Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Idoxifene (pyrrolidino-4-iodotamoxifen) is a novel selective estrogen receptor modulator (SERM). The compound has a 2.5-fold greater affinity for the estrogen receptor than tamoxifen (McCague et al., 1990; Chander et al., 1991) but is significantly less uterotrophic (Chander et al., 1991). Recent studies in our laboratories have demonstrated that idoxifene prevents and arrests the loss of bone mineral density (BMD) that occurs in the axial and appendicular skeleton in the ovariectomized rat, a model for human osteoporosis, and lowers serum cholesterol levels in the animals (Nuttall et al., 1998). These characteristics indicated that the compound might be effective in preventing or treating osteoporosis or increased cholesterol in postmenopausal women. The bone-protective effects observed with idoxifene in this animal model of osteoporosis suggested to us that the compound might have therapeutic effects in the commonly used animal model of inflammatory joint disease, the adjuvant arthritic rat. In this model, acute and chronic inflammation is accompanied by severe bone, cartilage, and joint destruction, and previous studies have shown that the administration of estrogen can alleviate the inflammatory events in adjuvant-induced disease (Schlaghecke et al., 1989). A probable mechanism by which estrogen regulates bone remodeling and prevents bone resorption is by modulating the production of cytokines such as interleukin (IL)-1, tumor necrosis factor- (TNF-), and IL-6 (Pacifici, 1996, 1998), and because these cytokines are very likely to be involved in adjuvant disease, this provided additional rationale for testing idoxifene.

The studies reported herein show that idoxifene and estrogen, administered prophylactically, effectively inhibit paw inflammation in both male and female rats with adjuvant-induced arthritis (AA) as well as improving BMD and bone mineral content (BMC) of the joint. Histological evaluation of the joints from female rats treated with idoxifene showed significant protection from adjuvant-induced joint destruction. Serum levels of the proinflammatory cytokine IL-6 were decreased in the arthritic rats following treatment with these compounds, indicating a potential mechanism of action in the disease model.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Animals. Male and female Lewis rats were obtained from Charles River Breeding Laboratories (Raleigh, NC). Within any given experiment, only animals of the same age were used. All experimental procedures were in accordance with National Institutes of Health guidelines and were reviewed by the SmithKline Beecham Animal Care and Use Committee.

Materials. Idoxifene (Fig. 1) was synthesized at SmithKline Beecham Pharmaceuticals (King of Prussia, PA). For in vivo experiments, idoxifene was administered p.o. in 0.5% gum tragacanth (Sigma Chemical Co., St. Louis, MO) at 0.3, 1, 3, or 10 mg/kg. 17-Estradiol was purchased from Sigma Chemical Co. and administered in olive oil s.c. at 5 mg/kg.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   Structure of idoxifene (pyrrolidino-4-iodotamoxifen).

AA. AA was induced by a single injection of 0.75 mg of Mycobacterium butyricum (Difco, Detroit, MI) suspended in paraffin oil into the base of the tail of Lewis rats aged 6 to 8 weeks (160-180 g). Hindpaw volumes were measured by a water displacement method on day 16 and/or day 21 (Webb and Griswold, 1984). Test compounds were homogenized in 0.5% gum tragacanth and administered orally in a volume of 10 ml/kg. Control animals were administered vehicle (gum tragacanth) alone. Change in paw volume is calculated as AA paw volume  non-AA paw volume for the AA control animals, and AA paw volume  treated AA paw volume for the treated groups. The data are presented as means + S.E. for 10 animals per group. Percentage of inhibition of hindpaw lesions was calculated as follows:
For statistical analysis, paw volumes of rats treated with idoxifene or estrogen were compared with the control animals by Student's t test with values of p < .05 considered significant.

BMD Measurement. BMD of the distal tibia was determined by dual energy X-ray absorptiometry with the Hologic QDR-1000 equipped with high-resolution scanning software as described previously (Bradbeer et al., 1996). Quality control of the instrument was carried out each day before sample analysis by scanning both a human anthropomorphic spine phantom (low resolution) and the lumbar portion of a rat spine (high resolution), both of which were embedded in methylmethacrylate. All high-resolution scans were carried out with the sample placed on top of an acrylic block 3.8 cm (1.5 in.) in depth. The X-ray beam was collimated to a diameter of 1.27 mm, and line spacing and point resolution were 0.25 and 0.127 mm, respectively. Scans were made of the distal tibia region of excised bones stored in 70% ethanol in a square plastic container. The depth of the liquid was constant for all samples and sufficient to cover the limbs by ~1.3 cm (~0.5 in.). BMD as well as BMC and bone area were determined for the distal tibia. The region of interest was defined as the area between a line drawn parallel to the proximal edge of the calcaneus and a second line drawn perpendicular to the long axis of the tibia midway between the first line and the point where the tibia meets the fibula. The point of connection between the fibula and tibia had to be approximated in some samples with low BMD. The width of the region of interest was kept constant between samples.

Histology. Tibiotarsal joints from randomly selected animals from the following 4 groups of rats were examined histologically: normal rats, AA control rats, AA rats treated with idoxifene at 10 mg/kg/day p.o., and AA rats treated with estrogen at 5 mg/kg/every other day s.c. Rats were sacrificed on day 22 of the disease by CO₂ administration; the hind legs were fixed in formalin and decalcified in Cal-Rite, and the feet were removed from the legs at the distal tibial diaphysis. After routine processing, the feet were embedded and coronal sections were cut in the plane midway through the tibiotarsal and tarsotarsal joints. Sections were stained with safranin O and counterstained with Fast Green.

Bioassay for IL-6. Serum samples were obtained when the rats were euthanized. IL-6 levels were determined with the previously described B9 bioassay (Aarden et al., 1985). Briefly, B9 cells (5 × 10³ cells/well in 96-well flat-bottomed plates) were cultured at 37°C with serial dilutions of rat serum in a final volume of 100 µl of RPMI-10. After 68 h, 0.5 µCi of [³H]thymidine was added and incubated for 6 h at 37°C. Cells were harvested and radioisotope incorporation was determined. IL-6 was quantified from a standard curve including known amounts of rat IL-6 (0.1-100 pg/ml). B9 proliferation was unaffected by any agents used in this study.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Inhibition of AA in Male Lewis Rats. In the rat model of AA, p.o. administration of idoxifene to male rats (0.3, 1, 3, and 10 mg/kg p.o.) from day 0 to day 16 to inhibited the development of immune-mediated hind leg inflammation. There was a dose-dependent inhibition of paw swelling at all concentrations of the compound. Inflammation was inhibited by 26% at 0.3 mg/kg (p < .05), 34% at 1 mg/kg (p < .01), 33% at 3 mg/kg (p < .01), and 51% inhibition at 10 mg/kg (p < .001) (Fig. 2A). This inhibition was sustained, and by day 21, the anti-inflammatory effect was 25% inhibition at 1 mg/kg (p < .01), 41% at 3 mg/kg (p < .001), and 47% at 10 mg/kg (p < .001) (Fig. 2B). The anti-inflammatory and antiarthritic activities of idoxifene were evaluated further by examining the BMC and BMD of the distal tibia region in treated AA rats. On day 22 when the rats were euthanized, hindlimbs were examined by X-ray absorptiometry. Compared with the AA controls, there was a significant normalization of BMD (39%, p < .001) and BMC (28%, p < .05) in the rats treated with 10 mg/kg/day idoxifene, indicating a protective effect on inflammation-mediated bone destruction and/or a direct effect on bone resorption proximal to the inflamed joint (Fig. 3, A and B).

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Dose-dependent suppression of hindpaw inflammation in rats with AA on day 16 (A) and day 21 (B) of arthritis by prophylactic administration of idoxifene. M. butyricum was injected on day 0 and idoxifene was administered 5 times per week for a total of 16 doses starting on day 0. Data are means ± S.E. of 10 animals per group. *p < .05, **p < .01, ***p < .001 compared with the untreated AA controls.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Bone densitometry evaluation of the distal tibia in AA rats treated with idoxifene. Rats were treated with idoxifene at 1, 3, and 10 mg/kg, 5 days per week, from day 0 to day 21. Values are the percentage of normal (assigned a value of 100%), mean ± S.E. of 10 animals per group. The actual BMD value was 0.2772 ± 0.0041 for normal rats and 0.1609 ± 0.01 for AA rats, which is a 42% decrease in BMD in the diseased animals. The actual BMC value was 0.0497 ± 0.0006 for normal rats and 0.0314 ± 0.0017 for AA rats, which is a 37% decrease in BMC in the diseased animals. *p < .05, ***p < .001.

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Comparison of idoxifene (10 mg/kg) with estrogen (5 mg/kg) in the AA rat on day 21 of disease. A, hindpaw volume. B, BMD; the actual BMD value was 0.2892 ± 0.003 for normal rats and 0. 1705 ± 0.007 for AA rats, which is a 41% decrease in BMD in the diseased animals. C, serum IL-6. Data are presented as means ± S.E. of 10 animals per group. ***p < .001.

Inhibition of AA in Female Lewis Rats. The evaluation of compounds for their ability to modulate disease in AA is typically performed in male Lewis rats because the incidence of disease is often higher than in females (Cannon et al., 1992). However, due to the nature of the compounds being evaluated it was deemed appropriate to demonstrate activity in female rats. In two separate experiments there was 100% incidence of disease in the female rats, and treatment with idoxifene at 10 mg/kg (administered prophylactically days 0-20) inhibited paw volume by 68% (p < .001) in the first experiment (data not shown) and 77% (p < .001) in the second experiment (Fig. 5A). There was a 56% normalization of BMD in the joints in the latter group of rats (Fig. 5B). Estrogen also was very effective in female rats, where a 71% inhibition of paw inflammation and 41% normalization of BMD was observed (Fig. 5, A and B).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Comparison of idoxifene (10 mg/kg) with estrogen (5 mg/kg) in the female AA rats on day 22 of disease. A, hindpaw volume. B, BMD; the actual BMD value was 0.2718 ± 0.003 for normal rats and 0.1761 ± 0.0095 for AA rats, which is a 35% decrease in BMD in the diseased animals. Data are presented as means ± S.E. of 10 animals per group. **p < .01, ***p < .001.

View larger version (125K): [in this window] [in a new window]   Fig. 6.   Histological evaluation of female rats. A, photomicrograph of a normal tibiotarsal joint from a non-AA control rat. Articular cartilage is stained pink, other tissues are stained blue/green. The articulation of the distal tibia and the proximal tarsus runs vertically through the center of the image. Note the normal joint architecture. B, detail of (A). The articular cartilage is stained consistently throughout its depth denoting no loss in proteoglycan (PG) content. C, photomicrograph of a coronal section from a vehicle control-treated AA rat. The inflammatory lesion is extensive and severe. The majority of the original bone in the tibia has been lost and replaced with granulation tissue and reactive woven bone. There has been extensive subchondral and surface erosion of the articular cartilage, which has been breached at numerous sites (arrowhead on left). There is also significantly reduced safranin O staining of the tibial articular cartilage, indicating extensive proteoglycan loss. This is in comparison to the more normal appearance of the articular cartilage of the tarsus (arrowhead on right) where consistent safranin O staining is apparent throughout the depth of the cartilage. Granulation tissue fills the joint space. D, detail of (C) showing numerous osteoclasts (arrows) at the subchondral surface. E, photomicrograph of a tibiotarsal joint from an AA rat administered 5 mg/kg estrogen. Joint architecture is histologically indistinguishable from the normal non-AA rat joint in (A). F, detail of (E) showing some mild roughening of the articular surface. G and H, photomicrographs of a tibiotarsal joint from an AA rat administered 10 mg/kg idoxifene. Again, joint architecture is strikingly similar to that of a normal noninflamed rat. Only the detailed photomicrograph shown in (H) demonstrates some reduction in safranin O staining of the articular cartilage of the tarsus, which may indicate some loss of matrix proteoglycan. Safranin O, Fast Green stained; original ×5 for A, C, E, and G; ×25 for B, D, F, and H.

Therapeutic Activity of Idoxifene and Estrogen in AA Rat. To determine the therapeutic efficacy of treatment with idoxifene and estrogen in male and female rats with AA, compounds were administered starting on day 10 of the disease process. Idoxifene at 10 mg/kg p.o. was administered daily and estrogen every other day at 5 mg/kg s.c., and dosing was continued until day 20. In experiments with this dosing protocol, hindpaw inflammation in male rats with AA was reduced by 37% (p < .001) with idoxifene and by 62% (p < .001) with estrogen (Table 2). In female rats, the inhibition was 25 and 46%, respectively.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Estrogen prevents the rapid bone loss in postmenopausal women and in clinically relevant models of osteoporosis such as the ovariectomized (OVX) rat (Kalu et al., 1985; Barzel, 1988). In addition to prevention of osteoporosis, the beneficial effects of estrogen on rheumatoid arthritis also have been apparent for some time. This is exemplified by findings that severity of disease is reduced during pregnancy (Lahita, 1985) and that estrogen-containing contraceptives have been shown to decrease the incidence of rheumatoid arthritis (RA) (Wingrave and Day, 1978). The role of estrogen in human autoimmune disease and in a number of animal models of disease is the subject of a recent review (Jansson and Holmdahl, 1998).

With AA as a model for human RA, we have examined the potential anti-inflammatory effects of several doses of the SERM idoxifene, as well as a single dose of estrogen. Previous studies by Schlaghecke et al. (1989), with male Long Evan rats, showed that estradiol given three times a week (2.5 and 5.0 mg/kg), either prophylactically or therapeutically, significantly inhibited the inflammation of AA. In these experiments, 0.5 mg/kg was not effective. In our studies, we have demonstrated that oral administration of idoxifene or administration of estrogen s.c. is effective in inhibiting the inflammatory response as well as in protecting joint integrity. The maximally effective dose for the inhibition of paw inflammation in male and female rats with a prophylactic dosing protocol (dosing initiated the day of adjuvant injection) was 10 mg/kg and ranged from 40 to 90% depending on the experiment. This was also the maximally tolerated dose, with higher doses causing some weight loss in the animals. For this reason, higher doses of drug were not evaluated. There was also a significant normalization of BMD and BMC at the 10 mg/kg dose in male and female AA rats. In the OVX rat model of osteoporosis, idoxifene was optimally effective at 1 mg/kg (Nuttall et al., 1998), which is lower than the optimally effective dose in the AA rat (10 mg/kg). However, significant inhibition of inflammation in the AA rat was observed at doses as low as 0.3 mg/kg/day (Fig. 2). In addition, the effects with idoxifene in the OVX rat model were observed following 8 to 17 weeks of dosing, whereas in the AA rat model the longest dosing protocol was 20 days. Also, unlike the OVX rat, adjuvant disease in the rat is a highly aggressive inflammatory disease, and it is not so surprising that higher doses of compound would be required for optimal efficacy.

The efficacious effects of estrogen have been shown in several animal models of arthritis besides AA. For example, the MRL-lpr mouse strain is used as a model for studying both systemic lupus erythematosus and RA. In this mouse strain, there is a spontaneous development of a mild arthritis that is enhanced by the intradermal injection of complete Freund's adjuvant. Ratkay et al. (1994) used this model to demonstrate that physiological levels of estradiol given on days 2, 3, 9, 15, and 20 postpartum delayed and reduced arthritic flares in the animals. In collagen-induced arthritis, a T-cell-mediated disease, 17-estradiol administration inhibited the development of disease in DBA/1 mice (Holmdahl et al., 1987) as did an endogenous metabolite of estradiol, 2-methoxyestradiol (Josefsson and Tarkowski, 1997). Because this metabolite also suppressed the in vitro proliferation of endothelial cells, Josefsson and Tarkowski (1987) suggested that the compound down-regulated angiogenesis and that this may be an important factor in its anti-inflammatory actions. We also have demonstrated that idoxifene is active in collagen-induced arthritis in DBA1 mice (D.E.G., unpublished data). In other studies, the production of nitric oxide from an endothelial cell line was shown to be inhibited by 2-methoxyestradiol. Nitric oxide is a proinflammatory mediator, and its inhibition could be of importance in vivo (Hayashi et al., 1998). These studies in arthritic diseases indicated that estrogen and/or its derivatives may be interesting candidates for the therapy of inflammatory diseases.

A potential mechanism for the antiosteoporotic effect of estrogens, at least in part, may be their ability to inhibit IL-6. For example, 17-estradiol has been shown to inhibit both TNF--induced IL-6 production and osteoclast development in primary bone cell cultures derived from neonatal murine calvariae (Girasole et al., 1992), and Ray et al. (1997) have shown that IL-6 gene expression in an endometrial adenocarcinoma cell line is repressed by 17-estradiol. In recent studies from our laboratories, Nuttall et al. (1998) have shown that IL-6 production from TNF--stimulated human osteosarcoma (MG-63) cells is inhibited by idoxifene. In the studies presented herein in the AA rat, we were able to show that treatment with idoxifene at 10 mg/kg and estrogen at 5.0 mg/kg resulted in a dramatic reduction in serum IL-6 on day 22 of the disease (Fig. 4C), although no inhibition of lipopolysaccharide-stimulated serum TNF- in normal rats was observed (D.E.G., unpublished data). It is possible, therefore, that the inhibition of an inflammatory cytokine such as IL-6 may play a role in the therapeutic effect of estrogens and SERMs such as idoxifene in inflammatory arthritides as well as in the antiosteoporotic effects of these compounds. IL-6 is a multifunctional cytokine produced by a range of cells, and it is suggested to be involved in the pathogenesis of a variety of diseases (Hirano et al., 1990). The role of IL-6 as an essential mediator of inflammatory responses has been unequivocally demonstrated in IL-6-deficient mice in which the mice were unable to mount a normal response to localized tissue damage generated by turpentine injection (Fattori et al., 1994), and blockage of the IL-6 receptor with monoclonal antibodies has been shown to ameliorate joint disease in murine collagen-induced arthritis (Takagi et al., 1998). IL-6 has been shown to be increased markedly in different biologic fluids in patients with autoimmune disease, particularly those with RA (Hirano et al., 1988; Houssiau et al., 1988; Swaak et al., 1988), and the level in various inflammatory compartments appears to be a sensitive marker of disease activity. Therefore, although the mechanism(s) of action of the anti-inflammatory and joint protective activities of idoxifene in the AA rat are not resolved, and with the understanding that different mechanisms may be involved, it is reasonable to consider that inhibition of IL-6 could be partially responsible for its activity and that SERM treatment may have a beneficial effect in RA.