mmol, 50 μCi/μl; Amersham) was added to each well. The plates were incubated overnight and then quantitated on a Wallac Microbeta counter. The data were averaged to calculate an IC50 and percent inhibition at 1 μM compound.
Three-day rat uterine antagonist
All rat experiments were conducted in accord with accepted standards of humane animal care and in accordance of the Animal Care and Use Committee at Eli Lilly & Co.
Female Sprague Dawley (SD) rats, six per group and 19–21 d of age, were orally treated with ethynyl estradiol (EE; 0.1 mg/kg) and 10, 1.0, 0.1, or 0.01 mg/kg SERM for 3 d. Test compounds were dissolved in 20% (wt/vol) -hydroxycyclodextrin in water and administered by oral gavage in a volume of 0.2 ml daily (15 min before the EE gavage). Groups of six rats were also given vehicle as a negative control, EE alone as a positive control, and EE plus LY117018 as a standard antagonist control. The animals were fasted overnight after the final dose. On the following morning, the animals were weighed and then killed (by carbon dioxide asphyxiation), and the uteri were rapidly collected (via a midline ventral incision), stripped of adipose tissue, luminal fluid removed by blotting onto absorbant paper, and weighed.
Uterine weight to body weight ratios (UWR) were calculated for each animal. The percentage inhibition of the estrogen-induced response was then calculated by the following formula: percent inhibition = 100 x (UWRE2 – UWRtest compound/UWRE2 – UWRcontrol). ED50 values were derived from a semilog regression analysis of the linear aspect of the dose-response curve.
Ovariectomized (OVX) rat assays
For the 42-d assay (see Figs. 2 and 6), virgin 6-month-old, SD female rats (Harlan, Indianapolis, IN) weighing about 270 g were bilaterally ovariectomized under isoflurane anesthesia, randomized to treatment groups, and dosed, beginning on the third day after surgery. For the 7-d assay (see Fig. 7), virgin 75-d-old, SD female rats (Harlan) weighing about 240 g were treated as above with dosing beginning 14 d after surgery (rats now 3 months of age). Groups from both experiments were orally dosed each day in a vehicle of 20% hydroxypropyl--cyclodextrin (Aldrich Chemical Co., Milwaukee, WI). Uteri were removed and analyzed as in the 3-d rat assay above. Quantitative determination of cholesterol levels was achieved by measurement of cholesterol esterase/cholesterol oxidase activity using a Hitachi-917 automated chemistry analyzer manufactured by Roche Diagnostics (Indianapolis, IN). LH was quantitated from sera using a rodent LH ELISA kit from Endocrine Technologies, Inc. (Newark, CA).
Ten-day hormone assay
Female Fischer 344 rats (9–10 wk old) were dosed with 30 times the ED50 dose value derived from the 3-d rat uterine antagonist model. Trunk blood was collected 2 h after the final dose. Serum levels of estradiol were measured by RIA, using an assay kit from DiaSorin (Saluggia, Italy).
Fourteen- and 35-d intact rat uterine efficacy
Mature, virgin, 3- (14-d assay) or 6-month-old (35-d assay) SD rats were administered compound by oral gavage daily. Animals were killed and the uteri collected and weighed as described for the 3-d rat uterine antagonist assay. For histology and morphometry in the 35-d study, uteri and ovaries were removed, weighed, and fixed in 10% neutral buffered formalin. Tissues were trimmed, processed, and embedded in paraffin blocks for histology. For trimming, transverse sections of the ovary and the midpoint of the left uterine horn were taken. Five-micrometer sections were stained with hematoxylin and eosin and evaluated by a board-certified veterinary pathologist for histologic changes. The tissue sections were examined by light microscopy by an American College of Veterinary Pathologists board-certified veterinary pathologist. Histologic changes were described, when applicable, according to their distribution, severity, and morphologic character (12) using standardized nomenclature for the rat and lesion definitions for SERM ovarian effects published elsewhere (13). The diagnosis of ovarian follicular cyst was reserved for prominent ovarian follicles that had a diameter of at least 2 mm, which was not observed in this study (14). Qualitative severity scores for the ovarian changes were assigned as follows: 1, minimal; 2, slight; 3, moderate; and 4, marked. Mean group lesion scores for individual groups were calculated by adding the individual animal severity scores and dividing by the total number of animals.
Morphometry was done on the uteri from all control (n = 6) and 0.5 mg/kg LY2066948-treated (n = 7) rats. Uteri were imaged at x25 magnification using a digital camera (Spot Diagnostic Instruments, Sterling Heights, MI) attached to a DMLB microscope (Leica Microsystems AG, Wetzlar, Germany). Cross-sectional area (square micrometers) of the uterine wall, endometrium, and myometrium were measured using Spot morphometry software. LH was measured by RIA as described in (13), progesterone (P4) was measured by active progesterone RIA (Diagnostic Systems Laboratories Inc., Webster TX). For the progesterone assay, the upper limit of detection was 60 ng/ml and the lower limit of detection was 0.3 ng/ml for a 25-μl sample volume.
Bone analysis
Vertebrae and femora were excised at necropsy and the midtransverse section of the lumbar vertebra L-4 (see Table 3) and distal femur metaphysis (see Fig. 6) was scanned in 50% ethanol/saline, using quantitative computed tomography (QCT) (Research M, Norland/Stratec, Ft. Atkinson, WI). Cross-sectional area, bone mineral content (BMC, milligrams), and volumetric bone mineral density (milligrams per cubic centimeter) were quantitated, using voxel dimensions of 148 x 148 x 500 μm as previously described (15).
Statistical analyses
One-way ANOVA was followed by Dunnett’s post hoc test when the one-way ANOVA was statistically significant. Dunnett’s test (using JMP 5.1 software, SAS Institute Inc., Cary, NC) was used to assess the statistical significance (P < 0.05) of treatment groups with respect to a nontreatment vehicle control group for all analyses.
Results
In vitro characterization
Compounds were first selected for high affinity binding to both estrogen receptor types (ER and ER). A cell-free competition-binding assay was used to measure the propensity to displace 3H-estradiol and generate Ki values for a library of SERM compounds. Our eventual lead compound, LY2066948-HCl (herein referred to as LY2066948), demonstrated high affinity binding to both human ERs, with a Ki- of 0.51 nM and Ki- of 1.36 nM. Table 1 lists ER binding affinities and other activities described below for LY2066948 and other SERMs including raloxifene, the mixed uterine agonist/antagonist tamoxifen (6), and the pure ER antagonist ICI 182780 (16).
The Ishikawa human uterine cell line was used to investigate potential uterine antagonist activity in vitro. Alkaline phosphatase activity was measured as an end point for relative estrogenic stimulation in both an agonist mode (in the presence of the SERM alone) and an antagonist mode (dose response for the SERM to block the stimulatory effect of a fixed dose of E2) (17, 18). In the antagonist mode, LY2066948 blocked the 1 nM E2 stimulatory response by 87.5% with an IC50 of 10.7 nM, compared with 53.4% inhibition and an IC50 of 421 nM for 4-hydroxytamoxifen, a first-generation SERM with known uterotropic activity (see Table 1). The agonist activity of LY2066948 alone was only 29% over control, compared with 123% with 4-hydroxytamoxifen. These data indicate that LY2066948 is a potent antagonist of uterine cell stimulation with significantly less agonist characteristics than 4-hydroxytamoxifen.
We also examined growth-inhibitory effects of LY2066948 in MCF-7 cells that were originally derived from a human breast adenocarcinoma and are frequently used in estrogenicity studies (19). These mammary cells express high numbers of ERs and demonstrate a dose-dependent increase in proliferation with increasing concentrations of estradiol. The growth effects of SERMs in this assay were examined by conducting the assay in both the agonist mode (absence of E2) and antagonist mode (presence of 10 pM E2). LY2066948 demonstrated no agonist activity up to 1000 nM (data not shown) but was able to antagonize E2 stimulation with an IC50 of 0.86 nM, confirming potent antiestrogenic activity in this cell line and suggesting estrogen antagonist potential in the mammary gland (Table 1).
In vivo uterine efficacy
To assess the potential of compounds to block uterine fibroid growth, we relied on the ability of compounds to oppose either exogenous or endogenous estradiol stimulation of the rat uterus as measured by changes in uterine wet weight. All of the following in vivo assays were carried out with various doses of test compound administered orally on a daily basis.
Uterine antagonism was measured in immature (3 wk old) female rats, which are highly sensitive to estrogenic stimulation of the uterus, given that their circulating estrogen levels are prepubertal (20). Administration of EE to immature rats produces a reliable increase (typically 3- to 4-fold) in uterine weight. In this assay, LY2066948 was found to be a potent uterine antagonist, completely inhibiting the estrogenic stimulus at 10 mg/kg and giving an ED50 value of 0.07 (± 0.02) mg/kg (Fig. 1 and Table 1). This potency compares favorably with raloxifene (21, 22), tamoxifen (partial agonist, partial antagonist), and the pure ER antagonist ICI 182780 (Table 1).
To verify that LY2066948 does not act as a mixed agonist/antagonist in the uterus after an extended period of treatment, the compound was administered to mature, OVX rats for 42 d and then uteri were analyzed. Castration dramatically decreases uterine weight and allows for sensitive measurements of potential estrogenic agonist activity. Slight uterine weight gains over OVX control were noted at all doses (no dose response seen), amounting to an 18% increase at the highest (10 mg/kg) dose administered and comparable with that seen with the raloxifene analog, LY117018 (23) (Fig. 2A). In this same experiment, we measured serum cholesterol levels because many SERMs have demonstrated a cholesterol-lowering effect (6, 22). LY2066948 lowered total serum cholesterol by 62–74% at tested doses, confirming this desirable SERM property (Fig. 2B).
Estrogen antagonism in the uterus of ovary-intact rats
LY2066948 was examined for the ability to antagonize estrogen effects on uterine wet weight in ovary-intact rats with circulating endogenous estrogen. Two different experiments were performed, differing in the duration of dosing. In the first experiment, 75-d-old rats were given LY2066948 for 14 d, resulting in a reduction in uterine wet weight of 37% at a dose of 1 mg/kg (data not shown). In the second study, 6-month-old rats were administered LY2066948 for 35 d at doses up to 3 mg/kg, greater than 40 times the ED50 of 0.07 mg/kg for uterine weight reduction in the immature rat study. Uterine antagonism was maximal at 1 mg/kg, with a 51% reduction in uterine wet weight (Fig. 3A), and the ED50 was estimated to be approximately 0.15 mg/kg. Morphometric measurements were carried out on all animals of the 0.5 mg/kg dose group and compared with the control group. At this dose of LY2066948, a uterine weight reduction of 43% was noted, compared with control. The total area of the uterine wall (endometrium and myometrium combined) was significantly reduced by 53% with nearly equal contributions in reduction of endometrial area (54%) and myometrial area (51%) (Fig. 3, B and C). Thus, these data confirm substantial antagonist activity in both the endometrial and myometrial compartments of the uterus in the presence of circulating levels of estradiol over several estrous cycles (one cycle is 4–5 d) in the rat.
Ovarian evaluation
To assess potential stimulatory effects of SERMs on the ovaries, we measured serum estradiol levels after 10 d of oral administration at doses exceeding the uterine ED50 dose by 30-fold. Compounds chosen for further analysis were those that produced less than 2-fold increases in serum estradiol concentrations. Nearly all compounds failed this ovarian safety hurdle with the exception of LY2066948, which had no significant effects on serum estradiol concentrations. Given this unique profile among the more than 120 compounds evaluated in this assay, LY2066948 became the focus for longer-term studies to further evaluate ovarian effects.
The effect of LY2066948 on the ovaries was assessed by evaluating changes in serum estrogen and LH, ovarian weight, and morphology after 35 d of oral administration. There was a dose-dependent increase in serum estradiol levels with doses greater than 0.5 mg/kg, and at 3.0 mg/kg, this increase was significant (Fig. 4). However, these levels did not exceed the normal mean proestrus level for SD rats of approximately 52 pg/ml (Cohen, I. R., unpublished observation). Serum LH was comparable with controls in rats administered LY2066948, although there was a trend toward LH elevation at higher doses. Mean ovarian weights were decreased in LY2066948-treated rats given 0.5 mg/kg or more, compared with controls (Table 2). The decreased ovarian weights corresponded to the moderate to marked depletion of corpora lutea (CL) that was observed in rats given LY2066948. Despite the general finding of corpora lutea depletion in LY2066948-treated rats, a few rats given 1 mg/kg or less had basophilic CL consistent with recent ovulation. Serum P4 levels were generally less than control, consistent with CL depletion in LY2066948-treated animals (see Table 2). However, P4 levels of some treated rats were of sufficient magnitude to indicate that ovulation had occurred.
The follicular prominence scores were generally comparable at doses between 0.05 and 1 mg/kg. Follicular prominence was characterized by dilated antral follicles that were lined by an orderly layer of attenuated granulosa cells. Oocytes within prominent follicles were uncommon and if present were often degenerate. The follicular prominence was consistent with retention of anovulatory follicles and in combination with the CL depletion indicated a prolongation of the ovarian cycle at the preovulatory phase.
Granulosa cell hyperplasia was observed in only a few animals. No rats had hyperplasia at a dose comparable with the ED50 for uterine antagonism in the immature rat assay, and only a single rat had minimal hyperplasia at the 0.5 mg/kg dose. When present, hyperplasia was usually graded minimal. The minimal grade was based on the small number of follicles affected (usually one or two), the focal nature of the hyperplasia, and the morphologic characteristics of the hyperplasia change; the hyperplasia consisted of papillary infoldings of the granulosa cells into the follicular lumen with subadjacent infoldings of thecal cells supported by a vascular bud. Ovarian cysts, a common manifestation of most SERMs, were not observed in rats given LY2066948 at doses up to 3 mg/kg.
We compared the ovarian effects of LY2066948 with those of rats treated with LY560212, a close analog of LY2066948 in which the methyl sulfone moiety was replaced with a fluorine atom. Whereas LY560212 is a very potent uterine antagonist (3-d uterine ED50 of 0.02 mg/kg), it also displays the more typical SERM profile of stimulating estradiol levels (4.6-fold, data not shown) in the 10-d hormone studies. In contrast to the modest ovarian effects seen with LY2066948, dosing with 0.45 mg/kg of LY560212 resulted in marked ovarian stimulation characterized by a 43% increase in ovarian weight and nearly a 1.5- and 3-fold higher histologic score for follicular prominence and granulosa cell hyperplasia, respectively (Table 2 and Fig. 5). The accentuated ovarian stimulation may have been due in part to the effects of increased levels of the trophic hormone, LH. LH was approximately 2-fold greater in LY560212 rats, compared with controls, but did not reach statistical significance (P = 0.13, Table 2).
Skeletal effects
Another end point of interest in the 35-d assay (uterine and ovarian effects shown in Figs. 3–5) was the analysis of bone to assure that a relevant dose of LY2066948 would not act as an estrogen antagonist in the skeleton and lead to bone loss in ovary intact rats. This was a concern because there have been reports of bone loss associated with another SERM, tamoxifen, in ovary-intact rats and also premenopausal women (24, 25). Cancellous bone from the fourth lumbar vertebra was evaluated by QCT analysis. QCT parameters included bone mineral density (BMD), BMC, and cross-sectional area. Vertebral BMD was significantly reduced in OVX animals relative to intact controls, but LY2066948 had no effect, compared with intact controls (Table 3). BMC and cross-sectional vertebral area were not significantly different between groups. These data showed no significant effects of LY2066948 at any dose on the bones of ovary-intact rats after 35 d of treatment. Whereas definitive effects of LY2066948 on the skeleton may require longer periods of treatment, these findings contrast with the significant loss of bone (femoral BMD and BMC) seen after 35 d of treatment with a GnRH agonist (Alzet pump-infused D-Trp6-LHRH), which was comparable with the loss seen in OVX rats (data not shown).
Skeletal efficacy of 0.01–10 mg/kg LY2066948 was also evaluated for 42 d in OVX animals and compared with 1 mg/kg of LY117018 (uterine data shown in Fig. 2). Treatment was initiated 3 d after ovariectomy to measure the ability of LY2066948 to prevent bone loss. Measurements were carried out on the distal femur metaphysis (cancellous bone measurement) and the femur midshaft (cortical bone measurement) (15), but the latter was not significantly different among any of the groups, including intact vs. OVX (data not shown). LY2066948 prevented ovariectomy-induced bone loss in the distal femur measured as preservation of both BMC and BMD, with effects comparable with those seen with LY117018 or raloxifene (26) (Fig. 6). Thus, LY2066948 acts as an ER agonist on bone in OVX rats and does not antagonize endogenous estrogen effects in the skeleton of ovary-intact rats.
Hypothalamic-pituitary effects
We were interested in determining whether LY2066948 had any effect on the hypothalamic-pituitary axis, given that estradiol normally feeds back on the hypothalamus to regulate pituitary gonadotrophins. We therefore treated OVX rats for 7 d with LY2066948 to see whether there was any effect on lowering the high levels of LH seen in the absence of estradiol. As shown in Fig. 7A, the high levels of serum LH were decreased by an average of 84% with estradiol treatment, whereas LY2066948 showed a significant but non-dose-dependent reduction (over the dose range tested) of about 40%, suggestive of a slight agonist activity for the hypothalamo-pituitary axis. In this same experiment, we measured uterine wet weight changes (Fig. 7B), which were consistent with the longer-term dosing of OVX rats shown in Fig. 2A, emphasizing the minimal uterine agonist activity of LY2066948.
Discussion
Structure-activity studies in a series of estrogen-dependent in vitro and in vivo assays have led to the discovery of a novel SERM that displays potent uterine antagonist activities, ER agonist effects on bone, and minimal ovarian stimulation in rats. A compound with such a pharmacological profile may be ideal for the treatment of uterine fibroids. In premenopausal women, SERMs such as clomiphene, tamoxifen, and raloxifene have shown ovarian stimulatory activity with the potential for ovarian cyst formation presumably through stimulation of the HPO axis (7, 8, 9, 10, 11). Currently, the main pharmacologic approach for treatment of uterine fibroids, GnRH analogs, is contraindicated for chronic use due to concerns over excessive bone loss with prolonged administration. A SERM with potent uterine antagonist activity that is safe with respect to the ovary and does not cause bone loss would represent an important therapeutic advance for premenopausal women with uterine fibroids.
For measurement of uterine antagonist efficacy and potency, we relied on the ability of the SERM to oppose estradiol-dependent uterine weight gains in rats. We used this model as a surrogate for changes in fibroid size, given the absence of animal models for uterine fibroids that truly represent human disease. Although there are two noteworthy models that have approximated human fibroids, it is unclear how relevant they are to human disease. For example, the Eker rat develops spontaneous uterine fibroids, but these are associated with a genetic mutation in the tuberous sclerosis-2 tumor suppressor gene that also predisposes the animals to renal cell carcinoma (27). This is not thought to be a mechanism involved in human predisposition to uterine fibroids. Moreover, the efficacy of fibroid reduction by a raloxifene analog was found to be equal to that of tamoxifen in the Eker rat (28), whereas raloxifene has been shown to be nonstimulatory in the human uterus and shows efficacy against human fibroids (29, 30) and tamoxifen is stimulatory and increases fibroid size (31, 32). Our rat uterine weight models clearly show that raloxifene is a much stronger antagonist in the uterus, compared with the partial agonist/antagonist character of tamoxifen. Another fibroid model of consideration is the hormone-dependent guinea pig model in which the animals are first ovariectomized and then stimulated with exogenous estradiol (33). Whereas this model has the advantage of estrogen dependence (like that seen with human fibroids) and raloxifene was found to be effective in decreasing fibroid size (34), the guinea pig fibroids occur mainly abdominally instead of in the uterus, a situation not normally seen in humans. Of note is the recent description of a guinea pig model in which a cocktail of nuclear receptor ligands was administered and in which fibroids were induced quickly in the correct uterine compartment (35). This model shows promise for assessing antileiomyoma efficacy; however, additional validation is necessary to determine whether this model predicts clinical efficacy.
Clinical data have clearly demonstrated that raloxifene decreases both uterine and leiomyoma size in postmenopausal patients (30). There are, however, conflicting data on the efficacy of SERMs in premenopausal women with uterine fibroids. In one study, raloxifene alone did not demonstrate efficacy, suggesting that the doses (60 and 180 mg/d for 24 wk) were not sufficient to antagonize the plasma estrogen levels in ovulatory women or that some other factor was responsible for differences in fibroid response in post and premenopausal women (36). Interestingly, when these authors administered GnRH agonist along with raloxifene to premenopausal women with fibroids, raloxifene had an additional effect of decreasing fibroid size, suggesting that depletion of estradiol via the GnRH agonist allowed raloxifene to antagonize fibroid growth (37). More recently, studies in premenopausal women treated with raloxifene concluded that raloxifene prevented the progression of fibroids and resulted in an average 22% reduction in fibroid volume after 3 months treatment with a dose of 180 mg/d (38). The difference in efficacy between this study and the study by Palomba et al. (36) was speculated to be due to differences in the average age of study subjects (36 vs. 40 yr), and therefore, potential differences in circulating estradiol levels, though other protocol dissimilarities bring this theory into question (39, 40).
The effects of tamoxifen on fibroid size have also been investigated, but no effect was observed in premenopausal women (41, 42). Furthermore, these studies demonstrated increased estradiol levels in treated patients and seven of 10 women developed ovarian cysts during the study. Both groups suggested that the increased hormonal levels and ovarian hyperstimulation were a result of indirect as well as direct effects of tamoxifen on the ovaries. There have also been studies or case reports in which tamoxifen was shown to increase fibroid growth as well as endometrial proliferation in postmenopausal women (32, 43, 44).
Ovarian stimulation and subsequent risk for ovarian cyst formation is an undesired effect of SERM treatment in premenopausal women. Preclinical rat and human clinical studies in premenopausal women have demonstrated that SERMs like raloxifene and tamoxifen stimulate the HPO axis and possibly the ovary directly (9, 13, 42). We found that doses of LY2066948 that produced potent uterine efficacy had limited ovarian effects. The effects on ovarian antral follicles and CL number were consistent with LY2066948-mediated prolongation of the follicular phase of the ovarian cycle. This pertubation of the follicular phase is likely due to loss of estrogen feedback on the hypothalmic-pituitary axis resulting in an absence of a LH surge (13). Without the LH surge, follicles will fail to ovulate and CL will decrease in number. The relative decrease in CL and increase in anovulatory follicles results in the ovarian morphology observed. Interestingly, there was clear histologic evidence that a few animals had ovulated during the 35-d in vivo study. Basophilic CL were diagnostic for recent (few days) ovulation. These data along with decreased progesterone levels suggest that this SERM may have a contraceptive effect, but the consistency of the effect is unknown and suggests the need for further studies.
There was minimal ovarian stimulation in LY2066948-treated rats. There were no ovarian cysts, and granulosa cell hyperplasia observed in a few animals was usually minimal. The ovarian effects seen at higher dose levels would likely be fully reversible, given previous findings with raloxifene (13). SERM-mediated ovarian stimulation in rats was proposed to be primarily the result of sustained, elevated LH levels due to loss of estrogen feedback mechanisms in the hypothalamic-pituitary axis (13). Therefore, it was not unexpected to find that LH levels were not significantly elevated in rats given LY2066948. Treatment of rats with LY560212 resulted in clear ovarian stimulation with an approximately 3-fold higher granulosa cell hyperplasia score, compared with rats treated with the highest dose of LY2066948. However, LH levels were only modestly elevated by LY560212 relative to LY2066948, suggesting that LY560212 (and other SERMs) exert direct effects on the ovaries as well as through the HPO axis (see below).
The reason(s) for the improved ovarian profile of LY2066948 are not clear. One contributing possibility may be low brain exposure of the compound, decreasing the chance that the compound can bind to hypothalamic estrogen receptors and stimulate the HPO axis. This hypothesis is consistent with the relative whole brain concentrations observed for LY2066948 in comparison with LY560212. The brain level of LY2066948 after oral administration to rats (10 mg/kg, n = 3/group) was found to be 103 ± 28 ng/g, and that of LY560212 was 2692 ± 211 ng/g after 6 h. This represents a more than 25 times increase in brain exposure of LY560212 relative to LY2066948. Many SERMs cause elevated levels of LH (and the highest stimulation of estrogen production), suggesting a hypothalamic GnRH mechanism or direct pituitary effects in animals and women (13, 45). Decreased LY2066948 exposure in the hypothalamic loci responsible for GnRH production may be responsible for limiting HPO stimulation relative to other SERMs. However, because it appears that at least part of the hypothalamus is outside the blood-brain barrier and the pituitary is clearly outside the blood-brain barrier (46), a relatively low brain exposure may not necessarily lead to neutral HPO effects. In fact, we did see evidence of an estrogen-like effect of LY2066948 on the hypothalamo-pituitary axis (Fig 7) in OVX rats, suggesting that even if the compound reaches the hypothalamus or pituitary in pharmacologic concentrations, its activity there might not lead to a block of hormonal feedback mechanisms.
Cells in the ovary express ERs (47), and estrogen stimulatory effects on follicular granulosa cell growth and differentiation have been widely reported (48). Therefore, it is possible that some SERMs have direct stimulatory effects on the ovaries, whereas LY2066948 does not or the effects are reduced. Reduction of direct ovarian effects could be a reflection of unique pharmacokinetic properties of the compound that limit exposure to the ovaries or unique conformation of the estrogen receptors induced by binding LY2066948. This conformational change may lead to unique cofactor/ER interactions. It is interesting to note that the ovary is a tissue in which ER expression appears to be high relative to ER (49). After examining the data from all of the SERMs used in our studies, we assessed whether there were differences in relative ER/ binding affinity that might correlate with the degree of ovarian stimulation, but no such correlation was found.
In summary, we have discovered a SERM, LY2066948, that shows good separation between desired pharmacology in the uterus and undesired pharmacology in the ovary. These studies have provided compelling data to support clinical evaluation of uterine antagonist efficacy as a possible treatment for gynecologic diseases. In addition to the desirable aspects of a safe pharmacologic treatment for uterine fibroids, other benefits over current therapies may be realized through reduced side effects mediated by estrogen agonism in selected tissues, allowing preservation of bone mass and normal vasomotor and cardiovascular homeostasis.
Acknowledgments
We thank Chahrzad Montrose-Rafizadeh, Sandy Cockerham, Ed Osbourne, Robert Amos, Pam Shetler, Ellen Rowley, and Harlan Cole for excellent help they have provided in the lead optimization biology labs and the animal rooms to make this study possible. We also thank Steve Iturria for statistical interpretation.
Footnotes
Abbreviations: BMC, Bone mineral content; BMD, bone mineral density; CL, corpora lutea; DCC, dextran-coated charcoal; E2, 17-estradiol; EE, ethynyl estradiol; ER, estrogen receptor; HPO, hypothalamo-pituitary-ovarian; Ki, inhibitory constant; OVX, ovariectomized; P4, progesterone; QCT, quantitative computed tomography; SD, Sprague Dawley; SERM, selective estrogen receptor modulator; UWR, uterine weight to body weight ratio.
References
Buttram Jr VC, Reiter RC 1981 Uterine leiomyomata: etiology, symptomatology, and management. Fertil Steril. 36:433–445
Cramer SF, Patel A 1990 The frequency of uterine leiomyomas. Am J Clin Pathol. 94:435–438
Stewart EA 2001 Uterine fibroids. Lancet 357:293–298
Friedman AJ, Hoffman DI, Comite F, Browneller RW, Miller JD 1991 Treatment of leiomyomata uteri with leuprolide acetate depot: a double-blind, placebo-controlled, multicenter study. The Leuprolide Study Group. Obstet Gynecol. 77:720–725
Draper MW, Chin WW 2003 Molecular and clinical evidence for the unique nature of individual selective estrogen receptor modulators. Clin Obstet Gynecol. 46:265–297
Sato M, Rippy MK, Bryant HU 1996 Raloxifene, tamoxifen, nafoxidine, or estrogen effects on reproductive and nonreproductive tissues in ovariectomized rats. FASEB J. 10:905–912
Huppert LC 1979 Induction of ovulation with clomiphene citrate. Fertil Steril. 31:1–8
Adashi EY 1984 Clomiphene citrate: mechanism(s) and site(s) of action—a hypothesis revisited. Fertil Steril. 42:331–344
Baker VL, Draper M, Paul S, Allerheiligen S, Glant M, Shifren J, Jaffe RB 1998 Reproductive endocrine and endometrial effects of raloxifene hydrochloride, a selective estrogen receptor modulator, in women with regular menstrual cycles. J Clin Endocrinol Metab. 83:6–13
Plouffe Jr L, Siddhanti S 2001 The effect of selective estrogen receptor modulators on parameters of the hypothalamic-pituitary-gonadal axis. Ann NY Acad Sci. 949:251–258
Mourits MJ, de Vries EG, Willemse PH, ten Hoor KA, Hollema H, Sluiter WJ, de Bruijn HW, van der Zee AG 1999 Ovarian cysts in women receiving tamoxifen for breast cancer. Br J Cancer. 79:1761–1764
Bucci TJ 1991 Evaluation of altered morphology. In: Haschek WM, Rousseaux CG, eds. Handbook of toxicologic pathology. San Diego: Academic Press; 23–35
Long GG, Cohen IR, Gries CL, Young JK, Francis PC, Capen CC 2001 Proliferative lesions of ovarian granulosa cells and reversible hormonal changes induced in rats by a selective estrogen receptor modulator. Toxicol Pathol. 29:403–410
Peluso J 1992 Morphologic and physiologic features of the ovary. In: Mohr U, Dungwort DL, Capen CC, eds. Pathobiology of the aging rat. Vol 1. Washington, DC: ILSI Press; 342–344
Sato M 1995 Comparative x-ray densitometry of bones from ovariectomized rats. Bone 17:157S–162S
DeFriend DJ, Howell A, Nicholson RI, Anderson E, Dowsett M, Mansel RE, Blamey RW, Bundred NJ, Robertson JF, Saunders C, Baum M, Walton P, Sutcliffe F, Wakeling AE 1994 Investigation of a new pure antiestrogen (ICI 182780) in women with primary breast cancer. Cancer Res. 15:408–414
Littlefield BA, Gurpide E, Markiewicz L, McKinley B, Hochberg RB 1990 A simple and sensitive microtiter plate estrogen bioassay based on stimulation of alkaline phosphatase in Ishikawa cells: estrogenic action of delta 5 adrenal steroids. Endocrinology 127:2757–2762
Bramlett KS, Burris TP 2003 Target specificity of selective estrogen receptor modulators within human endometrial cancer cells. J Steroid Biochem Mol Biol. 86:27–34
Katzenellenbogen B, Kendra K, Norman M, Berthois Y 1987 Proliferation, hormonal responsiveness, and estrogen receptor content of MCF-7 human breast cancer cells grown in the short-term and long-term absence of estrogens. Cancer Res. 47:4355–4360
Black L, Goode R 1980 Uterine bioassay of tamoxifen, trioxifene, and a new estrogen antagonist (LY117018) in rats and mice. Life Sci. 26:1453–1458
Dodge JA, Lugar CW, Fahey KJ, Cullinan GJ, Phillips DL, Short LL, Glasebrook AL, Sato M, Muehl BS, Kim JK, Yang NN, Magee DE, Cole HW, Bryant HU 1996 A novel estrogen receptor modulator with selectivity for bone over uterus in rats. J Bone Mineral Res. 8:T589
Palkowitz AD, Glasebrook AL, Thrasher KJ, Hauser KL, Short LL, Phillips DL, Muehl BS, Sato M, Shetler PK, Cullinan GJ, Pell TR, Bryant HU 1997 Discovery and synthesis of [6-Hydroxy-3-[4-[2-(1-piperidinyl)ethoxy]phenoxy]-2-(4-hydroxyphenyl)]bezo[b]thiophene: a novel, highly potent, selective estrogen receptor modulator. J Med Chem. 40:1407–1416
Kym P, Anstead G, Pinney K, Wilson SR, Katzenellenbogen J 1993 Molecular structures, conformational analysis, and preferential modes of binding of 3-aroyl-2-arylbenzo[b]thiophene estrogen receptor ligands: LY117018 and aryl azide photoaffinity labeling analogs. J Med Chem. 36:3910–3922
Sibonga JD, Evans GL, Hauck ER, Bell NH, Turner RT 1996 Ovarian status influences the skeletal effects of tamoxifen in adult rats. Breast Cancer Res Treat. 41:71–79
Powles T, Hickish T, Kanis J, Tidy A, Ashley S 1996 Effect of tamoxifen on bone mineral density measured by dual-energy x-ray absorptiometry in healthy premenopausal and postmenopausal women. J Clin Oncol. 4:78–84
Black L, Sato M, Rowley E, Magee D, Bekele A, Williams D, Cullinan G, Bendele R, Kauffman R, Bensch W, Frolik C, Termine J, Bryant HU 1994 Raloxifene (LY139481 HCl) prevents bone loss and reduces serum cholesterol without causing uterine hypertrophy in ovariectomized rats. J Clin Invest. 93:63–69
Everitt JI, Wolf DC, Howe SR, Goldsworthy TL, Walker C 1995 Rodent model of reproductive tract leiomyomata. Clinical and pathological features. Am J Pathol. 146:1556–1567
Walker CL, Burroughs KD, Davis B, Sowell K, Everitt JI, Fuchs-Young R 2000 Preclinical evidence for therapeutic efficacy of selective estrogen receptor modulators for uterine leiomyoma. J Soc Gynecol Investig. 7:249–256
Delmas PD, Bjarnason NH, Mitlak BH, Ravoux AC, Shah AS, Huster WJ, Draper M, Christiansen C 1997 Effects of raloxifene on bone mineral density, serum cholesterol concentrations, and uterine endometrium in postmenopausal women. N Engl J Med. 337:1641–1647
Palomba S, Sammartino A, Di Carlo C, Affinito P, Zullo F, Nappi C 2001 Effects of raloxifene treatment on uterine leiomyomas in postmenopausal women. Fertil Steril. 76:38–43
Fisher B, Costantino JP, Redmond CK, Fisher ER, Wickerham DL, Cronin WM 1994 Endometrial cancer in tamoxifen-treated breast cancer patients: findings from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14. J Natl Cancer Inst. 86:527–537
Schwartz LB, Rutkowski N, Horan C, Nachtigall LE, Snyder J, Goldstein SR 1998 Use of transvaginal ultrasonography to monitor the effects of tamoxifen on uterine leiomyoma size and ovarian cyst formation. J Ultrasound Med. 17:699–703
Porter K, Tsibris J, Nocosia S, Murphy J, O’Brien W, Rao P, Spellacy W 1995 Estrogen-induced guinea pig model for uterine leiomyomas: do the ovaries protect Biol Reprod. 52:824–832
Porter KB, Tsibris JC, Porter GW, Fuchs-Young R, Nicosia SV, O’Brien WF, Spellacy WN 1998 Effects of raloxifene in a guinea pig model for leiomyomas. Am J Obstet Gynecol. 179:1283–1287
Tsibris JC, Porter KB, Jazayeri A, Tzimas G, Nau H, Huang H, Kuparadze K, Porter GW, O’Brien WF, Spellacy WN 1999 Human uterine leiomyomata express higher levels of peroxisome proliferator-activated receptor , retinoid X receptor , and all-trans retinoic acid than myometrium. Cancer Res. 59:5737–5744
Palomba S, Orio Jr F, Morelli M, Russo T, Pellicano M, Zupi E, Lombardi G, Nappi C, Panici PL, Zullo F 2002 Raloxifene administration in premenopausal women with uterine leiomyomas: a pilot study. J Clin Endocrinol Metab. 87:3603–3608
Palomba S, Russo T, Orio Jr F, Tauchmanova L, Zupi E, Panici PL, Nappi C, Colao A, Lombardi G, Zullo F 2002 Effectiveness of combined GnRH analogue plus raloxifene administration in the treatment of uterine leiomyomas: a prospective, randomized, single-blind, placebo-controlled clinical trial. Hum Reprod. 17:3213–3219
Jirecek S, Lee A, Pavo I, Crans G, Eppel W, Wenzl R 2004 Raloxifene prevents the growth of uterine leiomyomas in premenopausal women. Fertil Steril. 81:132–136
Palomba S, Zullo F, Orio Jr F, Lombardi G 2004 Does raloxifene inhibit the growth of uterine fibroids Fertil Steril. 81:1719–1720
Jirecek S, Lee A, Eppel W, Wenzl R, Pavo I, Gerald C 2004 Reply: does raloxifene inhibit the growth of uterine fibroids Fertil Steril. 81:1720–1721
Lumsden MA, West CP, Baird DT 1989 Tamoxifen prolongs luteal phase in premenopausal women but has no effect on the size of uterine fibroids. Clin Endocrinol. 31:335–343
Sadan O, Ginath S, Sofer D, Rotmensch S, Debby A, Glezerman M, Zakut H 2001 The role of tamoxifen in the treatment of symptomatic uterine leiomyomata—a pilot study. Eur J Obstet Gynecol Reprod Biol. 96:183–186
Ugwumadu AH, Harding K 1994 Uterine leiomyomata and endometrial proliferation in postmenopausal women treated with the anti-oestrogen tamoxifen. Eur J Obstet Gynecol Reprod Biol. 54:153–156
Kang J, Baxi L, Heller D 1996 Tamoxifen-induced growth of leiomyomas. A case report. J Reprod Med. 41:119–120
Dukes M, Chester R, Yarwood L, Wakeling AE 1994 Effects of a nonsteroidal pure antioestrogen, ZM 189,154, on oestrogen target organs of the rat including bones. J Endocrinol. 141:335–341
Landas S, Fischer J, Wilkin LD, Mitchell LD, Johnson AK, Turner JW, Theriac M, Moore KC 1985 Demonstration of regional blood-brain barrier permeability in human brain. Neurosci Lett. 57:251–256
Brandenberger AW, Tee MK, Jaffe RB 1998 Estrogen receptor (ER-) and (ER-) mRNAs in normal ovary, ovarian serous cystadenocarcinoma and ovarian cancer cell lines: down-regulation of ER- in neoplastic tissues. J Clin Endocrinol Metab. 83:1025–1028
Palter SF, Tavares AB, Hourvitz A, Veldhuis JD, Adashi EY 2001 Are estrogens of import to primate/human ovarian folliculogenesis Endocr Rev. 22:389–424
Kuiper GG, Enmark E, Pelto-Huikko M, Nilsson S, Gustafsson JA 1996 Cloning of a novel receptor expressed in rat prostate and ovary. Proc Natl Acad Sci USA. 93:5925–5930(Andrew G. Geiser, Conrad )