Sclerostin antibody treatment enhances metaphyseal bone healing in rats



Sclerostin is the product of the SOST gene. Loss-of-function mutations in the SOST gene result in a high-bone-mass phenotype, demonstrating that sclerostin is a negative regulator of bone mass. Primarily expressed by osteocytes in bone, sclerostin is reported to bind the LRP5/6 receptor, thereby antagonizing canonical Wnt signaling and negatively regulating bone formation. We therefore investigated whether systemic administration of a sclerostin-neutralizing antibody would increase the regeneration of traumatized metaphyseal bone in rats. Young male rats had a screw inserted in the proximal tibia and were divided into six groups given 25 mg/kg of sclerostin antibody or control twice a week subcutaneously for 2 or 4 weeks. In four groups, the screws were tested for pull-out strength. At the time of euthanasia, a similar screw also was inserted in the contralateral tibia and pull-out tested immediately. Sclerostin antibody significantly increased the pull-out force by almost 50% compared with controls after 2 and 4 weeks. Also, the screws inserted at the time of euthanasia showed increased pull-out force. Micro–computed tomography (µCT) of the remaining two groups showed that the antibody led to a 30% increase in bone volume fraction in a region surrounding the screw. There also was a general increase in trabecular thickness in cancellous bone. Thus, as measured by the amount of bone and its mechanical resistance, the sclerostin antibody increased bone formation during metaphyseal repair but also in untraumatized bone. © 2010 American Society for Bone and Mineral Research.


Most fractures occur in osteoporotic cancellous bone in metaphyseal regions. If these fractures are stable, they heal predominantly by intramembranous bone formation.1 Screws and other fixation devices such as intramedullary nails may be inserted to stabilize the fractures. Pull-out strength and other measures of fixation correlate with the amount or structure of bone that surrounds the fixation device.2 The response to the trauma of inserting a screw in the cancellous bone appears similar to that of metaphyseal fracture repair. Because the formation of new bone around a screw determines the strength of its fixation—especially if the original fixation is weak—this can be used to estimate the vigor of the bone-regenerative response.

A fracture healing response is also involved in the early fixation of total joint replacements in metaphyseal bone. Since early fixation is coupled with the risk of late loosening,3 increased bone formation at early stages in the healing process might provide a better long-term outcome. The quality of the fixation can be influenced by drugs or treatments that have an effect on bone formation or resorption around the implant. Parathyroid hormone (PTH) and bisphosphonates can improve implant fixation in animals, and for bisphosphonates, this also has been shown in patients.4 However, bisphosphonates are not truly anabolic, and PTH can be given to humans only at low doses, with a weak effect on fracture repair.5

Sclerostin, a secreted glycoprotein, is the product of the SOST gene. Mutations of the SOST gene cause high bone mass in humans.6, 7 Similarly, a high-bone-mass phenotype is replicated in SOST knockout mice.8 Sclerostin is reported to bind the LRP5/6 receptor, thus antagonizing Wnt signaling and increasing β-catenin degradation.9 Sclerostin is expressed primarily in bone, exclusively in osteocytes, and it is believed to be a potent negative regulator of bone formation.10, 11 It is also an important mediator of the response to mechanical loading in bone.12 It was reported recently that sclerostin antibody treatment in a rat model of postmenopausal osteoporosis lead to an increase in bone formation, bone mass, and bone strength.13 It is reasonable to believe that inhibition of sclerostin would influence the healing of traumatized bone, considering the importance of β-catenin in fracture healing.14

We tested the hypothesis that systemic administration of a sclerostin antibody would increase the response to trauma and thereby the fixation of a screw implanted in cancellous nonosteoporotic bone in rats. We also used mechanical testing of untraumatized bone and measured bone mass and architecture by micro–computed tomography (µCT; Fig. 1) to evaluate anabolic effects in bone sites without injury.

Figure 1.

Illustrative µCT image of proximal rat tibia with inserted PMMA screw (not visible owing to its low radiodensity). Note the bone layer around the screw, responsible for part of the pull-out resistance. Some structural parameters for this bone are presented in Table 1, and the defined regions of interest are found in Fig. 2.

Materials and Methods

Antibody and implants

A “ratized” sclerostin antibody (Scl-AbIII) specifically designed for rat studies13 was provided by Amgen (Thousand Oaks, CA, USA). For mechanical evaluation, 1.6-mm-wide stainless steel screws (316L) were used. The threaded part of the screw is 2.8 mm long. Screws are custom made and fitted with a head that enables it to be mounted in a materials testing machine. The head protrudes 3.3 mm into the subcutaneous space. For µCT, screws of the same size were made out of polymethyl methacrylate (PMMA) to avoid metal artifacts. Previous studies have shown similar histology with stainless steel and PMMA implanted in a similar fashion as in this study.15

Animals and general layout

A total of 68 male 10-week-old Sprague-Dawley rats with a mean weight of 348 g (SD ±14 g) were used for two experiments. The first experiment was evaluated by mechanical testing and the second experiment by µCT. After surgery, the rats for mechanical testing were randomly assigned to four groups of 12 animals. Two groups received subcutaneous injections of 25 mg/kg of Scl-AbIII twice weekly for 2 or 4 weeks, with injections starting on day 3 after surgery. The other two groups were administered saline solution on the same occasions.

Rats designated for evaluation by µCT were similarly randomized after surgery into two groups of 10 to receive antibody treatment or saline as above for a treatment period of 4 weeks.

Surgical procedure

The surgery was performed under aseptic conditions. The rats were anesthetized with isoflurane gas. Antibiotics (20 mg/kg of oxytetracycline) and analgesic (0.043 mg/kg of buprenorphine) were given subcutaneously pre- and postoperatively. A 6- to 7-mm longitudinal incision was made along the rat tibia, and the periosteum was reflected proximally to the physis. A hole 1.2 mm in diameter was drilled with a handheld drill through one of the cortices, approximately 3 mm from the physis. A screw, either steel or PMMA, was inserted, and the skin was sutured. The animals were fully weight-bearing immediately after awakening from anesthesia.

All animals had screws inserted unilaterally in the right proximal tibia. For mechanical testing, the animals were euthanized after 2 or 4 weeks using carbon dioxide, and both tibias and the left femur were harvested. Animals used for µCT were euthanized after 4 weeks, and both tibias were harvested. The tibias were fixed for 48 hours in 4% formaldehyde and then kept in 70% ethanol until µCT measurement.

The study was approved by the Regional Ethics Committee for Animal Experiments, and institutional guidelines for the care and treatment of laboratory animals were followed. The rats were given free access to food and water during the experiment and were housed three rats per cage at 21°C in a room with 12 hours of light and 12 hours of dark.

Mechanical evaluation

All analyses were performed while the investigators were blinded for treatment. Harvested bones were kept moist by saline irrigation, and all bones were tested within 1 hour of harvesting. Screws in the right tibias from each treatment group were tested for pull-out strength in a computerized materials testing machine (100 R, DDL, Inc., Eden Prairie, MN, USA) at a cross-head speed of 0.1 mm/s. In the untraumatized contralateral tibia, a screw was inserted postmortem, similar to the insertion during the surgical procedure. This screw was tested for pull-out strength immediately in the same way as the screw in the traumatized leg. For this testing, the same screw was used for all samples. Bone and tissue were removed from the screw between each animal test.

The materials testing machine recorded the peak force and the energy uptake until the force had dropped to 90% of maximum. The stiffness then was obtained based on the slope of the force/displacement curve.

Evaluation by µCT

Both proximal tibias were imaged with an isotropic voxel size of 15 µm (Skyscan 1172, Version 1.5, Skyscan, Aarteselar, Belgium). The µCT scanner acquired topographic images of the bone at the energy settings of 100 kV and 100 µA using an aluminum filter of 0.5 mm and 10 repeated scans. The images were reconstructed using NRecon (Skyscan, Version by correcting for ring artifacts and beam hardening.

First, cylindrical regions of interest (ROI) with a diameter of 2.7 mm were defined coaxial with the screw. The first cylinder (total screw surrounding) extended from the most superficial circular section at the periosteal surface down to 0.5 mm beyond the deep end of the screw (Fig. 2). Second, the marrow portion of the screw surrounding was defined as the part of the cylinder extending from the deep end 1.5 mm toward the periosteal end. The cortical portion was defined as extending from the most superficial circular section 0.5 mm toward the marrow. This region included the entire cortex in all sections. Hence it reflected cortical thickness. The portion of the cylinder made up by the screw was subtracted from all measurements.

Figure 2.

Cylindrical regions of interest with a diameter of 2.7 mm were defined coaxial with the screw (white squares). The largest region of interest comprised both the entire intraosseous and cortical portions extending from the most superficial circular section at the periosteal surface down to 0.5 mm beyond the deep end of the screw (1a). The marrow portion of the cylinder was defined as the part of the cylinder extending from the deep end of the screw 1.5 mm toward the periosteum (1b). The cortical portion was defined as extending from the most superficial circular section 0.5 mm toward the marrow (1c). The portion of the cylinder made up by the screw was subtracted from all measurements. In both the operated and contralateral tibias, a segment was defined to extend from 1.5 to 3 mm below the growth plate. The trabecular bone within this segment (enclosed by a white line) was used to assess the structure of the bone that was not in the vicinity of the screw (2).

Furthermore, in both the operated and contralateral tibias, a ROI was defined to measure trabecular bone architecture in the bone that was not in the vicinity of the screw. It was defined to start 1.5 mm (100 images) below the growth plate and to continue for 1.5 mm distally (Fig. 2). Calibration of bone mineral density (BMD) was carried out according to the system manufacturer's protocol by scanning of a water phantom and two hydroxyapatite phantoms of known density (0.25 and 0.75 g/cm3). Mineralized bone tissue was assumed to have a BMD over 0.590 g/cm3, resulting in grayscale values of 70 to 255. Analyses of bone volume fraction (BV/TV) and trabecular thickness, number, and separation (Tb.Th, Tb.Sp, and Tb.N) were performed in CTAn (Version. Skyscan). All analyses were performed in a blinded manner.

Images were produced that show a section through the axis of the screw and of the tibial shaft. These images then were classified independently by PA and FA into two groups in a blinded manner depending on the degree of bone condensation around the screw.

Ash weight

The left femur was cleaned from soft tissue and mounted in a holder, enabling the most distal 6-mm segment of the femur to be cut off using a precise cutting saw. This segment then was incinerated at 900°C for 24 hours. The specimen was weighed before and after incineration.

Statistical analysis

Peak force was prespecified as the primary outcome variable. Results are presented as mean ± SD. Differences in the mechanical testing were tested using two-way ANOVA, with experimental time points (2 or 4 weeks) and treatment as fixed factors. Confidence intervals for differences between group means were calculated for treatment effects within the ANOVA using t tests. µCT data were analyzed by t test. A result was considered statistically significant when p < .05. SPSS statistical software was used (SPSS, Version 17.0, SPSS, Inc., Chicago IL, USA).


Two animals died before surgery. These animals were not replaced, and therefore, group sizes were not fully symmetrical. One femur fractured during preparation for ash weight determination and was discarded. One PMMA screw tibia cracked during harvesting and was excluded from µCT measurement. No other animals or data were excluded. The body weight had increased by 67 g (SD ±11 g) at 2 weeks and 116 g (SD ±30 g) at 4 weeks. This increase was not influenced by Scl-AbIII treatment.

Mechanical testing

Scl-AbIII treatment significantly increased the peak pull-out force by 38% and 56%, respectively, at 2 and 4 weeks after insertion compared with saline controls (Fig. 3A). Peak force increased from 2 to 4 weeks (two-way ANOVA, effect of time: p = .02) in both control and Scl-AbIII groups. Similarly, stiffness was increased significantly at both 2 and 4 weeks in the Scl-AbIII-treated group compared with controls (Fig. 3C). A significant increase in energy was observed with Scl-AbIII treatment (Fig. 3E).

Figure 3.

Mechanical test of screw fixation in the proximal tibia. The left column (A, C, E) represents screws inserted either 2 or 4 weeks before pull-out testing. The right column (B, D, F) represents screws inserted postmortem in the untraumatized contralateral tibia of the same animals. (A) Scl-AbIII treatment increased the peak force after both 2 weeks (a; p = .002 versus control) and 4 weeks (b; p = .003 versus control). (B) In the contralateral bone, the Scl-AbIII treatment increased the peak force after both 2 weeks (a; p = .002 versus control) and 4 weeks (b; p < .001 versus control). (C) Scl-AbIII treatment increased the stiffness of the fixation after both 2 weeks (a; p = .007 versus control) and 4 weeks (b; p = .042 versus control). (D) In the contralateral bone, the stiffness was slightly increased after 2 weeks (a; p = .05 versus control) and more so after 4 weeks (b; p = .003 versus control). (E) The energy required to remove the screw was not significantly changed after 2 weeks but was increased after 4 weeks (a; p < .001 versus control). (F) In the contralateral bone, the energy was increased after both 2 weeks (a; p < .001 versus control) and 4 weeks (b; p = .026 versus control). p values are corrected for assumption of unequal variances when necessary.

The screws inserted into the left tibia postmortem had less than a third of the pull-out force of those that had been inserted 2 or 4 weeks earlier. However, the fixation was still stronger in rats treated with Scl-AbIII, as demonstrated by significant increases in peak force, stiffness, and energy at both 2 and 4 weeks compared with controls (Fig. 3B, D, F).


The larger cylindrical ROI, comprising the entire intraosseous portion of the screw as well as the cortical portion, showed a 32% increase (p = .02) in bone volume fraction (BV/TV) and a 4.5% increase (p = .04) in average BMD in the Scl-AbIII-treated group compared with controls (Table 1).

Table 1. Bone Volume Fraction (BV/TV) and Mean Bone Mineral Density (BMD) Around the Screw After 4 Weeks of Treatment, Assessed by µCT in the Three Regions of Interest Described in Fig. 2 (1a–c)
Region of interestParameterTreatmentnMeanSDLoweraMeanaUpperap Value
  • a

    95% confidence interval for the difference between means of the Scl-AbIII and control groups.

Entire screw surrounding (1a)BV/TV (%)Control9286.96%31%56%.02
 BMD (g/cm3)Control91.120.050%4%8%.04
Marrow surrounding (1b)BV/TV (%)Control9256.1−2%23%47%.07
 BMD (g/cm3)Control91.100.020%3%6%.05
Cortical surrounding (1c)BV/TV (%)Control951126%28%50%.01
 BMD (g/cm3)Control91.140.0650%5%10%.04

The cancellous bone between the screw site and the growth plate showed an increase in mean BMD, as well as a dramatic increase in trabecular thickness (Table 2). No other parameters were affected. The unoperated tibia showed a slightly higher BMD than the operated tibia (pairwise comparison, p = .006). However, this was not affected by Scl-AbIII treatment. Blinded qualitative scoring was not able to identify treatment effects on bone condensation around the screws.

Table 2. Microstructural Parameters in the Cancellous Bone of the Proximal Tibia After 4 Weeks of Treatment, Assessed by µCT
Site of measurementParameterTreatmentnMeanSDLoweraMeanaUpperap Value
  • Note: The region of interest comprised trabecular bone in the proximal tibia, proximal to the screw implantation site, as described in Fig. 2.

  • a

    95% confidence interval for the difference between means of the Scl-AbIII and control groups.

Tibia with implanted screwBV/TV (%)Control9194.0–1%21%42%.6
 Tb.Th (µm)Control9924.122%27%32%<.001
 Tb.N (1/µm)Control92.10.48–25%–5%14%.6
 Tb.Sp (µm)Control927348–7%11%30%.2
 BMD (g/cm3)Control90.960.026%8%10%<.001
Untraumatized Contralateral tibiaBV/TV (%)Control10233.8–2%15%33%.09
 Tb.Th (µm)Control10933.126%29%33%<.001
 Tb.N (1/µm)Control102.40.40–26%–11%4%.1
 Tb.Sp (µm)Control1024442–6%13%32%.2
 BMD (g/cm3)Control100.980.035%7%9%<.001

Ash weight

Scl-AbIII treatment increased the ash weight of the distal femur by 18% [95% confidence interval (CI) 15–21%) compared with controls (Fig. 4).

Figure 4.

Ash weight of the cancellous femoral bone was increased by Scl-AbIII treatment at 2 weeks (a; p < .001 versus control) and 4 weeks (b; p < .001 versus control).


In this study, inhibition of sclerostin improved the fixation of a steel screw in the proximal tibia of young male rats. Insertion of an orthopedic implant into bone constitutes an injury, initiating bone repair. Thus implant fixation in this model may be considered to be the end result of a fracture healing response of the kind predominant in cancellous bone. Even after such a short time span as 2 and 4 weeks, Scl-AbIII treatment lead to an approximate 50% increase in pull-out force, as well as significant increases in bone stiffness and pull-out energy. Data from the contralateral tibia further show that Scl-AbIII had a general strengthening effect on the trabecular bone of the untraumatized proximal tibia. It appears that Scl-AbIII increased the holding capacity of the bone in a proportionate way, increasing pull-out force by about half in both traumatized and untraumatized bone. Because the traumatized bone exerted a healing response, with increased bone formation, the proportionate effect of Scl-AbIII led to a much larger increase in absolute terms.

µCT showed increased bone volume fraction and mineral density around the screw as a result of Scl-AbIII treatment and greatly increased trabecular thickness in the proximal tibia irrespective of surgery. This is consistent with the biomechanical data indicating a generalized response rather than a specifically increased repair response after trauma. The increase in bone volume fraction around the screw was somewhat smaller than the increase in pull-out force. A part of the effects of Scl-AbIII on pull-out force therefore might be explained by architectural improvements, that is, a thicker bone lamella adjacent to the implant, similar to the increased trabecular thickness seen in the untraumatized cancellous regions. Trabecular parameters were not reported in the immediate surroundings of the screw because after 4 weeks a trabecular structure has not yet formed. However, the increased bone volume fraction as well as increased mean BMD with Scl-AbIII treatment explains the increased mechanical strength. Moreover, Scl-AbIII increased the bone volume fraction in the cortical ROI. Since this region comprised the entire cortex plus some marrow space in all specimens, the increase in bone volume fraction is likely to reflect increased cortical thickness.

Our results are supported by an earlier study in which a sclerostin antibody was tested in aged osteopenic female rats.13 In that study, 5 weeks of treatment was enough to reverse the negative effects on BMD caused by 1 year of ovariectomy. The treatment was also able to restore and even improve the mechanical strength of femur and vertebrae. The authors concluded that inhibition of sclerostin increases bone formation and leads to increased trabecular and cortical thickness as well as a stronger bone.

Several drugs have been investigated for their ability to improve screw fixation by promoting bone formation or inhibiting bone resorption. PTH has a large effect on fixation, with an increase in bone strength and enhanced bone formation around screws.16 Bisphosphonates have been tested frequently, and systemic doses have a beneficial effect on screw fixation in animal models.17 Although bisphosphonates are technically antiresorptive and not considered to be anabolic agents, they increase the amount of bone formed around an implanted screw18 in much the same way as has been observed with PTH treatment.16 The positive effect of bisphosphonates on bone healing probably depends on the uncoupling between osteoblasts and osteoclast function during the early phase of healing. In contrast, the bone-forming effect of Scl-AbIII is due primarily to its action on the osteoblastic lineage.8, 13

Our results show that the large anabolic effect obtained from sclerostin antibody treatment in aged osteoporotic rats can be replicated in a young normal skeleton and can be beneficial not only in the setting of bone loss (osteoporosis) but also in the healing process of trabecular bone. In addition, the inhibition of sclerostin has a potential to increase fixation of implants and improve fracture healing. The inhibition of sclerostin has anabolic effects similar to intermittent PTH injections, which is interesting because SOST is a target gene for PTH, and sclerostin production is downregulated in the presence of PTH.19, 20 On the other hand, the previous rat data with PTH used higher doses than what can be given to humans. The effects of the Food and Drug Administration–accepted human doses of PTH appear to be weak.5 In contrast, owing to the specificity of sclerostin's function in the control of bone formation, it may be possible to use relatively higher doses of sclerostin antibodies in humans, although this awaits clinical investigation.

The inhibition of sclerostin is being investigated as a general treatment option for conditions where increased bone formation might be beneficial therapeutically (eg, osteoporosis). We have shown that Scl-AbIII treatment improves the mechanical fixation of steel screws in the cancellous bone of rats. This model can be regarded as a stable metaphyseal fracture. The treatment also improved the fixation of screws in nontraumatized bone, suggesting a general increase in the ability to form more bone rather than a specifically improved response to the fracture trauma. This was supported by morphometric data from µCT. The results point toward the potential of using sclerostin antibody therapy not just to alter general metabolic conditions in the skeleton but also to promote bone-healing responses.


PA serves as a consultant for Amgen and Eli Lilly. XL and HZK are employees of Amgen. FA and HI state that they have no conflicts of interest.


This research was funded by Amgen and the Swedish Research Council (VR). Scl-AbIII was provided by Amgen (Thousand Oaks, CA, USA). We thank Therese Andersson, Noomi Altgärde, Pernilla Eliasson, and Emma Mallis for assistance during animal surgery. We also thank Mats Christensson for the manufacturing of steel and PMMA screws used in this research.