E. A. Thorndike and A. S. Turner

Department of Clinical Sciences, Colorado State University, Ft.Collins ,CO 80523

Received 9/25/97 Accepted 4/10/98

5. OSTEOPOROSIS

5.1 Animal Models to Study Osteoporosis

Previously, rats, mice, dogs, swine, and non-human primates have been used to study the pathophysiology of bone loss (table 1). However, none of these animals perfectly mimic human osteoporosis because of obvious differences in reproductive physiology, hormone profiles and biomechanical bone characteristics. Even though most of these animal models do not experience a natural menopause, this state can be induced with some reliability by ovariectomy.

The most commonly used animal model for osteoporosis is the rat, and there is extensive literature studying the ovariectomized rat as a model for histomorphometric changes, biochemical markers, methodology for bone densitometry and evaluation of bone fragility (16-20). Rats are inexpensive, easy to house, and do not carry the societal concerns of the other models; their shorter life span also facilitates studies on the effects of aging on the bone. Because the rodent has been used so extensively in research of all types, much is known about bone turnover and the effect of diet on this process. Cortical thinning and increased fragility are well documented in aging rat and mouse bones, but it is unclear if this results in increased incidence of fractures.

There are however, disadvantages in the use of rodents as a model for osteoporosis. Rodents do not experience a natural menopause, but ovariectomy has been used to produce an artificial menopause. (16-20) Furthermore, although aged rodents have Haversian systems and ovariectomy results in a significant bone loss, they have a limited naturally occurring basic multicellular unit-based remodeling. Rats also have lamellar bone (although most is "fine-fibered"), trabecular remodeling, and no intracortical remodeling. (18-20) Longitudinal bone growth increases transiently after ovariectomy in long bones of rats, but this can be minimized by the use of aged rats (9-12 months old) or of skeletal sites where the longitudinal growth is greatly reduced (e.g. lumbar vertebrae) (18). Another limitation is the absence of impaired osteoblast function during the late stages of estrogen deficiency (18) which may be due to the decrease in bone fatigue experienced by small quadruped animals such as the rodent. Longer term studies which require several biopsies, or large blood samples, also are impossible in such a small animal. For a detailed review of other models of osteoporosis see reference 1 (1).

5.2 Sheep as a Model for Osteoporosis

Sheep are well suited for the study of osteoporosis for the following reasons: (1) bone loss associated with estrogen deficiency has been documented (21-25); (2) the hormone profiles of ewes are temporally and quantitatively similar to those of women (26); (3) the osteoblast product, osteocalcin, has been clearly defined in the sheep model (27-30); (4) the size of the sheep permits prosthetic implantation to evaluate performance in osteoporotic bone (31-33); and (5) the large size of these animals allows for the collection of substantial blood and urine samples, as well as multiple iliac crest biopsies (1,34).

Our group previously conducted studies of osteopenia in sheep following OVX which showed a significant decline in cancellous bone volume (BV/TV%) of the iliac crest following OVX (35) (figure 2). Subsequent studies (24,25) demonstrated bone loss (measured by dual energy X-ray absorptiometry (DXA)) in the lumbar vertebrae following OVX (figures 3). The activity of bone-specific alkaline phosphatase (BSAP), an isoenzyme involved in bone formation and skeletal mineralization, was found to be significantly increased in both the sham group and the group which underwent OVX and received estradiol implants (OVXE) (figure 4). This demonstrated that bone loss in the lumbar vertebrae of sheep following OVX was ameliorated by estrogen replacement therapy (ERT). The increase in activity of BSAP in the OVX group was highly indicative of accelerated bone turnover and was in agreement with the changes seen in aging adult humans, particularly postmenopausal women (36,37). Future studies using OVX sheep should investigate other newer biochemical markers indicative of type I collagen synthesis and degradation that are most likely to reflect bone turnover. Such markers include the carboxy-terminal propeptide (PICP) to reflect bone matrix synthesis and the carboxy-terminal telopeptide (ICTP) which reflects bone matrix degradation (37,38). These biochemical markers will be important in monitoring new therapeutic regimens for osteoporosis when using animal models.

Figure 2. Changes in cancellous bone volume (BV/TV%) of the iliac crest in 14 skeletally mature ewes (mean + S.E) six months after ovariectomy (OVX). * p = 0.0012; Wilcoxon Signed-Rank Test.

Figure 3. Changes in BMD of the 6th lumbar vertebra at 3 and 6 months expressed as percent change from baseline (Mean + S.E.) in sham (filled dots), ovariectomized (OVX, open squares) and ovariectomized estrogen-treated (OVXE, open triangles) ewes. Significant difference compared to OVX; * p < 0.10, **p < 0.05, *** p < 0.01. (See Ref # 25 for details).

Figure 4. Percentage change (mean + S.E.) of bone-specific alkaline phosphatase (BSAP). Estrogen treatment did not change BSAP at any time point compared to sham, however OVX significantly increased BSAP at 3, 4, 5 and 6 months compared to Sham and OVXE groups (* p < 0.05; based on ANOVA). See Ref # 24 for details).

The bony changes seen in the proximal femur following menopause have been studied intensively for many years. We documented a statistically different Singh index in the proximal femur in OVX ewes compared to both young and old sheep (35). For this model, precise measurements of excised sheep femora using DXA are possible (39). However, one distinct disadvantage of using a quadruped animal (including dogs) for measuring longitudinal (temporal) changes in BMD using DXA is the technical difficulty in positioning the animal to examine the femoral neck region. In longitudinal studies, it is critical that an identical region of interest is evaluated on a repeatable basis.

Sheep do not experience spontaneous menopause and extremely old animals are usually culled for economic reasons. Therefore finding animals with osteoporotic fractures of the spine and femoral neck is unlikely.

As densitometric techniques such as DXA do not furnish sufficient information about the quality of bone, an important endpoint when using all animal models of osteoporosis is the fragility of bone as a result of OVX. Thus, one of the important steps in the characterization of the aged OVX ewe as a model for osteoporosis will be determination of biomechanical changes of the bone.

Questions yet to be answered with this model of osteoporosis include:

1) At what time of the year should ovariectomy be performed to demonstrate the most rapid bone loss? In other words, should ovariectomy be performed in anestrus prior to when estrus cycles begin or during the breeding season when estrus cycles are occurring regularly?

2) Is skeletal fragility likely to occur after many years following ovariectomy?

3) Are there seasonal and circadian fluctuations in bone density, or fluctuations associated with parity, pregnancy and number of offspring and lactation?

4) What are the most suitable sites for measuring change in bone mass in sheep?

5) What extrinsic factors are likely to confound the bone loss following ovarian hormone deficiency? Such extrinsic factors may include breed, diet (e.g. Ca:P ratio) and exposure to sunlight (e.g. sunny or cloudy environment).

6) Will societal pressures (animal rights groups), lead to difficulties in using this model ?