- Email: [email protected]
Effect of trabecular bone contour on ultimate strength of lumbar vertebra after bilateral ovariectomy in rats
Bone Vol. 28, No. 6 June 2001:625– 633
Effect of Trabecular Bone Contour on Ultimate Strength of Lumbar Vertebra After Bilateral Ovariectomy in Rats S. IKEDA,1 H. TSURUKAMI,1 M. ITO,2 A. SAKAI,1 T. SAKATA,1 S. NISHIDA,1 S. TAKEDA,3 A. SHIRAISHI,4 and T. NAKAMURA1 1 Department of Orthopaedic Surgery, University of Occupational and Environmental Health, Kitakyushu, Japan, 2Department of Radiology, Nagasaki University School of Medicine, Nagasaki, Japan, 3Fuji Gotenba Research Laboratory, Chugai Pharmaceutical Co., Shizuoka, Japan, 4 Product Research Laboratory, Chugai Pharmaceutical Co., Tokyo, Japan
post-ovx period in rats. (Bone 28:625– 633; 2001) Elsevier Science Inc. All rights reserved.
To test the hypothesis that the effect of trabecular microarchitecture on bone strength varies with the duration of estrogen loss, we evaluated the relationship between threedimensional (3D) parameters for trabecular microarchitecture and bone minerals with the compressive load of the lumbar vertebra in rats. Female Sprague-Dawley rats (n ⴝ 190) were divided into 19 groups. Ten rats were killed at day 0. Half of the remaining rats underwent bilateral ovariectomy (ovx), and the others were subjected to sham surgery. Ten rats from each group were killed at 3, 7, 11, 14, 28, 42, 56, 70, and 84 days postsurgery. Urinary deoxypyridinoline and serum osteocalcin increased significantly in the ovx group from days 28 and 11, respectively, compared with the sham group. Bone mineral content (BMC) and bone mineral density (BMD) of the fifth lumbar body diminished from days 42 and 84, respectively, compared with the sham group. In ovx rats, trabecular bone volume (BV/TV), measured using 3D images of microcomputed tomography, diminished from day 28 compared with both baseline control and sham. The trabecular bone pattern factor (TBPf) and structure model index (SMI) increased from day 28 in the ovx group compared with both baseline control and sham. Ultimate compression loads diminished at day 28 compared with baseline control and decreased progressively thereafter. Neither of these parameters changed in the sham group during the same period. Within 4 weeks post-ovx, TBPf, SMI, and BV/TV correlated with load (p < 0.01). BMC and BMD correlated with load from 6 weeks post-ovx (p < 0.01). Stepwise regression analysis showed that TBPf was the most significant determinant of load within 4 weeks post-ovx (coefficient of determination [R2] ⴝ 0.669; p < 0.01). SMI correlated with TBPf (R2 ⴝ 0.968; p < 0.01). Moreover, R2 for ultimate load indicated higher values of 0.975 with TBPf and SMI. However, BMC was the most significant determinant of load from 6 weeks post-ovx (R2 ⴝ 0.511; p < 0.01), as it was in the sham group. These data suggest that changes in trabecular bone contour with increased bone turnover are critical for reducing lumbar bone strength during the early
Key Words: Microcomputed tomography (micro-CT); Threedimensional (3D) microarchitecture; Dual-energy X-ray absorptiometry (DXA); Trabecular bone pattern factor (TBPf); Structure model index (SMI); Bone metabolic markers. Introduction Estrogen deficiency in humans increases bone resorption and reduces trabecular bone mass,1 and increased osteoclastic activity results in perforation of the trabecular plates and disconnection of the trabecular continuity in cancellous bone.25 Previous studies have shown that both bone density and trabecular structure independently affect the mechanical strength of vertebral bone in adults and aged humans.5,27,28 In fact, iliac bone trabecular connectivity assessed by histological examination is the major predictor of vertebral fracture in male osteoporosis.20 However, the relationship between various bone indices, such as bone mineral density (BMD) and trabecular bone structure, to risk of bone fracture has not been fully established during the postmenopausal period.4,7 Several groups have investigated the serial changes in trabecular bone turnover, mass, structure, and strength after acute estrogen loss in ovariectomized rats. In lumbar bone, the number of trabecular osteoclasts was shown to increase significantly 3 days after ovariectomy (ovx), followed by a decrease in trabecular bone volume (BV/TV), as measured by histomorphometry, at 15 days post-ovx.29 Although trabecular thickness (Tb.Th) did not change after 3 months,24,29 nor did the ash content of the lumbar body decrease significantly at 4 weeks post-ovx,22 the ultimate compressive load was found to decrease at 3 months post-ovx, compared with sham-operated rats.22,24 These data suggest that deterioration of trabecular structure and bone mass reduction lead to bone fragility, although the contributions of these parameters to mechanical strength remain to be determined. Recently, a three-dimensional (3D) trabecular microarchitecture has been designed from 3D images of microcomputed tomography (micro-CT) in animal models and humans in vitro.8,12,15,23,26 Furthermore, in vivo 3D images of trabecular microarchitecture have been obtained in rats using X-ray tomographic microscopy,16 which identifies disruption of trabecular connectivity from 5 days post-ovx in the proximal tibia.19 Although these data indicate acute changes in 3D trabecular microarchitecture during the early post-ovx period, their contribu-
Address for correspondence and reprints: Satoshi Ikeda, M.D., Department of Orthopaedic Surgery, University of Occupational and Environmental Health, 1-1 Iseigaoka, Yahatanishi-ku, Kitakyushu 807-8555, Japan. E-mail: [email protected] © 2001 by Elsevier Science Inc. All rights reserved.
© 2001 by
625
8756-3282/01/$20.00 PII S8756-3282(01)00462-8
626
S. Ikeda et al. Effect of trabecular contour on bone strength
tions to mechanical strength of the vertebral bone have not yet been analyzed. In the present study, we obtained lumbar vertebrae from rats up to 3 months after ovx or sham surgery. After evaluating the 3D trabecular microarchitecture of the specimen by micro-CT, and bone minerals by dual-energy X-ray absorptiometry (DXA), we performed compression tests on the same specimen. To test the hypothesis that the contribution of trabecular microarchitecture to strength varies with the duration of the post-ovx period, we evaluated the relationship between several bone indices of mass and microarchitecture and compression load values by linear regression analysis. We then used stepwise regression analysis to identify the major determinants of the ultimate load during the early and late post-ovx periods. Materials and Methods Experimental Animals Female Sprague-Dawley (SD) rats (n ⫽ 190, 8 weeks of age) were purchased from Charles River Japan (Hino, Japan), and acclimated for 8 weeks prior to the start of the experiment. The rats had free access to tap water and to a commercial standard rat chow containing 1.18% calcium and 2.0 IU/g of vitamin D3 (CE-2; Japan Clea, Tokyo, Japan). Body weight was measured weekly. All rats were housed in metal cages (three or four rats per cage) in an air-conditioned environment (room temperature 23 ⫾ 2°C, humidity 55 ⫾ 10%) illuminated from 7:00 to 19:00. The experimental protocol was approved by the ethics review committee for animal experimentation of the University of Occupational and Environmental Health. Experimental Protocol At 4 months of age, rats were randomly assigned to 19 groups. Ten animals were killed at day 0 (baseline control). Half of the remaining rats underwent bilateral ovariectomy (ovx), whereas the others underwent sham surgery (sham). Ten rats from each group were killed at 3, 7, 11, 14, 28, 42, 56, 70, and 84 days postsurgery. Urine samples were collected from each rat over a period of 24 h before killing and stored at ⫺80°C until use. Cardiac blood samples were obtained at time of killing, immediately centrifuged, and serum samples stored at ⫺80°C for later analysis. After killing by exsanguination, the fifth lumbar vertebra (L-5) was dissected carefully, removed en bloc, and immediately stored at ⫺80°C until use. The effect of ovx was verified by involution atrophy of the uterus. Bone Metabolic Markers Serum osteocalcin concentrations were determined using an enzyme-linked immunosorbent assay (ELISA) kit employing rat osteocalcin antibody (Rat Osteocalcin RIA Biomedical Technologies, Inc., Stoughton, MA). Urinary deoxypyridinoline (DPD) cross-link excretion levels were determined using an ELISA kit (Pyrilinks-D Metra Biosystems, Inc., Mountain View, CA). The results were expressed as DPD/Cr ratios (nmol/L 䡠 mmol/L Cr).
Bone Vol. 28, No. 6 June 2001:625– 633
shifted automatically in the axial direction. The filtered 40 kVp X-ray spectrum peaks at 25 keV, allowing excellent bone vs. marrow contrast.15 The rotation and axial shift are manipulated separately with two stepping motors. A standard convolutionbackprojection procedure with a Shepp and Logan filter is used to reconstruct the CT images in 1024 ⫻ 1024 pixel matrices. The spatial resolution is typically defined by a 10% contrast level in the modulation transfer function (MTF), resulting in a in-plane spatial resolution of 28 m.15,26 The 3D trabecular microarchitecture was obtained for 200 reconstructed slices with an 8 m increment. The volume of interest (VOI) was determined for the entire region of the secondary spongiosa, excluding the endosteal surface and the region of the vertebral vessels. Calibration of the approximately 28 m resolution micro-CT for in vitro trabecular microarchitecture measurements has been reported previously.15 On 3D analysis, the tissue volume (TV; mm3) and trabecular bone volume (BV; mm3) were measured directly, and the fractional trabecular bone volume (BV/TV; %) was calculated. Trabecular thickness (Tb.Th;m), trabecular separation (Tb.Sp; m), and trabecular number (Tb.N; 1/mm) were measured directly on 3D images using the method described by Hildebrand et al.12,13 Thus, the parameters Tb.Th, Tb.Sp, and Tb.N are model-independent indices, not biased by eventual deviations of the actual structure. For nonmetric indices, the values of 3D structural indices, such as structure model index (SMI) and geometrical degree of anisotropy (DA), were computed using software included with the micro-CT machine. The parameter SMI represented the plate-rod characteristic of trabecular structure that was calculated by a differential analysis of a triangulated surface of a trabecular structure,12,14 and DA represents trabecular anisotropy defined as the ratio between the maximal and minimal radius of the mean intercept length (MIL),9,12 respectively. The trabecular bone pattern factor (TBPf; 1/mm), representing the ratio of concave:convex surfaces in 2D sections of trabeculae,11 was measured for each slice and the mean value was determined for each specimen. The coefficient variation (CV) values of BV/TV, Tb.Th, Tb.Sp, Tb.N, SMI, DA, and TBPf were 3.47%, 2.69%, 1.02%, 1.12%, 12.6%, 0.34%, and 9.36%, respectively. Bone Size and Bone Mineral Measurement The height of the L-5 vertebral body was measured with a micrometer. The L-5 vertebral body specimen was fixed with a clamp at the base of the transverse process in the holder of a diamond bandsaw (Exakt, Norderstedt, Germany). By removing both the cranial and caudal endplates planoparallel surfaces were obtained. From each vertebral body, a central cylinder specimen was obtained with planoparallel ends with a height of approximately 4.0 mm. The volume of the specimen was measured by the Archimedes method, using an electric volumetric apparatus (MK-550; Muromachi Machinery, Tokyo). Bone mineral content and density were measured in each specimen by DXA (DCS600; Aloka, Tokyo) using the small-animal scan mode with irradiation applied anteroposteriorly to the specimen. The values of bone mineral content (BMC; mg) and bone mineral density (BMD; mg/cm2) were obtained for each specimen. The CV values of BMC and BMD were 3.75% and 1.19%, respectively.
Structural Analysis Using Micro-CT After the removal of adhering tissues, the posterior elements, and transverse processes of the L-5 vertebral body, each specimen was scanned by micro-CT (CT20; Scanco Medical, Zurich, Switzerland). The CT20 is equipped with a microfocus X-ray tube with a focal spot of 10 m, producing a fan beam detected by a charge-couple device array with a turntable that can be
Mechanical Tests L-5 vertebral cylinder samples were placed centrally on the smooth surface of a steel disk (10 cm diameter) attached to the materials-testing machine (Tensilon UTA-1T; Orientec, Tokyo). A compression force was applied in the craniocaudal direction by using a steel disk (1.8 cm) at a nominal deformation rate of 2
Bone Vol. 28, No. 6 June 2001:625– 633
S. Ikeda et al. Effect of trabecular contour on bone strength
627
Figure 2. Serial changes in bone mineral indices of rat L-5 vertebral body. (A) BMD values. (B) BMC values. Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. *P ⬍ 0.05 vs. sham group at the same timepoint (Tukey–Kramer post hoc tests after two-way ANOVA).
Figure 1. Serial changes in the levels of bone metabolic markers. (A) Urinary deoxypyridinoline levels. (B) Serum osteocalcin levels. Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. *P ⬍ 0.05, **P ⬍ 0.01 vs. sham group at the same timepoint; ##P ⬍ 0.01 vs. baseline control (Tukey–Kramer post hoc tests after two-way ANOVA). 22
mm/min. Crosshead displacement was considered as specimen deformation. A load-deformation curve was displayed with a monitoring recorder linked to the tester in each specimen. The ultimate load (N) was measured directly from the load-deformation curve. Statistical Analysis All data were expressed as mean ⫾ SEM. Group data were first compared by two-factor factorial analysis of variance (ANOVA) for surgery and time. If analysis confirmed a significant difference between the ovx and sham groups, then the difference between values at each timepoint was assessed by Tukey– Kramer posthoc test. For data of bone metabolic markers, BMD, BMC, trabecular microarchitectural indices, and ultimate load, differences from baseline were also assessed. The relationships between ultimate load and each microarchitectural index, BMD,
BMC, and bone metabolic markers were analyzed by simple regression. Furthermore, factors that significantly influenced ultimate load were determined by stepwise regression among the factors of the microarchitectural indices, BMD, BMC, and bone metabolic markers. p ⬍ 0.05 was considered statistically significant. Results Mortality, Body Weight, and Bone Sizes Four rats died during surgery and three rats died postoperatively for unknown reasons. However, the mortality rate was similar among the groups. Body weight significantly increased in ovx rats at day 70 after surgery, compared with the sham group (data not shown). The height and volume of the L-5 vertebral body of ovx rats were not significantly different from those of shamoperated rats (data not shown). Urinary Deoxypyridinoline (DPD) and Serum Osteocalcin Levels Urinary DPD levels in the ovx groups increased significantly from 28 days and thereafter compared with the sham groups (Figure 1A). Serum osteocalcin levels were significantly higher at days 11, 28, and 42 post-ovx compared with the sham groups
628
S. Ikeda et al. Effect of trabecular contour on bone strength
Bone Vol. 28, No. 6 June 2001:625– 633
Figure 3. Three-dimensional trabecular microarchitectural images of rat L-5 vertebral body using micro-CT. (A) Baseline control group. (B) Sham group at 4 weeks postsurgery. (C) Sham group at 12 weeks postsurgery. (D) Ovx group at 4 weeks postovariectomy. (E) Ovx group at 12 weeks postovariectomy. Top: cranial side; bottom: inferior position.
and significantly higher at days 28 and 42 compared with baseline (Figure 1B), although values decreased gradually thereafter. BMD and BMC of L-5 Vertebral Body In ovx groups, L-5 BMD values decreased from day 42 and were significantly lower at day 84 compared with the sham group (Figure 2A). BMC values also decreased significantly at days 42 and 84 compared with the sham group (Figure 2B). However, these values were not significantly different from baseline. In the sham groups, BMD and BMC values were not significantly different from the respective baseline values. Micro-CT Images and Indices of 3D Trabecular Microarchitecture of L-5 Images of the L-5 vertebra in the sham groups did not appear to change throughout the experimental period (Figure 3A–C). However, the rod-like trabecular structure became dominant in the ovx groups at day 84 (Figure 3D, E). The BV/TV in the ovx groups decreased after surgery and was significantly lower from day 28 and thereafter compared with sham and baseline control groups (Figure 4A). Tb.Th in the ovx groups decreased significantly from day 42 compared with sham rats (Figure 4B). The values at days 56 and 70 were significantly lower than baseline values. Tb.Sp increased significantly from day 28 compared with the sham group and from day 42 compared with baseline (Figure 4C). Tb.N diminished significantly from day 28 compared with the sham group and from day 42 compared with baseline (Figure 4D). The SMI in ovx rats was significantly higher at day 28 and thereafter in comparison with sham and baseline controls (Figure 5A). DA was higher in ovx rats than in sham group at day 84, but not significantly different from baseline (Figure 5B). TBPf
increased significantly at day 28 days and thereafter in the ovx groups compared with sham and baseline controls (Figure 5C). Ultimate Compressive Load of L-5 The ultimate compressive load values of L-5 in the ovx groups gradually decreased postsurgery and became significantly lower from day 28 as compared with baseline (Figure 6). Values became smaller than those of sham-operated rats from day 42. Relationship Between Ultimate Load and Parameters of Bone and Turnover The results for ultimate load showed that the value decreased acutely by approximately 23% during 28 days post-ovx (Figure 6). On the other hand, the values essentially remained at a low level, showing only a 2% decrease during 42– 84 days post-ovx. The postsurgery period was divided into early (0 –28 days postsurgery) and late (42– 84 days postsurgery) periods for post hoc statistical analysis. Early postoperative period. In the sham group, BMD correlated significantly with ultimate load (Table 1). BMC and Tb.Th also correlated significantly with load. In ovx rats, TBPf, SMI, BV/TV, Tb.N, and Tb.Sp correlated significantly with load. Late postoperative period. In the sham group, BMD and BMC correlated significantly with the ultimate load. However, there was a weak relationship between Tb.N and ultimate load. In ovx rats, BMD and BMC correlated significantly with load. TBPf, Tb.N, BV/TV, Tb.Th, and SMI also correlated significantly with the load. Bone metabolic markers and DA did not correlate with the ultimate load during both early and late postoperative periods.
Bone Vol. 28, No. 6 June 2001:625– 633
S. Ikeda et al. Effect of trabecular contour on bone strength
629
Figure 4. Serial changes in three-dimensional microarchitectural indices of rat L-5 vertebral body analyzed by micro-CT. (A) Trabecular bone volume (BV/TV; %). (B) Trabecular thickness (Tb.Th; m). (C) Trabecular separation (Tb.Sp; m). (D) Trabecular number (Tb.N; 1/mm). Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. *P ⬍ 0.05, **P ⬍ 0.01 vs. sham group at the same timepoint. *P ⬍ 0.05, **P ⬍ 0.01 vs. baseline control (Tukey–Kramer post hoc tests after two-way ANOVA).
Factors Determining Ultimate Load
Discussion
In the early postoperative period, BMD was the most significant factor that influenced the ultimate load in sham-operated rats, as assessed by stepwise regression analysis (Table 2). In ovx groups, however, TBPf was the most significant factor that influenced the compressive load, followed by SMI. In the late period, BMC was the most significant factor that influenced load in both sham and ovx groups.
In the present study we demonstrated that the contribution of parameters of bone mass and trabecular microarchitecture to the ultimate load of the lumbar vertebra differed between the early and late periods after ovx in rats. Although BMD and BMC of the lumbar vertebra did not decrease significantly after ovx, the ultimate load values decreased in 4 weeks. Furthermore, BV/TV and Tb.N diminished and Tb.Sp increased during the same period. Nonmetric parameters, such as SMI and TBPf, also increased significantly during the same period. Although various parameters of bone mass and trabecular microarchitecture were related significantly to the compressive load in ovx groups, TBPf was the major determinant of load within 4 weeks after ovx as analyzed by stepwise regression. However, BMC had the strongest influence on load from 6 weeks post-ovx. In the sham groups, the effect of trabecular microarchitecture on the compressive load was limited, which depended mainly on BMD and BMC. TBPf and SMI did not correlate with compressive load in the sham group. Our study showed increases in two bone metabolic markers
Relationship Between Other Parameters and Determinants of Ultimate Load In the sham group, BMC and Tb.Th correlated significantly with BMD in the early postsurgery period (Table 3). In the late postsurgery period, BMD and Tb.Th correlated significantly with BMC. In the ovx groups, SMI, BV/TV, Tb.N, and Tb.Sp correlated significantly with TBPf in the early postsurgery period. Tb.Th, BMD, and BMC also correlated with TBPf in the same period. In the late postsurgery period, BMD, Tb.N, TBPf, BV/ TV, and Tb.Th correlated significantly with BMC.
630
S. Ikeda et al. Effect of trabecular contour on bone strength
Bone Vol. 28, No. 6 June 2001:625– 633
Figure 6. Serial changes in ultimate load of rat L-5 vertebral body. Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. *P ⬍ 0.05, **P ⬍ 0.01 vs. sham group at the same timepoint. **P ⬍ 0.01 vs. baseline control (Tukey–Kramer post hoc tests after two-way ANOVA).
Figure 5. Serial changes in direct nonmetric microarchitectural indices of rat L-5 vertebral body analyzed by micro-CT. (A) Structure model index (SMI) values. (B) Geometrical degree of anisotropy (DA). (C) Trabecular bone pattern factor (TBPf; 1/mm). Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. **P ⬍ 0.01 vs. sham group at the same timepoint; ##P ⬍ 0.01 vs. baseline control (Tukey– Kramer post hoc tests after two-way ANOVA).
after ovx. These findings are consistent with previous results reported in mature SD rats.19,29 In our study, serum osteocalcin concentrations increased in 7 days and reached a peak level at 23
days post-ovx, as observed by Yamaura et al.29 The difference in urinary DPD excretion between sham and ovx rats became significant within 4 weeks after surgery in this study, but the values in the ovx groups apparently began to increase from 11 days. Thus, the urinary DPD data are consistent with results reported by Lane et al.19; that is, an increase occurring from day 13 post-ovx. However, there was the discrepancy that the bone metabolic markers did not correlate TBPf and ultimate load in a significant manner. Egger et al.6 observed that the increases in excretion of pyridinium cross-links in rat closely reflects bone resorption in chronic but not acute bone metabolism. Lane et al.19 also observed that 3D trabecular disruption occurs early before the increase in urinary DPD after ovx in rats. Yamaura et al.29 reported that trabecular osteoclast number increase 2–3 days after ovx, but bone marker did not increase as rapidly. These data and the results of the present study both suggest that pyridinium cross-links are not sensitive enough to reflect acute increase in bone resorption in rats. Thus, we suggest that it possible that either TBPf, indicating the trabecular contour, or ultimate load did not correlate significantly with these metabolic markers. The term TBPf was first introduced by Hahn et al.,11 based on the hypothesis that abundant concave surfaces in 2D sections represent a well-connected spongy lattice, whereas abundant convex surfaces indicate a poorly connected trabecular lattice. Thus, TBPf represents trabecular connectivity, and a high value reflects the disruption of trabecular continuity in 2D sections.2,3,10,17 A shift from a plate- to a rod-like structure in trabecular bone post-ovx, as observed by 3D images in this study, is consistent with an increase in TBPf values. Our results show that DA increased significantly only at 3 months in the ovx groups as compared with the sham group. Thus, the geometrical degree of anisotropy of trabecular bone did not seem to be affected significantly by acute estrogen loss. Results of both simple and stepwise regression analyses indicated that the contour of the trabecular bone surface is important for the loss of lumbar bone strength post-ovx in rats. Specifically, changes in the surface contour probably contributed to the significant loss of compressive strength at 4 weeks postovx, when bone turnover was increased. Stepwise regression analysis showed that TBPf was the determinant factor for ultimate load, and the value was very closely related to SMI with a coefficient of determination (R2) value of 0.968. Moreover, R2 for ultimate load indicated a higher value of 0.975 with TBPf and
Bone Vol. 28, No. 6 June 2001:625– 633
S. Ikeda et al. Effect of trabecular contour on bone strength
631
Table 1. Coefficient of determination (R2) values for ultimate load between microarchitectural indices, bone mass, and bone metabolic markers analyzed by simple regression Sham
Microarchitectural indices BV/TV Tb.Th Tb.Sp Tb.N SMI DA TBPf Bone mass BMD BMC Bone metabolic markers Urinary DPD Serum osteocalcin
OVX
Early period (n ⫽ 48)
Late period (n ⫽ 38)
Early period (n ⫽ 48)
Late period (n ⫽ 39)
0.046 0.099* 0.018 0.080 0.046 0.001 0.056
0.112 0.100 0.082 0.154* 0.074 0.090 0.107
0.473** 0.176** 0.421** 0.440** 0.483** 0.001 0.488**
0.200** 0.197** 0.128* 0.223** 0.169* 0.060 0.305**
0.439** 0.101*
0.493** 0.500**
0.264** 0.155**
0.434** 0.503**
0.021 0.013
0.081 0.047
0.001 0.041
0.024 0.001
Early period was between 0 to 28 days, and late period was between 42 to 84 days postsurgery. *P ⬍ 0.05, **P ⬍ 0.01 (simple regression).
SMI. Thus, it is reasonable to presume that the feature of trabecular bone structure commonly assessed by TBPf and SMI, the surface contour, plays a critical role in the loss of bone strength. In the sham group, however, these nonmetric parameters did not change during the same period. In fact, although compressive load values did not correlate with these nonmetric parameters of trabecular microarchitecture, they did correlate with the parameters of bone mass, such as BMD and BMC. In the period from 6 weeks post-ovx, metric parameters, such as BV/TV and Tb.N, gradually decreased and Tb.Sp increased, but nonmetric parameters of SMI and TBPf remained stable as did the ultimate compressive load values. On the other hand, the parameters of trabecular microarchitecture values were so varied that we could not find any statistically significant changes in energy absorption, such as in one of the biomechanical parameters in this investigation (data not shown). We now believe that the changes in energy absorption may be affected not only by the increased turnover after ovx but also in the basal 3D microarchitecture before ovx. Taken together, we postulate that the compressive strength of the lumbar vertebra is determined mainly by the contour of trabecular bone when bone loss is accelerated with an increase in bone turnover. However, when trabecular bone loss is alleviated without marked changes in the contour of trabecular surface, the parameters of bone mass, such as BMD and BMC, are the major determinants of lumbar bone strength in rats. Compromised trabecular microarchitecture is an important and independent pathogenic factor in vertebral fracture in osteoporotic postmenopausal women and osteoporosic men.4 Using
necropsy sections of human vertebral bodies from subjects aged 17–90 years, Grote et al.10 demonstrated the presence of an age-dependent decrease in trabecular bone mass, primarily as a consequence of transformation from plates to rods and the loss of whole trabecular bone volume. Loss of trabecular connectivity as determined by 2D and 3D analyses also indicated changes in bone surface.19,21 Thus, it is possible that changes in the contour of trabecular bone surface by increased turnover affect the strength of the human vertebral body as well, consequently increasing the risk of spinal fracture in postmenopausal women with osteoporosis. We have previously demonstrated that the 28 m resolution micro-CT was suitable for the evaluation of 3D parameters of trabecular bone mass and structure using human iliac bone biopsy specimens.15 Laib and Ru¨egsegger18 also demonstrated that a 3D peripheral quantitative computed tomography machine with a similar resolution value was suitable for evaluation of trabecular bone structure in vivo. Thus, it is possible to evaluate the surface contour of trabecular bone, although it is often difficult to use this technique to visualize irregularities on the surface of individual trabecula and remodeling pits that may represent structurally weak sites when the bone is loaded. Thus, based on the results of the present study alone, we cannot estimate the role of specific changes in trabecular bone surfaces by local remodeling activity or accumulation of trabecular discontinuity. To further estimate the effect of trabecular bone contour on strength, visualization of bone surfaces with a higher level of spatial resolution is necessary. In conclusion, we have demonstrated that TBPf was the most
Table 2. Determinants of ultimate compressive load analyzed by stepwise regression Early period
Late period
Group
Index
R2
P value
Index
R2
P value
Sham
BMD BMD⫹BMC TBPf TBPf⫹SMI
0.258 0.331 0.669 0.975
⬍0.05 ⬍0.01 ⬍0.01 ⬍0.01
BMC BMC⫹BMD BMC BMC⫹BMD
0.447 0.739 0.511 0.772
⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01
OVX
Early period was between 0 to 28 days, and late period was between 42 to 84 days postsurgery. Number of animals was 48 in the sham and OVX groups during the early period. Number of animals was 38 in the sham group and 39 in the OVX group during the late period.
632
S. Ikeda et al. Effect of trabecular contour on bone strength
Bone Vol. 28, No. 6 June 2001:625– 633
Table 3. Coefficient of determination (R2) values for each of the determinant factors between microarchitectural indices, bone mass, and bone metabolic markers analyzed by simple regression Sham
Microarchitectural indices BV/TV Tb.Th Tb.Sp Tb.N SMI DA TBPf Bone mass BMD BMC Bone metabolic markers Urinary DPD Serum osteocalcin
OVX
Early period (n ⫽ 48)
Late period (n ⫽ 38)
Early period (n ⫽ 48)
Late period (n ⫽ 39)
R2 for BMD
R2 for BMC
R2 for TBPf
R2 for BMC
0.044 0.163** 0.003 0.040 0.037 0.002 0.048
0.037 0.118* 0.018 0.053 0.023 0.056 0.053
0.958** 0.470** 0.753** 0.765** 0.968** 0.003 —
0.125* 0.122* 0.055 0.167* 0.060 0.003 0.163*
— 0.320**
0.750** —
0.273** 0.086*
0.780** —
0.004 0.032
0.008 0.060
0.001 0.053
0.003 0.026
Early period was between 0 to 28 days, and late period was between 42 to 84 days postsurgery. *P ⬍ 0.05, **P ⬍ 0.01 (simple regression)
significant determinant of reduced compressive strength of lumbar bone in rats during the early post-ovx period and that SMI was closely related to TBPf. However, in the late post-ovx period, BMC was the major determinant of bone strength. These results imply that changes in trabecular bone contour with increased bone turnover are critical in reducing lumbar bone strength in the early period after ovx in rats.
10.
11.
12.
Acknowledgments: The authors thank Dr. Nobukazu Okimoto, Dr. Yuichi Okazaki, Dr. Soshi Uchida, Dr. Shinya Tanaka, Dr. Hiroshi Wadayama, Junko Nomura, and Erika Kobayashi. This work was supported in part by grants-in-aid from the Research Society for Metabolic Bone Diseases and the Ministry of Health and Welfare of Japan.
13.
14.
References 1. Bouillon, R., Burkhardt, P., Christiansen, C., Fleisch, H. A., Fujita, T., Gennari, C., Martin, T. J., Mazzuoli, G., Melton, L. J., Ringe, J. D., et al. Consensus development conference: Prophylaxis and treatment of osteoporosis. Osteopor Int 1:114 –117; 1991. 2. Chappard, D., Legrand, E., Basle, M. F., Fromont, P., Racineux, J. L., Rebel, A., and Audran, M. Altered trabecular architecture induced by corticosteroids: A bone histomorphometric study. J Bone Miner Res 11:676 – 685; 1996. 3. Croucher, P. I., Garrahan, N. J., and Compston, J. E. Assessment of cancellous bone structure: Comparison of strut analysis, trabecular bone pattern factor, and marrow space star volume. J Bone Miner Res 11:955–961; 1996. 4. Dempster, D. W. The contribution of trabecular architecture to cancellous bone quality. J Bone Miner Res 15:20 –23; 2000. 5. Dempster, D. W., Ferguson-Pell, M. W., Mellish, R. W., Cochran, G. V., Xie, F., Fey, C., Horbert, W., Parisien, M., and Lindsay, R. Relationships between bone structure in the iliac crest and bone structure and strength in the lumbar spine. Osteopor Int 3:90 –96; 1993. 6. Egger, C. D., Muhlbauer, R. C., Felix, R., Delmas, P. D., Marks, S. C., and Fleisch, H. Evaluation of urinary pyridinium crosslink excretion as a marker of bone resorption in the rat. J Bone Miner Res 9:1211–1219; 1994. 7. Faulkner, K. Bone matters: Are density increases necessary to reduce fracture risk? J Bone Miner Res 15:183–187; 2000. 8. Feldkamp, L. A., Goldstein, S. A., Parfitt, A. M., Jesion, G., and Kleerekoper, M. The direct examination of three-dimensional bone architecture in vitro by computed tomography. J Bone Miner Res 4:3–11; 1989. 9. Goulet, R. W., Goldstein, S. A., Ciarelli, M. J., Kuhn, J. L., Brown, M. B., and
15.
16. 17.
18.
19.
20.
21.
22.
Feldkamp, L. A. The relationship between the structural and orthogonal compressive properties of trabecular bone. J Biomech 27:375–389; 1994. Grote, H. J., Amling, M., Vogel, M., Hahn, M., Posl, M., and Delling, G. Intervertebral variation in trabecular microarchitecture throughout the normal spine in relation to age. Bone 16:301–308; 1995. Hahn, M., Vogel, M., Pompesius-Kempa, M., and Delling, G. Trabecular bone pattern factor—a new parameter for simple quantification of bone microarchitecture. Bone 13:327–330; 1992. Hildebrand, T., Laib, A., Mu¨ller, R., Dequeker, J., and Ru¨gsegger, P. Direct three-dimensional morphometric analysis of human cancellous bone: Microstructural data from spine, femur, iliac crest, and calcaneus. J Bone Miner Res 14:1167–1174; 1999. Hildebrand, T. and Ru¨gsegger, P. A new method for the model-independent assessment of thickness in three-dimensional images. J Microsc 185:67–75; 1997. Hildebrand, T. and Ru¨gsegger, P. Quantification of bone microarchitecture with the structure model index. Compreh Meth Biomech Biomed Eng 1:15–23; 1997. Ito, M., Nakamura, T., Matsumoto, T., Tsurusaki, K., and Hayashi, K. Analysis of trabecular microarchitecture of human iliac bone using microcomputed tomography in patients with hip arthrosis with or without vertebral fracture. Bone 23:163–169; 1998. Kinney, J. H., Lane, N. E., and Haput, D. In vivo, three-dimensional microscopy of trabecular bone. J Bone Miner Res 10:264 –270; 1995. Kippo, K., Hannuniemi, R., Isaksson, P., Lauren, L., Osterman, T., Peng, Z., Tuukkanen, J., Kuurtamo, P., Vaananen, H. K., and Sellman, R. Clodronate prevents osteopenia and loss of trabecular connectivity in estrogen-deficient rats. J Bone Miner Res 13:287–296; 1998. Laib, A. and Ru¨gsegger, P. Calibration of trabecular bone structure measurements of in vivo three-dimensional peripheral quantitative computed tomography with 28-micron-resolution microcomputed tomography. Bone 24:35–39; 1999. Lane, N. E., Thompson, J. M., Haupt, D., Kimmel, D. B., Modin, G., and Kinney, J. H. Acute changes in trabecular bone connectivity and osteoclast activity in the ovariectomized rat in vivo. J Bone Miner Res 13:229 –236; 1998. Legrand, E., Chappard, D., Pascaretti, C., Duquenne, M., Krebs, S., Rohmer, V., Basle, M.-F., and Audran, M. Trabecular bone microarchitecture, bone mineral density, and vertebral fractures in men with osteoporosis. J Bone Miner Res 15:13–19; 2000. Mellish, R. W. E., Ferguson-Pell, M. W., Cochran, G. V. B., Lindsay, R., and Dempster, D. W. A new manual method for assessing two dimensional cancellous bone structure: Comparison between iliac crest and lumbar vertebra. J Bone Miner Res 6:689 – 696; 1991. Mosekilde, Li., Danielsen, C. C., and Knudsen, U. B. The effect of aging and
Bone Vol. 28, No. 6 June 2001:625– 633
23.
24.
25.
26.
ovariectomy on the vertebral bone mass and biomechanical properties of mature rats. Bone 14:1– 6; 1993. Mu¨ller, R., Van Campenhout, H., Van Damme, B., Van Der Perre, G., Dequeker, J., Hildebrand, T., and Ru¨gsegger, P. Morphometric analysis of human bone biopsies: A quantitative structural comparison of histological sections and micro-computed tomography. Bone 23:59 – 66; 1998. Ohnishi, H., Nakamura, T., Narusawa, K., Murakami, H., Abe, M., Barbier, A., and Suzuki, K. Bisphosphonate tiludronate increases bone strength by improving mass and structure in established osteopenia after ovariectomy in rats. Bone 21:335–343; 1997. Parfitt, A. M., Mathews, C. H. E., Villanueva, A. R., Kleerekoper, M., Frame, B., and Rao, D. S. Relationships between surface, volume and thickness of iliac trabecular bone in aging and in osteoporosis. J Clin Invest 72:1396 –1409; 1983. Ru¨egsegger, P., Koller, B., and Mu¨ller, R. A microtomographic system for the nondestructive evaluation of bone architecture. Calcif Tissue Int 58:24 –29; 1996.
S. Ikeda et al. Effect of trabecular contour on bone strength
633
27. Thomsen, J. S., Ebbessen, E. N., and Mosekilde, Li. Relationships between static histomorphometry and bone strength measurements in human iliac crest bone biopsies. Bone 22:153–163; 1998. 28. Vesterby, A., Mosekilde, Li., Gundersen, H. J., Melsen, F., Holme, K., and Sorensen, S. Biologically meaningful determinants of the in vitro strength of lumbar vertebrae. Bone 12:219 –224; 1991. 29. Yamaura, M., Nakamura, T., Tsurukami, H., Hijioka, A., Narusawa, K., Ohnishi, H., Ohta, T., and Hosoda, K. Local bone turnover in the metaphysis of the proximal tibia and the lumbar vertebra during the early periods after ovariectomy in rats. Calcif Tissue Int 58:52–59; 1996.
Date Received: August 25, 2000 Date Revised: January 22, 2001 Date Accepted: January 29, 2001
Effect of Trabecular Bone Contour on Ultimate Strength of Lumbar Vertebra After Bilateral Ovariectomy in Rats S. IKEDA,1 H. TSURUKAMI,1 M. ITO,2 A. SAKAI,1 T. SAKATA,1 S. NISHIDA,1 S. TAKEDA,3 A. SHIRAISHI,4 and T. NAKAMURA1 1 Department of Orthopaedic Surgery, University of Occupational and Environmental Health, Kitakyushu, Japan, 2Department of Radiology, Nagasaki University School of Medicine, Nagasaki, Japan, 3Fuji Gotenba Research Laboratory, Chugai Pharmaceutical Co., Shizuoka, Japan, 4 Product Research Laboratory, Chugai Pharmaceutical Co., Tokyo, Japan
post-ovx period in rats. (Bone 28:625– 633; 2001) Elsevier Science Inc. All rights reserved.
To test the hypothesis that the effect of trabecular microarchitecture on bone strength varies with the duration of estrogen loss, we evaluated the relationship between threedimensional (3D) parameters for trabecular microarchitecture and bone minerals with the compressive load of the lumbar vertebra in rats. Female Sprague-Dawley rats (n ⴝ 190) were divided into 19 groups. Ten rats were killed at day 0. Half of the remaining rats underwent bilateral ovariectomy (ovx), and the others were subjected to sham surgery. Ten rats from each group were killed at 3, 7, 11, 14, 28, 42, 56, 70, and 84 days postsurgery. Urinary deoxypyridinoline and serum osteocalcin increased significantly in the ovx group from days 28 and 11, respectively, compared with the sham group. Bone mineral content (BMC) and bone mineral density (BMD) of the fifth lumbar body diminished from days 42 and 84, respectively, compared with the sham group. In ovx rats, trabecular bone volume (BV/TV), measured using 3D images of microcomputed tomography, diminished from day 28 compared with both baseline control and sham. The trabecular bone pattern factor (TBPf) and structure model index (SMI) increased from day 28 in the ovx group compared with both baseline control and sham. Ultimate compression loads diminished at day 28 compared with baseline control and decreased progressively thereafter. Neither of these parameters changed in the sham group during the same period. Within 4 weeks post-ovx, TBPf, SMI, and BV/TV correlated with load (p < 0.01). BMC and BMD correlated with load from 6 weeks post-ovx (p < 0.01). Stepwise regression analysis showed that TBPf was the most significant determinant of load within 4 weeks post-ovx (coefficient of determination [R2] ⴝ 0.669; p < 0.01). SMI correlated with TBPf (R2 ⴝ 0.968; p < 0.01). Moreover, R2 for ultimate load indicated higher values of 0.975 with TBPf and SMI. However, BMC was the most significant determinant of load from 6 weeks post-ovx (R2 ⴝ 0.511; p < 0.01), as it was in the sham group. These data suggest that changes in trabecular bone contour with increased bone turnover are critical for reducing lumbar bone strength during the early
Key Words: Microcomputed tomography (micro-CT); Threedimensional (3D) microarchitecture; Dual-energy X-ray absorptiometry (DXA); Trabecular bone pattern factor (TBPf); Structure model index (SMI); Bone metabolic markers. Introduction Estrogen deficiency in humans increases bone resorption and reduces trabecular bone mass,1 and increased osteoclastic activity results in perforation of the trabecular plates and disconnection of the trabecular continuity in cancellous bone.25 Previous studies have shown that both bone density and trabecular structure independently affect the mechanical strength of vertebral bone in adults and aged humans.5,27,28 In fact, iliac bone trabecular connectivity assessed by histological examination is the major predictor of vertebral fracture in male osteoporosis.20 However, the relationship between various bone indices, such as bone mineral density (BMD) and trabecular bone structure, to risk of bone fracture has not been fully established during the postmenopausal period.4,7 Several groups have investigated the serial changes in trabecular bone turnover, mass, structure, and strength after acute estrogen loss in ovariectomized rats. In lumbar bone, the number of trabecular osteoclasts was shown to increase significantly 3 days after ovariectomy (ovx), followed by a decrease in trabecular bone volume (BV/TV), as measured by histomorphometry, at 15 days post-ovx.29 Although trabecular thickness (Tb.Th) did not change after 3 months,24,29 nor did the ash content of the lumbar body decrease significantly at 4 weeks post-ovx,22 the ultimate compressive load was found to decrease at 3 months post-ovx, compared with sham-operated rats.22,24 These data suggest that deterioration of trabecular structure and bone mass reduction lead to bone fragility, although the contributions of these parameters to mechanical strength remain to be determined. Recently, a three-dimensional (3D) trabecular microarchitecture has been designed from 3D images of microcomputed tomography (micro-CT) in animal models and humans in vitro.8,12,15,23,26 Furthermore, in vivo 3D images of trabecular microarchitecture have been obtained in rats using X-ray tomographic microscopy,16 which identifies disruption of trabecular connectivity from 5 days post-ovx in the proximal tibia.19 Although these data indicate acute changes in 3D trabecular microarchitecture during the early post-ovx period, their contribu-
Address for correspondence and reprints: Satoshi Ikeda, M.D., Department of Orthopaedic Surgery, University of Occupational and Environmental Health, 1-1 Iseigaoka, Yahatanishi-ku, Kitakyushu 807-8555, Japan. E-mail: [email protected] © 2001 by Elsevier Science Inc. All rights reserved.
© 2001 by
625
8756-3282/01/$20.00 PII S8756-3282(01)00462-8
626
S. Ikeda et al. Effect of trabecular contour on bone strength
tions to mechanical strength of the vertebral bone have not yet been analyzed. In the present study, we obtained lumbar vertebrae from rats up to 3 months after ovx or sham surgery. After evaluating the 3D trabecular microarchitecture of the specimen by micro-CT, and bone minerals by dual-energy X-ray absorptiometry (DXA), we performed compression tests on the same specimen. To test the hypothesis that the contribution of trabecular microarchitecture to strength varies with the duration of the post-ovx period, we evaluated the relationship between several bone indices of mass and microarchitecture and compression load values by linear regression analysis. We then used stepwise regression analysis to identify the major determinants of the ultimate load during the early and late post-ovx periods. Materials and Methods Experimental Animals Female Sprague-Dawley (SD) rats (n ⫽ 190, 8 weeks of age) were purchased from Charles River Japan (Hino, Japan), and acclimated for 8 weeks prior to the start of the experiment. The rats had free access to tap water and to a commercial standard rat chow containing 1.18% calcium and 2.0 IU/g of vitamin D3 (CE-2; Japan Clea, Tokyo, Japan). Body weight was measured weekly. All rats were housed in metal cages (three or four rats per cage) in an air-conditioned environment (room temperature 23 ⫾ 2°C, humidity 55 ⫾ 10%) illuminated from 7:00 to 19:00. The experimental protocol was approved by the ethics review committee for animal experimentation of the University of Occupational and Environmental Health. Experimental Protocol At 4 months of age, rats were randomly assigned to 19 groups. Ten animals were killed at day 0 (baseline control). Half of the remaining rats underwent bilateral ovariectomy (ovx), whereas the others underwent sham surgery (sham). Ten rats from each group were killed at 3, 7, 11, 14, 28, 42, 56, 70, and 84 days postsurgery. Urine samples were collected from each rat over a period of 24 h before killing and stored at ⫺80°C until use. Cardiac blood samples were obtained at time of killing, immediately centrifuged, and serum samples stored at ⫺80°C for later analysis. After killing by exsanguination, the fifth lumbar vertebra (L-5) was dissected carefully, removed en bloc, and immediately stored at ⫺80°C until use. The effect of ovx was verified by involution atrophy of the uterus. Bone Metabolic Markers Serum osteocalcin concentrations were determined using an enzyme-linked immunosorbent assay (ELISA) kit employing rat osteocalcin antibody (Rat Osteocalcin RIA Biomedical Technologies, Inc., Stoughton, MA). Urinary deoxypyridinoline (DPD) cross-link excretion levels were determined using an ELISA kit (Pyrilinks-D Metra Biosystems, Inc., Mountain View, CA). The results were expressed as DPD/Cr ratios (nmol/L 䡠 mmol/L Cr).
Bone Vol. 28, No. 6 June 2001:625– 633
shifted automatically in the axial direction. The filtered 40 kVp X-ray spectrum peaks at 25 keV, allowing excellent bone vs. marrow contrast.15 The rotation and axial shift are manipulated separately with two stepping motors. A standard convolutionbackprojection procedure with a Shepp and Logan filter is used to reconstruct the CT images in 1024 ⫻ 1024 pixel matrices. The spatial resolution is typically defined by a 10% contrast level in the modulation transfer function (MTF), resulting in a in-plane spatial resolution of 28 m.15,26 The 3D trabecular microarchitecture was obtained for 200 reconstructed slices with an 8 m increment. The volume of interest (VOI) was determined for the entire region of the secondary spongiosa, excluding the endosteal surface and the region of the vertebral vessels. Calibration of the approximately 28 m resolution micro-CT for in vitro trabecular microarchitecture measurements has been reported previously.15 On 3D analysis, the tissue volume (TV; mm3) and trabecular bone volume (BV; mm3) were measured directly, and the fractional trabecular bone volume (BV/TV; %) was calculated. Trabecular thickness (Tb.Th;m), trabecular separation (Tb.Sp; m), and trabecular number (Tb.N; 1/mm) were measured directly on 3D images using the method described by Hildebrand et al.12,13 Thus, the parameters Tb.Th, Tb.Sp, and Tb.N are model-independent indices, not biased by eventual deviations of the actual structure. For nonmetric indices, the values of 3D structural indices, such as structure model index (SMI) and geometrical degree of anisotropy (DA), were computed using software included with the micro-CT machine. The parameter SMI represented the plate-rod characteristic of trabecular structure that was calculated by a differential analysis of a triangulated surface of a trabecular structure,12,14 and DA represents trabecular anisotropy defined as the ratio between the maximal and minimal radius of the mean intercept length (MIL),9,12 respectively. The trabecular bone pattern factor (TBPf; 1/mm), representing the ratio of concave:convex surfaces in 2D sections of trabeculae,11 was measured for each slice and the mean value was determined for each specimen. The coefficient variation (CV) values of BV/TV, Tb.Th, Tb.Sp, Tb.N, SMI, DA, and TBPf were 3.47%, 2.69%, 1.02%, 1.12%, 12.6%, 0.34%, and 9.36%, respectively. Bone Size and Bone Mineral Measurement The height of the L-5 vertebral body was measured with a micrometer. The L-5 vertebral body specimen was fixed with a clamp at the base of the transverse process in the holder of a diamond bandsaw (Exakt, Norderstedt, Germany). By removing both the cranial and caudal endplates planoparallel surfaces were obtained. From each vertebral body, a central cylinder specimen was obtained with planoparallel ends with a height of approximately 4.0 mm. The volume of the specimen was measured by the Archimedes method, using an electric volumetric apparatus (MK-550; Muromachi Machinery, Tokyo). Bone mineral content and density were measured in each specimen by DXA (DCS600; Aloka, Tokyo) using the small-animal scan mode with irradiation applied anteroposteriorly to the specimen. The values of bone mineral content (BMC; mg) and bone mineral density (BMD; mg/cm2) were obtained for each specimen. The CV values of BMC and BMD were 3.75% and 1.19%, respectively.
Structural Analysis Using Micro-CT After the removal of adhering tissues, the posterior elements, and transverse processes of the L-5 vertebral body, each specimen was scanned by micro-CT (CT20; Scanco Medical, Zurich, Switzerland). The CT20 is equipped with a microfocus X-ray tube with a focal spot of 10 m, producing a fan beam detected by a charge-couple device array with a turntable that can be
Mechanical Tests L-5 vertebral cylinder samples were placed centrally on the smooth surface of a steel disk (10 cm diameter) attached to the materials-testing machine (Tensilon UTA-1T; Orientec, Tokyo). A compression force was applied in the craniocaudal direction by using a steel disk (1.8 cm) at a nominal deformation rate of 2
Bone Vol. 28, No. 6 June 2001:625– 633
S. Ikeda et al. Effect of trabecular contour on bone strength
627
Figure 2. Serial changes in bone mineral indices of rat L-5 vertebral body. (A) BMD values. (B) BMC values. Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. *P ⬍ 0.05 vs. sham group at the same timepoint (Tukey–Kramer post hoc tests after two-way ANOVA).
Figure 1. Serial changes in the levels of bone metabolic markers. (A) Urinary deoxypyridinoline levels. (B) Serum osteocalcin levels. Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. *P ⬍ 0.05, **P ⬍ 0.01 vs. sham group at the same timepoint; ##P ⬍ 0.01 vs. baseline control (Tukey–Kramer post hoc tests after two-way ANOVA). 22
mm/min. Crosshead displacement was considered as specimen deformation. A load-deformation curve was displayed with a monitoring recorder linked to the tester in each specimen. The ultimate load (N) was measured directly from the load-deformation curve. Statistical Analysis All data were expressed as mean ⫾ SEM. Group data were first compared by two-factor factorial analysis of variance (ANOVA) for surgery and time. If analysis confirmed a significant difference between the ovx and sham groups, then the difference between values at each timepoint was assessed by Tukey– Kramer posthoc test. For data of bone metabolic markers, BMD, BMC, trabecular microarchitectural indices, and ultimate load, differences from baseline were also assessed. The relationships between ultimate load and each microarchitectural index, BMD,
BMC, and bone metabolic markers were analyzed by simple regression. Furthermore, factors that significantly influenced ultimate load were determined by stepwise regression among the factors of the microarchitectural indices, BMD, BMC, and bone metabolic markers. p ⬍ 0.05 was considered statistically significant. Results Mortality, Body Weight, and Bone Sizes Four rats died during surgery and three rats died postoperatively for unknown reasons. However, the mortality rate was similar among the groups. Body weight significantly increased in ovx rats at day 70 after surgery, compared with the sham group (data not shown). The height and volume of the L-5 vertebral body of ovx rats were not significantly different from those of shamoperated rats (data not shown). Urinary Deoxypyridinoline (DPD) and Serum Osteocalcin Levels Urinary DPD levels in the ovx groups increased significantly from 28 days and thereafter compared with the sham groups (Figure 1A). Serum osteocalcin levels were significantly higher at days 11, 28, and 42 post-ovx compared with the sham groups
628
S. Ikeda et al. Effect of trabecular contour on bone strength
Bone Vol. 28, No. 6 June 2001:625– 633
Figure 3. Three-dimensional trabecular microarchitectural images of rat L-5 vertebral body using micro-CT. (A) Baseline control group. (B) Sham group at 4 weeks postsurgery. (C) Sham group at 12 weeks postsurgery. (D) Ovx group at 4 weeks postovariectomy. (E) Ovx group at 12 weeks postovariectomy. Top: cranial side; bottom: inferior position.
and significantly higher at days 28 and 42 compared with baseline (Figure 1B), although values decreased gradually thereafter. BMD and BMC of L-5 Vertebral Body In ovx groups, L-5 BMD values decreased from day 42 and were significantly lower at day 84 compared with the sham group (Figure 2A). BMC values also decreased significantly at days 42 and 84 compared with the sham group (Figure 2B). However, these values were not significantly different from baseline. In the sham groups, BMD and BMC values were not significantly different from the respective baseline values. Micro-CT Images and Indices of 3D Trabecular Microarchitecture of L-5 Images of the L-5 vertebra in the sham groups did not appear to change throughout the experimental period (Figure 3A–C). However, the rod-like trabecular structure became dominant in the ovx groups at day 84 (Figure 3D, E). The BV/TV in the ovx groups decreased after surgery and was significantly lower from day 28 and thereafter compared with sham and baseline control groups (Figure 4A). Tb.Th in the ovx groups decreased significantly from day 42 compared with sham rats (Figure 4B). The values at days 56 and 70 were significantly lower than baseline values. Tb.Sp increased significantly from day 28 compared with the sham group and from day 42 compared with baseline (Figure 4C). Tb.N diminished significantly from day 28 compared with the sham group and from day 42 compared with baseline (Figure 4D). The SMI in ovx rats was significantly higher at day 28 and thereafter in comparison with sham and baseline controls (Figure 5A). DA was higher in ovx rats than in sham group at day 84, but not significantly different from baseline (Figure 5B). TBPf
increased significantly at day 28 days and thereafter in the ovx groups compared with sham and baseline controls (Figure 5C). Ultimate Compressive Load of L-5 The ultimate compressive load values of L-5 in the ovx groups gradually decreased postsurgery and became significantly lower from day 28 as compared with baseline (Figure 6). Values became smaller than those of sham-operated rats from day 42. Relationship Between Ultimate Load and Parameters of Bone and Turnover The results for ultimate load showed that the value decreased acutely by approximately 23% during 28 days post-ovx (Figure 6). On the other hand, the values essentially remained at a low level, showing only a 2% decrease during 42– 84 days post-ovx. The postsurgery period was divided into early (0 –28 days postsurgery) and late (42– 84 days postsurgery) periods for post hoc statistical analysis. Early postoperative period. In the sham group, BMD correlated significantly with ultimate load (Table 1). BMC and Tb.Th also correlated significantly with load. In ovx rats, TBPf, SMI, BV/TV, Tb.N, and Tb.Sp correlated significantly with load. Late postoperative period. In the sham group, BMD and BMC correlated significantly with the ultimate load. However, there was a weak relationship between Tb.N and ultimate load. In ovx rats, BMD and BMC correlated significantly with load. TBPf, Tb.N, BV/TV, Tb.Th, and SMI also correlated significantly with the load. Bone metabolic markers and DA did not correlate with the ultimate load during both early and late postoperative periods.
Bone Vol. 28, No. 6 June 2001:625– 633
S. Ikeda et al. Effect of trabecular contour on bone strength
629
Figure 4. Serial changes in three-dimensional microarchitectural indices of rat L-5 vertebral body analyzed by micro-CT. (A) Trabecular bone volume (BV/TV; %). (B) Trabecular thickness (Tb.Th; m). (C) Trabecular separation (Tb.Sp; m). (D) Trabecular number (Tb.N; 1/mm). Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. *P ⬍ 0.05, **P ⬍ 0.01 vs. sham group at the same timepoint. *P ⬍ 0.05, **P ⬍ 0.01 vs. baseline control (Tukey–Kramer post hoc tests after two-way ANOVA).
Factors Determining Ultimate Load
Discussion
In the early postoperative period, BMD was the most significant factor that influenced the ultimate load in sham-operated rats, as assessed by stepwise regression analysis (Table 2). In ovx groups, however, TBPf was the most significant factor that influenced the compressive load, followed by SMI. In the late period, BMC was the most significant factor that influenced load in both sham and ovx groups.
In the present study we demonstrated that the contribution of parameters of bone mass and trabecular microarchitecture to the ultimate load of the lumbar vertebra differed between the early and late periods after ovx in rats. Although BMD and BMC of the lumbar vertebra did not decrease significantly after ovx, the ultimate load values decreased in 4 weeks. Furthermore, BV/TV and Tb.N diminished and Tb.Sp increased during the same period. Nonmetric parameters, such as SMI and TBPf, also increased significantly during the same period. Although various parameters of bone mass and trabecular microarchitecture were related significantly to the compressive load in ovx groups, TBPf was the major determinant of load within 4 weeks after ovx as analyzed by stepwise regression. However, BMC had the strongest influence on load from 6 weeks post-ovx. In the sham groups, the effect of trabecular microarchitecture on the compressive load was limited, which depended mainly on BMD and BMC. TBPf and SMI did not correlate with compressive load in the sham group. Our study showed increases in two bone metabolic markers
Relationship Between Other Parameters and Determinants of Ultimate Load In the sham group, BMC and Tb.Th correlated significantly with BMD in the early postsurgery period (Table 3). In the late postsurgery period, BMD and Tb.Th correlated significantly with BMC. In the ovx groups, SMI, BV/TV, Tb.N, and Tb.Sp correlated significantly with TBPf in the early postsurgery period. Tb.Th, BMD, and BMC also correlated with TBPf in the same period. In the late postsurgery period, BMD, Tb.N, TBPf, BV/ TV, and Tb.Th correlated significantly with BMC.
630
S. Ikeda et al. Effect of trabecular contour on bone strength
Bone Vol. 28, No. 6 June 2001:625– 633
Figure 6. Serial changes in ultimate load of rat L-5 vertebral body. Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. *P ⬍ 0.05, **P ⬍ 0.01 vs. sham group at the same timepoint. **P ⬍ 0.01 vs. baseline control (Tukey–Kramer post hoc tests after two-way ANOVA).
Figure 5. Serial changes in direct nonmetric microarchitectural indices of rat L-5 vertebral body analyzed by micro-CT. (A) Structure model index (SMI) values. (B) Geometrical degree of anisotropy (DA). (C) Trabecular bone pattern factor (TBPf; 1/mm). Open squares: sham group; filled circles: ovx group. Values are mean ⫾ SEM. **P ⬍ 0.01 vs. sham group at the same timepoint; ##P ⬍ 0.01 vs. baseline control (Tukey– Kramer post hoc tests after two-way ANOVA).
after ovx. These findings are consistent with previous results reported in mature SD rats.19,29 In our study, serum osteocalcin concentrations increased in 7 days and reached a peak level at 23
days post-ovx, as observed by Yamaura et al.29 The difference in urinary DPD excretion between sham and ovx rats became significant within 4 weeks after surgery in this study, but the values in the ovx groups apparently began to increase from 11 days. Thus, the urinary DPD data are consistent with results reported by Lane et al.19; that is, an increase occurring from day 13 post-ovx. However, there was the discrepancy that the bone metabolic markers did not correlate TBPf and ultimate load in a significant manner. Egger et al.6 observed that the increases in excretion of pyridinium cross-links in rat closely reflects bone resorption in chronic but not acute bone metabolism. Lane et al.19 also observed that 3D trabecular disruption occurs early before the increase in urinary DPD after ovx in rats. Yamaura et al.29 reported that trabecular osteoclast number increase 2–3 days after ovx, but bone marker did not increase as rapidly. These data and the results of the present study both suggest that pyridinium cross-links are not sensitive enough to reflect acute increase in bone resorption in rats. Thus, we suggest that it possible that either TBPf, indicating the trabecular contour, or ultimate load did not correlate significantly with these metabolic markers. The term TBPf was first introduced by Hahn et al.,11 based on the hypothesis that abundant concave surfaces in 2D sections represent a well-connected spongy lattice, whereas abundant convex surfaces indicate a poorly connected trabecular lattice. Thus, TBPf represents trabecular connectivity, and a high value reflects the disruption of trabecular continuity in 2D sections.2,3,10,17 A shift from a plate- to a rod-like structure in trabecular bone post-ovx, as observed by 3D images in this study, is consistent with an increase in TBPf values. Our results show that DA increased significantly only at 3 months in the ovx groups as compared with the sham group. Thus, the geometrical degree of anisotropy of trabecular bone did not seem to be affected significantly by acute estrogen loss. Results of both simple and stepwise regression analyses indicated that the contour of the trabecular bone surface is important for the loss of lumbar bone strength post-ovx in rats. Specifically, changes in the surface contour probably contributed to the significant loss of compressive strength at 4 weeks postovx, when bone turnover was increased. Stepwise regression analysis showed that TBPf was the determinant factor for ultimate load, and the value was very closely related to SMI with a coefficient of determination (R2) value of 0.968. Moreover, R2 for ultimate load indicated a higher value of 0.975 with TBPf and
Bone Vol. 28, No. 6 June 2001:625– 633
S. Ikeda et al. Effect of trabecular contour on bone strength
631
Table 1. Coefficient of determination (R2) values for ultimate load between microarchitectural indices, bone mass, and bone metabolic markers analyzed by simple regression Sham
Microarchitectural indices BV/TV Tb.Th Tb.Sp Tb.N SMI DA TBPf Bone mass BMD BMC Bone metabolic markers Urinary DPD Serum osteocalcin
OVX
Early period (n ⫽ 48)
Late period (n ⫽ 38)
Early period (n ⫽ 48)
Late period (n ⫽ 39)
0.046 0.099* 0.018 0.080 0.046 0.001 0.056
0.112 0.100 0.082 0.154* 0.074 0.090 0.107
0.473** 0.176** 0.421** 0.440** 0.483** 0.001 0.488**
0.200** 0.197** 0.128* 0.223** 0.169* 0.060 0.305**
0.439** 0.101*
0.493** 0.500**
0.264** 0.155**
0.434** 0.503**
0.021 0.013
0.081 0.047
0.001 0.041
0.024 0.001
Early period was between 0 to 28 days, and late period was between 42 to 84 days postsurgery. *P ⬍ 0.05, **P ⬍ 0.01 (simple regression).
SMI. Thus, it is reasonable to presume that the feature of trabecular bone structure commonly assessed by TBPf and SMI, the surface contour, plays a critical role in the loss of bone strength. In the sham group, however, these nonmetric parameters did not change during the same period. In fact, although compressive load values did not correlate with these nonmetric parameters of trabecular microarchitecture, they did correlate with the parameters of bone mass, such as BMD and BMC. In the period from 6 weeks post-ovx, metric parameters, such as BV/TV and Tb.N, gradually decreased and Tb.Sp increased, but nonmetric parameters of SMI and TBPf remained stable as did the ultimate compressive load values. On the other hand, the parameters of trabecular microarchitecture values were so varied that we could not find any statistically significant changes in energy absorption, such as in one of the biomechanical parameters in this investigation (data not shown). We now believe that the changes in energy absorption may be affected not only by the increased turnover after ovx but also in the basal 3D microarchitecture before ovx. Taken together, we postulate that the compressive strength of the lumbar vertebra is determined mainly by the contour of trabecular bone when bone loss is accelerated with an increase in bone turnover. However, when trabecular bone loss is alleviated without marked changes in the contour of trabecular surface, the parameters of bone mass, such as BMD and BMC, are the major determinants of lumbar bone strength in rats. Compromised trabecular microarchitecture is an important and independent pathogenic factor in vertebral fracture in osteoporotic postmenopausal women and osteoporosic men.4 Using
necropsy sections of human vertebral bodies from subjects aged 17–90 years, Grote et al.10 demonstrated the presence of an age-dependent decrease in trabecular bone mass, primarily as a consequence of transformation from plates to rods and the loss of whole trabecular bone volume. Loss of trabecular connectivity as determined by 2D and 3D analyses also indicated changes in bone surface.19,21 Thus, it is possible that changes in the contour of trabecular bone surface by increased turnover affect the strength of the human vertebral body as well, consequently increasing the risk of spinal fracture in postmenopausal women with osteoporosis. We have previously demonstrated that the 28 m resolution micro-CT was suitable for the evaluation of 3D parameters of trabecular bone mass and structure using human iliac bone biopsy specimens.15 Laib and Ru¨egsegger18 also demonstrated that a 3D peripheral quantitative computed tomography machine with a similar resolution value was suitable for evaluation of trabecular bone structure in vivo. Thus, it is possible to evaluate the surface contour of trabecular bone, although it is often difficult to use this technique to visualize irregularities on the surface of individual trabecula and remodeling pits that may represent structurally weak sites when the bone is loaded. Thus, based on the results of the present study alone, we cannot estimate the role of specific changes in trabecular bone surfaces by local remodeling activity or accumulation of trabecular discontinuity. To further estimate the effect of trabecular bone contour on strength, visualization of bone surfaces with a higher level of spatial resolution is necessary. In conclusion, we have demonstrated that TBPf was the most
Table 2. Determinants of ultimate compressive load analyzed by stepwise regression Early period
Late period
Group
Index
R2
P value
Index
R2
P value
Sham
BMD BMD⫹BMC TBPf TBPf⫹SMI
0.258 0.331 0.669 0.975
⬍0.05 ⬍0.01 ⬍0.01 ⬍0.01
BMC BMC⫹BMD BMC BMC⫹BMD
0.447 0.739 0.511 0.772
⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01
OVX
Early period was between 0 to 28 days, and late period was between 42 to 84 days postsurgery. Number of animals was 48 in the sham and OVX groups during the early period. Number of animals was 38 in the sham group and 39 in the OVX group during the late period.
632
S. Ikeda et al. Effect of trabecular contour on bone strength
Bone Vol. 28, No. 6 June 2001:625– 633
Table 3. Coefficient of determination (R2) values for each of the determinant factors between microarchitectural indices, bone mass, and bone metabolic markers analyzed by simple regression Sham
Microarchitectural indices BV/TV Tb.Th Tb.Sp Tb.N SMI DA TBPf Bone mass BMD BMC Bone metabolic markers Urinary DPD Serum osteocalcin
OVX
Early period (n ⫽ 48)
Late period (n ⫽ 38)
Early period (n ⫽ 48)
Late period (n ⫽ 39)
R2 for BMD
R2 for BMC
R2 for TBPf
R2 for BMC
0.044 0.163** 0.003 0.040 0.037 0.002 0.048
0.037 0.118* 0.018 0.053 0.023 0.056 0.053
0.958** 0.470** 0.753** 0.765** 0.968** 0.003 —
0.125* 0.122* 0.055 0.167* 0.060 0.003 0.163*
— 0.320**
0.750** —
0.273** 0.086*
0.780** —
0.004 0.032
0.008 0.060
0.001 0.053
0.003 0.026
Early period was between 0 to 28 days, and late period was between 42 to 84 days postsurgery. *P ⬍ 0.05, **P ⬍ 0.01 (simple regression)
significant determinant of reduced compressive strength of lumbar bone in rats during the early post-ovx period and that SMI was closely related to TBPf. However, in the late post-ovx period, BMC was the major determinant of bone strength. These results imply that changes in trabecular bone contour with increased bone turnover are critical in reducing lumbar bone strength in the early period after ovx in rats.
10.
11.
12.
Acknowledgments: The authors thank Dr. Nobukazu Okimoto, Dr. Yuichi Okazaki, Dr. Soshi Uchida, Dr. Shinya Tanaka, Dr. Hiroshi Wadayama, Junko Nomura, and Erika Kobayashi. This work was supported in part by grants-in-aid from the Research Society for Metabolic Bone Diseases and the Ministry of Health and Welfare of Japan.
13.
14.
References 1. Bouillon, R., Burkhardt, P., Christiansen, C., Fleisch, H. A., Fujita, T., Gennari, C., Martin, T. J., Mazzuoli, G., Melton, L. J., Ringe, J. D., et al. Consensus development conference: Prophylaxis and treatment of osteoporosis. Osteopor Int 1:114 –117; 1991. 2. Chappard, D., Legrand, E., Basle, M. F., Fromont, P., Racineux, J. L., Rebel, A., and Audran, M. Altered trabecular architecture induced by corticosteroids: A bone histomorphometric study. J Bone Miner Res 11:676 – 685; 1996. 3. Croucher, P. I., Garrahan, N. J., and Compston, J. E. Assessment of cancellous bone structure: Comparison of strut analysis, trabecular bone pattern factor, and marrow space star volume. J Bone Miner Res 11:955–961; 1996. 4. Dempster, D. W. The contribution of trabecular architecture to cancellous bone quality. J Bone Miner Res 15:20 –23; 2000. 5. Dempster, D. W., Ferguson-Pell, M. W., Mellish, R. W., Cochran, G. V., Xie, F., Fey, C., Horbert, W., Parisien, M., and Lindsay, R. Relationships between bone structure in the iliac crest and bone structure and strength in the lumbar spine. Osteopor Int 3:90 –96; 1993. 6. Egger, C. D., Muhlbauer, R. C., Felix, R., Delmas, P. D., Marks, S. C., and Fleisch, H. Evaluation of urinary pyridinium crosslink excretion as a marker of bone resorption in the rat. J Bone Miner Res 9:1211–1219; 1994. 7. Faulkner, K. Bone matters: Are density increases necessary to reduce fracture risk? J Bone Miner Res 15:183–187; 2000. 8. Feldkamp, L. A., Goldstein, S. A., Parfitt, A. M., Jesion, G., and Kleerekoper, M. The direct examination of three-dimensional bone architecture in vitro by computed tomography. J Bone Miner Res 4:3–11; 1989. 9. Goulet, R. W., Goldstein, S. A., Ciarelli, M. J., Kuhn, J. L., Brown, M. B., and
15.
16. 17.
18.
19.
20.
21.
22.
Feldkamp, L. A. The relationship between the structural and orthogonal compressive properties of trabecular bone. J Biomech 27:375–389; 1994. Grote, H. J., Amling, M., Vogel, M., Hahn, M., Posl, M., and Delling, G. Intervertebral variation in trabecular microarchitecture throughout the normal spine in relation to age. Bone 16:301–308; 1995. Hahn, M., Vogel, M., Pompesius-Kempa, M., and Delling, G. Trabecular bone pattern factor—a new parameter for simple quantification of bone microarchitecture. Bone 13:327–330; 1992. Hildebrand, T., Laib, A., Mu¨ller, R., Dequeker, J., and Ru¨gsegger, P. Direct three-dimensional morphometric analysis of human cancellous bone: Microstructural data from spine, femur, iliac crest, and calcaneus. J Bone Miner Res 14:1167–1174; 1999. Hildebrand, T. and Ru¨gsegger, P. A new method for the model-independent assessment of thickness in three-dimensional images. J Microsc 185:67–75; 1997. Hildebrand, T. and Ru¨gsegger, P. Quantification of bone microarchitecture with the structure model index. Compreh Meth Biomech Biomed Eng 1:15–23; 1997. Ito, M., Nakamura, T., Matsumoto, T., Tsurusaki, K., and Hayashi, K. Analysis of trabecular microarchitecture of human iliac bone using microcomputed tomography in patients with hip arthrosis with or without vertebral fracture. Bone 23:163–169; 1998. Kinney, J. H., Lane, N. E., and Haput, D. In vivo, three-dimensional microscopy of trabecular bone. J Bone Miner Res 10:264 –270; 1995. Kippo, K., Hannuniemi, R., Isaksson, P., Lauren, L., Osterman, T., Peng, Z., Tuukkanen, J., Kuurtamo, P., Vaananen, H. K., and Sellman, R. Clodronate prevents osteopenia and loss of trabecular connectivity in estrogen-deficient rats. J Bone Miner Res 13:287–296; 1998. Laib, A. and Ru¨gsegger, P. Calibration of trabecular bone structure measurements of in vivo three-dimensional peripheral quantitative computed tomography with 28-micron-resolution microcomputed tomography. Bone 24:35–39; 1999. Lane, N. E., Thompson, J. M., Haupt, D., Kimmel, D. B., Modin, G., and Kinney, J. H. Acute changes in trabecular bone connectivity and osteoclast activity in the ovariectomized rat in vivo. J Bone Miner Res 13:229 –236; 1998. Legrand, E., Chappard, D., Pascaretti, C., Duquenne, M., Krebs, S., Rohmer, V., Basle, M.-F., and Audran, M. Trabecular bone microarchitecture, bone mineral density, and vertebral fractures in men with osteoporosis. J Bone Miner Res 15:13–19; 2000. Mellish, R. W. E., Ferguson-Pell, M. W., Cochran, G. V. B., Lindsay, R., and Dempster, D. W. A new manual method for assessing two dimensional cancellous bone structure: Comparison between iliac crest and lumbar vertebra. J Bone Miner Res 6:689 – 696; 1991. Mosekilde, Li., Danielsen, C. C., and Knudsen, U. B. The effect of aging and
Bone Vol. 28, No. 6 June 2001:625– 633
23.
24.
25.
26.
ovariectomy on the vertebral bone mass and biomechanical properties of mature rats. Bone 14:1– 6; 1993. Mu¨ller, R., Van Campenhout, H., Van Damme, B., Van Der Perre, G., Dequeker, J., Hildebrand, T., and Ru¨gsegger, P. Morphometric analysis of human bone biopsies: A quantitative structural comparison of histological sections and micro-computed tomography. Bone 23:59 – 66; 1998. Ohnishi, H., Nakamura, T., Narusawa, K., Murakami, H., Abe, M., Barbier, A., and Suzuki, K. Bisphosphonate tiludronate increases bone strength by improving mass and structure in established osteopenia after ovariectomy in rats. Bone 21:335–343; 1997. Parfitt, A. M., Mathews, C. H. E., Villanueva, A. R., Kleerekoper, M., Frame, B., and Rao, D. S. Relationships between surface, volume and thickness of iliac trabecular bone in aging and in osteoporosis. J Clin Invest 72:1396 –1409; 1983. Ru¨egsegger, P., Koller, B., and Mu¨ller, R. A microtomographic system for the nondestructive evaluation of bone architecture. Calcif Tissue Int 58:24 –29; 1996.
S. Ikeda et al. Effect of trabecular contour on bone strength
633
27. Thomsen, J. S., Ebbessen, E. N., and Mosekilde, Li. Relationships between static histomorphometry and bone strength measurements in human iliac crest bone biopsies. Bone 22:153–163; 1998. 28. Vesterby, A., Mosekilde, Li., Gundersen, H. J., Melsen, F., Holme, K., and Sorensen, S. Biologically meaningful determinants of the in vitro strength of lumbar vertebrae. Bone 12:219 –224; 1991. 29. Yamaura, M., Nakamura, T., Tsurukami, H., Hijioka, A., Narusawa, K., Ohnishi, H., Ohta, T., and Hosoda, K. Local bone turnover in the metaphysis of the proximal tibia and the lumbar vertebra during the early periods after ovariectomy in rats. Calcif Tissue Int 58:52–59; 1996.
Date Received: August 25, 2000 Date Revised: January 22, 2001 Date Accepted: January 29, 2001