FDA guidelines and animal models for osteoporosis

Bone Vol. 17, No. 4, Supplement October |995:125S-133S

ELSEVIER

FDA GUIDELINES AND ANIMAL MODELS FOR OSTEOPOROSIS D. D. Thompson, H. A. Simmons, C. M. Pirie, and H. Z. Ke Department of Metabolic Diseases, Central Research Division, Pfizer Inc., Groton, CT 06340, USA

ABSTRACT The recent FDA Guidelines For Preclinical and Clinical Evaluation of Agents Used in the Treatment or Prevention of Postmenopausal Osteoporosis (1994) delineate specific preclinical animal models to demonstrate the efficacy and safety of new, potential agents for osteoporosis therapy. The Guidelines recommend that agents be evaluated in two animal species, including the ovariectomized (OVX) rat and in a second non-rodent model. We have performed a series of studies to determine whether the recommended OVX rat models, endpoints, and study design adequately address the efficacy and safety of therapeutic agents for the treatment or prevention of osteoporosis. Our study results indicate that the rat OVX model mimics postmenopausal cancellous bone loss when examined over relatively short periods of time. These data illustrate that cancellous bone turnover increases following OVX and this increased bone turnover produces bone loss. Estrogen completely blocks the activation of bone turnover and bone loss. Thus, our data suggest that the rat OVX model in the proximal tibia, distal femur, and lumbar vertebrae mimics conditions in the postmenopausal woman and is suitable for the evaluation of potential therapeutic agents for the prevention of osteoporosis. However, when the duration of the studies extends to 12 months as suggested by the Guidelines, the indices of cancellous bone turnover return to the value of sham controls, although the trabecular bone volume remains lower than that of sham controls in OVX rats. Therefore, it is difficult to determine the effects of potential therapeutic agents on the bone turnover in estrogen deficient conditions. Further, in long term OVX, the effects of increased skeletal size on the cancellous and cortical bone compartments of the skeleton become less clear. The effect that the increased bone area has on biomechanical strength parameters remains to be determined. Because bone strength and quality are two key endpoints to the animal pharmacology studies, it is essential to determine how the increase in skeletal size impacts the bone strength measurements. It therefore may be reasonable to restrict studies of potential therapeutic agents in OVX rats to a duration of less than 12 months and probably no more than 6 months. Combined with previously published data, our preclinical data on OVX rat model suggest that the proposed animal models by the Guidelines are appropriate for evaluation of agents to prevent bone loss in postmenopausal women. However, modification of endpoints and study duration of the rat OVX studies contained in the Guidelines is recommended. Furthermore, we found that the Guidelines should be modified to evaluate agents for the treatment of osteoporosis. Also, the Guidelines fail to account for osteoporosis therapy which uses combination or cyclical regimens. INTRODUCTION The recent FDA Guidelines For Preclinical and Clinical Evaluation of Agents Used in the Treatment or Prevention of Postmenopausal Osteoporosis (1994) recommend specific animal models for the evaluation of new, potential agents for osteoporosis therapy. The Guidelines recognize that no single animal model precisely mimics the human condition of osteoporosis. Consequently, the Guidelines recommend that agents be evaluated in two animal species, including the ovariectomized rat and in a second non-rodent model. An essential feature of the Guidelines is the recognition that appropriate animal models provide information on bone quality and structure not obtainable in patients participating in clinical trials. Animal models of osteoporosis are, therefore, pivotal for anticipating a compound's Address for correspondence and reprints: Dr. D. D. Thompson, Department of Metabolic Diseases, Pfizer Central Research, Groton, CT 06340. © 1995 by Elsevier Science Inc.

125S

8756-3282/95/$9.50 SSDI 8756-3282(95)00285-L

126S

D . D . Thompson et al. FDA guidelines and animal models

Bone Vol. 17, No. 4, Supplement October 1995:125S-133S

efficacy and safety as measured by bone quality. Thus, the proper use of the animal models and interpretation of the data are critical to the successful development of new therapy in this area. In this paper, we summarized available preclinical data and limitations of these animal models as it relate to the Guidelines. Furthermore, we discussed whether the recommended animal models, endpoints, and study design adequately address the efficacy and safety of therapeutic agents for the treatment or prevention of osteoporosis.

Do the Recommended Preclinical Animal Models Address the Efficacy and Safety of Agents for the Treatment and Prevention of Osteoporosis? The Guidelines equate the treatment and the prevention of osteoporosis and therefore recommend the same models, endpoints, and study design for an agent that is designed to either prevent or to treat osteoporosis. It is important to distinguish between these two therapeutic approaches for osteoporosis as each require different validation in preclinical models. In the prevention of osteoporosis, it is essential to prove an agent's ability to prevent the activation of bone turnover leading to bone loss, whereas in the treatment of osteoporosis, it is necessary to prove an agent's ability to restore bone and rebuild bone structure to an osteopenic skeleton. Thus, the animal model, the research design, and the endpoints used would be different for the different therapeutic requirements. The Guidelines recommend that new therapeutic agents undergo evaluation in the ovariectomized (OVX) rat model for 12 months following surgical removal of the ovaries. A second non-rodent model, preferably an ovariectomized primate, is to be used where the compound is given for 24 months. The design of these studies are clearly aimed at evaluating the ability of test compounds to prevent OVX-induced activation of bone turnover leading to bone loss. Estrogen replacement in either the OVX rat or primate has been shown to prevent increased bone turnover and bone loss (1-2, 6-8). Thus, these models are well-validated for assessing compound's ability to prevent bone loss in estrogen deficient conditions but are inappropriate to evaluate a compound's ability to restore bone mass to an osteopenic skeleton and therefore treat osteoporosis. Evaluation of test agents under consideration for the treatment of osteoporosis is more effectively evaluated in models of osteopenia where significant bone loss has already occurred. This could be accomplished in ovariectomized rats or primates if the evaluation of a compound begins after the animals have undergone significant bone loss. Thus, the animal models recommended in the Guidelines are appropriate for agents which prevent estrogen deficiency bone loss, but inappropriate for agents which are targeted for the treatment of osteoporosis or restoration of bone mass. Another potential therapy for osteoporoses would utilize combined or sequential therapy, where an anabolic agent is administered to an established osteopenic skeleton for restoring bone mass and rebuilding bone structure, then a second antiresorptive agent is given to maintain the new bone mass and bone structure induced by the anabolic agent (4,5). This treatment concept has been labeled as the "loss/restore/maintain" or "LRM" strategy by Jee et al (4,5). Also, the ADFR osteoporosis treatment strategy proposed by Frost (3) utilizes a two agents, one to activate (A) bone turnover, a second to depress (D) bone resorption, and then allow (Free) bone formation, thereafter, repeat (R) the above cycle. Clearly, the current Guidelines fail to adequately address the appropriate animal model, study design, and endpoints to assess efficacy and safety of agents used in these more complicated treatment scenarios. Thus, the Guidelines delineate animal models to evaluate agents that prevent bone loss due to estrogen deficiency, but fail to adequately delineate appropriate preclinical models to effectively evaluate a compound's ability to restore bone mass or evaluate combined or cyclical therapies. Are the Preclinical Animal Study Designs and Endpoints Appropriate for E v a l u a t i o n of Agents for the Prevention of Postmenopausal Bone Loss? It is well recognized that at menopause, bone turnover in women is increased and rapid bone loss is recorded. Do the OVX rat and primate models adequately mimic skeletal changes in women following menopause and therefore serve as models of postmenopausal osteoporosis? RAT O V A R I E C T O M Y (OVX) M O D E L

A. Immediate Effects of OVX on Rat Skeleton The fact that female rats lose cancellous bone in the proximal tibia and lumbar vertebrae following OVX as a direct result of estrogen deficiency is well documented and recently reviewed by Wronski and Yen (8), Kalu (6), and Frost and Jee (2). The currently accepted mechanism of bone loss in estrogen deficient rat skeletons is by an imbalance in bone turnover, where bone resorption exceeds bone formation.

Bone Vol. 17, No. 4, Supplement October 1995:125S-133S

D.D. Thompson et al. FDA guidelines and animal models

Initially, we determined the effects of O V X on the rat skeleton in the first 2 weeks following surgery. At day 14, bone turnover and bone resorption are increased and bone loss is occurring (Fig. la-d). Trabecular bone volume is significantly decreased in O V X rats by day 14 due to the combination of thinning and loss of individual trabeculae (Fig. le and f). Treatment with 17~l-estradiol completely prevents the O V X - i n d u c e d increases in bone resorption, bone turnover, and bone loss in cancellous bone sites. Thus, the early events recorded in cancellous bone of the proximal tibiae in O V X rats appear similar to that reported in early postmenopausal women, where increased bone turnover and loss of bone mass are reported. a.

Osteoclast

Number

4o

'!i _

:[

30

E

---t

E

I

Trabecular

b.

Bone

Volume

"-"-I

,----

o~ 20 10

'

I

'

'

'

'

5

o

c.

I

'

'

'

Mineral

0

'

10

15

Apposition

,

'

'

Rate

'

I

'

'

'

'

5

0

d.

Bone

I

'

'

'

'

I0

Formation

Rate/BV

3

°°°t

~2,

~400

~ J

m

/

E

0

'

'

'

~

0

I

'

'

5

'

x

I

'

'

10

e. T r a b e c u l a r

'

'

15

0

5

10

f. T r a b e c u l a r

Number

2O

4o

16-

30-

15

Width

.

BB

E 12" E '~ 8 -

Q ~

~20 10-

4

0

, 0

'

'

'

I

. . . .

I

. . . .

5 10 DAYS POST-SURGERY

0

15

. . . .

0

I

'

'

'

'

I

'

'

5 10 DAYS POST-SURGERY

'

'

15

FIG. 1. Effects of sham-operation (solid line) and OVX (dotted line) on proximal tibial cancellous bone histomorphometry in 5-month-old female rats 10 and 14 days post-surgery. *: p < 0.05 vs. sham. B. L o n g - T e r m E f f e c t s o f O V X o n R a t S k e l e t o n The Guidelines r e c o m m e n d that the O V X rat study be conducted for at least 12 months. The rationale for this is that the rat skeleton should be exposed to a compound's effects for a duration that would be equivalent to about 4 years in humans indexed to the average bone turnover rates in humans (100-200 days) and rats (40 days). However, few data exist in O V X rats documenting skeletal changes over 12 months. Thus, it is unclear if the rat skeleton retains similarities to the p o s t m e n o p a u s a l w o m e n ' s skeleton when evaluated over this extended period of time.

127S

128S

D . D . Thompson et al. FDA guidelines and animal models

Bone Vol. 17, No. 4, Supplement October 1995:125S- 133S

a. Whole Body BMC 20 i i _ . ~ j _ _ = _ ~ _ _ - ~ - - -. -- - - - - - - ~ = ..~,-~-=.3J=--~-=: -~ -

16 E

12

i .e__L~J=. -

..

.=

r3

8

i 1 i

0

'

I

0

1

2

3

4

5

6

7

8

9 10 11

12

b. Whole Body Bone Area 120

1

1

;

lOO8004

< E

60-" 40-" 200 0

2

3

4

5

6

7

I

8

9 10

11

12

8

9 10

11

12

c. Whole Body BMD 0.2 0.15E "~ E

0.1

¢3 0 . 0 5 -

0

1

2

3

4

5

6

7

d. Body Weight 800 70060O 500 E 400 (9 300 200 100 0

i j-4 -°'-.

I

0

2

5

i

i _i ' . ~

I

3 4 5 6 7 8 Months Post-Surgery

9 10 11

12

Bone Vol. 17, No. 4, Supplement October 1995:125S-133S

D.D. Thompson et al. FDA guidelines and animal models

FIG. 2. Effect of sham-operation (solid line) and OVX (dotted line) on whole body bone mineral content (BMC), bone area, bone mineral density (BMD), and body weight in 5-month-old rats. *: p < 0.05 vs. sham since this time point.

To evaluate the long-term effects of OVX on the rat skeleton, studies were conducted in sham-operated or OVX rats beginning at 3 months of age. Whole body and lumbar bone mineral content, bone area, and bone mineral density were determined in the sham-operated or OVX rats by dual energy x-ray absorptiometry (DEXA) (Hologic QDR-1000/W) over 12 months. After 12 months, lean and fat body mass were also determined by DEXA. The rats were then sacrificed, and bones harvested for histomorphometric analyses. Findings from this study in OVX rats showed significant differences from those reported in postmenopausal women. The most noteworthy difference was the significantly greater increase in bone mineral content (BMC) and bone area (BA) in OVX rats compared to sham-operated controls over the 12 month duration of this study (Figs. 2a and b). These differences in BMC and BA in OVX rats became apparent by 1 month post-OVX. The greater increase in total bone area in the OVX rat relative to the increase in bone mineral content yielded a bone mineral density that was significantly less than observed in sham-operated controls (Fig. 2c). The changes in bone area in OVX rats are in sharp contrast to those recorded in postmenopausal women where no changes in bone area have been reported. Thus, whole body bone mineral data from rat OVX studies lasting one month or more must be interpreted with caution. One of the methods cited in the Guidelines for bone mass quantity is bone ash determination. The data from our investigations suggest that the use of bone ash data to quantify bone mass changes due to OVX is inappropriate. Also, our data stress the importance of including both baseline controls and longitudinal measurements in long-term OVX rat studies. Body weight was significantly increased in ovariectomized rats compared to sham-operated rats in the first month following surgery (Fig. 2d). The body mass analyses by DEXA at 12 months showed that the increased body weight was entirely due to increased body fat with the ovariectomized rats having, on average, about 250 additional grams of fat compared to the sham-operated controls (Fig. 3a). The lean body mass was slightly reduced, albeit significantly, in OVX rats compared to sham-operated controls (Fig. 3b). These data demonstrate that the significantly increased body weight gain found in OVX rats was due entirely to increased body fat and not increased lean body mass.

a. Fat Body Mass 400

600 400 200

b. Lean Body Mass

t

300 E200 (3 100 0 SHAM

OVX

SHAM

OVX

FIG. 3. Fat and lean body mass in sham-operated and OVX rats at 12 months post-surgery. *: p < 0.05 vs. sham.

Histomorphometric analyses of cancellous bone in the proximal tibia demonstrate trabecular bone loss in OVX rats after 12 months compared to sham-operated controls (Fig. 4a). However, by 12 months no differences in bone turnover remain suggesting that a new steady state in bone remodeling has been achieved (Fig. 4b). Similarly no differences in mineral apposition rate or osteoclast number were recorded in OVX rats 12 months post surgery (data not shown). Since proximal tibial trabecular bone volume remained decreased while whole body bone mineral was significantly increased, there must be a net bone gain in some skeletal sites (cortical bone?) in OVX rats 12 months post-surgery. It is important to identify these skeletal sites as they are clearly inappropriate sites for mechanical testing and bone quality determination. These data suggest that the characterization of this long-term model is incomplete and suitability of this model for FDA Guideline is unclear.

129S

130S

D . D . Thompson et al. FDA guidelines and animal models

Bone Vol. 17, No. 4, Supplement October 1995:125S-133S

a. Trabecular Bone Volume 40

b. Bone Formation Rate/BV 200

3O o

o~ 20

~oo ~100 ~c

10

T

0-

I

OVX

SHAM

OVX

SHAM

FIG. 4. Effects of OVX on proximal tibial cancellous bone volume and bone formation rate-bone volume (BV) referent at 12 m o n t h s post-surgery in female rats. *: p < 0.05 vs. sham.

a. Trabecular Bone Volume 5 month old rat "k

~r

4O

b. Trabecular Bone Volume 19 month old rat

30-

20lOl

SHAM

t

I

I

OVX+E2

c. Bone Turnover Rate 5 month old rat

1200

400

l

OVX

SHAM

300

®

®

OVX+E2

OVX

d. Bone Turnover Rate 19 month old rat

200 100

*

0

l

SHAM

T

I

OVX

OVX+E2

SHAM

OVX

OVX+E2

FIG. 5. Differential effects of OVX and estrogen (E2, 30 ~g/kg/day) in proximal tibial cancellous bone volume and bone turnover rate in 5- and 19-month-old female rats at 4 or 8 weeks post-surgery, respectively. *: p < 0.05 vs. OVX.

C. Effects of OVX in G r o w i n g Versus Aged Fem al e Rats The Guidelines fail to address the recommended ages of rats to be used in these studies. We conducted studies to evaluate age-related differences in the skeletal responses of rats to OVX. Studies were conducted in growing (5 months of age) and aged (19 months of age) rats and following shamoperation, OVX+vehicle, or OVX+estradiol (30 lag&g/day) for either 4 weeks (growing) or 8 weeks (aged). Bone mineral density and bone histomorphometric analyses were conducted in bones from the rats at the termination of the study.

Bone Vol. 17, No. 4, Supplement October 1995:125S-133S

D . D . Thompson et al. FDA guidelines and animal models

a. Proximal Tibia Metaphysis Bone Turnover Rate

b. Lumbar Vertebra Bone Turnover Rate

300

300

200

200 0J

~100

100 , SHAM C,

OVX

SHAM

Proximal Tibia Metaphysis Trabecular Bone Volume

OVX

OVX+E2

d. Lumbar Vertebra Trabecular Bone Volume

40

o~ 20

m

0

T OVX+E2

40

t

"k

"/c

~c

o~ 20 10 0

I

SHAM e.

~r

30

OVX

OVX+E2

i

SHAM

Proximal Tibia Metaphysis Trabecular Thickness

i

OVX

OVX+E2

f. Lumbar Vertebra Trabecular Thickness 12t 81

E

E 41

SHAM

OVX

]

OVX+E2

SHAM

g. Proximal Tibia Metaphysis Trabecular Number 5

E:t

OVX

I

OVX+E2

h. Lumbar Vertebra Trabecular Number 5

~c

"A"

E3E ~2-

E

/

0

I

SHAM

4-

OVX

OVX+E2

I

SHAM

OVX

OVX+E2

131S

132S

D.D. Thompson et al. FDA guidelines and animal models

Bone Vol. 17, No. 4, Supplement October 1995:125S-133S

FIG. 6. Differential effects of OVX and estrogen (E2, 30 I.tg/kg/day) in cancellous bone histomorphometry of proximal tibial metaphyses and lumbar vertebra in 19-month-old female rats at 8 weeks post-surgery, respectively. *: p < 0.05 vs. OVX.

Significant differences were identified in the skeletal response to OVX in growing versus aged females rats. At the proximal tibial site, growing animals respond to OVX by losing bone more rapidly than do aged animals (Fig. 5a,b). Also, the magnitude of the response to OVX is greater in young rats than in old rats (Fig. 5a,b). However, a significant factor is that both growing and aged rats retain the capability to positively respond to estrogen replacement treatment. Administration of estrogen to estrogen deficient animals, whether young or aged, completely protected against bone loss (Fig. 5a,b). The sham-operated aged rats had reduced proximal tibial trabecular bone volume (16%) compared with the growing rats (29%) (Figure 5a and b). Following OVX, the aged rats had a trabecular bone volume of about 10% (a reduction of about 38% compared to sham) after 8 weeks compared with the growing rat of about 14% (a reduction of about 50% compared to sham) after 4 weeks (Fig. 5a,b). In both the growing and aged OVX rats, estradiol was able to completely block trabecular bone loss. Bone turnover rate in the OVX growing rat increased 3.5-fold after 4 weeks, whereas no significant increase was observed in the aged rats (Fig. 5c and d). These data indicate that the growing rat is more responsive to OVX than the aged rat. Furthermore, our data may indicate that 8 weeks of OVX in 19month-old rats is not suitable for the observation of bone turnover in proximal tibial metaphyseal cancellous bone. D. Differential Response in Skeletal Sites to OVX in Aged Female Rats The Guidelines do not specify which sites in the OVX rat skeleton are optimal for analysis of prospective therapeutic agents for osteoporosis, and few data exist comparing the effects of OVX in different regions of the skeleton. Therefore, the effects of OVX in different skeletal sites, including proximal tibia, distal femur, and lumbar spine, were investigated using bone mineral density and bone histomorphometric analyses. Below, data comparing the differential responses in the proximal tibia and lumbar vertebrae in 19 month old rats are presented. The proximal tibiae and lumbar vertebrae in OVX rats respond similarly to estrogen deficiency as measured by loss of trabecular bone and positive response to estrogen treatment (Fig. 6a-d). However, there were significant differences in bone turnover rate in response to OVX. Bone turnover rate showed no significant difference between sham and OVX in proximal tibia, while it increased significantly in OVX compared to sham controls in lumbar vertebrae at 8 weeks following OVX. Another difference between the two sites is in the manner in which cancellous bone is lost. In the proximal tibia, the trabeculae are thinner than in the vertebrae and following OVX, trabecular number significantly decreased while thickness is only slightly affected (Fig. 6e-h). This contrasts with the lumbar vertebrae where trabecular thinning is significantly decreased in the OVX rats while trabecular number is unchanged. Thus, the tibia loses bone by eliminating trabeculae and the vertebrae loses bone by thinning trabeculae. Both the proximal tibia and lumbar vertebrae of OVX rats respond to estrogen. LARGE A N I M A L M O D E L S The Guidelines also recommend that new potential agents for the prevention of osteoporosis be evaluated in a "second non-rodent species (i.e., larger remodeling species) which will be left to the discretion of the sponsor although there is evidence that the dog may not be a good model" (p.2). The OVX primate model was utilized for the evaluation of alendronate (1,7). These data showed transient declines in bone mineral content following OVX relative to the sham-operated group. The decline in bone mineral control due to OVX was blocked by high doses of alendronate. Histomorphometric analyses revealed that alendronate blocked the increased bone turnover induced by OVX and prevented the cancellous bone loss. Thus, this model shows similarities with the ovariectomized rat model in that bone turnover is increased following OVX and that agents which block bone turnover suppress bone loss. The relevance of this model to postmenopausal bone loss in women remains to be elucidated. Because the significant loss of bone mineral content following OVX was transient and the shamoperated primates gained bone mass throughout the two year study, it remains to be confirmed that permanent bone loss due to estrogen deficiency would be achieved if older adult primates were utilized instead of growing primates as used in alendronate study. Additional studies have evaluated bone changes following OVX in sheep and pigs. These models remain to be validated for the evaluation of potential agents for the prevention of osteoporosis.

Bone Vol. 17, No. 4, Supplement October 1995:125S-133S

D . D . Thompson et al. FDA guidelines and animal models

CONCLUSION The current FDA Guidelines For Preclinical and Clinical Evaluation of Agents Used in the Treatment or Prevention of Postmenopausal Osteoporosis (1994) recommend that new potential agents be evaluated in two animal models. First, the rat OVX model and a second model that utilizes a large animal species with a skeleton that remodels. The rat OVX model is useful for the evaluation of potential agents for the prevention of postmenopausal osteoporosis but not for agents for the treatment of osteoporosis. New guidelines must be developed to adequately address agents for the treatment of osteoporosis and for therapy which uses combinations of drugs. Limitations of the rat OVX model restrict the evaluations to cancellous bone sites in the skeleton. For the evaluation of cortical bone responses to therapy, the OVX rat model may not be appropriate since the rat does not mimic the postmenopausal women in this respect. The duration of studies to evaluate potential agents must be considered carefully, taking into account the long-term responses of the rat to OVX. A study duration of 12 months in OVX rat may be an inappropriate model for postmenopausal osteoporosis. The endpoints that are included in studies that investigate new agents must consider the intact skeletal response to OVX. Whole bone ashing and biomechanical analyses in OVX rat skeletons may not provide adequate information to evaluate efficacy and safety of compounds as the cross sectional area of bone increases in OVX and is reduced to sham levels with estrogen treatment. Therefore, the validity of these endpoints in the OVX rat is unclear. The choice of a second model for evaluation of new agents is more difficult. Currently, the primate OVX model would be preferred based on the proven efficacy of alendronate in this model, however, the sheep and pig OVX remain to be validated and may prove equally useful in the future. REFERENCES 1. Balena, R., Toolan, B. C., Shea, M., Markatos, A., Myers, E. R., Lee, S. C., Opas, E. E., Seedor, J. G., Klein, H., Frankenfield, D., Quartuccio, H., Fiorvanti, C., Clair, J., Brown, E., Hayes, W. C., and Rodan, G.. The effects of two-year treatment with the aminobisphosphonate alendroante on bone metabolism, bone histomorphometry, and bone strength in ovariectomized nonhuman primates. J Clin Invest 92:2577-2586; 1993. 2. Frost, H. M., and Jee, W. S. S. On the rat model of human osteopenias and osteoporosis. Bone and Mineral 18:227-236; 1992. 3. Frost, H. F. Treatment of osteoporosis by manipulation of coherent bone cell populations. Clin Orthop Rel Res 143:227-244; 1979. 4. Jee, W. S. S., Tang, L., Ke, H. Z., Sttterberg, R. B., Kimmel, D. B. Maintaining restored bone with bisphosphonate in the ovariectomized rat skeleton: Dynamic histomorphometry of changes in bone mass. Bone 14:493-498; 1993. 5. Jee, W. S. S., Ma, Y. F., Li, M., Liang, X. G., Lin, B. Y., Li, X. J., Ke, H. Z., Mori, S., Setterberg, R. B., Kimmel, D. B. Sex steroids and prostaglandins in bone metabolism. Zeigler R., Pfeilschifter J., and Brautigam M. eds. Sex Steroids and Bone. Springer-Verlag, Berline Germany, 1994; 119-150. 6. Kalu, D. N. The ovariectomized rat model of postmenopausal bone loss. Bone and Mineral 15:175-192,1991. 7. Thompson, D. D., Seedor, J. G., Quartuccio, H., Solomon, H., Fioravanti, C., Davidson, J., Klein, H., Jackson, R., Clair, J., Frankenfield, D., Brown, E., Simmons, H. A., and Rodan, G. A. The bisphosphonate, Alendronate, prevents bone loss in ovariectomized baboon. J Bone Miner Res 7:951-960; 1992. 8. Wronski, T. J., and Yen, C.-F. The ovariectomized rat model as an animal model for postmenopausal bone loss. Cells Materials suppl. 1:69-74; 1991.

133S