Overview, Vol 13, Issue 10

Only doubt is certain and disbelief worth believing.
Without this courage there can be no learning.
Believe nothing.

"The quarterly journal Progress in Osteoporosis began in October 1993 as Advances in Osteoporosis. Its purpose was to provide readers without easy access to the literature with summaries of the most important literature. We now inhabit a virtual world. Information is instantaneously accessible to all at the tap of a keyboard; understanding is not. In the spirit captured by the anonymous author*, the purpose of this publication is to provide critical evaluation of the most important literature and so to provoke discussion. It is our intention to promote dialogue which examines the quality of information published and so its credibility. The opinions expressed are my own and do not necessarily reflect those of the International Osteoporosis Foundation."

We invite readers to comment on and discuss this journal entry at the bottom of the page.

PDF Version

FREE DOWNLOAD of PIO Volume 13-10 Figures


Several advances have been made in therapeutics and the most interesting work presented at the 2013 ASBMR Meeting in Baltimore is summarized below. Further work of interest will be presented in the next issue of Progress in Osteoporosis. As always, the interpretation of the data is based on the contents of the abstracts and presentations and must remain tentative until the work has undergone peer review and can be scrutinized when it is in published so that having written “…all your piety nor wit shall lure it back to cancel half a line, nor all your tears blot out a word of it" (Rubaiyat of Omar Khayyam).

Modeling unopposed by resorptive remodeling

There is net loss of bone from the skeleton during aging because bone loss from the intracortical, endocortical, trabecular components of the inner (endosteal) envelope is greater than any periosteal bone formation taking place concurrently. However, if endosteal bone loss was prevented, any continued periosteal bone formation during advancing age will no longer be offset and the total mineralized bone matrix volume should increase.

Ominsky et al (1) examined fluorochrome labeling at 6, 12 and 16 months in proximal femur sections from OVX adult cynomologus monkeys treated with vehicle (n=20) or 25 mg/kg QM denosumab (DMAb) (n=14). Despite suppressed remodeling, femoral neck BMD increased from 5.9% at month 6 to 11.3% above baseline at month 16. One explanation for the continued increase in BMD is more complete secondary mineralization of bone matrix no longer resorbed because remodeling is suppressed. While this is plausible given that secondary mineralization takes several years to reach completion, the increase in BMD should become asymptotic and this does not seem to be the case in clinical studies in human subjects.

In this study of cynomologus monkeys, labeling upon trabeculae was low consistent with reduced remodeling upon trabecular surfaces but labels were detected upon the superior endocortical surface and the inferior periosteal surface. These labels occurred over smooth cement lines consistent with a modeling (not remodeling) dependent bone formation during denosumab treatment.

The authors do not suggest denosumab is an anabolic agent but rather speculate that that in the face of suppressed remodeling, existing slow bone modeling becomes detectable and might partly explain the continued increase in BMD at the hip in human subjects. Evidence for the occurrence of modeling dependent bone formation in humans is not strong and this may be because we have not explored this possibility sufficiently. It is also of interest to determine whether brief episodes of increased endogenous PTH associated with repeated v dosing may contribute to bone formation. In addition, a remaining question is what proportion of the increase in BMD is the result of more complete secondary mineralization.

Figure 1. Femoral neck from cynomologus monkeys treated with DMAb. Little trabecular labeling consistent with remodeling suppression. However, upon the superior endocortical surface and inferior periosteal surface labels are detected consistent with new bone deposition upon previously unremodeled bone. Reproduced from J Bone Miner Res 28 (Suppl 1) with permission of the American Society of Bone and Mineral Research.


Denosumab Reduces Cortical Porosity of the Proximal Femur

Zebaze et al (2) measured cortical porosity of the femoral neck in women who received placebo (n=22) or 60 mg DMAb (n=28) every 6 months from hip images obtained at baseline and year 3. Cortical porosity correlated positively with serum CTX (p=0.017) and negatively with hip strength estimated using finite element analysis (p=0.027). DMAb reduced porosity across the entire cortex and in each compartment producing a net treatment effect (DMAb–placebo) of -1.8% (inner transitional zone), -5.6% (outer transitional zone), and -7.9% (compact-appearing cortex) (all p<0.001).

The reduction in cortical porosity is partly the result of perturbation of surface level remodeling. When an antiresorptive is administered, the many remodeling sites excavated prior to treatment refill while fewer new resorption sites appear simultaneously. The effect is a net reduction in porosity and a net increase in BMD. Secondary mineralization of osteons formed by remodeling months to several years earlier is likely to contribute because secondary mineralization takes months to years to complete. Another possibility is age-related modeling based bone formation upon the intracortical and endocortical surfaces may contribute unopposed by continued remodeling (as discussed above). Whether the reduction in porosity explains some or most of the reduced fracture rates reported using denosumab is yet to be established.

Figure 2. Percent change from baseline after three years in the placebo and DMAb treated subjects. *Compared with baseline and placebo. Reproduced from J Bone Miner Res 28 (Suppl 1) with permission of the American Society of Bone and Mineral Research.







Combining Denosumab and Intermittent PTH

Leder et al (3) hypothesized that administration of PTH 1-34 may act primarily as a bone forming agent when given concurrently with DMAb. DMAb is a very powerful remodeling suppressant as it reduces the work and lifespan of existing osteoclasts and prevents the birth of new osteoclasts. If resorption is inhibited then remodeling sites that have completed their resorption phase and are in their formation phase may be stimulated by PTH. In addition, PTH may stimulate bone formation upon quiescent bone surfaces by promoting differentiation of lining cells while little, if any, of PTH action to increase remodeling would occur. Two studies in human subjects and a third in rats examined this interesting hypothesis.

In the study by Leder et al (3), 100 postmenopausal women were randomized to teriparatide (TPTD) (20 µg sc daily), DMAb (60 mg sc/6 months), or both. At 24 months, spine BMD (mean±SD) increased by 12.7±5.1% with combined treatment, by 9.5±5.9% with TPTD, and by 8.3±3.4% with DMAb. Femoral neck BMD also increased more using combined treatment (6.4±3.8%) than TPTD (2.8±3.6%, P=0.002) or DMAb (4.1±3.8%, P=0.028) alone. BMD at the one-third distal radius increased similarly in the DMAb (2.0±3.7%) and combined groups (2.4±3.1%) but decreased by 1.7±4.6% in the TPTD group (P<0.001).

Figure 3. Regional changes in BMD following 2 years TPTD, DMAb or both (see text for details). Reproduced from J Bone Miner Res 28 (Suppl 1) with permission of the American Society of Bone and Mineral Research.





From the same group of investigators, Tsai et al (4) report improvements in bone microarchitecture using combined therapy. At the tibia, DTot increased more using combined therapy (3.1±2.1%) than TPTD (0.0±2.3%) or DMAb (2.2±2.0%) alone. Ct.Th also increased more using combined therapy (5.4±3.8%) than using single therapy. The combination appeared to prevent the 5.6±10.3% increase in porosity with TPTD. Similar observations were reported at the distal radius. No between-group differences in trabecular microarchitecture were observed.

As reported in the studies in human subjects, Tokuyama et al (5) examined 3 month old OVX mice assigned to anti-RANKL monoclonal antibody (5 mg/kg injection), TPTD (80 mg/kg/d injection), or both. Combined therapy produced a greater increase in BMD at distal femur, shaft and spine than antibody alone. Cortical bone volume increased in combined and PTH groups compared with antibody use alone.

These studies support the notion that combining a powerful antiresorptive agent like a RANKL inhibitor may benefit bone structure. It remains feasible that the increase in porosity reported using PTH may be factitious. Newly deposited bone may be regarded as ‘porosity’ due to its lower tissue mineralization density. This new bone may be ‘seen’ as void or ‘nonbone’ during image analysis as voxels containing under mineralized attenuate photons less and the attenuation level may fall within the range nominated to be below the threshold for ‘bone’ and so may be designated as ‘porosity’. If inhibition of osteoclast function results in a net improved anabolic effect of PTH then it is interesting to speculate that combining odanocatib with PTH may function in a similar fashion as this drug appears to reduce the resorptive activity of osteoclasts perhaps without influencing bone formation adversely.

Prolonged Treatment With Denosumab
Are fracture rates reduced?

Papapoulos et al (6) report low fracture rates in subjects receiving 8 years 60 mg DMAb Q6M (3 years in FREEDOM and 5 years in extension n=1382) or the crossover group given 3 years placebo in FREEDOM then 5 years DMAb in the extension study (n=1296). Incidence of new vertebral and non-vertebral fracture remain low throughout the extension; hip fracture incidence during year 8 was 0.2% and 0.1% for the long-term and crossover groups, respectively. BMD increased by 18.5% (spine) and 8.2% (total hip) in the long-term group and by 13.8% (spine) and 4.8% (total hip) in the crossover group during 5 years.

The question is whether the low fracture rates are the result of treatment or sampling bias produced by inclusion of healthier compliers. As it is unethical to withhold therapy for 8 years, there was no control group, an understandable limitation but one that leaves this question unanswered. The continued increase in BMD at the spine may be the result of facet joint and intervertebral arthritic changes. The increase at the hip is unlikely to be confounded in this way (see above).

Figure 4. BMD continues to increase at the spine and total hip in the long term group and increased in parallel fashion in the cross-over group of the extension of the FREEDOM trial. Reproduced from J Bone Miner Res 28 (Suppl 1) with permission of the American Society of Bone and Mineral Research.








Denosumab in Men and Histomorphometry

In the phase 3 ADAMO trial in men with low BMD, 1 year of DMAb (60 mg/6 months) increased BMD and reduced serum C-terminal telopeptide. Dempster et al (7) reported 29 subjects (n=17 DMAb; n=12 placebo) participated in a bone biopsy substudy. Qualitative bone histology showed normally mineralized lamellar bone. Structural indices, including cancellous bone volume and trabecular number and surface, were similar; 12/17 (71%) DMAb-treated and 12/12 (100%) placebo-treated subjects had double labels in trabecular and/or cortical compartments. In 6 DMAb-treated and 12 placebo-treated subjects, static and dynamic remodeling indices were lower in DMAb-treated than placebo-treated subjects. The authors infer that treatment of men with low BMD with DMAb for one year resulted in qualitatively normal bone with reduced bone turnover.


Cathespin K inhibitors are potentially important additions to therapeutic options for osteoporosis. The results of the pivotal antifracture efficacy study are not yet available but independent review suggests the relevant antifracture endpoints have been met. The remaining question is whether this class of drug is safe. Time will tell. These agents prevent collagen degradation and so prevent bone resorption. They also may have beneficial effects on bone formation as osteoclast mediated bone resorption is inhibited but osteoclast numbers are normal or increased and the cells may participate in resorption-formation coupling at the level of the basic multicellular unit by producing local factors permissive for bone formation (8).

Chen et al (9) either sham-operated (n=20) or orchidectomized (ORX) (n=24/group) mature male rabbits. After 7.5 months, ORX-animals had reduced lumbar vertebral areal BMD (LV aBMD) by 5% vs. sham. ORX rabbits were randomized to vehicle (Veh), alendronate (ALN) (300 mg/kg/wk, sc) or odanacatib (ODN) (1.5 and 6 mg/kg) for 14 months. LV.BMD increased in ORX-rabbits treated with ODN 1.5 mg/kg by 9-19% and 6 mg/kg by 6-27% vs. Veh, and compared to an increase by 9% in the ALN group. Persistent reductions of the bone resorption marker helical peptide of collagen type I (HP) were observed in animals treated with each ODN dose (43-71% and 47-62%), compared to ALN (24-50%) vs. Veh. Bone specific alkaline phosphatase was reduced (16-34%) in the ALN group, maintained with ODN 1.5 mg/kg and trended up (16-30%) with ODN 6 mg/kg compared to Veh. Veh-treated ORX rabbits had decreased BV/TV by 19% vs. sham. ALN tended to increase BV/TV by 13% (p=0.124) vs. Veh. ODN at both doses increased BV/TV by 55-57% (p<0.001) vs. Veh, and 37-39% above ALN. ODN reduced lumbar spine MS/BS by 41-44%, BFR/BS by 51-52% and MAR by10 15%. Compared to ALN and Veh, trabecular osteoclast surface and number were higher in both ODN groups. Interpretation is difficult without comparisons versus baseline or a sham-operated group because the Veh-treated OVX group lose bone.

In the same study, Pennypacker et al (10) evaluated the effects of ODN on bone quality of the lumbar vertebrae (LV) and femur of male rabbits (age 11 months) that were sham-operated (N=20) or ORX (N=24/group) 7.5 months prior to treatment. ORX rabbits were randomized to Veh, ODN (1.5 or 6 mg/kg/d), or ALN (300 mg/kg/wk, sc) for 14 months. ORX resulted in ~9% bone loss in LV3-4 aBMD vs. sham (p<0.01). Relative to Veh, ODN increased LV aBMD by 21% and 26% for each respective dose (p<0.001), and 9.9% compared to ALN (p<0.01). pQCT-based LV5-6 trabecular (Tb) vBMC and vBMD increased in ODN groups (up to 88%, p<0.001) relative to vehicle. Similarly, ALN increased Tb.vBMC and BMD by 11% and 8%, respectively (both p< 0.05). ODN at both doses increased Tb thickness (32-51%, p<0.001). Significant increases in peak load, apparent strength, yield load, and yield stress, stiffness of LV5-6 were noted between ALN to Veh and ODN to Veh or sham. From compression testing of LV, peak load from all groups correlated with pQCT-vBMC (R=0.9383, p<0.001). ODN also dose-dependently increased aBMD of the total hip and distal femur, but did not change central femur (CF) aBMD compared to Veh. ODN 6 mg/kg increased cortical thickness and area by ~7% (p<0.01), and reduced endocortical perimeter by 4% (p<0.05) vs. Veh. ODN-treated CF tended to have higher peak load and increases in work-to-failure (25%, p< 0.05) and toughness (29%, p<0.01) vs. Veh. No changes in biomechanical parameters of CF were noted in the ALN group. CF peak load vs. pQCT-vBMC correlated (R=0.7555, p<0.001). ODN increased bone mass and strength in the lumbar spine to levels above Veh and sham, as well as maintaining biomechanical properties at femoral sites of ORX male rabbits.

Cusick et al (11) studied the effects of ODN on fracture repair in a radial osteotomy model during reparative (6 weeks) and remodeling (25weeks) phases. Rabbits were randomized and pretreated for 12 weeks then osteotomy was performed. Treatment was Veh, ODN (1.5 or 6 mg/kg/d) or ALN (0.3 mg/kg/wk, sc). Postsurgery, ODN was continued (ODN/ODN) or discontinued (ODN/Veh). Fracture callus in ODN/ODN, ODN/Veh, ALN groups were no different to Veh. ODN/ODN and ALN increased pQCT-based total vBMD at 6 weeks (9-10%) and 25weeks (26-33%) without altering callus area vs. Veh. ODN/ODN enhanced callus vBMC by 15-18% at 6 weeks and 49-69% at 25 weeks vs. Veh. Mature callus vBMC with ALN was unaltered at 6 weeks and increased 9-18 at 25 weeks vs. Veh. Biomechanical tests of fractured and intact radii in all groups at 6 weeks were not different to Veh. At 25 weeks, peak load across the healing sites in ODN/Veh and ALN groups were the same as that in Veh, peak load in ODN/ODN groups increased by 18% and 25% for each dose, reaching intact contralateral levels. The authors infer that ODN does not delay fracture union and callus remodeling.

Fujii and Tanaka (12) synthesized peptidomimetic compounds which inhibit cathepsin K. SI-591 inhibited human cathepsin K with an IC50 of 1.4 nM, 8- to 320-fold more selective for cathepsin K than other human cathepsins. In vitro, osteoclast-like cells resorption pits formed on the slices. SI-591 (0.01-10 mM) dose dependently inhibited the release of CTX-I and decreased excavation of pits. In OVX adult female F344 rats, SI-591 dose-dependently inhibited urinary CTX-I release and prevented BMD loss at the vertebrae and femur. OVX rats were treated with vehicle or SI-591 (4-25 mg/kg p.o., b.i.d.) for 12 weeks from 12 weeks following surgery. SI-591 dose dependently inhibited urinary deoxypyridinoline release and prevented BMD loss in the vertebrae and femur. In OVX cynomolgus monkeys (aged 9-16 years) treated with a vehicle or SI-591 (4-36 mg/kg p.o. b.i.d.) for 6 months following surgery, SI-591 inhibited the NTX release and prevented BMD loss.


1. Ominsky M, Boyce RG, Kosteniuk P, et al. Nf1 Continuous modeling-based bone formation: A novel mechanism that could explain the sustained increases in hip bone mineral density (BMD) with denosumab treatment. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=89fd1f4a-ff73-4a62-a.... Accessed November 15, 2013.

2. Zebaze RM, Libanati C, McClung MR, et al. Denosumab reduces hip cortical porosity in women with osteoporosis. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=d383c95f-4672-4dc0-8.... Accessed November 15, 2013.

3. Leder B, Uihlein A, Tsai J, et al. The DATA Extension Study: 2 years of combined denosumab and teriparatide in postmenopausal women with osteoporosis: A randomized controlled trial. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=c6aff623-b458-4737-8.... Accessed November 15, 2013.

4. Tsai J, Uihlein A, Zhu Y, et al. Comparative effects of teriparatide, denosumab, and combination therapy on peripheral compartmental bone density and microarchitecture: The DATA-HRpQCT Study. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=3ec3b6d6-5992-4935-8.... Accessed November 15, 2013.

5. Tokuyama N, Masuda H, Hirose H, et al. Individual and combining effects of anti-RANKL monoclonal antibody and teriparatide in ovariectomized mice. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=bf5558ff-c0f9-4b8d-9.... Accessed November 15, 2013.

6. Papapoulos S, Lippuner K, Roux C, et al. Eight years of denosumab treatment in postmenopausal women with osteoporosis: Results from the first five years of the FREEDOM Extension. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=4e003dea-3d2d-4ae5-9.... Accessed November 15, 2013.

7. Dempster D, Kendler DL, Hall J, et al. Effect of denosumab treatment on bone histology and histomorphometry in men with low bone mineral density. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=83536378-0e27-431e-9.... Accessed November 15, 2013.

8. Lotinun S, Kiviranta R, Matsubara T, et al. Osteoclast-specific cathepsin K deletion stimulates S1P-dependent bone formation. J Clin Invet 2013;123:666.

9. Chen C, Pennypacker B, Belfast M, Duong Ll. Efficacy of odanacatib versus alendronate in the treatment of bone loss in orchidectomized male rabbits. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=bed76f94-88dd-46df-8.... Accessed November 15, 2013.

10. Pennypacker B, Chen C, Cusick T, Samadam R, Duong L. Treatment with odanacatib increases bone mass and maintains normal biomechanical properties in orchidectomized male rabbits. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=c864c4d1-5659-45a6-b.... Accessed November 15, 2013.

11. Cusick T, Samadfam R, Duong L. Effects of odanacatib on early and late stage fracture healing in an adult rabbit radial osteotomy model. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=1f011475-ff42-432d-b.... Accessed November 15, 2013.

12. Fujii T, Tanaka Y. Introduction of a new class of cathepsin K inhibitor with peptidomimetic structure, SI-591. J Bone Miner Res 2013;28 (Suppl 1). Available at http://www.asbmr.org/asbmr-2013-abstract-detail?aid=3a2ddee4-20c0-4bb8-9.... Accessed November 15, 2013.