The authors state that they have no conflicts of interest.
Selective Modification of Bone Quality by PTH, Pamidronate, or Raloxifene†
Article first published online: 8 DEC 2008
Copyright © 2009 ASBMR
Journal of Bone and Mineral Research
Volume 24, Issue 5, pages 800–808, May 2009
How to Cite
Brennan, T. C., Rizzoli, R. and Ammann, P. (2009), Selective Modification of Bone Quality by PTH, Pamidronate, or Raloxifene. J Bone Miner Res, 24: 800–808. doi: 10.1359/jbmr.081227
- Issue published online: 4 DEC 2009
- Article first published online: 8 DEC 2008
- Manuscript Accepted: 5 DEC 2008
- Manuscript Revised: 17 NOV 2008
- Manuscript Received: 18 JUL 2008
- bone strength;
- osteoporosis treatments
Bone strength, a determinant of resistance to fracture, depends on BMD, geometry, microarchitecture, bone turnover rates, and properties of the bone at the material level. Despite comparable antifracture efficacy, anti-catabolics and bone anabolic agents are likely to modify the various determinants of bone strength in very different ways. Eight weeks after ovariectomy (OVX), 8-mo-old osteoporotic rats received pamidronate (APD; 0.6 mg/kg, 5 days/mo, SC), raloxifene (3 mg/kg, 5/7 days, tube feeding), PTH(1–34) (10 μg/kg, 5/7 days, SC), or vehicle for 16 wk, and we measured vertebral BMD, maximal load, stiffness and energy, microarchitecture, and material properties by nanoindentation, which allows the calculation of the elastic modulus, tissue hardness, and working energy. Markers of bone turnover, plasma osteocalcin, and urinary deoxypyridinoline (Dpd) were also determined. PTH induced greater maximal load than APD or raloxifene, as well as greater absorbed energy, BMD, and increased bone turnover markers. PTH markedly increased trabecular bone volume and connectivity to values higher than sham. Animals treated with APD had BV/TV values significantly higher than OVX but lower than sham, whereas raloxifene had no effect. Tissue hardness was identical in PTH-treated and OVX untreated controls. In contrast, APD reversed the decline in strength to levels not significantly different to sham, reduced bone turnover, and increased hardness. Raloxifene markedly increased material level cortical hardness and elastic modulus. These results show the different mechanisms by which anti-catabolics and bone anabolics reduce fracture risk. PTH influences microarchitecture, whereas bisphosphonates alter material-level bone properties, with probable opposite effects on remodeling space. Raloxifene primarily improved the material stiffness at the cortical level.
Postmenopausal osteoporosis, primarily caused by the steep reduction in estrogen after menopause,(1) is characterized by an increase in bone remodeling leading to decreased bone mass and bone strength and increased fracture risk.(2) Bone strength is dependent on several parameters including bone microarchitecture and geometry, mineral content, and intrinsic bone tissue quality. Currently, there are several options for the treatment of postmenopausal osteoporosis that generally fall under the headings of anti-catabolic or anabolic (bone-forming) agents.(3) Despite having comparable antifracture efficacies, these two classes of therapeutic agents are likely to have distinct effects on the modulation of the various determinants of bone strength.
Antiresorptive agents such as bisphosphonates and selective estrogen receptor modulators (SERMs) are the most commonly prescribed treatments for osteoporosis in postmenopausal women. Bisphosphonates act directly on osteoclasts to inhibit resorption(4) and thus slow the increased rate of remodeling which is seen in postmenopausal osteoporosis. This allows more time for secondary mineralization and an increased BMD,(5,6) ultimately leading to a reduced fracture risk.
Estrogen replacement therapy reduces the accelerated bone remodeling, prevents further bone loss, and increases bone mass, particularly at sites that are substantially comprised of trabecular bone.(7,8) It is well known, however, that estrogen replacement therapy causes various adverse side effects and often fear of cancer may preclude the use of estrogen. A class of related compounds called SERMs bind directly to the estrogen receptors and provide similar bone protective effects.(9) Among them, raloxifene has been shown to decrease bone turnover in postmenopausal women, prevent bone loss, and reduce the risk of vertebral fractures without any significant influence on nonvertebral fractures.(10)
Intact PTH(1–84) and its shorter peptide analog, teriparatide [human recombinant PTH(1–34)], constitute a new class of anabolic therapy in the treatment of osteoporosis. Teriparatide has a positive effect on BMD in the lumbar spine and the femoral neck,(11) bone quality,(12,13) and microarchitecture(14) and reduces the risk of new(11) and primary(15) fractures because of its ability to stimulate the formation of new bone. PTH must be considered a useful alternative in the treatment of severe osteoporosis, both in men and women.(12) PTH has also been shown to be efficacious in the treatment of glucocorticoid-induced and other secondary osteoporosis.(12,16)
Whereas bisphosphonates, SERMs, and PTH all have proven therapeutic antifracture uses, the determinants of bone strength that they each modify to achieve these effects are likely to be different and through distinct mechanisms of action. We studied and compared the effects of these anti-osteoporotic treatments on maximal load (N), stiffness (N/mm), and energy (N*mm) with an axial compression test, BMD, and microarchitecture by μCT and properties of material stiffness, in particular, tissue hardness (mPa), elastic modulus (gPa), and working energy at the material level were measured by nanoindentation. The levels of urinary deoxypyridinoline (Dpd) and serum osteocalcin were also determined as markers of bone turnover.
The ovariectomized (OVX) rat model is well documented to mimic the effects of the sharp decline in estrogen on skeletal remodeling seen in postmenopausal women.(17–21) It is known that OVX increases bone turnover, with an overall negative effect on bone microarchitecture and bone mass(21); however, whereas the effects of OVX, OVX and treatment with PTH,(22) or OVX and treatment with raloxifene(23) on intrinsic bone material properties have been reported in monkeys, there have been no reports as yet that have simultaneously compared the effects of these various treatments at the material level in OVX rats.
MATERIALS AND METHODS
Female Sprague-Dawley rats (Charles River, L'Abresle, France), housed individually at 25°C with a 12:12-h light-dark cycle, were strictly pair-fed a laboratory diet (Provima Lacta, Cossonay, Switzerland) containing 0.8% phosphorous and 1.1% calcium throughout the experimental period. All experimental procedures received approval from the Animal Ethics Committee of the Geneva University Faculty of Medicine. Six-month-old rats underwent transabdominal OVX or sham operation (SHAM, n = 10) under anesthesia with intraperitoneal ketamine hydrochloride (100 mg/7 kg body weight). At age 8 mo, or 8 wk postoperational at a time when the effects of OVX on BMD have been shown to be fully evident, OVX rats were randomly allocated to one of four groups of 10 rats. Rats were treated with either pamidronate (APD; 0.6 mg/kg, 5 d/mo, SC), raloxifene (3 mg/kg, 5/7 days, tube feeding), teriparatide (10 μg/kg, 5/7 days, SC), or vehicle for 16 wk. Blood was collected from the aorta at time of death for serum osteocalcin determination. DXA measurements were performed just before treatment was started (i.e., 8 wk after OVX) and at 8-wk intervals during the treatment period. The day before death, rats were placed in metabolic cages, and urine was collected over an 18-h period for the determination of Dpd.
Areal BMD and BMC measurements
Areal BMD and BMC were measured before death by DXA using a regular Hologic QDR-1000 instrument adapted to measurements for small animals.(24) An ultra-high-resolution mode (line spacing: 0.0254 cm; resolution: 0.0127 cm) was used with a collimator of 0.9 mm in diameter. During the measurements, animals were anesthetized with ketamine hydrochloride (100 mg/kg body weight, IP). BMD, BMC, and scanned bone area were recorded at the level of the lumbar spine. The stability of the instrument was controlled by scanning a phantom six times per week.
Bone mechanical properties
The lumbar spine was excised immediately after death and frozen at −20°C in plastic bags. Bones were slowly thawed at 7°C overnight and warmed to room temperature before mechanical testing. Vertebrae L3 and L4 were isolated from the lumbar spine at the level of the intervertebral discs. Vertebral pedicles were dissected out, taking care not to damage the cortical shell. Because caudal and cranial surfaces of rat vertebral bodies are not parallel, 1 mm of the caudal and cranial sections of each vertebrae was embedded in methylmethacrylate (Technovit 4701; Heraeus Kulzer, Wehrheim, Germany) to ensure regular distribution of the compressive forces without compromising vertebral body architecture. Using silicon tubes and metallic fixation, we first imbedded 1 mm of the vertebral body in methylmethacrylate. Then, the vertebra with the methylmethacrylate was slipped inside a silicon tube with an open window, allowing the visual control to imbed another 1 mm of the other side of vertebral body in methylmethacrylate. Thus, the vertebral body was imbedded on both sides along 1 mm. Between the different steps of preparation, each specimen was kept immersed in physiological saline. The mechanical resistance to failure was tested using a servo-controlled electromechanical system (Instron 1114; Instron, High Wycombe, UK), with the actuator displaced at 2 mm/min. Maximal load (N), stiffness (slope of the linear part of the curve, representing the elastic deformation, N/mm), and energy (surface under the curve, N*mm) were calculated.
Nanoindentation tests were used to evaluate the intrinsic mechanical properties of trabecular and cortical bone tissue. They were performed using a nano-hardness tester (NHT; CSM Instruments, Peseux, Switzerland). In this test, force-displacement data of a pyramidal diamond indenter that is pressed into a material are recorded as previously described.(25) Briefly, the indenter tip is loaded at a given depth into the sample and the load is held constant, leading to a creeping of the material below the tip. This results in a complex combination of elastic and postyield deformation, which is analyzed to obtain elastic modulus and tissue hardness as well as working energy. Fugure 1 shows a typical load displacement curve obtained during indentation of a pyramidal diamond indenter that is pressed into the bone sample. This allows for the calculation of tissue hardness, elastic modulus, and working energy. Tissue hardness is interpreted as the mean pressure the material can resist and is calculated as the ratio of maximum force to contact area: .
Elastic modulus is defined by the initial slope of the unloading section of the curve, or part 3 of Fig. 1, and is calculated using the following equation: .
Working energy is calculated as the surface under the load-displacement curve (shaded area of the load-displacement curve in Fig. 1).
For the nanoindentation tests, the L4 vertebral body of each rat was dissected from the intervertebral discs, embedded in polymethylmethacrylate, and cut transversely through the middle as previously described.(25) Samples were rehydrated following a standardized protocol for 16 h in saline solution. The mechanical tests included five indentations on the cortical shell and five indentations of the trabecular node at the posterior end of each vertebral body. Indents were set to a 900-nm depth with an approximate strain rate of ϵ = 0.066 1/s for both loading and unloading. Full rehydration occurs at this distance necessary for the nanoindenter of <1 μm from the surface of the sample and is stable for up to a period of 60 h.(25) It has also been previously reported that rehydrated bone gives significant results, whereas dry samples under the same conditions do not.(25) At maximum load, a 5-s holding period was applied. The limit of the maximal allowable thermal drift was set to 0.1 nm/s. All tests were performed by a technician blinded to the treatment of each group.
Bone mass and architecture of the vertebral bodies were analyzed in a high-resolution μCT system (μCT 40; Scanco Medical, Basserdorf Switzerland) as previously described.(26) Briefly, 3D images of the vertebral body were acquired with a voxel size of 20 μm in all spatial directions. The embedded vertebral body was secured in a cylindrical canister in air. The resulting grayscale images were segmented using a low-pass filter to remove noise, and a fixed threshold was used to extract the mineralized bone phase. The trabecular and cortical sections of the vertebral body were separated with semiautomatically drawn contours. From the binarized images, structural indices were also assessed including the structural model index (SMI), which provides a quantitative assessment of the architecture of the bone. Relative bone volume (BV/TV), trabecular number (Tb.N), thickness (Tb.Th), and separation (Tb.S) in trabecular bone were also calculated by measuring the direct 3D distances.(26)
Osteocalcin was measured using radioimmunoassay (RIA) reagents from Biomedical Technologies (Stoughton, MA, USA). Total urinary Dpd was calculated using a kit from Metra Biosystems (Mountain View, CA, USA) according to manufacturer's instructions after acid hydrolysis of urine collected from rats maintained in metabolic cages over an 18-h period and just before death.
All results are presented as means ± SE. A one-way ANOVA, followed by Fisher's protected least significance difference (PLSD) post-test, was used to determine significant differences between groups.
Determinants of vertebral body bone strength
The effects of OVX or OVX followed by treatment with pamidronate, raloxifene, or PTH(1–34) on mechanical properties were studied at the level of the vertebral body, which contains cortical and trabecular bone. The compression test, performed at the level of the intact L4 body, indicated that, compared with sham-operated control rats, OVX induced a significant decrease in maximal load (p < 0.01; Fig. 2) and in total energy (p < 0.05; Table 1). The maximal load in rats treated with PTH was markedly higher than in OVX (+47%, p < 0.001) and higher than in sham control animals (+25%, p < 0.001). PTH also induced significant increases in total energy compared with both OVX (p < 0.001) and sham (p < 0.01) controls. APD significantly increased the maximal load by 33% compared with OVX (p < 0.001). Raloxifene did not induce a significant increase in maximal load versus OVX, and neither APD nor raloxifene provoked significant increases in total energy compared with OVX (Table 1). Areal BMD was markedly increased by PTH, ADP, and raloxifene compared with OVX (p < 0.001, p < 0.001, and p < 0.05, respectively; Table 1), whereas only PTH increased BMD versus sham (p < 0.05; Table 1.)
Effect of OVX or OVX and treatments on vertebral body microarchitecture
A marked decrease in trabecular bone volume was observed after OVX (p < 0.001 compared with sham). OVX also reduced trabecular thickness and number (p < 0.001 and p < 0.01, respectively) while increasing the spacing between the trabecular network (p < 0.01 compared with sham). All treatments tested increased BV/TV to significantly higher values than in OVX animals, particularly PTH, with values higher than sham-operated rats (p < 0.001; Table 2; Fig. 3). PTH also induced striking increases in trabecular thickness and number (p < 0.001 for both values compared with respective OVX rats), with a concurrent and significant decrease in trabecular spacing (p < 0.001). APD increased trabecular number (p < 0.05) and decreased spacing (p < 0.01), although neither values were as potent as PTH. Raloxifene was unable to restore or improve these parameters of microarchitecture after the detrimental effects of OVX (Table 2).
Effect of OVX or OVX and treatments on material-level properties of the vertebral body
Elastic modulus (+11.3%, p < 0.05) and tissue hardness (+16.7%, p < 0.01) were significantly increased in cortical bone by raloxifene compared with OVX. PTH increased cortical tissue hardness by 15.6% (p < 0.05) yet failed to make an impact on the effect of OVX on all other parameters measured. APD also had no effect compared with OVX (Table 3). Nanoindentation was also performed at the level of the trabecular bone, where raloxifene increased working energy (+8.0%, p < 0.05). PTH significantly decreased tissue hardness (−20.6%, p < 0.001), whereas APD and raloxifene had no effect.
Biochemical changes after OVX or OVX and treatments
Markers of bone turnover were measured in the urine and serum of rats in all groups, and it was observed that, after OVX, markers of bone turnover were increased compared with sham-operated rats. Serum osteocalcin levels rose by >30% (p < 0.05) and, although not significant, there was an obvious trend toward increased Dpd excretion (Table 4). The observed increase in bone remodeling markers after OVX was prevented by APD, with levels of urinary Dpd excretion even significantly lower than those found in sham-operated animals (p < 0.05; Table 4). Raloxifene also reduced the OVX-induced rise in serum osteocalcin (p < 0.01; Table 4). PTH, a known bone anabolic agent, caused a further, and significantly higher compared with OVX animals (p < 0.001), increase in osteocalcin levels and bore no effect on OVX-induced rises in the marker of bone resorption, Dpd.
Treatments for osteoporosis aim to increase bone strength and reduce the risk of fracture.(7) Two major approaches for correcting the negative effects of estrogen loss on fracture risk and bone strength after menopause have been widely adopted for the treatment of osteoporosis: anti-catabolic and anabolic therapies. Although both approaches have consistently proved therapeutically useful, their mechanisms of action are glaringly different, which led us to hypothesize that these different agents varyingly alter the different properties or determinants of bone strength even if their endpoint, a reduction in fracture risk, remains the same. In addition, although areal BMD measurements remain an established diagnostic tool for osteoporosis, improvement of BMD is insufficient when it comes to explaining the efficacy of an agent to reduce fracture risk.(27,28) We studied the ability of three currently available treatments, two inhibitors of bone resorption and one anabolic agent, to alter several different properties of bone strength, a prominent determinant of resistance to fracture, in 8-mo-old osteoporotic rats. Bone strength describes the greatest force, or load, which can be applied to a bone before a fracture occurs. Maximal load was directly measured using a compression test on the vertebral bodies. At the end of the study period, OVX induced a marked and significant decrease in maximal load compared with sham-operated and rats. It is known that the strength of the long bones and femoral neck is not significantly influenced by a loss of estrogen in rats, because the external diameter expands, compensating for the decrease in bone mass and deterioration of trabecular architecture.(21) This diameter expansion and subsequent preservation of bone strength is not observed in vertebral bodies, where the strength of the bone is not determined by the diameter but rather the trabecular bone microarchitecture, mineral content and density, intrinsic bone tissue quality, and other biomechanical parameters, as well as the rate of remodeling.(1,29–31)
The decrease in maximal load observed in OVX rats was prevented with APD treatment. APD treatment induced a large increase in the stiffness of the vertebral bodies, even in relation to the other treatments, although all three therapies significantly increased the vertebral mechanical stiffness compared with OVX. Our results also indicate that APD effectively showed an overall positive influence on trabecular microarchitecture in the rat vertebrae. Although APD did not significantly increase either cortical or trabecular tissue hardness compared with OVX, the values obtained from the APD-treated animals were not significantly different from sham-operated rats; in fact, the values were closer to sham values than OVX. It could thus be said that APD returned the values to levels close to that of control rats. It may also be possible that bisphosphonate treatment requires a longer period than 16 wk to induce more than modest effects on tissue hardness. Serum levels of osteocalcin were decreased in the APD-treated rats. It is well known that bisphosphonates reduce bone remodeling rates, preventing the hyperactive remodeling seen after the loss of estrogen in postmenopausal woman, with the primary action of osteoclast inhibition.(4) Human studies have also shown that an increase in the mean degree of mineralization of both compact and cancellous bone is observed after 2, 3,(32,33) and 5(34) yr of bisphosphonate treatment without affecting the size and orientation of the mineral particles. It is likely that the observed increases in the mean degree of mineralization are a product of decreased remodeling induced by the use of bisphosphonates. A decrease in activation frequency would allow for a longer secondary mineralization process, whereby more bone structural units reach a state of maximum mineralization and an overall increase in bone strength is thus observed.(32)
We also examined another inhibitor of bone resorption, raloxifene, a commonly prescribed SERM, known to reduce vertebral fracture risk and improve BMD in postmenopausal women.(35,36) The results showed similarities between raloxifene and APD in preventing the significant decreases in maximal load, bone mass, and microarchitecture after OVX. Whereas we found a significant increase in BMD, the result was rather moderate compared with the APD. This finding parallels observations from canine studies that showed a significant improvement of the mechanical properties at the material level after 1-year treatment with raloxifene despite no changes in aBMD, vBMD, BV/TV, or percent ash compared with the vehicle group.(37) This same group were able to show that these material-level improvements occurred at both cortical and trabecular bone sites.(38) In the clinical setting, a relatively mild effect of raloxifene on BMD has also been observed despite a very noteworthy vertebrae antifracture efficacy.(10,39) The observed improvements in microarchitecture may, at least in part, underlie the ability of raloxifene to increase bone strength and reduce fracture risk, yet it seems that the substantial beneficial effect on intrinsic bone material quality, both at the cortical and trabecular level, plays an important role in the mechanism by which this agent acts as a therapeutic tool in the treatment of osteoporosis. There were similarities between the abilities of the two anti-catabolics to substantially improve the material level properties, unlike PTH, yet the effects of raloxifene, as measured by nanoindentation, were by far the most pronounced. These results complement the mounting evidence that raloxifene enhances bone strength by improving tissue quality.(40) The mechanism by which raloxifene exerts its beneficial effect at the material level is not well understood; however, it has been hypothesized that raloxifene may alter the organic matrix,(38) and in particular, collagen, which is known to contribute to the biomechanical(41) and intrinsic properties of bone.(42) It should also be noted that SERMs are known to have a protective effect on osteocytes in vitro(43); however, no convincing evidence of this has been observed in vivo in humans.(44)
PTH induced the highest increase in maximal load of all treatments tested; in fact, maximal load was increased by almost 50% compared with OVX and significantly higher than in sham. The increases observed in trabecular bone volume, thickness, and number and decrease in spacing were remarkable for the PTH-treated group. As an anabolic agent, PTH improves trabecular microarchitecture by inducing bone formation to a greater extent, and earlier, than its effects on bone resorption are observable.(45) Previous studies, including in humans, have also reported noticeable improvements in bone microarchitecture as measured by μCT.(12,22,46,47) The mechanism by which PTH induces bone formation includes a stimulation of type I collagen protein expression. A recent report, however, provides evidence that there may be a lack of mature, mineralized matrix with the use of such an agent that acts to increase bone remodeling.(28) Isomerization of type I collagen is linked to decreased fracture risk, and although the exact mechanism of action is unclear, it has been suggested that this biochemical change is part of the maturation of the collagen molecules resting in the matrix and affects the degree of mineralization. PTH(1–84) was less effective than alendronate in stimulating the isomerization of type I collagen. It is possible that the increased remodeling induced by PTH caused an overall decrease in the maturation of bone collagen,(28) and this may be reflected by the decreased hardness found at the trabecular level. Such reports provide beneficial information regarding changes in the properties of the bone matrix caused by different anti-osteoporotic treatments independent of effects on BMD, and may, at least in part, explain the effect of PTH to improve maximal load, bone mass, and microarchitecture while somewhat diminishing modulus and hardness at the trabecular level in this study. It is interesting that we observed opposite effects of PTH on tissue hardness in cortical versus trabecular bone. It is well known that remodeling does not occur in rat cortical bone. This means very little new tissue is deposited and therefore no changes in hardness are detected, whereas the remodeling process that is stimulated by PTH in the trabecular bone leads to a substantial deposition of new bone, reducing the hardness of the bone until the tissue matures.
To our knowledge, this is the first report on the direct comparison of biomechanical, microstructural, and biochemical mechanisms of action by which three widely available treatments for osteoporosis increase bone strength. Furthermore, the results suggest strong mechanistic similarities within the therapeutic classes as summarized in Table 5. APD and raloxifene are both known to substantially reduce bone resorption by osteoclastic activity, and whereas the molecular specifics of each drug differs, the effects they have on the parameters observed in this study are relatively similar and certainly more closely related to each other than to PTH. Specifically, the two anti-catabolic treatments have a much greater positive influence on tissue at the material level, suggesting a greater degree of mineralization, whereas the antifracture efficacy of PTH seems to be largely caused by the substantial deposition of new bone, which perhaps lacks some mineral quality seen in animals treated with the other two tested drugs. Having said this, it is likely that these effects will be reversed after treatment is discontinued and the newly deposited bone matures. Determining an agent's ability to augment bone strength remains a major factor in its therapeutic utility for the treatment of osteoporosis.
The authors thank S. Clement and S. Vouillamoz for animal management and I. Badoud for biomechanical testing and technical assistance.
- 71998 Preclinical evaluation of new therapeutic agents for osteoporosis. In: MeunierPJ (ed.) Osteoporosis: Diagnosis and management. Martin Dunitz, London, UK, 257–273., ,
- 15for the Treatment of Osteoporosis with Parathyroid Hormone Study Group 2007 Effect of recombinant human parathyroid hormone (1–84) on vertebral fracture and bone mineral density in postmenopausal women with osteoporosis: A randomized trial. Ann Intern Med 146: 326–339., , , , , , , , ,
- 232002 Effects of raloxifene on bone density, biomarkers, and histomorphometric and biomechanical measures in ovariectomized cynomolgus monkeys. J North Am Menopause Soc 9: 320–328., , , , ,
- 322002 Effects of bisphosphonates on matrix mineralisation. J Musculoskelet Neuronal Interact 6: 538–543.,
- 392004 Alendronate produces greater effects than raloxifene on bone density and bone turnover in postmenopausal women with low bone density: Results of EFFECT (Efficacy of FOSAMAX versus EVISTA Comparison Trial) International. J Intern Med 255: 503–511., , , , , , ,