Long-term osteopenic changes in cancellous bone structure in ovariectomized rats.код для вставкиСкачать
THE ANATOMICAL RECORD 2361433-441 (1993) Long-Term Osteopenic Changes in Cancellous Bone Structure in Ovariectomized Rats S.C. MILLER AND T.J. WRONSKI Division of Radiobiology, School of Medicine, University of Utah, Salt Lake City, Utah (S.C.M.); Department of Physiological Sciences, College of Veterinary Medicine, University of Florida, Gainesville, Florida (T.J.W.) ABSTRACT Cancellous bone mass decreases following ovariectomy in rodents, providing a useful model for post-menopausal bone loss in humans. This study describes and quantifies the longer-term changes in cancellous bone structure in the ovariectomized (OVX) rat. Rats were OVX or sham-OVX at 100 days of age and bones were collected 540 days later. Lumbar vertebral bodies were prepared for microradiography and structural analyses (nodal analyses and star volume analyses) of cancellous bone. Proximal humerii were prepared for scanning electron microscopy (SEMI. Microradiography confirmed the loss of cancellous bone from the central spongiosa regions of the vertebral bodies and the humerii in the OVX rats. Changes in trabecular structural elements included relative increases in the number of free to free, cortical to free, cortical to node struts and decreases in the node to node struts in the OVX animals compared with controls. There were increases in average lengths of the node to free, node to node, and free to free trabecular struts in the OVX animals. The marrow star volume was increased in the OVX animals indicating a greater trabecular separation in these animals compared with controls. Viewed by SEM, metaphyseal trabeculae in the controls consisted of rods and plates but in the OVX animals the remaining trabeculae were mostly longitudinal rods with smaller transverse connecting rods. The remaining bone in the OVX animals was found in the lateral metaphyseal areas and is consistent with maintenance of the structural capacity of the bone. These long-term changes in cancellous bone structure are likely due to the continuation of functional skeletal loading but a decrease in gonadal hormones resulting in a decreased necessity to maintain a skeletal mineral store for reproduction (e.g., pregnancy and lactation). o 1993 Wiley-Liss, Inc. Key words: Bone, Cancellous bone, Osteopenia, Ovariectomy, Morphometry, Rats Osteoporosis is a disease characterized by a decrease in bone mass (osteopenia), strength and functional capacity resulting in fractures and skeletal pain (Alioa et al., 1985). The primary determinants of skeletal strength are bone mass and the organization and structure or local architecture of the osseous tissues. The possible relationship of the structure of cancellous bone with mechanical competence of this tissue was noted by Arnold et al. (1966) who observed that during aging of human bone there were perforations of individual bone spicules resulting in the loss of connectivity between trabeculae. It was also observed that as plate-like trabeculae became perforated, there was a successive remodeling into more rod-like structures (Arnold, 1980). It was proposed that structural changes in the trabecular lattice may result in a greater decrease in bone strength than would be predicted solely from bone mass measurements (Bell et al., 1967). More recent experimental studies have confirmed the relationship 0 1993 WILEY-LISS, INC. between changes in cancellous bone structure and strength (Kleerekoper et al., 1985; Mosekilde et al., 1987). Gonadal hormones have substantial influences on skeletal metabolism. For example, the decline in skeletal mass after the cessation of ovarian function in humans is well recognized (e.g., post-menopausal osteoporosis).Estrogen replacement is a common therapy to slow the rate of bone loss in oophorectomized or postmenopausal women. Bone loss also occurs in rodent models of gonad hormone deficiency. The ovariectomized (OVX) rat, for example, is a very useful model, replicating many events associated with the develop- Received October 28, 1992; accepted December 29, 1992. Address reprint requests to Dr. Scott C. Miller, Division of Radiobiology, Building 586, University of Utah, Salt Lake City, UT 84112. 434 S.C. MILLER AND T.J. WRONSKX ment of post-menopausal osteoporosis in humans (Kalu, 1991). Following OVX in rats there is a rapid decrease in cancellous bone mass, particularly evident in the metaphyseal regions of the long bones (Wronski et al., 1985,1986; Turner et al., 1988; Kalu et al., 1989; Miller et al., 1991). This initial phase of rapid bone loss is associated with increased bone turnover (increases in bone formation and resorption), but with an excess in bone resorption resulting in a decreased cancellous bone volume, similar to the perimenopausal period in humans (Stepan et al., 1987; Wronski et al., 1986, 1989b).At later times after OVX, cancellous bone volume continues to decline, but at a slower rate (Wronski et al., 1989a,b).There are also corresponding decreases in the bone turnover rate over those observed early after OVX. This study describes and quantifies structural characteristics of cancellous bone in aged rats that were ovariectomized as young adults. This was done to determine the long-term structural changes in cancellous bone that occur in the absence of ovarian hormones. The results from this study demonstrate that 1.5 years after OVX, in addition to substantial differences in cancellous bone mass, there are substantial changes in cancellous bone structure. These changes may reflect the maintenance of mechanical properties on the bone in the absence of ovarian hormones. MATERIALS AND METHODS Experimental Design Twenty-four female Sprague-Dawley rats (Charles River Co., Cambridge, MA) were obtained at 85 days of age. The animals were caged individually in a room with a 13 hours on and 11 hours off light/dark cycle. They were weight randomized into experimental and control groups. At 100 days of age, 12 animals were ovariectomized (OVX) and the other 12 were shamOVX using the dorsal approach (Waynforth, 1980) while under ketasetlxylazine anesthesia. To minimize the increase in body weight associated with OVX (Wronski et al., 1987), the food consumption of OVX rats was restricted to that of the control rats (pairfeeding). The animals were fed a standard rodent chow. The same number of control and OVX rats was scheduled for sacrifice at 540 days post-OVX. Some animals died from mammary tumors, renal calculi and respiratory infections, therefore, the sample size at 540 days post-OVX consisted of 7 controls and 10 OVX rats. Tissue Preparation At 540 days after OVX, the animals were killed and the lumbar vertebral bodies and humerii were collected. The lumbar vertebral bodies were fixed in 70% ethanol and further dehydrated in absolute ethanol. The vertebral bodies were separated a t the intervertebra1 disks and embedded individually in methyl methacrylate. One vertebral body from each animal was cut in the sagittal plane and another cut in cross-section using a bone saw. The sections were ground to 100 (*m in thickness and microradiographed. One humerus from each of the OVX and sham-OVX groups was also prepared for microradiography. In this case the humerus was cut in the sagittal plane. N-N F-F Fig. 1 . Diagram illustrating the types of bone “struts” determined by image analysis. The intersection of the various types of struts are termed “nodes.”The types of struts include cortical node to free end (C-F), cortical node to node (C-N),node to free end (N-F), node to node (N-N), and free end to free end (F-F). Image Analysis One sagittal section microradiograph and two crosssection microradiographs taken in the middle of the lumbar vertebral bodies from each animal were selected for trabecular structural analyses. Using a television microscope image analysis system (KSS Scientific Consultants, Magna, UT) interfaced with a microcomputer, as previously described (Miller et al., 1989), the entire cancellous bone areas in the crossand longitudinal-sections of the lumbar vertebral bodies were quantified. Using commercial software (KSS Scientific Consultants), the structure of the cancellous bone was quantified by “nodal analysis” and “star volume analysis.” Nodal analyses are useful for defining the connectivity and structure of bone spicules in spongiosa regions from 2 dimensional images (Garrahan et al., 1986; Compston et al., 1987), such as from microradiographs. The microradiographs are captured from the television microscope image analysis system and converted to binary form. The areas to be quantified (using both the nodal analysis and star volume software) are selected and the bone image is skeletonized, obtaining the symmetrical axis, as previously described (Compston et al., 1989). The symmetrical axis is used to identify the trabecular patterns, or “struts” based on the connections of these struts with other struts a t “nodes” or free-ends. The following types of struts are defined, as illustrated in Figure 1, free end to free end (free-free), node to free end (node-free),node to node (node-node),cortex to free (cortex-free),and cortex to node (cortex-node).The data CANCELLOUS BONE IN OVARIECTOMIZED RATS Fig. 2. Microradiographs. a: Cross section of the lumbar vertebral body from a control, non-ovariectomized rat. x 18. b: Cross section of the lumbar vertebral body from an OVX rat. x 18. c: Longitudinal section of the lumbar vertebral body from a control rat. x 10. d Longitudinal section of the lumbar vertebral body from an OVX rat. x 10.e: Sagittal section of the proximal humerus from a control rat. x 8. E Sagittal section of the proximal humerus from an OVX rat. x 8. 435 436 S.C. MILLER AND T.J. WRONSKI Fig. 3. SEM of the proximal humerus from an OVX rat. Little, if any, cancellous bone is found in the central region of the metaphysis (M). The remaining trabeculae are found in the lateral regions (arrowheads). E, epiphysis. GP, growth plate. X 15. Fig. 4. Proximal humerus from a control animal. Compared with the OVX animals (Fig. 3), there is considerable cancellous bone in the metaphyseal region. There is less cancellous bone in the central metaphyseal region than in the lateral region in these aging animals. X 15. are expressed as the percent of each strut type. The length of each strut was also quantified and is expressed as the average length. Changes in the distribution andor the lengths of each type of strut are not only indicators of trabecular connectivity but also cancellous bone architecture. Star volume analyses are useful in measuring the separation between trabeculae in spongiosa regions (Vesterby et al., 1989).The marrow star volumes were obtained from the same binary images of the microradiographs as described above for the nodal analyses of the cancellous bone. Marrow star volumes are defined as the mean volume calculated from computer-generated lines emanating from a point in the marrow in all directions to cancellous bone boundaries. If any of the lines intersect an artificial boundary, such as the edge of the field, the entire volume was excluded. Marrow star area measurements were taken in a uniform pattern of points over the spongiosa regions. The average marrow star volumes measured from each sampling point, were measured in 2 dimensions, as areas but were converted to volumes and expressed as mm’. Increases in the size of the marrow star volumes, and thus the marrow space between trabeculae is considered as an indicator of the removal or perforation of trabeculae. Scanning Electron Microscopy All but one proximal humerii from each group were trimmed open and rendered anorganic by treatment with 5%sodium hypochlorite (commercialbleach). This dissolved all organic material, including the marrow, exposing the mineral surface of the osseous tissues. The specimens were dehydrated in ethanol, critical point dried, coated with gold, and viewed in a JEOL JSM-35 scanning electron microscope at an accelerating voltage of 20 kV. The scanning microscopy of the metaphyseal spongiosa was done to assess the threedimensional changes in bone structure and to correlate these changes with the indices of bone structure obtained from two-dimensional microradiographic images used for morphology and structural analyses. RESULTS Microradiography At 1.5 years after ovariectomy, there were substantial differences in cancellous bone amount and structure at an axial skeletal site (lumbar vertebral body) and an appendicular site (proximal humerus) (Fig. 2). There were reductions in the amount of cancellous bone in the OVX animals, consistent with previous studies using ovariectomized rodents (reviewed in 437 CANCELLOUS BONE IN OVARIECTOMIZED RATS TABLE 1. Type and length of trabecular struts in longitudinal- and cross-sections of the lumbar vertebral bodies from aged control and OVX rats Longitudinal-sections Control ovx Type of strut (% of total ? SD) Free-free 6.4 2 4.2 14.8 2 8.4** Node-free 37.9 t 8.3 38.9 t 4.6 46.5 t 10.8 26.9 t 9.l**** Node-node Cortex-free 2.6 f 0.9 5.2 t 3.1** Cortex-node 8.1 t 2.2 14.1 2.1**** Length of strut (pm 2 SD) 102 f 27 125 t 20* Free-free 117 t 12 137 t 15**** Node-free 211 k 8 278 2 29**** Node-node 250 2 54 224 ? 27 Cortex-free 161 f 44 196 2 42 Cortex-node * Cross-sections Type of strut (% of total ? SD) 3.6 f 1.9 Free-free 27.8 6.9 Node-free 49.3 t 6.0 Node-node 2.1 1.2 Cortex-free Cortex-node 17.2 2 5.7 Length of strut (pm f SD) 130 60 Free-free 134 f 27 Node-free Node-node 241 t 59 Cortex-free 232 44 163 t 25 Cortex-node * * * Group Cross-sections Contro1 ovx Longitudinal sections Control ovx Marrow star volume (mm3 SD) * 0.110 0.069 0.948 f 0.714* 0.249 2 0.126 1.419 ? 0.738* *Significantly greater than controls, P < 0.005. aration which would result in an increase in marrow star volume are also clearly evident in the microradio1.8**** graphs (Fig. 2). 8.7 2 30.6 f 3.4 31.2 t 6.3**** 3.8 t 0.8**** 26.3 t 8.3** * TABLE 2. Marrow star volumes determined by image analysis of microradiographs of cross- and longitudinal-sections of the lumbar vertebral bodies from aged control and OVX rats 170 t 17* 252 2 61**** 312 t 31*** 253 t 52 167 t 17 *Significantly different from controls, P < 0.05. **Significantly different from controls, P < 0.025. ***Significantly different from controls, P < 0.01. ****Significantly different from controls, P < 0.005. Miller et al., 1991; Kalu, 1991). This decrease in cancellous bone was particularly evident in the central lumbar vertebral body (Fig. 2a-d) and the central metaphysis of the humerus (Fig. 2e,f). Cancellous bone was not observed below the metaphyseal primary spongiosa in the center of the proximal humerus in any of the OVX animals (Fig. Zf), whereas it was consistently found in all of the intact, control animals (Fig. Be). Scanning Electron Microscopy There was a dramatic difference in the structure and amount of metaphyseal spongiosa in the humerii of the OVX animals (Fig. 3) when compared with controls (Fig. 4). The marked reduction in cancellous bone observed in the OVX animals compared with controls correspond to the microradiographic images (Figs. 2e,f). The metaphyseal cancellous bone in the control animals consisted of plates and rods in a honeycomb-type pattern. While the lateral metaphyseal trabeculae were oriented preferentially along the longitudinal axis of the bone, the central metaphyseal trabeculae were less ordered. In these aging control animals, the plate-type trabeculae were most evident in the lateral portions of the metaphysis (Figs. 4 , 5 ) while in the central metaphysis the spicules were thin with fewer connections (Fig. 6) than in the more lateral portions (Fig. 5). There was no cancellous bone in the central metaphyseal region below the primary spongiosa in the OVX animals (Figs. Zf, 3). The primary spongiosa in these old animals formed a very thin mineralized plate adjacent to the growth plate (Fig. 7). In the lateral metaphyseal region in the OVX animals, the bone spImage Analyses icules were usually shaped as long struts spanning Image analyses were done on digitized, Z-dimen- from beneath the epiphyseal growth plate, at the prisional images taken from the microradiographs of the mary spongiosa, t o the metaphyseal cortex (Figs. 3, 7sagittal and cross-sections of the lumbar vertebral bod- 10). These bone spicules were generally quite straight, ies. There were substantial differences in the types of had a fairly uniform diameter and ran along the lonstruts (defined in Fig. 1)and their relative sizes in the gitudinal axis of the bone. The longitudinal trabecular OVX rats compared with the age-matched, controls struts were often interconnected by rod-like transverse (Table 1).In both the cross and sagittal sections of the struts (Fig. 9) which were usually much smaller in divertebral bodies, there were significant increases in ameter than the connecting longitudinal trabeculae. the relative percentages of free-free, cortex-free, and The longitudinal trabeculae had a broad base a t the cortex-node trabecular struts, with a significant de- insertion into the mineralized tissue adjacent to the crease in the percentage of node-node-type struts. growth plate (Fig. 10). There were also significant increases in the average DISCUSSION lengths of the free-free, node-free, and node-node trabecular struts. There were profound differences in cancellous bone The bone marrow star volume analyses were also volume and structure in the animals that had been performed from the images taken from the sagittal and OVX for 1.5 years compared with the sham-OVX anicross-sections of the lumbar vertebral bodies (Table 2). mals. A decrease in cancellous bone mass with the loss The marrow star volumes were substantially increased of ovarian function was expected as this is one of the in the OVX rats compared with controls when mea- defining features of the OVX rat model (Saville, 1969; sured in both planes. The increases in trabecular sep- Wronski et al., 1985; Kalu, 1991). There were, how- 438 S.C. MILLER AND T.J. WRONSKI Fig. 5. Lateral cancellous bone in the proximal humerus from a control animal. The cancellous bone is structured as a broad trabeculae plates (P)and rods (R). X45. Fig. 6. In the central metaphyseal regions of the control animals there are fewer plates, as observed primarily in the lateral regions (Fig. 5), and more bone spicules. Many of the spicules in the aging animals are very thin (arrowheads)and often are not well connected with other spicules. x 60. ever, dramatic differences in cancellous bone structure between the OVX and sham-operated controls. Structural analyses of two-dimensional sagittal and crosssectional plane images of the lumbar vertebral cancellous bone demonstrate less bone connectivity in the OVX animals compared with the controls. There were relative increases in the free-free and cortex-free type trabecular struts, but a decrease in the node-node type struts. The lengths of the free-free, node-free and nodenode type struts were greater in the OVX animals compared with controls. These increases in trabecular strut length could be attributed to a decrease in connectivity, and thus nodes, between the struts. Decreases in trabecular connectivity have also been reported in cancellous bone from aging humans using both morphological (Arnold et al., 1966; Arnold, 1980) and morphometric approaches (Compston et al., 1987; 1989). The star volume analyses of the cancellous bone of the lumbar vertebral bodies also indicated substantial differences in cancellous bone architecture in the OVX animals compared with controls. Bone marrow star volumes were increased in images obtained from cross and sagittal sections. This indicates a greater trabecular separation in the OVX animals compared with controls. Yoshida et al. (1991) reported that after 9 months following OVX in rats there was substantially de- creased trabecular number, but not trabecular thickness in cancellous bone areas. This was particularly evident in central spongiosa areas of the bone, similar to that observed in this study, This loss of individual trabeculae results in a larger marrow star volume, reported here. The loss of individual structural elements by removal of individual trabeculae and perhaps perforation of others, can result in less continuity and connectivity between the remaining structural elements, as indicated in this study from the nodal analyses. An increase in marrow star volume due to the removal and perforation of individual bone trabeculae occurs in aging human cancellous bone (Vesterby et al., 1989). The differences in the 3-dimensional architecture of metaphyseal cancellous bone in the OVX compared with control rats was evident from the scanning electron micrographs of the proximal humerii. There was virtually no cancellous bone in the central metaphyseal region in the OVX animals, consistent with the reported loss of central metaphyseal bone following OVX in other long bones including the femur (Miller et al., 1991) and tibia (Wronski et al., 1985; Yoshida et al., 1991). The remaining metaphyseal cancellous bone was found in the lateral metaphysis spanning from the mineralized tissue adjacent to the epiphyseal growth plate (the remnant of the primary spongiosa) to the CANCELLOUS BONE I N OVARIECTOMIZED RATS 439 Fig.7. The bone spicules remaining in the lateral metaphysis of the proximal humerii from the OVX animals consist primarily of long and relatively straight longitudinal struts. These longitudinal struts usually span from the mineralized tissues adjacent to the growth plate (GP)to the metaphyseal cortex (0. x 35. Fig. 8 . Detail of the cylindrical longitudinal struts extending from the mineralized primary spongiosa to the cortex in the humeral metaphysis from an OVX animal. No cancellous bone was found in the central metaphyseal regions (*). x 55. metaphyseal cortex. In the control animals much of the cancellous bone consisted of plates and rods whereas in the OVX animals most of the bone consisted of fairly straight rods. The remodeling of the trabecular plates into rods in the OVX animals and no doubt to some extent in the aging controls, is consistent with pattern of cancellous bone loss in the aging and osteoporotic human where plate-type trabeculae are initially converted to rod-type trabeculae by perforation of the trabecular plates (Arnold, 1970; Singh, 1977; Parfitt et al., 1983). Skeletal tissues are adaptable to changes in usage and mechanical loading (Rubin, 1984; Lanyon et al., 1986). For example, in humans and experimental animals an increase in mechanical usage and loading increases bone modeling and remodeling to increase bone mass. While there were substantial changes in the volume and structure of cancellous bone in the OVX rat, the bone probably remains responsive to mechanical stimuli. Endurance training attenuates bone loss in rats following OVX (Donahue et al., 1988) while skeletal underloading accentuates bone loss (Okumura et al., 1988; Bagi et al., 1992). Yoshida et al. (1991) suggested that trabecular loss following OVX occurs in regions where the mechanical stresses are lowest. The loss of trabeculae in the central regions of the cancellous bone following OVX reflects these differences in biomechanical properties. The results from the present study support this hypothesis. The longitudinal orientation of the rods in the OVX animals is consistent with the propagation of skeletal loading forces through the longitudinal axis of the bone from the articular surfaces, through the epiphysis and epiphyseal plate to the metaphyseal and diaphyseal cortex. The thinner transverse rods that connected the larger longitudinal trabeculae were often much thinner than the rods they connected. As much of the normal loading force in bone is along the longitudinal axis, the thinner transverse rods may serve primarily as stabilizers for the weightand force-bearing longitudinal rods. This is consistent with the appearance of rod-like bone spicules that connect trabecular plates and thicker vertical columns in models of aging human bone (Jensen et al., 1990). The loss of ovarian function following ovariectomy removes or reduces the endocrine signals that maintain skeletal mass for reproductive purposes. In rodents there are changes in the maternal skeletal metabolism during pregnancy providing a calcium reservoir for fetal skeletal mineralization and during lactation for milk production (Miller et al., 1986).During the first reproductive cycle in rats, there is a substantial loss of central metaphyseal cancellous bone (Miller et al., 1986), quite similar to the bone loss observed following OVX. These observations provide further evidence that the central metaphyseal bone found in virgin female rats is more metabolically active and S.C. MILLER AND T.J. WRONSKI 440 Fig. 9. The longitudinal struts in the metaphyses of the OVX animals are often interconnected with smaller, cylindrical transverse struts (arrows). x 55. Fig. 10. Detail of the insertion of a longitudinal strut into the mineralized tissue of the primary spongiosa adjacent to the growth plate. The base of the longitudinal strut is spread over the mineralized piate(*). x loo. LITERATURE CITED may not be essential for structural integrity. While a decrease in cancellous bone mass in the OVX or lactat- Aloia, J.F., S.H. Cohn, A. Vaswani, J.K. Yeh, K. Yuen, and K. Ellis 1985 Risk factors for postmenopausal osteoporosis. Am. J . Med., ing rat may influence the mechanical strength of these 78:95-100. tissues, under normal loading conditions the structural J.S. 1980 Trabecular patterns and shapes in aging and oscapacity of the skeletons is not significantly compro- Arnold, teoporosis. In: Bone Histomorphometry. W.S.S. Jee and A.M. mised. There is no increase in spontaneous fractures in Parfitt, eds. Laboratoire Armour-Montagu, Levallois, pp. 297308. this or other studies using OVX rats, or other OVX Arnold, J.S., M.H. Bartley, S.A. Tont, and D.P. Jenkins 1966 Skeletal animal models, such as dogs. in aging and disease. Clin. Orthop. Rel. Res., 49:17-38. In summary, this study examined the differences in Bagi,changes C.M., S.C. Miller, B.M. Bowman, G.L. Blomstrom, and E.P. cancellous bone structure in long-term OVX rats comFrance 1992 Differences in cortical bone in overloaded and underloaded femdrs from ovariectomized rats: comparison of bone pared with sham-OVX controls. As expected, there was morphometry with torsional testing. Bone, 13:35-40, 1992. a marked decrease in cancellous bone mass in the OVX G.H., 0. Dunbar, J.S. Beck, and A. Gibb 1967 Variations in animals. There were, however, substantial differences Bell,strength of vertebrae with age and their relation to osteoporosis. in cancellous bone architecture in the OVX rats. These Calcif. Tissue Res., 1:75-86. included a loss of trabecular connectivity, an increase Coppston, J.E., R.W.E. Mellish, P. Croucher, R. Newcombe, and N.J. Garrahan. 1989 Structural mechanisms of trabecular bone loss in in trabecular separation and a loss of plate-like trabecman. Bone Miner., 6:339-350. ular elements. While there was a considerable decrease Compston, J.E., R.W.E. Mellish, and N.J. Garrahan 1987 Age-related in cancellous bone mass in the OVX animals, remodchanges in iliac crest trabecular microanatomic bone structure in man. Bone, 8:289-292. eling maintained its structural and functional capacity. The changes in skeletal mass and structure in the Donahue, H.J., R.S. Mazzeo, and S.M. Horvath 1988 Endurance training and bone loss in calcium-deficient and ovariectomized rats. OVX rats were likely due to the decrease in gonadal Metabolism, 37:741-744. hormones in the presence of normal mechanical load- Garrahan, N.J., R.W.E. Mellish, and J.E. Compston 1986 A new ing forces. method for the two-dimensional analysis of bone structure in huACKNOWLEDGEMENTS This study was supported by U.S. Public Health Service grants DE-07006 (S.C.M.) and AG-06484 (T.J.W.) from the National Institutes of Health. man iliac crest biopsies. J . Microsc., 142:341-349. Jensen, K.S., L. Mosekilde, and L. Mosekilde 1990 A model of vertebral trabecular bone architecture and its mechanical properties. Bone, 11:417-423. Kalu, D.N. 1991 The ovariectomized rat model of postmenopausal bone loss. Bone Miner., 15175-192. CANCELLOUS BONE IN OVARIECTOMIZED RATS Kleerekoper, M., A.R. Villanueva, J. Stanciu, D. Sudhaker Rao, and A.M. Parfitt 1985 The role of three-dimensional trabecular microstructure in the pathogenesis of vertebral compression fractures. Calcif. Tissue Int., 37594-597. Lanyon, L.E., C.T. Rubin, and G. Baust 1986 Modulation of bone loss during calcium insufficiency by controlled dynamic loading. Calcif. Tissue Int., 38209-216. Miller, M.A., T.H. Omura, and S.C. Miller 1989 Increased cancellous bone remodeling during lactation in beagles. Bone, 10:279-285, 1989. Miller, S.C., B.M. Bowman, M.A. Miller, andC.M. Bagi 1991 Calcium absorption and osseous organ-, tissue-, and envelope-specific changes following ovariectomy in rats. Bone, 12439-446, 1991. Mosekilde, L.,L. Mosekilde, and C.C. Danielsen 1987 Biomechanical competence of vertebral trabecular bone in relation to ash density and age in normal individuals. Bone, 839-85. Okumura, H., T. Yamamuro, R. Kasai, T. Hayashi, K. Tada, and Y. Nishi 1988 Effect of la-hydroxyvitamin D3 on osteoporosis induced by immobilization combined with ovariectomy in rats. Bone, 8351-355. Parfitt, A.M., C.H.E. Mathews, A.R. Villanueva, and M. Kleerekoper 1983 Relationship between surface, volume, and thickness of iliac trabecular bone in aging and in osteoporosis. J. Clin. Invest., 721396-1409. Rubin, C.T. 1984 Skeletal strain and the functional significance of bone architecture. Calcif. Tissue Int. 36511-518. S.aville, P.D. 1969 Changes in skeletal mass and fragility with castration in the rat: a model of osteoporosis. J. Am. Geriatr. SOC., 17:155-164. Singh, I. 1978 The architecture of cancellous bone. J. Anat., 127305310. 441 Stepan, J.J., J. Pospichal, J. Presi, and V. Pacovsky 1987 Bone loss and biochemical indices of bone remodeling in surgically induced postmenopausal women. Bone, 8:279-284. Turner, R.T., G.K. Wakely, K.S. Hannon and N.H. Bell 1988 Tamoxifen inhibits osteoclast-mediated resorption of trabecular bone in ovarian hormone-deficient rats. Endocrinology, 12211461150. Vesterby A,, H.J.G. Gundersen, and F. Melsen 1989 Star volume of marrow space and trabeculae of the first lumbar vertebra: sampling efficiency and biological variation. Bone, 10:7-13. Waynforth, H.B. 1980 Experimental and Surgical Technique in the Rat. Academic Press, London. Wronski, T.J., L.M. Dann, and S.L. Horner 1989a Time course of vertebral osteopenia in ovariectomized rats. Bone, 10:295-301. Wronski, T.J., L.M. Dann, K.S. Scott, and M. Cintron 1989b Longterm effects of ovariectomy and aging on the rat skeleton. Calcif. Tissue. Int., 45:360-366. Wronski, T.J., P.L. Lowry, C.C. Walsh, C.C., and L.A. Ignaszewski 1985 Skeletal alterations in ovariectomized rats. Calcif. Tissue Int., 37324-328. Wronski. T.J.. P.A. Schenck. M. Cintron. and C.C. Walsh 1987 Effect of body weight on osteopenia in ovariectomized rats. Calcif. Tissue Int., 40:155-159. Wronski, T.J., C.C. Walsh, and L.A. Ignaszewski 1986 Histologic evidence for osteopenia and increased bone turnover in ovariectomized rats. Bone, 7119-123. Yoshida, S., T. Yamamuro, H. Okumura, and H. Takahashi 1991 Microstructural changes in osteopenic trabeculae in the rat. Bone 12185-194.