Effects of Long-Term Estrogen Replacement Therapy on Osteoarthritis Severity in Cynomolgus Monkeys

Transcription

1 ARTHRITIS & RHEUMATISM Vol. 46, No. 7, July 2002, pp DOI /art , American College of Rheumatology Effects of Long-Term Estrogen Replacement Therapy on Osteoarthritis Severity in Cynomolgus Monkeys Kimberley D. Ham, 1 Richard F. Loeser, 2 Bruce R. Lindgren, 3 and Cathy S. Carlson 1 Objective. To determine the effects of long-term estrogen replacement therapy (ERT) on the severity of osteoarthritis (OA) of the knee joint in surgically postmenopausal (bilaterally ovariectomized) female monkeys. A secondary aim was to evaluate the effect of soy phytoestrogen (SPE) treatment on the severity of OA. Methods. Feral adult female cynomolgus macaques were ovariectomized bilaterally and then randomly divided into 3 age- and weight-matched treatment groups. For 3 years, the first group received ERT with conjugated equine estrogens, the second group received SPE, and the third group received no treatment (controls). At necropsy, histologic lesions of OA were graded, and the area and thickness of cartilage and subchondral bone were measured. The data were summarized by principal components analysis, and the resulting factors and individual variables were compared using analysis of variance and analysis of covariance (age and weight as covariates). Results. Cartilage lesions of OA were significantly less severe in the animals given ERT compared with those in the control group. This treatment effect remained significant when adjusted for age and weight. The factor representing subchondral bone was significantly higher, but the number of osteophytes was lower, in the ERT group compared with the control group. SPE Supported by grants from the NIH (RR-14099), the US Department of Agriculture ( ), and the American Federation for Aging Research. 1 Kimberley D. Ham, BA, Cathy S. Carlson, DVM, PhD: College of Veterinary Medicine, University of Minnesota, St. Paul; 2 Richard F. Loeser, MD: Rush Medical College, Rush Presbyterian St. Luke s Medical Center, Chicago, Illinois; 3 Bruce R. Lindgren, MS: School of Public Health, University of Minnesota, Minneapolis. Address correspondence and reprint requests to Kimberley D. Ham, BA, Department of Veterinary PathoBiology, College of Veterinary Medicine, University of Minnesota, 262 Veterinary Diagnostic Laboratory Building, 1333 Gortner Avenue, St. Paul, MN hamx0012@tc.umn.edu. Submitted for publication September 17, 2001; accepted in revised form March 25, treatment had no significant effect on cartilage or bone lesions of OA. Conclusion. These results demonstrate that longterm ERT significantly reduces the severity of OA lesions in this animal model. Estrogen deficiency that occurs with menopause has been associated with an increased incidence and severity of osteoarthritis (OA) in women (1). Before the age of 45, women have a lower prevalence of OA than men do, but after age 55, OA is observed more frequently in women (2,3). The rate of incident OA of the knee in the Framingham Study, the largest longitudinal study of the disease in humans, was reported to be 1.7 times higher in older women than in older men (4). Another study reported a 4 times higher prevalence of OA in year-old women compared with men of the same age group (5). Most observational studies suggest that postmenopausal women taking estrogen replacement therapy (ERT) have a reduced risk of radiographic hip and knee OA compared with those not taking ERT (6 10). In a meta-analysis of 4 prevalence studies in postmenopausal women, using a combined end point of knee and hip OA, a clear reduction in OA with the use of ERT was shown (pooled odds ratios 0.76, 95% confidence interval ) (6). Not all studies, however, report a positive effect of ERT on OA. In one study of postmenopausal women with cardiac disease, 4 years of estrogen plus progestin therapy had no significant effect on knee pain compared with placebo (11). In addition, conflicting results have been reported for studies using animal models, with some estrogen-treated ovariectomized animals showing an increased incidence of OA and others a decreased incidence (12,13). These conflicting results may be due, in part, to methodologic differences among the studies. In addition to the possible benefits of estrogen for OA, there is abundant evidence that ERT decreases 1956

2 LONG-TERM ERT DECREASES OA SEVERITY IN CYNOMOLGUS MONKEYS 1957 the risk of atherosclerosis (14) and osteoporosis (15,16). These diseases tend to be clinically silent prior to a major life-altering event, such as a heart attack or a hip fracture, possibly explaining, in part, the low compliance rate in women taking ERT (only 8% of naturally postmenopausal women take ERT [17]). The lack of compliance is primarily due to negative side effects, such as continuing menstrual periods due to the need for accompanying progestin therapy and fear of an increased risk of breast cancer (18,19). In contrast to atherosclerosis and osteoporosis, OA is a chronic and painful disease that interferes with a person s ability to function on a daily basis. If it could be definitively demonstrated that ERT slows the progression of OA, women may have more of an incentive to take estrogen supplements, and compliance among those for whom ERT has been prescribed may improve. The primary purpose of the present study was to evaluate the effects of ERT on OA severity; however, the study design also allowed the evaluation of the effects of soy phytoestrogens (SPEs) on this disease. Dietary SPEs have been investigated as a potential alternative to traditional ERT. Since they are nutritional supplements derived from natural plant products, rather than pharmacologic therapy, they may be more acceptable than ERT to the patient population. SPEs possess estrogen-like activity and are known to combine with estrogen receptors, although at a much lower affinity (in binding affinity experiments, 250 1,000-fold molar excesses were required to give 50% inhibition of specific binding of 3 H-labeled 17 -estradiol) (20). Although not as potent as 17 -estradiol, SPEs have been shown to produce the same maximum receptor response in vitro when the tested dose was high enough (1,000 times the concentration required for a maximum receptor response with 17 -estradiol) (20). Although SPEs are not as well characterized as estrogen, studies suggest that they may provide similar health benefits, such as a decreased risk of coronary artery disease (21) and osteoporosis (22 24). In the present study, cynomolgus macaques, a well-characterized animal model of OA (25), were used to evaluate the effects of ERT on OA severity. Cynomolgus macaques are the most commonly used nonhuman primate in biomedical research (26) and have been used to study atherosclerosis (27), osteoporosis (28), and diabetes mellitus (29) as well as OA (25). It has been demonstrated that ERT in surgically postmenopausal cynomolgus monkeys has significant effects on serum lipids and significantly decreases atherosclerotic lesions (17,21). Furthermore, it has been well established that ERT in this animal model prevents bone loss after ovariectomy (30). This nonhuman primate model of OA is extremely useful because of the close phylogenetic relationship between monkeys and humans. Moreover, cynomolgus macaques develop OA naturally as they age, and the histologic lesions that develop in the tibial plateaus and femoral condyles are similar to those that occur in humans with OA (25,31,32). As in humans, OA can occasionally occur in young adult cynomolgus macaques and can spare certain older individuals (Carlson CS, et al: unpublished observations). A previous crosssectional study showed that, as in humans, the prevalence and severity of the disease in cynomolgus macaques increase with increasing age (31). In addition, older adult monkeys with severe radiographic OA have been observed to exhibit a stiff gait and/or a limp, suggesting that the joint changes observed histologically may become symptomatic (Carlson CS, et al: unpublished observations). In the present study, we determined the effects of long-term ERT on the progression and severity of OA in surgically postmenopausal (bilaterally ovariectomized) female monkeys. We hypothesized that ERT would reduce the severity of the histologic lesions of OA of the knee joint in these animals. A secondary aim was to evaluate the effect of SPE treatment on the severity of OA and, similarly, we hypothesized that SPE treatment also would reduce the severity of OA. At the termination of the study, the knee joints were carefully characterized morphologically to determine the effects of these therapies at the tissue level. MATERIALS AND METHODS Animals. The animals used in this study were subjects in an experiment designed to study the effects of estrogen deficiency, exogenous estrogen treatment, and SPE treatment on atherosclerosis of the coronary arteries (21). The study included 180 feral adult female cynomolgus macaques imported from Indonesia (Charles River Research Primates, Port Washington, NY). The monkeys were kept in social groups of 4 or 5 in indoor/outdoor pens, measuring meters, with concrete floors. At the termination of the study, the mean age of the animals, as estimated by dentition (33), was 12.0 years (range years). The average lifespan of cynomolgus monkeys in captivity is years. Study design. The animals were given water ad libitum and fed a moderately atherogenic diet (40% of calories from fat) for 26 months in order to induce lesions of atherosclerosis (21). This diet resembles the average human diet found in

3 1958 HAM ET AL Table 1. Semiquantitative histologic grading scale for tibial plateaus Histologic parameter Feature Articular cartilage structure Grade 0 Articular surface smooth and intact Grades 1 4 Minimal/mild superficial fibrillation (less than one-tenth of the articular cartilage thickness) involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Grades 5 8 Fibrillation and/or clefts and/or loss of cartilage affecting the superficial one-third or less of the articular cartilage thickness involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Grades 9 12 Fibrillation and/or clefts and/or loss of cartilage affecting more than one-third, but less than two-thirds, of the articular cartilage thickness involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Grades Fibrillation and/or clefts and/or loss of cartilage affecting more than two-thirds and up to the full thickness of the articular cartilage involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Grades Fibrillation and/or clefts and/or loss of cartilage extending into calcified cartilage and/or subchondral bone involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Loss of toluidine blue staining Grade 0 None Grades 1 4 Loss of staining affecting 20% of the articular cartilage depth and involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Grades 5 8 Loss of staining affecting 20% but 50% of the articular cartilage depth and involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Grades 9 12 Loss of staining affecting 50% but 100% of the articular cartilage depth and involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Grades Full-thickness loss of staining involving 20%, 20 50%, 50% but 100%, and 100% of the plateau, respectively Chondrocyte clones* Grade 0 None Grade 1 Clone candidates (clusters of 3 5 chondrocytes) Grades clones, 5 10 clones, and 10 clones, respectively Subchondral bone thickness Grade m Grades 1 and to 450 m involving 50% and 50% of the plateau, respectively Grades 3 and to 750 m involving 50% and 50% of the plateau, respectively Grades 5 and m involving 50% and 50% of the plateau, respectively Osteophytes, maximum no. of tidemarks, Counted vessels crossing the tidemark, and chondrified vessels * A cluster of 6 or more chondrocytes. industrialized Western societies such as the US. At the end of this 26-month period, the monkeys were ovariectomized bilaterally to simulate menopause and then randomly divided into 3 age- and weight-matched treatment groups, using a stratified randomization scheme (21). One group of animals (n 60) received no treatment (controls), one group (n 60) received ERT with conjugated equine estrogens (Premarin; Wyeth-Ayerst Laboratories, Philadelphia, PA), and one group (n 60) received SPE. The treatments were administered in the diet, with ERT being given at a dosage equivalent to mg/day in humans and SPE (SUPRO 670-HG soy protein isolate, containing mg of genistein, mg of daidzein, and 0.08 mg of glycitein per gm of soy protein isolate; generously provided by Protein Technologies, St. Louis, MO) being given at a dosage approximately equivalent to 129 mg/day in humans (21). The treatments were given for 36 months, during which all 3 groups were fed a moderately atherogenic diet consisting of 120 calories per kg of body weight per day (approximately equivalent to 1,800 calories/day for the average woman in the US) (21). The diet of the control and ERT groups contained a soy protein isolate that had been extracted by an aqueousalcohol wash to deplete the SPEs. The isolate was added to the diet to keep the amount of soy protein consistent among the 3 groups. The alcohol-extracted soy protein (SUPRO 670-IF; generously provided by Protein Technologies) contained 0.04 mg of genistein, 0.01 mg of daidzein, and 0.01 mg of glycitein per gram of isolate (21). To evaluate the effectiveness of the dietary ERT treatment, blood was collected into evacuated tubes containing EDTA, and plasma estrogen concentrations were measured using a commercially available radioimmunoassay kit (Diagnostic Products, Los Angeles, CA). Plasma estrogen levels were measured 3 times during the 3-year postmenopausal treatment period. Plasma SPE levels were measured once, at 34 months after initiation of treatment; high-performance liquid chromatography mass spectrometry methods were used (21). Plasma SPE levels were not measured in the ERT and control groups because previous studies had shown that animals consuming a similar diet had SPE levels below the level of detection.

4 LONG-TERM ERT DECREASES OA SEVERITY IN CYNOMOLGUS MONKEYS 1959 Table 2. Age and weight of the monkeys at the end of the 3-year study, by treatment group* Covariable Control (n 52) Treatment group ERT (n 54) SPE (n 60) Age, mean SD years Weight, mean SD kg * ERT estrogen replacement therapy (conjugated equine estrogens); SPE soy phytoestrogen. Necropsy and tissue preparation. Tissues from 166 of the 180 monkeys (60 receiving SPE, 54 receiving ERT, and 52 controls) were available for evaluation in this study. One animal was excluded from the evaluation because of trauma to a knee joint. Tissues were not collected from the remaining 13 animals because of lack of technical assistance on the day these animals were necropsied. At necropsy, both knee joints were collected and the body weight (in kg) was recorded. A midcoronal section of the right tibial plateau was obtained, decalcified, embedded in paraffin, sectioned at 6 m, and stained with toluidine blue. Histologic sections were randomized and relabeled to blind the evaluator to the treatment groups. Each lateral and medial tibial plateau was graded using a semiquantitative histologic grading scale that was developed to evaluate lesions of OA (Table 1). The femoral condyles were not evaluated because in previous OA studies using this animal model, the lesions at these sites were consistently less severe than those of the tibial plateaus (25,31,32). Grading included assessment of the severity of articular cartilage fibrillation, extent of loss of toluidine blue staining, and severity of subchondral bone thickening. In addition, vessels crossing the tidemark, chondrified vessels, chondrocyte clones, osteophytes, and tidemarks were counted. The perimeters of the articular cartilage, calcified cartilage, and subchondral bone were measured using the Osteomeasure histomorphometry system (Osteometrics, Atlanta, GA). The tissues were measured in a mm field in the center of the tibial plateau using a 2 objective. This area included 80% of the articular cartilage and subchondral bone of the tibial plateau. From these data, widths and areas were calculated. Statistical analysis. The histologic grading and measurement data were summarized using principal components analysis (factor analysis). Two separate analyses were run, one using the data from the ERT and control groups only, and the other using the data from all 3 treatment groups. Using an orthogonal rotation method (varimax), these initial linear combinations of the variables were rotated to create independent factors. The subset of variables that had a factor loading greater than the absolute value of 0.3 was used to represent each factor. The resulting factors and individual variables were compared using analysis of variance (ANOVA). Although the average weights and ages of the monkeys were similar among the groups (Table 2), the factors and individual variables were compared using analysis of covariance (ANCOVA), with age and weight as covariables because these are well-recognized risk factors for OA. P values less than 0.05 were considered statistically significant. Simple linear regression (Excel 2000; Microsoft, Redmond, WA) was used to evaluate the individual variables as a function of age and weight. Both lateral and medial tibial plateaus were evaluated; however, the scores and measurements revealed more severe disease in the medial plateaus, as seen in previous studies. Table 3. Results of the factor analysis* Factor %of variance explained Variable Factor loading Treatment group Control (n 52) SPE (n 60) ERT (n 54) Factor 1 (OA severity) 28 AC structure, score Chondrocyte clones, score Loss of staining, score Vessels crossing tidemark, no Chondrified vessels, no Osteophytes, no AC thickness, m , , , Factor 2 (SCB thickness) 17 SCB area, mm SCB thickness, m SCB score Factor 3 (AC thickness) 15 AC area, mm AC thickness, m , , , Chondrified vessels, no Vessels crossing tidemark, no Tidemarks, no Factor 4 (CC thickness) 12 CC thickness, m CC area, mm Tidemarks, no * Composition of the factors when only the estrogen replacement therapy (ERT) and control groups were used in the factor analysis. Values are the mean SD. See Table 1 for a description of the scores. SPE soy phytoestrogen; OA osteoarthritis; AC articular cartilage; SCB subchondral bone; CC calcified cartilage.

5 1960 HAM ET AL Therefore, only the data for the medial plateaus were evaluated statistically. RESULTS Plasma hormone concentrations revealed increased 17 -estradiol levels in the ERT group (mean pmoles/liter, compared with 31.6 in the control group and 25.7 in the SPE group), which were stable across time, as well as detectable SPE levels in the SPE group (mean 776 nmoles/liter) (21). On average, the animals in all 3 treatment groups gained weight over the 3-year treatment period (mean weight gain kg). However, there was not a significant difference in weight gain among the 3 treatment groups. Description of factors. Principal components analysis using the data from the control group and the ERT group identified 4 factors that explained 71% of the variability in the data (Table 3). The differences in OA severity were described by factor 1 (weighted primarily by scores for articular cartilage structure, number of chondrocyte clones, and loss of toluidine blue staining). The differences in subchondral bone thickness were explained by factor 2 (weighted primarily by subchondral bone area and subchondral bone thickness). Variability in articular cartilage thickness was explained by factor 3 (weighted primarily by articular cartilage area and articular cartilage thickness). Factor 4 described differences in the calcified cartilage thickness (weighted primarily by calcified cartilage thickness and calcified cartilage area). When the factor analysis was repeated using the data from all 3 treatment groups, the results were nearly identical (4 factors weighted by the same variables and explaining 70% of the variability in the data). Results of analyses comparing the ERT group with the control group. Factor 1 scores (weighted by variables for OA severity) were significantly lower in the ERT group than in the control group (P 0.012), and the effect remained significant when adjusted for age and weight (P 0.044) (Table 4). In addition, when adjusted for treatment group and weight, factor 1 was significantly affected by age (P 0.003) (i.e., as age increased, OA severity increased). All variables that contributed highly to factor 1 were lower in the ERT group than in the control group, with the exception of articular cartilage thickness, which was higher in the ERT group (Table 3). None of the individual variables contributing to factor 1, however, was significantly different between the 2 treatment groups. Factor 2 scores (weighted by subchondral bone thickness and area) were significantly higher in the ERT group than the control group (P 0.033). The effect Table 4. Results of ANOVA and ANCOVA, comparing factor scores for only the ERT and control groups* Factor P for ANOVA Variable P for ANCOVA Factor 1 (OA severity) Treatment Age Weight Factor 2 (SCB thickness) Treatment Age Weight Factor 3 (AC thickness) Treatment Age Weight * Factor 4 is not shown because there were no significant findings. ANOVA analysis of variance; ANCOVA analysis of covariance; ERT estrogen replacement therapy (conjugated equine estrogens); OA osteoarthritis; not significant; SCB subchondral bone; AC articular cartilage. The given variable adjusted for other variables. remained significant when adjusted for age and weight (P 0.026) (Table 4). In contrast, the average number of osteophytes present was lower in the ERT group than in the control group (Table 3). However, this variable (as well as all comparisons of individual variables contributing to factor 2) was not significantly different between the 2 groups. Factor 3 (weighted by articular cartilage thickness and area) was not significantly affected by treatment (Table 4), and there were no significant treatment differences among the individual variables contributing to factor 3 (Table 3). However, although not statistically significant, the mean articular cartilage thickness and area were higher in the ERT group than in the control group (Table 3), and several of the variables that describe lesions of OA and contributed to factor 3 (numbers of chondrified vessels, vessels crossing the tidemark, and tidemarks) were higher in the control group than in the ERT group. Factor 3 was significantly affected by age when adjusted for treatment group and weight (P 0.038) and was significantly affected by weight when adjusted for treatment group and age (P 0.048) (Table 4). Regression analysis revealed that the articular cartilage thickness and area decreased with increasing age (P for thickness, and P 0.01 for area) and increased with increasing weight (P for both thickness and area). Factor 4 (weighted by calcified cartilage thickness and area) was not significantly affected by treatment, age, or weight. Results of analyses comparing all 3 treatment groups. Among the 4 factors, only factor 1 (weighted by variables for OA severity) was significantly affected by

6 LONG-TERM ERT DECREASES OA SEVERITY IN CYNOMOLGUS MONKEYS 1961 Table 5. Results of ANOVA and ANCOVA, comparing factor scores for the ERT, SPE, and control groups* Factor P for ANOVA Variable P for ANCOVA Factor 1 (OA severity) Treatment Age Weight Factor 3 (AC thickness) Treatment Age Weight * Factors 2 and 4 are not shown because there were no significant findings. ANOVA analysis of variance; ANCOVA analysis of covariance; ERT estrogen replacement therapy (conjugated equine estrogens); SPE soy phytoestrogen; OA osteoarthritis; not significant; AC articular cartilage. The given variable adjusted for other variables. treatment when all 3 groups were compared using ANOVA (P 0.032). The effect did not remain significant when adjusted for age and weight using ANCOVA (Table 5). When adjusted for treatment group and weight, factor 1 was significantly affected by age (P 0.001). When adjusted for treatment group and age, factor 3 (weighted by articular cartilage area and thickness) was significantly affected by weight (P 0.001). Like the ERT group, the SPE group had lower factor 1 scores (weighted by variables for OA severity) and higher factor 2 scores (weighted by subchondral bone thickness and area) than the control group, but the differences were not statistically significant (data not shown). With the exception of the articular cartilage area (which was lowest in the SPE group), the average values for the individual variables that contributed to factors 1, 2, and 3 were intermediate for the SPE group (between the values for the ERT and control groups) (Table 3). DISCUSSION To our knowledge, this is the first study to demonstrate the protective effect of long-term ERT at the histologic level in an animal model of naturally occurring OA. Treatment with ERT significantly reduced articular cartilage lesions of OA in this monkey model. Unique features of the present study were its long-term duration (3 years) and inclusion of a relatively high number of monkeys (n 166). The age range of the monkeys used in this study ( years) would be approximately equivalent to an age range of years in humans. Our previous studies on the effect of age on OA severity in these animals indicated that animals in this age range would be in an early stage of the disease process (31), thus explaining the lack of severe histologic lesions of OA in these animals. Little is known about the pathogenesis of early OA and, to our knowledge, no previous controlled studies have been done to examine the role of estrogen in the early stages of the naturally occurring disease. Factor 1 scores (weighted by variables for OA severity) were significantly lower in the ERT group compared with the control group; however, the mechanisms behind this protective effect of ERT remain unclear. Estrogen receptors have been documented in chondrocytes from humans and from some animal species, including cynomolgus macaques (34 37). Furthermore, previous studies have shown that in cell culture, estrogen receptors on articular cartilage chondrocytes from cynomolgus macaques are transcriptionally functional (37), suggesting a possible direct effect of estrogen on chondrocytes. In addition, ERT may stimulate the insulin-like growth factor (IGF) system within joint tissues. Previous studies of ovariectomized monkeys demonstrated that, compared with untreated controls, animals given ERT had higher synovial fluid levels of IGF-1 and IGF-2 as well as 2 IGF binding proteins, IGFBP-1 and IGFBP-3 (38). Furthermore, chondrocytes isolated from monkeys receiving ERT produced higher levels of IGFBP-2 and had a higher level of proteoglycan synthesis than cells from untreated controls (37). The exact mechanism by which ERT stimulates IGF system components in joint tissues and whether this is related to the observed decrease in cartilage lesions seen in monkeys receiving ERT are not clear and require further study. When the ANOVA for factor 1 was adjusted for treatment group and weight, there was a highly significant effect of age on OA severity; this was accentuated when data from all 3 groups were examined. Because age is well recognized as the number one risk factor for OA (39), it is not surprising that it had a greater impact on factor 1 (P 0.003) than ERT treatment did (P 0.044). Although the data demonstrate that ERT does not have a strong enough protective effect to completely override the effect of age on the knee joint, the results provide evidence that ERT significantly reduces the articular cartilage lesions of OA. Interestingly, weight did not have a statistically significant effect on OA severity in the present study. This has also been found in previous studies of OA in these animals (31) and is most likely due to the fact that weights among the animals were similar, with no animals that could be characterized as obese. The present study demonstrated significantly

7 1962 HAM ET AL higher factor 2 scores (weighted by subchondral bone thickness and area) of the animals given ERT compared with the controls. ERT is known to maintain bone mineral density in postmenopausal women due to inhibition of bone turnover (40), but to our knowledge, subchondral bone has not been studied specifically. Because the role of subchondral bone in the progression of OA has not been well characterized, it is not clear whether the thickness difference observed in this study (most likely due to the prevention of subchondral bone loss in the ERT group) is protective or detrimental with regard to OA severity. Increased subchondral bone density is often observed radiographically in humans with OA, and increased subchondral bone thickness has been observed early in the OA disease process in cynomolgus monkeys. The monkey studies, however, have all been ex vivo, and thus, it is not known whether a change in subchondral bone thickness predicts the progression of OA and/or directly influences the severity of articular cartilage lesions. In the present study, the higher factor 2 scores in the ERT group did not prevent a concurrent reduction in the severity of cartilage lesions in this group. Bony lesions that have been directly linked to OA severity are the formation of osteophytes and changes in the shape of the joint (41). In the present study, the average number of osteophytes was lower in the ERT group than in the control group, suggesting that the bony lesions of OA were less severe with ERT. Although factor 3 (weighted by articular cartilage thickness and area) was not significantly affected by treatment in the present study, the mean articular cartilage thickness of monkeys in the ERT group was 100 m thicker than that of monkeys in the control group. A recent study using magnetic resonance imaging (MRI) revealed that postmenopausal women taking ERT for 5 years had thicker articular cartilage compared with those who were not taking ERT (42). The OA lesions in animals treated with SPEs were not significantly different from those in the untreated control group. This may be due to the fact that SPEs have a lower affinity for the estrogen receptor than does 17 -estradiol (20). It is possible that at a higher dosage, SPE may have provided protection against OA, since the severity scores in this group were intermediate between those of the control and ERT groups. These same SPE-treated animals had significant differences in serum lipid concentrations and a nearly significant reduction of atherosclerosis lesions (P 0.10) compared with the control animals (21). The results of the present study provided no evidence for any detrimental effects of SPEs. This study did not evaluate the effects of progesterone, which is often given with estrogen in humans receiving ERT. Progesterone must also be considered when evaluating studies in humans. A previous study in humans found that there was no association between knee symptoms and ERT among women who were given estrogen plus progesterone, but the severity of OA lesions of the knee was not directly assessed (11). In addition, a study of synovial fluid levels of IGF-1 (which may be protective against OA) in ovariectomized monkeys revealed that IGF-1 levels were higher with estrogen treatment than with estrogen plus progesterone (38). Together, these studies suggest that progesterone might modulate the beneficial effects of estrogen on joint tissues. The results of the present study showed a significant protective effect of estrogen on cartilage lesions of OA, but further studies are needed to determine whether concurrent progesterone treatment would have modulated those effects. There are several limitations of the present study, the most important of which is a lack of baseline information about the knee joints of these animals. Taking a baseline biopsy, however, would not have been feasible and would most likely not have been informative. Baseline radiographs would have identified animals with severe preexisting disease; however, none were identified at the termination of the study, which indicates that none of the animals had severe OA at baseline. Finally, high-detail MRI techniques, which would provide more information on articular cartilage and bone morphology, were not available at the time this study was initiated. The present study also lacks clinical data; however, subtle signs of joint pain would have been difficult to distinguish in these animals, since they were housed in groups. During the study, none of the animals was observed to be clinically lame. Finally, the fact that these animals were fed an atherogenic diet may be considered by some to be a study limitation. This diet, however, is similar to that consumed by humans in Western societies. Furthermore, the diet consumed by these animals was carefully controlled and was similar among all treatment groups. We have observed lesions of OA in newly imported, wild-caught cynomolgus monkeys that have never eaten an atherogenic diet, suggesting that this diet is not a major predisposing factor in the disease. The effect of an atherogenic diet on OA lesion severity, however, has not been systematically studied in this model. In summary, the results of the present study demonstrate that long-term ERT decreases the articular

8 LONG-TERM ERT DECREASES OA SEVERITY IN CYNOMOLGUS MONKEYS 1963 cartilage lesions of knee OA in cynomolgus monkeys. These results, coupled with the results of previous epidemiologic studies in humans showing that postmenopausal women taking ERT have a lower prevalence of OA (43), provide a strong argument that ERT may be a simple way to decrease the severity of OA in estrogen-deficient women, particularly for those individuals who do not have multiple risk factors. Therefore, in addition to reducing the risk of osteoporosis and atherosclerosis, the reduction in the severity of OA may be an additional reason for postmenopausal women to take ERT. ACKNOWLEDGMENTS The authors thank Dr. Thomas B. Clarkson for providing the knee joints for this study and Mary Anthony for providing the details of the clinical trial. We also thank Hermina Borgerink for technical assistance in preparing the histology sections and Jean Gardin for assistance with tissue collection and processing. REFERENCES 1. Wluka AE, Cicuttini FM, Spector TD. Menopause, oestrogens and arthritis. Maturitas 2000;35: Oliveria SA, Felson DT, Reed JI, Cirillo PA, Walker AM. Incidence of symptomatic hand, hip, and knee osteoarthritis among patients in a health maintenance organization. Arthritis Rheum 1995;38: Felson DT, Nevitt MC. The effects of estrogen on osteoarthritis. Curr Opin Rheumatol 1998;10: Felson DT, Zhang Y, Hannan MT, Naimark A, Weissman BN, Aliabadi P, et al. The incidence and natural history of knee osteoarthritis in the elderly: the Framingham Osteoarthritis Study. Arthritis Rheum 1995;38: Meachim G, Pedley RB. Implications of a sex difference in osteoarthrosis [letter]. Ann Rheum Dis 1980;39: Nevitt MC, Felson DT. Sex hormones and the risk of osteoarthritis in women: epidemiological evidence. Ann Rheum Dis 1996;55: Zhang Y, McAlindon TE, Hannan MT, Chaisson CE, Klein R, Wilson PWF, et al. Estrogen replacement therapy and worsening of radiographic knee osteoarthritis: the Framingham Study. Arthritis Rheum 1998;41: Nevitt MC, Cummings SR, Lane NE, Genant HK, Pressman AR. Current use of oral estrogen is associated with a decreased prevalence of radiographic hip OA in elderly white women [abstract]. Arthritis Rheum 1994;37 Suppl 9:S Samanta A, Jones S, Regan M, Wilson S, Doherty M. Is osteoarthritis in women affected by hormonal changes or smoking? Br J Rheumatol 1993;32: Spector TD, Nandra D, Hart DJ, Doyle DV. Is hormone replacement therapy protective for hand and knee osteoarthritis in women?: The Chingford Study. Ann Rheum Dis 1997;56: Nevitt MC, Felson DT, Williams EN, Grady D, for the Heart and Estrogen/Progestin Replacement Study Research Group. The effect of estrogen plus progestin on knee symptoms and related disability in postmenopausal women: the Heart and Estrogen/ Progestin Replacement Study, a randomized, double-blind, placebo-controlled trial. Arthritis Rheum 2001;44: Rosner IA, Goldberg VM, Moskowitz RW. Estrogens and osteoarthritis. Curr Opin Rheumatol 1986;10: Turner AS, Athanasiou KA, Zhu CF, Alvis MR, Bryant HU. Biochemical effects of estrogen on articular cartilage in ovariectomized sheep. Osteoarthritis Cartilage 1997;5: Mendelsohn ME, Karas RH. Mechanisms of disease: the protective effects of estrogen on the cardiovascular system. N Engl J Med 1999;340: Altkorn D, Vokes K. Treatment of postmenopausal osteoporosis. JAMA 2001;285: Doren M, Samsioe G. Prevention of postmenopausal osteoporosis with oestrogen replacement therapy and associated compounds: update on clinical trials since Hum Reprod 1995;6: Clarkson TB, Anthony MS, Williams JK, Honore EK, Cline JM. The potential of soybean phytoestrogens for postmenopausal hormone replacement therapy. Exp Biol Med 1998;217: Ravnikar VA. Compliance with hormone therapy. Am J Obstet Gynecol 1987;156: Ravnikar VA. Compliance with hormone replacement therapy: are women receiving the full impact of hormone replacement therapy preventive health benefits? Womens Health Issues 1992; 2: Miksicek RJ. Interaction of naturally occurring nonsteroidal estrogens with expressed recombinant human estrogen receptor. J Steroid Biochem 1994;49: Clarkson TB, Anthony MS, Moran TM. Inhibition of postmenopausal atherosclerosis progression: a comparison of the effects of conjugated equine estrogens and soy phytoestrogens. J Clin Endocrinol Metab 2001;86: Picherit C, Bennetau-Pelissero C, Chanteranne B, Lebecque P, Davicco MJ, Barlet JP, et al. Soybean isoflavones dose-dependently reduce bone turnover but do not reverse established osteopenia in adult ovariectomized rats. J Nutr 2001;131: Picherit C, Coxam V, Bennetau-Pelissero C, Kati-Coulibaly S, Davicco MJ, Lebecque P, et al. Daidzein is more efficient than genistein in preventing ovariectomy-induced bone loss in rats. J Nutr 2000;130: Anderson JJ, Ambrose WW, Garner SC. Biphasic effects of genistein on bone tissue in the ovariectomized, lactating rat model. Proc Soc Exp Biol Med 1998;217: Carlson CS, Loeser RF, Jayo MJ, Weaver DS, Adams MR, Jerome CP. Osteoarthritis in cynomolgus macaques: a primate model of naturally occurring disease. J Orthop Res 1994;12: Held JR, Wolfle TL. Imports: current trends and usage. Am J Primatol 1994;34: Clarkson TB, Hughes CL, Klein KP. The nonhuman primate model of the relationship between gonadal steroids and coronary heart disease. Prog Cardiovasc Dis 1995;38: Jerome CP, Carlson CS, Register TC, Bain FT, Jayo MJ, Weaver DS, et al. Bone functional changes in intact, ovariectomized, and ovariectomized, hormone-supplemented adult cynomolgus monkeys (Macaca fascicularis) evaluated by serum markers and dynamic histomorphometry. J Bone Miner Res 1994;9: Wagner JD, Carlson CS, O Brien TD, Anthony MS, Bullock BC, Cefalu WT. Diabetes mellitus and islet amyloidosis in cynomolgus monkeys. Lab Anim Sci 1996;46: Jerome CP, Peterson PE. Nonhuman primate models in skeletal research. Bone 2001;29: Carlson CS, Loeser RF, Purser CB, Gardin JF, Jerome CP. Osteoarthritis in cynomolgus macaques. III. Effects of age, gender, and subchondral bone thickness on the severity of disease. J Bone Miner Res 1996;11: Carlson CS, Loeser RF, Johnstone B, Tulli HM, Dobson DB, Caterson B. Osteoarthritis in cynomolgus macaques. II. Detection of modulated proteoglycan epitopes in cartilage and synovial fluid. J Orthop Res 1995;13:

9 1964 HAM ET AL 33. Bowen WH. Determination of age in monkeys (Macaca irus) on the basis of dental development. Lab Anim Sci 1970;4: Ushiyama T, Ueyama H, Inoue K, Ohkubo I, Hukuda S. Expression of genes for estrogen receptors alpha and beta in human articular chondrocytes. Osteoarthritis Cartilage 1999;7: Sheridan PJ, Aufdemorte TB, Holt GR, Gates GA. Cartilage of the baboon contains estrogen receptors. Rheumatol Int 1985;5: Dayani N, Corvol MT, Robel P, Eychenne B, Moncharmont B, Tsagris L, et al. Estrogen receptors in cultured rabbit articular chondrocytes: influence of age. J Steroid Biochem 1988;31: Richmond RS, Carlson CS, Register TC, Shanker G, Loeser RF. Functional estrogen receptors in adult articular cartilage: estrogen replacement therapy increases chondrocyte synthesis of proteoglycans and insulin-like growth factor binding protein 2. Arthritis Rheum 2000;43: Fernihough JK, Richmond RS, Carlson CS, Cherpes T, Holly JMP, Loeser RF. Estrogen replacement therapy modulation of the insulin-like growth factor system in monkey knee joints. Arthritis Rheum 1999;42: Hamerman D. Aging and osteoarthritis: basic mechanisms. J Am Geriatr Soc 1993;41: Rodan GA, Martin TJ. Therapeutic approaches to bone diseases. Science 2000;289: Spector TD, Dacre JE, Harris PA, Huskisson EC. Radiological progression of osteoarthritis: an 11 year follow up study of the knee. Ann Rheum Dis 1992;51: Wluka AE, Davis SR, Bailey M, Stuckey SL, Cicuttini FM. Users of oestrogen replacement therapy have more knee cartilage than non-users. Ann Rheum Dis 2001;60: Felson DT, Nevitt MC. Estrogen and osteoarthritis: how do we explain conflicting study results? Prev Med 1999;28: New Address, Phone, and Fax Numbers for Arthritis & Rheumatism Editorial Office The Arthritis & Rheumatism editorial office has a new mailing address and new telephone and fax numbers, effective immediately. Correspondence, including manuscript submissions, should now be sent to: David S. Pisetsky, MD, PhD Editor, Arthritis & Rheumatism Mailing address: Box 3806, DUMC Durham, NC Courier address: 2200 W. Main St., Suite B210 Durham, NC Phone: (919) Fax: (919) jennifer.deyton@duke.edu