A Single Percutaneous Injection of Recombinant Human Bone Morphogenetic Protein-2 Accelerates Fracture Repair

Einhorn, Thomas A. MD; Majeska, Robert J PhD; Mohaideen, Ahamed MD; Kagel, Eric M. MD; Bouxsein, Mary L. PhD; Turek, Thomas J.; Wozney, John M. PhD

Journal of Bone & Joint Surgery - American Volume:
Scientific Article

Background: Recombinant human bone morphogenetic protein-2 (rhBMP-2), surgically implanted with a matrix material, has been shown to induce bone formation and enhance fracture repair. The purpose of this investigation was to test the hypothesis that a single, local, percutaneous injection of rhBMP-2 would accelerate fracture-healing in a standard rat femoral fracture model.

Methods: Fractures were created, following intramedullary pinning, in the femora of 144 male Sprague-Dawley rats. The animals were divided into three groups of forty-eight each. Six hours after the fracture, one group received an injection of 80 μg of rhBMP-2 in 25 μL of buffer vehicle, one received an injection of 25 μL of buffer vehicle alone, and one did not receive an injection. Twelve animals from each of these three groups were killed at one, two, three, and four weeks after treatment, and the femora were harvested for torsional biomechanical testing. An additional cohort of seventy-two animals, in which a fracture was also created, was divided into the same three treatment groups; six animals from each of these groups was killed at one, two, three, and four weeks; and the femora were processed for qualitative histological analysis.

Results: Torsional biomechanical testing indicated that the stiffness of the rhBMP-2-treated fractures was twice that of both control groups at the two, three, and four-week time-points. The strength of the rhBMP-2-treated fractures was 34% greater than that of the buffer-treated controls (p = 0.03) at three weeks and, at four weeks, was 60% and 77% greater than that of the buffer-treated controls and that of the untreated controls, respectively (p < 0.005). At four weeks, the stiffness and strength of the rhBMP-2-treated fractures were equal to those of the intact contralateral femora, whereas the buffer-treated and untreated fractures were significantly weaker than the intact femora. At two and three weeks, large areas of bone formation, typically spanning the fracture, were observed histologically in the rhBMP-2-treated sites. In contrast, the control fractures exhibited primarily soft cartilaginous callus at these time-points. By four weeks, remodeling of the hard callus and recorticalization were observed in the rhBMP-2-treated fracture sites, whereas cartilage and/or soft tissue was still present in the control fracture sites.

Conclusions: These data demonstrate that a single, local, percutaneous injection of rhBMP-2 accelerates fracture repair in this rat femoral fracture model. This effect appears to result from a combination of the induction of bone formation at the fracture site and acceleration of the rate at which the fracture callus matures.

Clinical Relevance: The ability of an injection of rhBMP-2 to accelerate fracture repair provides a rationale for its use in fractures that do not require operative treatment or that are to undergo operative treatment without direct exposure of the fracture site. In addition, the finding that an injectable compound may produce a bone-graft-like effect suggests the need for additional studies to explore its application to other types of skeletal surgery.

Author Information

Thomas A. Einhorn, MD; Boston University Medical Center, 720 Harrison Avenue, Suite 808, Boston, MA 02118-2393. E-mail address: thomas.einhorn@bmc.org

Robert J. Majeska, PhD; Department of Orthopaedics, Mount Sinai Medical Center, One Gustave L. Levy Place, Box 1188, New York, NY 10029-6574

Ahamed Mohaideen, MD; The Center for Bone and Joint Surgery of Palm Beaches, 10131 West Forest Hill Boulevard, Suite 206, West Palm Beach, FL 33414

Eric M. Kagel, MD; San Jose Orthopaedic Associates, 725 East Santa Clara Street, Suite 204, San Jose, CA 95112

Mary L. Bouxsein, PhD; Thomas J. Turek; John M. Wozney, PhD; Wyeth Research, 87 Cambridge Park Drive, Cambridge, MA 02140

Article Outline

Advances in fracture management have led to better outcomes for patients with skeletal injuries. The development of sophisticated methods for the handling of soft tissues, the refinement of closed and minimally invasive techniques for interlocking intramedullary nailing and plate fixation, and the introduction of biophysical modalities such as electromagnetic field and ultrasound stimulation have reduced the rate of nonunions. However, because up to 10% of the 6.2 million fractures sustained annually in the United States have some difficulty in healing, there remains a need for technologies to ensure rapid skeletal repair 1. Moreover, while the process of fracture-healing is considered to be biologically optimal, the ability to accelerate the repair process and thereby enable patients to return to their jobs and activities of daily living earlier would have a substantial positive impact on society.

Bone morphogenetic proteins (BMPs) are a family of bioactive factors responsible for the bone-inductive activity of bone matrix 2-4. While bone contains a mixture of these molecules, individual BMP proteins have been demonstrated to induce a complex series of cellular events culminating in bone formation. As a result, members of this family of proteins have been developed as osteoinductive compounds for a variety of applications. BMPs primarily function by causing the differentiation of mesenchymal cells into bone-forming and cartilage-forming cells that may result in endochondral or direct ossification, depending on the environment in which they are placed 5,6. They are believed to be important physiologic mediators of fracture repair, and indeed the expression of several of the BMPs is induced during the fracture-repair process 7-10.

The BMPs that have been most widely studied for their ability to induce bone regeneration include recombinant human BMP-2 (rhBMP-2) and rhBMP-7 (OP-1 [osteogenic protein-1]), both of which are manufactured by a biotechnology process utilizing mammalian cell expression. Preclinical and clinical research has demonstrated that these materials induce bone formation and healing of bone defects and accelerate fracture repair when they are surgically implanted with an appropriate carrier material 6,11. For example, rhBMP-2 combined with an absorbable collagen sponge has been shown to generate bone formation and induce healing of critical-sized defects in a canine radius model 12 as well as to accelerate fracture repair by 33% in a rabbit ulnar osteotomy model 13,14. rhBMP-2 combined with an absorbable collagen sponge induces clinically relevant amounts of bone formation in humans, such as those treated with alveolar ridge augmentation because additional bone is required for placement of dental implants 15. rhBMP-2 combined with an absorbable collagen sponge has also been shown to obviate the need for autogenous bone graft in spinal interbody fusions 16. Moreover, in a large, multinational clinical study, implantation of rhBMP-2 combined with an absorbable collagen sponge accelerated healing of open tibial fractures and reduced the rate of second interventions associated with this difficult fracture type 17. Similarly, OP-1 combined with a collagenous bone matrix carrier has been reported to be as effective as autogenous iliac bone graft for the treatment of fracture nonunion 18.

While these studies have shown the utility of open surgical application of BMP, percutaneous injection of the osteoinductive factor would have obvious clinical advantages. In the present study, we tested the hypothesis that fracture repair is accelerated by a percutaneous injection of rhBMP-2.

Back to Top | Article Outline

Materials and Methods

Preliminary Studies

Prior to initiation of the main study, two pilot investigations were conducted. The first was performed to determine the optimal dose of rhBMP-2 to be used in the main study. Standard mid-diaphyseal fractures were created in seventy-two rats, as described in the Operative Procedure section below. Six hours after creation of the fractures, 0, 5, 20, or 80 μg of rhBMP-2 (in 10 μL of buffer) was injected into the fracture site of eighteen rats each. Six hours after the fracture was chosen as a clinically meaningful time-point for the injection because patients typically do not receive definitive treatment of a fracture for several hours after it has been sustained. In addition, we have observed that, in the rat, the fracture hematoma appears to consolidate by this time. Cohorts of six rats from each group were killed at one, two, and three weeks after treatment for histological analysis. Minimal changes in the nature of the fracture callus were observed at all time-points after the 5-μg rhBMP-2 injection. Injection of 20 or 80 μg of rhBMP-2 resulted in more robust cartilaginous callus formation and accelerated more extensive maturation of the soft callus into osseous callus. This effect was most pronounced with the 80-μg dose. Consequently, 80 μg of rhBMP-2 was utilized for subsequent experiments.

As application in a carrier/matrix material typically augments the activity of BMP, a second pilot investigation was done to compare the effect of injection of rhBMP-2 in buffer with that of injection in a collagenous carrier. After creation of the fractures, three groups of thirty rats each were given an injection of rhBMP-2 in buffer, rhBMP-2 in a bovine type-I collagen dispersion, or the collagen dispersion alone. Cohorts of fifteen animals were killed at one and two weeks postoperatively. Ten limbs were evaluated with torsional biomechanical testing (see below), and five were studied histologically. Minimal differences in the biomechanical findings were observed among the groups at the one-week time-point. At two weeks, both groups of rhBMP-2-treated limbs were stiffer than the control limbs, with the rhBMP-2/buffer group being superior to the rhBMP-2/collagen group. The histological findings at the two-week time-point supported the biomechanical results. The rhBMP-2/buffer group exhibited mineralized woven bone formation that often bridged the fracture, whereas the collagen dispersion groups did not. Rather, bone formation at the fracture site was commonly interrupted by marked fibrous ingrowth from the callus exterior to the fracture margins in the collagen groups (with or without rhBMP-2). On the basis of these results, rhBMP-2 in its aqueous formulation buffer was utilized in the main fracture-repair study.

Back to Top | Article Outline
Study Design

One hundred and forty-four male Sprague-Dawley rats were divided into three groups of forty-eight each. Following the production of standard, closed mid-diaphyseal femoral fractures, 80 μg of rhBMP-2 in 25 μL of buffer vehicle was injected in one group, 25 μL of buffer vehicle alone was injected in another, and one group did not undergo an injection. Twelve animals from each of these three groups were killed at one, two, three, and four weeks after treatment, and the femora were harvested for biomechanical evaluation. A fracture was created in an additional cohort of seventy-two animals, which was then divided into the same three treatment groups. Six animals from each group were killed at one, two, three, and four weeks, and the femora were processed for qualitative histological analysis.

Back to Top | Article Outline
Operative Procedure

Protocols for all operative procedures were approved by the Institutional Animal Care and Use Committee. Adult male Sprague-Dawley rats that weighed approximately 450 g were anesthetized with ketamine (90 mg/kg, intramuscularly) and xylazine (10 mg/kg, intramuscularly). The operative site was shaved and was prepared with Betadine (povidone-iodine) and ethanol. A longitudinal incision was made just medial to the patellar tendon and extended from the femoral condyle to the proximal part of the tibia. The patella and associated tendon were reflected laterally, exposing the knee joint capsule and distal part of the femur. With use of a dental handpiece and a 1-mm drill, a hole was made between the condyles and extended into the medullary canal. A 0.45-mm Steinmann pin was introduced into the medullary canal retrograde through the pilot hole. Once the pin engaged the proximal part of the femur, it was retracted 2 to 3 mm, cut off, and then countersunk below the condylar surface. The integrity of the intramedullary pinning was established by palpation of the femur.

Fractures were created with use of a blunt guillotine driven by a dropped weight. This procedure results in closed fractures with a predominantly transverse configuration that are associated with minimal damage to the local soft tissues 19. The fractures were assessed with radiography, and animals with excessive comminution of the fracture site were excluded from the study.

rhBMP-2 (Wyeth Research, Cambridge, Massachusetts) produced with use of a mammalian cell culture process was diluted to a concentration of 3.2 mg/mL in 5-mM L-glutamic acid, 2.5% glycine, 0.5% sucrose, 5-mM NaCl, and 0.01% polysorbate 80, pH 4.5. Approximately six hours after production of the fractures, the animals were again sedated, the fracture site was located by palpation, and 25 μL of buffer alone or of buffer with 80 μg of rhBMP-2 was injected percutaneously into the intramedullary gap of the fracture with a 0.5-mL Hamilton syringe with a 27-gauge needle. Groups of seventy-two animals were killed at one, two, three, and four weeks after treatment. Femora, including unfractured contralateral bones, were harvested and processed for histological or biomechanical evaluation as described below.

Back to Top | Article Outline
Biomechanical Testing and Statistical Methods

The femora were prepared for biomechanical testing by extraction of the intramedullary pins and removal of the distal femoral physis. The ends of the harvested femora were then embedded in Cerrobend alloy (Cerrometal Products, Bellafonte, Pennsylvania), mounted in the grips of a servo-actuated DC motor torsion tester, and loaded to failure at a rotation of 1°/sec. Torsional stiffness (Nm/deg), strength (Nm), and total energy to failure (Nm-deg) were calculated from the torque-rotation plots. A two-factor analysis of variance, with the time-point after treatment and the type of treatment as the independent variables, was used to evaluate the effect of rhBMP-2 on fracture repair. When there was a significant interaction between the type of treatment and the time-point, a one-factor analysis of variance at each time-point was used to evaluate the effect of treatment. The level of significance was p < 0.05. Analysis of variance indicated that there were no differences among the biomechanical values of the intact femora at the different time-points, and therefore these groups were combined. The location of the failure was noted according to the classification described by White et al. 20. Differences in the location of the fracture following torsional biomechanical testing were evaluated with use of the chi-square statistic.

Back to Top | Article Outline
Histological Analysis

The femora were harvested, fixed in 70% alcohol, dehydrated with ascending degrees of alcohol in a tissue processor, and embedded in methylmethacrylate. A microtome was used to cut the plastic blocks into 5-μm sections, which were then deplasticized and were stained with Goldner trichrome. Qualitative evaluation of the fracture callus and associated cortical bone was performed.

Back to Top | Article Outline


Recovery from the surgery and subsequent healing proceeded uneventfully. The percutaneous injection was easily performed, with minimal leakage of the solution from the site of injection.

Back to Top | Article Outline
Biomechanical Findings

One week after the fracture, all biomechanical parameters in all treatment groups were uniformly low relative to those of the intact controls ( Table I ). The strength of both the rhBMP-2-treated and the buffer-treated fractures was significantly less than that of the surgical controls (no injection), perhaps because of physical manipulation of the fracture site during the injection procedure. With the exception of this finding, there were no differences between the buffer-treated and untreated fractures throughout the course of the study.

By the second week, the rhBMP-2-treated fractures were approximately twice as stiff as the fractures in both control groups (p < 0.005). This effect persisted throughout the course of the study. While the rhBMP-2-treated fractures had only a slightly higher failure torque than the controls at the two-week time-point, they were 34% stronger than the buffer-treated fractures at three weeks (p = 0.03). At four weeks, the rhBMP-2-treated fractures were 60% and 77% stronger than the fractures in the buffer-treated and untreated control groups (p < 0.005). Thus, rhBMP-2 had a positive effect on fracture repair, which was observed as early as fourteen days after the injection.

The rhBMP-2-treated fracture sites also returned to normal stiffness and strength faster than the fracture sites in the control groups ( Fig. 1 ). At four weeks, the biomechanical properties of the rhBMP-2-treated fractures were equivalent to those of the intact controls, whereas the buffer-treated and untreated fractures were approximately half as stiff and strong. Accelerated fracture-healing was also demonstrated by the location of failure during the biomechanical testing. In the first two weeks after treatment, the biomechanical testing caused all femora to fail through the fracture site ( Table II ). However, rhBMP-2 treatment significantly influenced the location of the fractures during biomechanical testing at four weeks (chi-square = 9.47, p = 0.009). Specifically, whereas 85% and 92% of the limbs in the buffer-treated and untreated groups failed through the original fracture site, only 42% of the rhBMP-2-treated limbs failed through the original fracture site; the remainder failed at least partially through the intact bone adjacent to the fracture.

Back to Top | Article Outline
Histological Findings

The histological findings at the fracture sites supported the biomechanical evidence of acceleration of fracture repair by rhBMP-2. Substantial differences in the rhBMP-2-treated and control sites were already visible by one week after treatment. The fracture calluses in the rhBMP-2-treated limbs were generally larger than those in the controls ( Fig. 2 ). Cartilaginous callus formation was visible in the rhBMP-2-treated fracture sites ( Fig. 3 ), whereas no cartilage formation was yet visible in the control fractures. Periosteal bone formation was present proximal and distal to the fracture site in both the treated and the control specimens. However, in the rhBMP-2-treated limbs, large areas of endochondral bone formation were present superior to this reactive bone formation ( Fig. 4 ). The bone in these supraperiosteal areas was highly vascular. At two weeks, the control callus showed periosteal bone formation at the edges, with substantial regions of cartilage formation at the fracture site and some fibrous tissue invasion in the center. In contrast, the rhBMP-2-treated sites typically contained a large area of trabecular bone formation, often bridging the defect. While trabecular bone spanning the fracture was often observed, cartilage and/or loose connective tissue typically still intervened between the cortical ends. In addition, this large area of bone formation in the rhBMP-2-treated femora was often asymmetrical (i.e., present on only one side of the fracture), suggesting localization of the rhBMP-2 effect at its injection site ( Fig. 2 , weeks 2 and 3). The callus was routinely larger in the rhBMP-2-treated fractures, with larger areas of cartilaginous and/or osseous callus.

At three weeks, progression of endochondral ossification of the callus was observed in the control samples, although substantial cartilage and fibrous tissue remained at the fracture site. In contrast, the rhBMP-2-treated fractures contained osseous callus, now well-integrated with the cortical bone and spanning both periosteal and endosteal surfaces. By four weeks, remodeling of the osseous callus was already visible in the rhBMP-2-treated sites. Both cortices were typically bridged, and remodeling of the trabecular osseous callus into cortical bone was evident. The control samples demonstrated increased amounts of osseous callus at four weeks, but the fracture sites were still interrupted by areas of cartilage, with little visible bone-remodeling.

There was no evidence of malignant transformation in any of the cell types in the calluses in the rhBMP-2-treated groups. Histological examination for malignant transformation was not performed at other skeletal sites.

Back to Top | Article Outline


This study demonstrated that a percutaneous injection of rhBMP-2 substantially accelerates and enhances the fracture-healing process in a rat model. Torsional biomechanical testing showed that, after only two weeks of healing, the rhBMP-2-treated fractures had more than twice the stiffness of the control specimens (treated with buffer only or untreated). A significant increase in strength was observed subsequently, at the three-week time-point. A study by White et al. 20 suggested that fracture-healing can be divided into four stages on the basis of biomechanical properties. Fracture sites have low stiffness and low strength in the first stage, high stiffness and low strength in the second stage, and high stiffness and high strength in the third stage. In the fourth stage, stiffness and strength are sufficiently high so that if the bone is refractured, the fracture line propagates to areas of previously unfractured, intact bone. Our finding that rhBMP-2 treatment leads to increased stiffness followed by increased strength suggests that the healing sequence follows the pattern described by White et al. but proceeds at an accelerated rate. In addition, failure of the rhBMP-2-treated bones during biomechanical testing transitioned from failure through the fracture site during the first three weeks to failure through the fracture site and the intact bone in many of the animals at four weeks. In a study that demonstrated acceleration of healing by operative implantation of rhBMP-2 with a collagen sponge in a rabbit ulnar osteotomy model, stiffness and strength increased concurrently 13. In the rabbit model, the ulnar osteotomy was stabilized by the radius, whereas in the current study an intramedullary pin was used to stabilize the fractured femur. Thus, the observed difference in the biomechanical healing patterns may be due to differences in the stability of the fractures. Alternatively, there may be a difference associated with the healing of an osteotomy site as opposed to a fracture produced by blunt trauma.

One limitation of our study is that the untreated fractures never healed within the time course of the investigation. Thus, it is difficult to quantitate the magnitude of the acceleration of fracture-healing by rhBMP-2. However, previous experiments 19 and extrapolation of the biomechanical data ( Fig. 1 ) indicate that it is likely that the control fractures would have healed at approximately five to six weeks after the fracture. In contrast, the rhBMP-2-treated fractures healed between three and four weeks after the fracture. Thus, the time to healing was decreased by 20% to 50%. This finding is in keeping with the 33% acceleration of healing observed following implantation of rhBMP-2 in an absorbable collagen sponge in the rabbit ulnar osteotomy model 13.

Our histological evaluation of the time-course of healing suggests that rhBMP-2 may augment fracture repair by means of several interrelated mechanisms. First, the bone-inductive effect of rhBMP-2 is evident. Areas of bone formation were observed within the fracture callus as early as seven days after injection of the rhBMP-2. This endochondral bone formation is spatially distinct from the periosteal bone formation observed in both the treated and the control fractures at this early time-point. In addition, substantial areas of trabecular bone spanned the rhBMP-2-treated fractures at two to three weeks. Thus, percutaneous injection of rhBMP-2 produces the same effect as open placement of autogenous bone graft: it creates a large mass of bone at the fracture site, which enhances the ability of the fracture to heal. Second, when large areas of cartilaginous callus were present in the control defects (at two to three weeks), substantially less cartilage and correspondingly larger quantities of bone were present in the rhBMP-2-treated fractures. This suggests that rhBMP-2 accelerates endochondral bone formation and causes more rapid conversion of cartilaginous callus to bone. In support of this hypothesis, in vitro studies have demonstrated that rhBMP-2 can cause not only differentiation of mesenchymal cells into chondrocytes and osteoblasts, but also conversion of cells with the chondroblastic phenotype into those with an osteoblastic phenotype 21. Although these findings suggest direct effects on cellular function and acceleration of a normal physiological process, there was no evidence of malignant transformation in any of the calluses stimulated by rhBMP-2.

The increase in torsional biomechanical stiffness and strength correlated with the histological observation of increasing amounts of bone bridging the fracture site. For example, at three weeks, the rhBMP-2-treated limbs had large amounts of trabecular bone spanning the fracture in contrast to the control limbs, in which soft tissue remained interposed; thus, the biomechanical properties were enhanced in the rhBMP-2-treated limbs. However, stiffness and strength did not return to normal levels (those in the unfractured limbs) before the osseous callus had remodeled and the cortical structure had been reestablished.

BMPs are typically applied in a carrier or matrix material. A carrier has several functions, including providing a format for surgical delivery of the osteoinductive protein, maintaining the BMP at the site of application for sufficient time for the bone-inductive process to occur, and perhaps providing an environment in which bone formation can take place 22. In the present study, application of rhBMP-2 without a matrix clearly accelerated fracture repair in the rats. In addition, the presence of large areas of bone formation within the fracture callus as early as one to two weeks after treatment indicates that the rhBMP-2 had a bone-inductive capacity even without a carrier application system. A similar observation was made in a study of rhBMP-7 (OP-1), in which rhBMP-7 in an aqueous buffer improved healing in a goat tibial fracture model but rhBMP-7 combined with a collagenous matrix did not 23. The preliminary studies described in the Materials and Methods section of the present report suggested that the use of a carrier decreased the ability of rhBMP-2 to enhance fracture repair in this particular model. On the other hand, acceleration of healing by rhBMP-2 combined with an injectable collagenous carrier has been observed in a rabbit model 24. It is possible that the presence of particular types of matrix material may physically disrupt the formation of the fracture callus in the rapidly healing rat model.

Studies of local retention have indicated that, when rhBMP-2 is delivered in aqueous buffer, it is detectable at the injection site for approximately seven days 25. The use of carrier systems can extend this residence time to several weeks. While aqueous rhBMP-2 without a carrier was effective in accelerating fracture repair in this rodent model, fracture repair in the rat is rapid compared with that in primate models and that in the clinical situation 26. Thus, one must be cautious about extending these observations to situations in which healing is slower. As the BMP is applied only once, it is likely that application of BMP in an aqueous form may not result in a sufficiently long residence time for the osteoinductive protein to provide optimal bone induction and augmentation of fracture repair in the wide range of potential clinical situations.

Bioactive factors other than BMPs have also been evaluated to assess their abilities to accelerate fracture repair in animal models 27. Although transforming growth factor-β (TGF-β) is not osteoinductive, it has been found to stimulate osteogenesis when injected subperiosteally. This effect is likely a result of proliferation of bone cells and/or an increase in their synthesis of bone extracellular matrix 28. In studies of rats and rabbits, a single administration of TGF-β did not augment fracture repair (and, at the highest dose, appeared to inhibit it); multiple or continuous administration appeared to be necessary for a positive effect 29-31. The practical difficulties associated with this type of treatment regimen appear to have limited the clinical utility of TGF-β for this application.

Fibroblast growth factor-2 (FGF-2) has also been evaluated to assess its effects on bone in several settings, including fracture repair 32-34. A single application of FGF-2 results in an increase in the size of the fracture callus, with a subsequent increase in biomechanical strength, in dog and nonhuman primate models 35-37. The magnitude of the acceleratory effect is difficult to discern from these studies, however, as single-time end points were used. FGF-2 most likely stimulates proliferation of the mesenchymal precursor cells within the periosteum, which then increases the volume of the fracture callus. Thus, FGF-2 would likely have to be administered early in the fracture-repair process and at a site with adequate periosteum. It is also possible that FGF-2 exerts an angiogenic effect that results in increased callus size. It is unclear if this effect is dependent on the timing of the administration after the fracture.

Platelet-derived growth factor (PDGF) is a general mitogen for a variety of cell types. A preliminary study of rabbits indicated that application of PDGF accelerated the healing of tibial osteotomy sites 38. However, the lack of subsequent in vivo studies suggests that the positive effect of PDGF on fracture-healing is limited.

Unlike the above growth factors, rhBMP-2 is a differentiation factor that causes differentiation of mesenchymal progenitor cells into bone and cartilage-forming cells. Thus, in addition to augmenting the ongoing fracture-repair process, it has the capacity to form new volumes of bone at the site of its application.

In summary, using a well-established model of fracture repair in the rat, we showed that a single percutaneous injection of rhBMP-2 accelerates healing. If these findings can be translated into clinical applications, the impact of such an advance could be substantial. Not only might this technology obviate the need for harvesting autologous bone in certain settings in which bone graft is now needed, but the availability and ease of administration of an injectable bone-inductive compound could be used to shorten the time to healing and restore skeletal function in patients in whom normal fracture-healing is anticipated.

Note: The authors thank Howard Seeherman for insightful comments on the manuscript.

Investigation performed at Boston University Medical Center, Boston, Massachusetts, Mount Sinai Medical Center, New York, New York, and Wyeth Research, Cambridge, Massachusetts

In support of their research or preparation of this manuscript, one or more of the authors received grants or outside funding from Wyeth (formerly Genetics Institute). In addition, one or more of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity (Wyeth). Also, a commercial entity (Wyeth) paid or directed, or agreed to pay or direct, benefits to a research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.

A commentary is available with the electronic versions of this article, on our web site (http://www.jbjs.org) and on our quarterly CD-ROM (call our subscription department, at 781-449-9780, to order the CD-ROM).

1. Praemer A, Furner S, Rice DP. Musculoskeletal conditions in the United States. Park Ridge, IL: American Academy of Orthopaedic Surgeons; 1992. p 85-124.
2. Wozney JM, Rosen V, Celeste AJ, Mitsock LM, Whitters MJ, Kriz RW, Hewick RM, Wang EA. Novel regulators of bone formation: molecular clones and activities. Science , 1988;242: 1528-34.
3. Wang EA, Rosen V, D'Alessandro JS, Bauduy M, Cordes P, Harada T, Israel DI, Hewick RM, Kerns KM, LaPan P, Luxenberg DH, McQuid D, Moutsatsos IK, Nove J, Wozney JM. Recombinant human bone morphogenetic protein induces bone formation. Proc Natl Acad Sci USA , 1990;87: 2220-4.
4. Sampath TK, Coughlin JE, Whetsone RM, Banach D, Corbett C, Ridge RJ, Özkaynak E, Oppermann H, Rueger DC. Bovine osteogenic protein is composed of dimers of OP-1 and BMP-2A, two members of the transforming growth factor-beta superfamily. J Biol Chem , 1990;265: 13198-205.
5. Sakou T. Bone morphogenetic proteins: from basic studies to clinical approaches. Bone , 1998;22: 591-603.
6. Wozney JM, Rosen V. Bone morphogenetic protein and bone morphogenetic protein gene family in bone formation and repair. Clin Orthop , 1998;346: 26-37.
7. Bostrom MP. Expression of bone morphogenetic proteins in fracture healing. Clin Orthop , 1998;355 Suppl: 116-23.
8. Bostrom MP, Lane JM, Berberian WS, Missri AA, Tomin E, Weiland A, Doty SB, Glaser D, Rosen VM. Immunolocalization and expression of bone morphogenetic proteins 2 and 4 in fracture healing. J Orthop Res , 1995;13: 357-67.
9. Onishi T, Ishidou Y, Nagamine T, Yone K, Imamura T, Kato M, Sampath TK, ten Dijke P, Sakou T. Distinct and overlapping patterns of localization of bone morphogenetic protein (BMP) family members and a BMP type II receptor during fracture healing in rats. Bone , 1998;22: 605-12.
10. Yaoita H, Orimo H, Shirai Y, Shimada T. Expression of bone morphogenetic proteins and rat distal-less homolog genes following rat femoral fracture. J Bone Miner Metab , 2000;18: 63-70.
11. Cook SD, Rueger DC. Osteogenic protein-1: biology and applications. Clin Orthop , 1996;324: 29-38.
12. Sciadini MF, Johnson KD. Evaluation of recombinant human bone morphogenetic protein-2 as a bone-graft substitute in a canine segmental defect model. J Orthop Res , 2000;18: 289-302.
13. Bouxsein ML, Turek TJ, Blake CA, D'Augusta D, Li X, Stevens M, Seeherman HJ, Wozney JM. Recombinant human bone morphogenetic protein-2 accelerates healing in a rabbit ulnar osteotomy model. J Bone Joint Surg Am , 2001;83: 1219-30.
14. Luppen CA, Blake CA, Ammirati KM, Stevens ML, Seeherman HJ, Wozney JM, Bouxsein ML. Recombinant human bone morphogenetic protein-2 enhances osteotomy healing in glucocorticoid-treated rabbits. J Bone Miner Res , 2002;17: 301-10.
15. Boyne PJ, Marx RE, Nevins M, Triplett G, Lazaro E, Lilly LC, Alder M, Nummikoski P. A feasibility study evaluating rhBMP-2/absorbable collagen sponge for maxillary sinus floor augmentation. Int J Periodontics Restorative Dent , 1997;17: 11-25.
16. McKay B, Sandhu HS. Use of recombinant human bone morphogenetic protein-2 in spinal fusion applications. Spine , 2002;27(16 Suppl 1): 66-85.
17. BMP-2 Evaluation in Surgery for Tibial Trauma (BESTT) Study Group, Govender S, Csimma C, Genant HK, Valentin-Opran A, Amit Y, Arbel R, Aro H, Atar D, Bishay M, Börner MG, Chiron P, Choong P, Cinats J, Courtenay B, Feibel R, Geulette B, Gravel C, Haas N, Raschke M, Hammacher E, van der Velde D, Hardy P, Holt M, Josten C, Ketterl RL, Lindeque B, Lob G, Mathevon H, McCoy G, Marsh D, Miller R, Munting E, Oevre S, Nordsletten L, Patel A, Pohl A, Rennie W, Reynders P, Rommens PM, Rondia J, Rossouw WC, Daneel PJ, Ruff S, Rüter A, Santavirta S, Schildhauer TA, Gekle C, Schnettler R, Segal D, Seiler H, Snowdowne RB, Stapert J, Taglang G, Verdonk R, Vogels L, Weckbach A, Wentzensen A, Wisniewski T. Recombinant human bone morphogenetic protein-2 for treatment of open tibial fractures: a prospective, controlled, randomized study of four hundred and fifty patients. J Bone Joint Surg Am , 2002;84: 2123-34.
18. Friedlaender GE, Perry CR, Cole JD, Cook SD, Cierny G, Muschler GF, Zych GA, Calhoun JH, LaForte AJ, Yin S. Osteogenic protein-1 (bone morphogenetic protein-7) in the treatment of tibial nonunions. J Bone Joint Surg Am , 2001;83 Suppl 1(Pt 2): 151-8.
19. Bonnarens F, Einhorn TA. Production of a standard closed fracture in laboratory animal bone. J Orthop Res , 1984;2: 97-101.
20. White AA 3rd, Panjabi MM, Southwick WO. The four biomechanical stages of fracture repair. J Bone Joint Surg Am , 1977;59: 188-92.
21. Rosen V, Nove J, Song JJ, Thies RS, Cox K, Wozney JM. Responsiveness of clonal limb bud cell lines to bone morphogenetic protein 2 reveals a sequential relationship between cartilage and bone cell phenotypes. J Bone Miner Res , 1994;9: 1759-68.
22. Li RH, Wozney JM. Delivering on the promise of bone morphogenetic proteins. Trends Biotechnol , 2001;19: 255-65.
23. Blokhuis TJ, den Boer FC, Bramer JA, Jenner JM, Bakker FC, Patka P, Haarman HJ. Biomechanical and histological aspects of fracture healing, stimulated with osteogenic protein-1. Biomaterials , 2001;22: 725-30.
24. Bax BE, Wozney JM, Ashhurst DE. Bone morphogenetic protein-2 increases the rate of callus formation after fracture of the rabbit tibia. Calcif Tissue Int , 1999;65: 83-9.
25. Seeherman H, Wozney J, Li R. Bone morphogenetic protein delivery systems. Spine , 2002;27(16 Suppl 1): 16-23.
26. Seeherman H. The influence of delivery vehicles and their properties on the repair of segmental defects and fractures with osteogenic factors. J Bone Joint Surg Am , 2001;83 Suppl 1(Pt 2): 79-81.
27. Barnes GL, Kostenuik PJ, Gerstenfeld LC, Einhorn TA. Growth factor regulation of fracture repair. J Bone Miner Res , 1999;14: 1805-15.
28. Bostrom MP, Asnis P. Transforming growth factor beta in fracture repair. Clin Orthop , 1998;355 Suppl: 124-31.
29. Lind M, Schumacker B, Soballe K, Keller J, Melsen F, Bunger C. Transforming growth factor-beta enhances fracture healing in rabbit tibiae. Acta Orthop Scand , 1993;64: 553-6.
30. Nielsen HM, Andreassen TT, Ledet T, Oxlund H. Local injection of TGF-beta increases the strength of tibial fractures in the rat. Acta Orthop Scand , 1994;65: 37-41.
31. Critchlow MA, Bland YS, Ashhurst DE. The effect of exogenous transforming growth factor-beta2 on healing fractures in the rabbit. Bone , 1995;16: 521-7.
32. Kawaguchi H, Kurokawa T, Hanada K, Hiyama Y, Tamura M, Ogata E, Matsumoto T. Stimulation of fracture repair by recombinant human basic fibroblast growth factor in normal and streptozotocin-diabetic rats. Endocrinology , 1994;135: 774-81.
33. Bland YS, Critchlow MA, Ashhurst DE. Exogenous fibroblast growth factors-1 and -2 do not accelerate fracture healing in the rabbit. Acta Orthop Scand , 1995;66: 543-8.
34. Kato T, Kawaguchi H, Hanada K, Aoyama I, Hiyama Y, Nakamura T, Kuzutani K, Tamura M, Kurokawa T, Nakamura K. Single local injection of recombinant fibroblast growth factor-2 stimulates healing of segmental bone defects in rabbits. J Orthop Res , 1998;16: 654-9.
35. Nakamura T, Hara Y, Tagawa M, Tamura M, Yuge T, Fukuda H, Nigi H. Recombinant human basic fibroblast growth factor accelerates fracture healing by enhancing callus remodeling in experimental dog tibial fracture. J Bone Miner Res , 1998;13: 942-9.
36. Radomsky ML, Aufdemorte TB, Swain LD, Fox WC, Spiro RC, Poser JW. Novel formulation of fibroblast growth factor-2 in a hyaluronan gel accelerates fracture healing in nonhuman primates. J Orthop Res , 1999;17: 607-14.
37. Kawaguchi H, Nakamura K, Tabata Y, Ikada Y, Aoyama I, Anzai J, Nakamura T, Hiyama Y, Tamura M. Acceleration of fracture healing in nonhuman primates by fibroblast growth factor-2. J Clin Endocrinol Metab , 2001;86: 875-80.
38. Nash TJ, Howlett CR, Martin C, Steele J, Johnson KA, Hicklin DJ. Effect of platelet-derived growth factor on tibial osteotomies in rabbits. Bone , 1994;15: 203-8.
Copyright 2003 by The Journal of Bone and Joint Surgery, Incorporated