SEARCH

SEARCH BY CITATION

Keywords:

  • sclerostin;
  • antibody;
  • bone formation;
  • bone strength;
  • primate

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

The development of bone-rebuilding anabolic agents for treating bone-related conditions has been a long-standing goal. Genetic studies in humans and mice have shown that the secreted protein sclerostin is a key negative regulator of bone formation. More recently, administration of sclerostin-neutralizing monoclonal antibodies in rodent studies has shown that pharmacologic inhibition of sclerostin results in increased bone formation, bone mass, and bone strength. To explore the effects of sclerostin inhibition in primates, we administered a humanized sclerostin-neutralizing monoclonal antibody (Scl-AbIV) to gonad-intact female cynomolgus monkeys. Two once-monthly subcutaneous injections of Scl-AbIV were administered at three dose levels (3, 10, and 30 mg/kg), with study termination at 2 months. Scl-AbIV treatment had clear anabolic effects, with marked dose-dependent increases in bone formation on trabecular, periosteal, endocortical, and intracortical surfaces. Bone densitometry showed that the increases in bone formation with Scl-AbIV treatment resulted in significant increases in bone mineral content (BMC) and/or bone mineral density (BMD) at several skeletal sites (ie, femoral neck, radial metaphysis, and tibial metaphysis). These increases, expressed as percent changes from baseline were 11 to 29 percentage points higher than those found in the vehicle-treated group. Additionally, significant increases in trabecular thickness and bone strength were found at the lumbar vertebrae in the highest-dose group. Taken together, the marked bone-building effects achieved in this short-term monkey study suggest that sclerostin inhibition represents a promising new therapeutic approach for medical conditions where increases in bone formation might be desirable, such as in fracture healing and osteoporosis. © 2010 American Society for Bone and Mineral Research


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Bone-related conditions cause significant morbidity worldwide, particularly among the elderly, and include systemic bone loss (eg, postmenopausal osteoporosis), focal bone loss, and traumatic fractures.1 Bone is continuously remodeled, with osteoclasts resorbing bone and osteoblasts forming new bone. Various therapeutics are currently used for the treatment of low bone mass, with the vast majority being antiresorptive agents (eg, bisphosphonates) that act by decreasing the rate of osteoclast-mediated bone resorption, thereby preventing further bone loss and skeletal weakening.2, 3 However, because these agents cannot replace bone that has been lost, there has been a long-standing goal to develop therapeutics that can stimulate bone formation to increase bone mass and bone strength. It is thought that such bone-rebuilding anabolics could provide important treatment options not only for low-bone-mass conditions but also for fracture healing and other orthopedic applications. Currently, the only approved bone anabolic agents for osteoporosis are full-length and truncated parathyroid hormone (PTH), both of which are administered by daily subcutaneous injection.2, 3

The human high-bone-mass genetic disease sclerosteosis is caused by lifelong absence of the osteocyte-secreted protein sclerostin and is characterized by increased osteoblast-mediated bone formation.4–8 Similar to the human condition, sclerostin knockout mice have robust increases in bone formation, bone mass, and bone strength.9 In ovariectomized (OVX) rats, a model of postmenopausal osteoporosis, short-term administration of a sclerostin-neutralizing monoclonal antibody (mAb) had strong anabolic effects with marked increases in bone formation on trabecular, periosteal, endocortical, and intracortical surfaces.10 Bone mass and bone strength were increased to levels that exceeded those of age-matched non-OVX control rats, suggesting that antibody-mediated sclerostin inhibition might be a viable approach for the treatment of bone-related disorders. Furthermore, the powerful anabolic response seen in these aged OVX rats showed that sclerostin plays a pivotal role in the regulation of bone formation even during later stages of life when the incidence of bone-related disorders is highest.

In a mouse model of colitis, short-term treatment with a sclerostin mAb increased bone formation and bone strength, thereby countering the effects of accelerated bone loss caused by chronic inflammation.11 The results of that study showed that inhibition of sclerostin has beneficial effects on bone in the physiologic setting of inflammation-induced bone loss. In rodent models of fracture healing, sclerostin mAb treatment resulted in increased callus density and bone strength at fracture sites and accelerated bone repair.12

Nonhuman primates are considered to be the most appropriate species to approximate human bone biology owing to similarities in cortical bone remodeling and response to estrogen withdrawal.13–15 To further explore the clinical potential of sclerostin inhibition, we treated cynomolgus monkeys with a humanized sclerostin-neutralizing mAb (Scl-AbIV). Our broad aim in this study was to investigate the effects of antibody-mediated sclerostin inhibition in monkeys using a dosing interval and dose levels similar to what generally might be tested in early human trials for antibody therapeutics. To this end, the effects of once-monthly dosing of Scl-AbIV on serum bone turnover biomarkers, bone density, bone histomorphometry, and bone strength were examined in adolescent gonad-intact female monkeys for 2 months.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Animals, treatment, and experimental design

Female cynomolgus monkeys (Macaca fascicularis) aged 3 to 5 years (mean 3.9 years) were received from Primate Products, Inc. (Miami, FL, USA). Animals were socially housed in wall-mounted cages (2 to 3 animals/cage) equipped with an automatic watering system. All animals had access to 2050C Certified Global Primate Diet (PMI Nutrition International, Inc.) twice daily, containing 0.93% calcium, 0.75% phosphorus, and 8.0 IU of vitamin D/g, as well as daily food supplements including fresh fruit. The animal room environment was controlled, with settings targeted at temperature 24 ± 3°C, humidity 50 ± 20%, photoperiod 12 hours of light and 12 hours of dark, and 12 air changes per hour.

An acclimation period of 6 weeks was allowed between animal receipt and the start of Scl-AbIV administration. Only animals considered in good health, with minimal skeletal abnormalities (by radiograph), and normal serum/urine chemistry panels were used in the study. Vehicle (PBS, n = 4) or a sclerostin-neutralizing monoclonal antibody (Scl-AbIV) at 3 (n = 2), 10 (n = 3), or 30 mg/kg (n = 3) was administered subcutaneously in a volume of 3 mL/kg on days 1 and 29. These groups were balanced by baseline body weight and confirmed to have similar mean baseline lumbar bone mineral density (BMD) and content (BMC) by dual-energy X-ray absorptiometry (DXA). The study was terminated on day 61 when bones were harvested for analysis. Scl-AbIV is a high-affinity (KD less than 15 pM against cynomolgus sclerostin) humanized IgG2 mAb produced in Chinese hamster ovary (CHO) cells. The study was performed at Charles River Laboratories Montreal, Senneville, Quebec, Canada. Protocols and procedures involving animals were conducted in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) accredited facility in accordance with the requirements and guidelines of the US National Research Council and the Canadian Council on Animal Care and complied with the protocols approved by the Charles River Montreal Institutional Animal Care and Use Committee. All endpoints were collected at Charles River Laboratories Montreal except for the serum ELISAs for Scl-AbIV and Crosslaps, which were performed at Amgen. Inc.

Serum levels of sclerostin antibody and biochemical markers of bone turnover

Blood samples were collected prior to dosing on days 1 and 29 and 0.5, 1, 2, 4, 6, 13, 20, and 27 days after each dose. Animals were food-deprived overnight (approximately 12 hours) prior to blood sampling on all occasions, except for predose on day 1 and day 29 and for the 0.5 day postdose bleeds, where food deprivation was for approximately 7 hours. Scl-AbIV concentrations were measured by sandwich ELISA in all treated serum samples. The bone-formation markers osteocalcin (OC) and intact N-terminal propeptide of type 1 procollagen (P1NP) were measured in all serum samples by radioimmunoassay (OC: DSL-6900, Diagnostic System Laboratoriescountry; P1NP: Intact P1NP, Orion Diagnostica). The bone-resorption marker serum C-telopeptide (CTx) was quantified in all serum samples using a Serum Crosslaps ELISA (Nordic Biosciences).

Histomorphometry

Double fluorochrome labels were administered intravenously on days 14 and 24 (25 mg/kg of tetracycline) and on days 47 and 57 (8 mg/kg of calcein). The L2 vertebrae, right proximal tibiae, and right femoral diaphyses were partially trimmed and fixed overnight in neutral buffered 10% formalin, transferred into 70% alcohol, dehydrated, and embedded in methyl methacrylate. For trabecular bone evaluation, sections were prepared from the proximal tibia along the craniofrontal plane and from the body of the second lumbar vertebra (L2) along the median plane. Static parameters were determined from 5-µm-thick sections stained with Goldner's trichrome, whereas dynamic parameters were assessed in 7-µm-thick unstained sections. Cortical bone histomorphometric parameters were determined from 20- to 40-µm-thick ground, unstained sections. The dynamic bone-formation parameters, mineralizing surface (MS; calculated as double-label surface + half the single-label surface), and bone-formation rate (BFR) were generated for both sets of labels and normalized to bone surface (BS) measured at the study end; mineral apposition rate (MAR) was calculated as the mean separation between each set of labels divided by the label interval (10 days). These dynamic parameters were generated on trabecular, periosteal, endocortical, and intracortical surfaces. Static endpoints in trabecular regions included bone volume as a percent of total volume (BV/TV), trabecular thickness, osteoclast surface, and osteoblast surface.

Bone densitometry by DXA and pQCT

Prior to scanning, monkeys were first sedated by injection with a cocktail of glycopyrrolate, ketamine hydrochloric acid (HCl), and xylazine, followed by anesthesia with isoflurane gas prior to and during scanning. Areal BMC and BMD of the lumbar spine (L1–L4), right proximal femur (including femoral neck subregion), and right distal radius (one-third and ultradistal) were measured using DXA (Hologic QDR 2000+) at baseline (predose 1) and on days 26 and 61. The precision [coefficient of variation (CV, %)] of DXA scanning with repositioning ranged from 0.8% at the lumbar spine to 4.5% at the femoral neck.

Peripheral quantitative computed tomography (pQCT; Stratec XCT Research SA scanner, Software Version.5.40B) was used to measure volumetric bone mineral content (vBMC), volumetric BMD (vBMD), and geometric parameters of the right distal radius and right proximal tibia at baseline and on days 26 and 61. Metaphyseal data were generated as an average from three scans separated by 0.5 mm beginning at 3% of the length for the distal radius and at the tibial-fibular junction for the proximal tibia (contour mode 2, peelmode 20, trabecular area: 30% for radius, 40% for tibia). A diaphyseal scan was taken at approximately 15% of the bone length for the distal radius and 12% for the proximal tibia (peelmode 2, cortmode 2: threshold 0.930 cm−1). Nominal voxel size was 0.35 mm at the proximal tibia and 0.2 mm at the distal radius. The precision (CV, %) of pQCT scanning with repositioning was 0.2% to 1.1% at the proximal tibia and 0.2% to 4.0% at the distal radius. Cross-sectional moment of inertia (CSMI) for in vivo diaphyseal scans was calculated from periosteal and endocortical circumferences based on a circular geometry assumption.

Biomechanical testing

Lumbar vertebrae (L3–L4) and left femurs were stored at –20 °C prior to biomechanical testing. Vertebral endplates and spinous processes were removed to obtain a specimen with planoparallel ends using a diamond saw. Bone densitometry scans were performed prior to biomechanical testing in the vertebral midplane and femoral diaphysis by pQCT, as described earlier. Nominal voxel size for pQCT was reduced to 0.15 mm for the femur and 0.2 mm for L3–L4 vertebral bodies.

Destructive strength testing was performed using an MTS 858 Mini Bionix servo-hydraulic test system (MTS Systems Corporation) with data collected using Testworks (Version.3.8A) for Teststar (Version.4.0c) software (MTS Systems Corporation). The femoral diaphysis was tested to failure in three-point bending (displacement rate 1 mm/s). L3 and L4 vertebral samples were tested to failure in compression (displacement rate 20 mm/min), with data reported as an average of both vertebral tests. Peak load was recorded as the maximum of the load-displacement curve, and stiffness was the slope of the linear portion. Energy to failure was calculated as the area under the curve to the breaking point for the three-point bending tests and to peak load for the compression tests. Ultimate stress, modulus, and toughness were calculated from the bending tests based on pQCT-based cortical geometry data, as described previously.16

Statistical analysis

Results are expressed as the mean ± SE. Statistical analyses were conducted using Release 9.1 of SAS/STAT (SAS Institute Inc.), and each statistical test was conducted at the 0.05 level of significance. A one-way ANOVA was applied to each data set: change from baseline for biomarker data, percent change from baseline for in vivo DXA/pQCT data, and raw data for histomorphometric and biomechanical endpoints. The Brown and Forsythe variation of Levene's test (Brown and Forsythe, 1974) was applied to confirm the variance homogeneity among groups. The Dunnett multiple-comparison procedure was used to compare each dose level of Scl-AbIV to the vehicle control group. For biomechanical parameters, baseline lumbar BMC was used as a covariate in the analyses. Linear regression analyses were performed within GraphPad PrismVersion 5.01 (GraphPad Software Inc.) across all groups.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Effect of sclerostin antibody treatment on biochemical markers of bone turnover

To determine the effects of sclerostin inhibition in nonhuman primates, a sclerostin-neutralizing mAb, Scl-AbIV, was administered to intact adolescent female cynomolgus monkeys. Once-monthly injections of Scl-AbIV (ie, two injections, day 1 and day 29) at three dose levels (3, 10, and 30 mg/kg) were given, and the study was terminated at 2 months (day 61). At all doses used, Scl-AbIV did not alter body weight gain or general health. Peak Scl-AbIV concentrations at each dose level were found to occur within the first week following dosing, after which mAb levels declined. For the 3 mg/kg dose group, Scl-AbIV was at undetectable levels (<10 ng/mL) at both the 4-week postdosing timepoints (Fig. 1A). The bimodal serum Scl-AbIV concentration profile over 2 months resulted in the observed bimodal increases in the serum bone-formation markers P1NP and osteocalcin. Changes in serum P1NP peaked 2 weeks after the first injection and 1 week after the second injection, with significant increases found in all three dose groups (Fig. 1B). The serum osteocalcin profile was shifted about a week later than the P1NP profile, with the highest percent increase being found in the 30 mg/kg dose group (Fig. 1C). The peak changes in P1NP and osteocalcin were in excess of 50% above baseline values. By 4 weeks after the first and second Scl-AbIV doses, serum P1NP and osteocalcin had returned to baseline levels, consistent with the observed clearance of Scl-AbIV from the circulation over time. No clear effect on the serum bone-resorption marker CTx was found with any dose of Scl-AbIV, although the 10 mg/kg dose group was significantly lower than the vehicle control group 4 days after the first dose (Fig. 1D).

thumbnail image

Figure 1. Increases in serum markers of bone formation (ie, P1NP and osteocalcin) following administration of sclerostin monoclonal antibody Scl-AbIV. Cynomolgus monkeys were injected subcutaneously on days 1 and 29 with vehicle (VEH) or Scl-AbIV at 3, 10, or 30 mg/kg, and serum was collected at various time points. (A) Serum concentration of Scl-AbIV. Serum levels, as change from baseline, for (B) P1NP, (C) osteocalcin, and (D) the bone-resorption marker CTx. Data represents mean ± SE; n = 2 to 4 per group. *p < .05 versus vehicle control.

Download figure to PowerPoint

Histomorphometry

To measure bone formation following each of the two once-monthly administrations of Scl-AbIV (dosed on days 1 and 29), newly mineralizing bone was labeled in vivo with tetracycline (days 14 and 24) and calcein (days 47 and 57). Bones were harvested at study termination on day 61. Histomorphometric analysis was performed at trabecular (L2 vertebra, proximal tibia) and cortical bone sites (femoral diaphysis). Extensive labeling of bone surfaces showed that Scl-AbIV markedly stimulated bone formation in both the trabecular and cortical bone compartments. Even with this increased rate of mineralized bone deposition, the newly formed bone was found to be lamellar in appearance, and no osteoid accumulation or marrow fibrosis was observed.

Qualitatively, at the highest dose (30 mg/kg), Scl-AbIV treatment resulted in increased trabecular thickness, greater labeled surface, and greater separation between labels compared with vehicle controls, in which few regions with all four labels were identifiable (Fig. 2A–H). Quantitatively, Scl-AbIV dose-dependently increased mineralizing surface (MS/BS), mineral apposition rate (MAR), and bone-formation rate (BFR/BS). Specifically, MS/BS was increased significantly at the 30 mg/kg dose in the first month of dosing at the L2 vertebra (+345%; Fig. 2K) and in the second month of dosing at the proximal tibia (+262%; Fig. 2L) compared with vehicle. MAR and BFR/BS were increased significantly at the 30 mg/kg dose during both months at both trabecular sites (+28% to 57% for MAR, +124% to 490% for BFR/BS; Fig. 2M–P). The increased bone formation with Scl-AbIV treatment in trabecular bone was associated with significant dose-dependent increases in trabecular thickness (Fig. 2J) and nonsignificant but dose-dependent increases in trabecular bone volume (Fig. 2I). Trabecular tunneling was not observed in sections from Scl-AbIV-treated cynomolgus monkeys, and trabecular number remained unchanged at both trabecular sites (data not shown). Wall thickness was nonsignificantly greater (20% to 38%) in both the proximal tibia and the L2 vertebra after treatment with 10 or 30 mg/kg compared with vehicle controls (data not shown). At the end of the study, osteoblast surface on trabecular bone was not significantly different for Scl-AbIV treatment versus control. This was the case for osteoblast surface both at the proximal tibia (4.4% ± 1.8% for 30 mg/kg Scl-AbIV versus 4.2% ± 1.9% for vehicle) and at the L2 vertebra (0.53% ± 0.17% for 30 mg/kg Scl-AbIV versus 1.52% ± 0.45% for vehicle). This finding for osteoblast surface was consistent with the declines noted for P1NP and osteocalcin and showed that the osteoblast stimulating effect of Scl-AbIV waned as the antibody was cleared from the circulation. Similar to the results for osteoblast surface, no significant difference was found at study end for osteoclast surface on trabecular bone for Scl-AbIV treatment versus control (data not shown).

thumbnail image

Figure 2. Scl-AbIV administration increases trabecular bone volume and bone formation. Trabecular regions in the L2 vertebrae and right proximal tibias from vehicle- and Scl-AbIV-treated monkeys were analyzed for static and dynamic histomorphometry. Goldner's trichrome–stained sections in the top panel (4× objective) illustrate the increases in trabecular bone volume (blue) and trabecular thickness for the 30 mg/kg Scl-AbIV dosing (B, D) versus vehicle (A, C) in the L2 vertebra and proximal tibia at study termination (2 months). Fluorescent microscopic images in panels E–H (20× objective) show the surfaces where new bone was forming when the tetracycline (orange; injected on days 14 and 24) and calcein (green; injected on days 47 and 57) labels were administered. Compared with vehicle-treated controls (E, G), Scl-AbIV (F, H) increased the amount of labeled surfaces as well the distance between labels at both skeletal sites. Quantitative histomorphometric analyses show dose-dependent increases in (I) trabecular bone volume (BV/TV) and (J) trabecular thickness after 2 months of Scl-AbIV treatment. These changes were associated with dose-dependent increases in (K, L) mineralizing surface (MS/BS), (M, N) mineral apposition rate (MAR), and (O, P) bone-formation rate (BFR/BS) at the L2 vertebra and proximal tibia. D14/24 = day 14 and 24 tetracycline labels; D47/57 = day 47 and 57 calcein labels. Data represent mean ± SE; n = 2 to 4 per group. *p < .05 versus vehicle control.

Download figure to PowerPoint

The femoral diaphyseal cortex contained more calcein and tetracycline labels on the endocortical and periosteal surfaces after treatment with 30 mg/kg Scl-AbIV compared with vehicle controls (Fig. 3A). Histomorphometric analysis showed dose-dependent increases in MS/BS, MAR, and BFR/BS for both surfaces after each dose, with greater percent increases in MS/BS and BFR/BS found for the endocortical surface. Specifically, at the 30 mg/kg dose, there were significant increases in endocortical MS/BS (+706% to 748%) and BFR/BS (+895% to 1013%; Fig. 3C, G), as well as nonsignificant increases in periosteal MS/BS (+72% to 122%) and BFR/BS (+171% to 233%; Fig. 3B, F). Of note, the absolute magnitudes of MS/BS and BFR/BS were similar for the periosteal and endocortical surfaces when compared across each of the three Scl-AbIV dose levels. The greater treatment-related percent increases found for MS/BS and BFR/BS at the endocortical surface compared with the periosteal surface were due to the relatively low absolute values for these parameters at the endocortical surface in vehicle-treated control animals. For MAR, the dose-dependent increases observed were similar for the periosteal and endocortical surfaces. Analysis of the intracortical region of the femoral diaphysis at 2 months showed that cortical porosity, as a percentage of bone volume, was unchanged at the 3 and 10 mg/kg doses and slightly elevated in the 30 mg/kg group compared with the low levels found in controls (1.92% ± 0.04% versus 0.96% ± 0.18%; p < .05). Bone formation on intracortical surfaces was increased significantly during month 2 at the 10 and 30 mg/kg doses (+393% and +465%, respectively; data not shown).

thumbnail image

Figure 3. Scl-AbIV administration increases cortical bone formation at the femur midshaft. (A) Fluorescent microscopy images of the femur midshaft illustrate the Scl-AbIV-mediated increase in tetracycline (orange; days 14 and 24) and calcein labels (green; days 47 and 57) on the endocortical and periosteal surfaces. The white boxes in the whole-cortex images are magnified 5× in the panels to the right. Quantitative data for the periosteal surface are in panel B, D and F and for the endocortical surface in C, E and G as noted on the y axes. Quantitative histomorphometric analyses show dose-dependent increases in (B, C) mineralizing surface (MS/BS), (D, E) mineral apposition rate (MAR), and (F, G) bone-formation rate (BFR/BS) at both the periosteal and endocortical surfaces. Data represent mean ± SE; n = 2 to 4 per group. *p < .05 versus vehicle control.

Download figure to PowerPoint

Serum bone-formation-marker concentrations were associated with local bone-formation rates at trabecular (proximal tibia) and endocortical surfaces (femoral diaphysis). Month 1 BFR/BS was significantly and positively correlated with P1NP and osteocalcin on days 13 and 20; month 2 BFR/BS was significantly and positively correlated with P1NP and osteocalcin on days 13, 20, 35, and 42 and day 48 osteocalcin (day 42 correlations are shown in Fig. 4).

thumbnail image

Figure 4. Serum bone-formation marker levels reflect local bone-formation rates in trabecular and cortical sites. Linear regression analyses were performed across vehicle and Scl-AbIV (3, 10, and 30 mg/kg) treatment groups between the day-42 determination of serum levels of bone-formation biomarkers and histomorphometric bone-formation rates (BFR/BS). BFR/BS represents bone-formation rates measured from the calcein labels administered on days 47 and 57. (A) P1NP versus BFR/BS at the trabecular surface of the proximal tibia. (B) P1NP versus BFR/BS at the endocortical surface of the femoral midshaft. (C) Osteocalcin versus BFR/BS at the trabecular surface of the proximal tibia. (D) Osteocalcin versus BFR/BS at the endocortical surface of the femoral midshaft. The coefficients of determination (r2) are indicated on each graph; all the dotted regression lines had slopes that were significantly different from zero (p < .05).

Download figure to PowerPoint

In vivo densitometry

Two injections of Scl-AbIV at the 30 mg/kg dose in adolescent cynomolgus monkeys resulted in significant increases, as percent change from baseline after 2 months, in areal BMC for the whole body (24% versus 6.4% for vehicle) and femoral neck (35.2% versus 5.4% for vehicle; Table 1). Nonsignificant increases, which were greatest in the high-dose Scl-AbIV group, also were found by DXA at the lumbar spine and distal radius for both areal BMC and areal BMD. pQCT analysis (Table 2) showed dose-related increases that attained significance in the 30 mg/kg dose group at the metaphyses of the distal radius and proximal tibia for both total volumetric bone mineral content (vBMC; radius: 19.8% versus 0.8% for vehicle; tibia: 27.3% versus −1.0% for vehicle) and volumetric BMD (vBMD; radius: 14.2% versus 2.6% for vehicle; tibia: 18.8% versus 2.9% for vehicle). In the trabecular subregion of the metaphysis, dose-related increases in vBMD were seen at both sites, reaching significance at the 30 mg/kg dose in the proximal tibia (34.9% versus −1.7% for vehicle). Cortical vBMC after 2 months was nonsignificantly increased with each Scl-AbIV dose compared with controls for the diaphysis of both radius and tibia. Mean cortical thickness, cortical area, and cross-sectional moment of inertia (CSMI) were increased at the radial and tibial diaphyses, with greater than 9% increases above baseline at the tibia after 2 months in the 10 and 30 mg/kg Scl-AbIV groups. For both sites, these changes were associated with consistent, dose-dependent increases in periosteal circumference. No consistent treatment-related changes were found for endocortical circumference and cortical vBMD at either site.

Table 1. Percent Change in DXA BMD and BMC in Cynomolgus Monkeys After 2 Months of Scl-AbIV Treatment
DXA ENDPOINT (% Change from Baseline)VEHScl-AbIV
3 mg/kg10 mg/kg30 mg/kg
  1. Data expressed as mean ± SE for percent change versus baseline after 2 months. *Significantly different from VEH; p < .05.

Whole Body BMD1.6 ± 3.04.4 ± 5.410.8 ± 3.29.4 ± 2.8
Whole Body BMC6.4 ± 3.05.8 ± 6.219.2 ± 6.724.0 ± 2.2 *
Lumbar Spine BMD1.7 ± 1.89.8 ± 1.44.2 ± 3.811.1 ± 3.0
Lumbar Spine BMC2.8 ± 2.715.0 ± 0.38.1 ± 6.616.5 ± 6.2
Femoral Neck BMD4.6 ± 1.810.2 ± 10.911.5 ± 5.819.5 ± 3.4
Femoral Neck BMC5.4 ± 5.117.3 ± 11.610.5 ± 2.835.2 ± 7.2 *
Ultra-Distal Radius BMD2.7 ± 4.08.5 ± 0.96.2 ± 5.615.1 ± 1.0
Ultra-Distal Radius BMC5.3 ± 3.79.7 ± 5.511.3 ± 8.319.8 ± 4.4
Table 2. Percent Change in pQCT Endpoints in Cynomolgus Monkeys After 2 Months of Scl-AbIV Treatment
pQCT ENDPOINT (% Change from Baseline)DISTAL RADIUSPROXIMAL TIBIA
VEHScl-AbIVVEHScl-AbIV
3 mg/kg10 mg/kg30 mg/kg3 mg/kg10 mg/kg30 mg/kg
  • Data expressed as mean ± SE for percent change versus baseline after 2 months.

  • *

    Significantly different from VEH; p < .05.

Metaphysis Total Area−1.7 ± 1.32.2 ± 2.6−0.7 ± 2.14.7 ± 3.2−3.6 ± 2.15.7 ± 9.95.8 ± 2.67.0 ± 2.4
Metaphysis Total vBMC0.8 ± 1.64.1 ± 5.17.7 ± 1.719.8 ± 7.2 *−1.0 ± 3.29.4 ± 15.017.4 ± 5.927.3 ± 6.2 *
Metaphysis Total vBMD2.6 ± 2.01.8 ± 2.48.5 ± 2.214.2 ± 3.4 *2.9 ± 3.53.2 ± 4.510.9 ± 3.718.8 ± 4.7 *
Metaphysis Trabecular vBMD−3.2 ± 4.013.1 ± 26.021.7 ± 6.834.3 ± 14.4−1.7 ± 4.98.4 ± 18.721.1 ± 6.734.9 ± 8.2 *
Diaphysis Cortical Area2.8 ± 1.27.2 ± 8.55.0 ± 3.310.0 ± 4.21.2 ± 3.211.5 ± 13.512.1 ± 3.012.6 ± 3.7
Diaphysis Cortical vBMC2.4 ± 0.77.3 ± 7.23.3 ± 1.98.8 ± 2.52.5 ± 3.29.7 ± 9.611.0 ± 3.813.6 ± 2.8
Diaphysis Cortical vBMD−0.3 ± 0.50.2 ± 1.2−1.6 ± 1.4−0.9 ± 1.61.3 ± 0.7−1.2 ± 3.3−1.1 ± 0.91.0 ± 1.0
Diaphysis Cortical Thickness2.1 ± 0.97.3 ± 7.11.6 ± 1.14.3 ± 2.40.6 ± 3.413.4 ± 14.210.8 ± 4.310.2 ± 2.6
Periosteal Circumference1.1 ± 0.81.6 ± 2.73.0 ± 1.95.1 ± 1.90.5 ± 0.60.7 ± 1.93.0 ± 1.63.8 ± 1.1
Endocortical Circumference0.0 ± 1.4−3.5 ± 1.24.0 ± 2.55.9 ± 2.80.7 ± 1.6−5.2 ± 3.9−1.3 ± 4.2−0.7 ± 0.8
Moment of Inertia4.6 ± 2.98.7 ± 12.512.3 ± 7.822.2 ± 8.92.3 ± 3.59.1 ± 13.516.1 ± 5.918.8 ± 6.1

The increases in the bone-formation markers osteocalcin and P1NP resulting from Scl-AbIV treatment correlated well with the increases in BMC and BMD measured at the end of the study. Because these densitometric changes reflect the cumulative increases in bone formation that occurred throughout the study, we used area under the curve (AUC, baseline-corrected) as a variable to represent the cumulative biomarker response. The regressions for serum osteocalcin are shown in Fig. 5. Increases in lumbar spine and femoral neck areal BMD were significantly and positively correlated with serum osteocalcin AUC. Trabecular vBMD at the metaphyses of the proximal tibia and distal radius also were significantly and positively correlated with serum osteocalcin AUC. At purely cortical sites in the tibial and radial diaphyses, cortical vBMC was significantly correlated with serum osteocalcin AUC. Serum P1NP AUC also was significantly correlated with these densitometric changes (r2 = 0.45–0.61; data not shown). Thus, at both cortical and trabecular sites, by both DXA and pQCT, the increases in BMD and bone mass after Scl-AbIV treatment were correlated with the cumulative increases in the serum bone-formation biomarkers osteocalcin and P1NP.

thumbnail image

Figure 5. Increases in serum osteocalcin reflect increases in BMD and bone mass at trabecular and cortical sites. Linear regression analyses were performed across groups between bone densitometric endpoints and the bone-formation biomarker serum osteocalcin. The baseline-corrected area under the curve (AUC) calculated from the serum osteocalcin profile (Fig. 1C) was positively correlated with the increases in areal BMD at the (A) lumbar spine and (B) femoral neck. Serum osteocalcin AUC also was positively correlated with metaphyseal trabecular vBMD at the (C) proximal tibia and (D) the distal radius, as well as with diaphyseal cortical vBMC at (E) the tibia and (F) the radius. The coefficients of determination (r2) are indicated on each graph; all the dotted regression lines had slopes that were significantly different from zero (p < .05).

Download figure to PowerPoint

Bone strength

To assess vertebral bone strength, compression testing of the third and fourth lumbar vertebrae was used to measure peak load, stiffness, and energy to failure, expressed as an average of L3 and L4 data. Two once-monthly injections of Scl-AbIV at the 30 mg/kg dose resulted in increases in lumbar vertebral bone strength, with significant increases found for peak load (+97%) and energy to failure (+183%), compared with vehicle-treated controls (Fig. 6). A three-point bending test was used to assess bone strength at the femoral diaphysis. Nonsignificant increases at the 30 mg/kg dose were found for peak load, stiffness, and energy to failure (Table 3). Intrinsic (material) properties were derived from the femur test using ex-vivo pQCT geometric data. Two months of Scl-AbIV administration did not significantly alter parameters of intrinsic strength such as ultimate strength, elastic modulus, and toughness (Table 3). Consistent with this finding, there was a strong correlation between the femoral diaphysis vBMC and peak load (r2 = 0.98; Fig. 7A). At the lumbar spine, L3–L4 peak load also was highly correlated with L3–L4 total vBMC across groups (r2 = 0.92; Fig. 7B). At both sites, the slopes and intercepts of the regression lines were statistically similar (p > .4) for the vehicle group alone compared with those for the pooled Scl-AbIV data.

thumbnail image

Figure 6. Scl-AbIV administration increases bone strength at lumbar vertebrae. The bone-strength parameters (A) peak load, (B) stiffness, and (C) energy to failure were generated from destructive compression tests of trimmed third lumbar (L3) and fourth lumbar (L4) vertebral bodies. Data represent mean ± SE of the averaged L3–L4 results; n = 2 to 4 per group. *p < .05 versus vehicle control.

Download figure to PowerPoint

Table 3. Femoral Diaphysis Bending Strength in Cynomolgus Monkeys After 2 Months of Scl-AbIV Treatment
FEMORAL DIAPHYSIS STRENGTH ENDPOINTVEHScl-AbIV
3 mg/kg10 mg/kg30 mg/kg
  1. Data expressed as mean ± SE.

Peak Load (N)1008 ± 102917 ± 1211005 ± 811285 ± 241
Stiffness (N/mm)888 ± 97838 ± 106873 ± 841040 ± 192
Energy to Failure (N*mm)3600 ± 28225233190 ± 7434994 ± 904
Ultimate Strength (MPa)200 ± 14210 ± 2199 ± 8216 ± 10
Elastic Modulus (MPa)8932 ± 97910122 ± 3998929 ± 2188751 ± 616
Toughness (MPa)9.10 ± 0.508.488.53 ± 1.8912.69 ± 1.62
thumbnail image

Figure 7. Bone mass and bone strength remain tightly correlated after 2 months of sclerostin inhibition. Linear regression analyses were performed across vehicle and Scl-AbIV (3, 10, and 30 mg/kg) treatment groups between bone mass (pQCT vBMC) and peak load at the (A) femoral diaphysis and (B) L3–L4 vertebrae. The coefficients of determination (r2) are indicated on each graph; both the dotted regression lines had slopes that were significantly different from zero (p < .0001).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

We used a high-affinity sclerostin-neutralizing monoclonal antibody (Scl-AbIV) to study, for the first time, the effects of sclerostin inhibition on bone turnover, bone mass, and bone strength in cynomolgus monkeys over a 2-month time period. Once-monthly antibody administration at three dose levels (3, 10, and 30 mg/kg) was specifically chosen as the dosing regimen to investigate the anabolic effects of intermittent sclerostin inhibition and to determine the pharmacokinetic and pharmacologic profile of Scl-AbIV in nonhuman primates. Serum levels of Scl-AbIV and bone-formation markers showed a similar bimodal pattern of increase and decline following each of the two once-monthly administrations of antibody. These results demonstrated that antibody-mediated sclerostin inhibition resulted in a rapid increase in anabolic activity that returned toward baseline as the antibody was cleared from the circulation. In addition, the immediate restimulation of this anabolic pathway upon a second administration of antibody demonstrated the reproducibility and durability of the anabolic response. Consistent with the return to baseline levels observed for serum markers of bone formation, the osteoblast surface in the Scl-AbIV treatment groups at study end was not significantly different from that of vehicle-treated controls. Collectively, these data highlight the substantial and rapid modulation of bone formation that can be achieved with pharmacologic regulation of sclerostin activity.

Histomorphometric analysis showed that the key bone-formation indices MS/BS, MAR, and BFR/BS were dose-dependently elevated on trabecular surfaces as well as on the periosteal and endocortical surfaces of cortical bone. These effects were seen during both the first and the second months of the study. In addition, bone formation (BFR/BS) was dose-dependently increased on intracortical surfaces during month 2. The observed increases in MAR with Scl-AbIV suggest that there was an increase in the amount of bone matrix deposited per active osteoblast. The marked increases in active bone-forming surface (MS/BS) indicated that sclerostin inhibition resulted in increased recruitment, activation, and/or functional longevity of osteoblasts.

Despite the clear increase in anabolic activity with Scl-AbIV, no increase was found in the bone-resorption marker serum CTx, suggesting that with sclerostin inhibition there is an absence of the coupling that typically exists between osteoblast-mediated bone formation and osteoclast-mediated bone resorption during bone remodeling. The lack of increased bone resorption in Scl-AbIV-treated primates is consistent with previous results in sclerostin knockout mice and in OVX rats treated with a sclerostin antibody.9, 10 The imbalance of bone formation versus resorption with sclerostin inhibition may reflect direct activation of bone formation on quiescent surfaces (bone modeling) without prior activation of resorption (bone remodeling). This mode of anabolism may differ from that found in parathyroid hormone (PTH) treatment studies, in which bone-resorption markers in humans17 and primates18 were increased 1 month after administration. Histomorphometric analyses in humans have demonstrated that the increases in bone-forming surfaces with PTH treatment were primarily through remodeling-based mechanisms in short-19 and longer-term studies.20 Examination of the impact of sclerostin inhibition on bone resorption and bone modeling warrants further investigation.

Bone densitometry showed that the robust increases in bone formation with Scl-AbIV treatment resulted in significant increases in BMC and/or BMD at several skeletal sites (ie, femoral neck, radial metaphysis, and tibial metaphysis). These increases, expressed as percent changes from baseline, were 11 to 29 percentage points higher than those found in the vehicle-treated group after only 2 months. In addition, nonsignificant increases were found in BMC and BMD at the lumbar spine and in BMC at both the radial and tibial diaphyses. The largest densitometric increases were found in the highest-dose group (30 mg/kg), but of note, even in the lowest-dose group (3 mg/kg), BMD and BMC almost always were numerically increased relative to vehicle controls.

Despite the small study size and its short duration, the rapid increases in bone mass that were achieved were large enough to be translated into significant increases in bone strength. In the highest-dose group, significant increases in peak load and energy to failure were found at the lumbar vertebrae, whereas nonsignificant increases in bone strength were found at the femoral diaphysis, a purely cortical site. Linear regression analysis revealed very strong correlations between bone mass and bone strength at the femoral diaphysis and at the lumbar vertebrae for the Scl-AbIV and vehicle groups. The similarities between regression lines for the vehicle group alone compared with the pooled Scl-AbIV treatment groups indicated that inhibition of sclerostin preserved the normal relationship between bone mass and bone strength and did not result in changes in the fundamental material properties of bone. Rather, the Scl-AbIV-mediated increases in bone strength were a direct result of the marked increases in bone mass.

Serum biomarkers reflect dynamic changes in bone formation and bone resorption that occur throughout the skeleton, whereas histomorphometry provides insight into the specific compartments where these bone turnover changes occur. In this study, positive correlations were found between serum bone-formation biomarker levels and local bone-formation rates (BFR/BS) on trabecular and cortical bone surfaces. Positive correlations also were found between increases in serum bone-formation biomarkers and the increases seen in BMD and bone mass. These results indicate that in the setting of short-term sclerostin inhibition, increased serum markers of bone formation reflect the anabolic effects on both trabecular and cortical bone surfaces, thus validating their importance and use in future studies.

In conclusion, we have shown for the first time that administration of a sclerostin-neutralizing monoclonal antibody in cynomolgus monkeys produced a strong anabolic response in both trabecular and cortical bone, each of which play an important role in overall skeletal strength. Serum bone-formation markers rapidly increased and then returned to baseline, concordant with the rise and fall in serum sclerostin antibody levels. This finding highlights the dynamic nature of the powerful anabolic axis that sclerostin controls. Furthermore, the significant increases in bone formation, bone mass, and bone strength found in this study underscore sclerostin's pivotal role in negatively regulating the anabolic output of the osteoblast lineage in primates. In this regard, a report from a recent study in humans indicated that administration of a sclerostin-neutralizing monoclonal antibody in healthy postmenopausal women resulted in dose-dependent increases in biochemical markers of bone formation.21 Finally, the marked bone-building effects achieved in this short-term monkey study suggest that sclerostin inhibition represents a promising new therapeutic approach for medical conditions where increases in bone formation might be desirable, such as in fracture healing and osteoporosis.

Disclosures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

MSO, BS, JG, YG, JC, KG, BT, JC, RD, LZ, MDH, DLL, WSS, and CP are employees of Amgen, Inc., and have received stock and stock options from Amgen, Inc. GD (Amgen, Inc.) recently passed away. DJL, AJH, AGP, ARM, and MKR are employees of UCB-Celltech and have received stock and stock options from UCB-Celltech. FV, JJ, and SYS received funding from Amgen, Inc., and UCB-Celltech for this study.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

This research was funded by Amgen, Inc., and UCB-Celltech. The authors thank Daniel Burns (Amgen, Inc.) and Thomas Gruetzner (Charles River Laboratories) for biomarker measurements, as well as the imaging, histomorphometry, and biomechanics technical teams at Charles River Laboratories.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References