Recombinant Human Bone Morphogenetic Protein-2 Accelerates Healing in a Rabbit Ulnar Osteotomy Model

Background: Approximately 5% to 20% of fractures have delayed or impaired healing. Therefore, it is desirable to develop new therapies to enhance fracture-healing that can be used in conjunction with traditional treatment methods. The purpose of this study was to evaluate the ability of a single application of recombinant human bone morphogenetic protein-2 to accelerate fracture-healing in a rabbit ulnar osteotomy that heals spontaneously. Methods: Bilateral mid-ulnar osteotomies (approximately 0.5 to 1.0 mm wide) were created in seventy-two skeletally mature male rabbits. The limbs were assigned to one of three groups: those treated with an absorbable collagen sponge containing recombinant human bone morphogenetic protein-2, those treated with an absorbable collagen sponge containing buffer, and those left untreated. In the first two groups, an 8 20-mm strip of absorbable collagen sponge containing either 40 g of recombinant human bone morphogenetic protein-2 or buffer only was wrapped around the osteotomy site. The rabbits were killed at two, three, four, or six weeks after surgery. In addition, twenty-four age-matched rabbits were used to provide data on the properties of intact limbs. The retention of recombinant human bone morphogenetic protein-2 at the osteotomy site was determined with scintigraphic imaging of 125I-labeled recombinant human bone morphogenetic protein-2. After the rabbits were killed, the limbs were scanned with peripheral quantitative computed tomography to assess the area and mineral content of the mineralized callus. The limbs were then tested to failure in torsion, and undecalcified specimens were evaluated histologically. Results: Gamma scintigraphy of 125I-recombinant human bone morphogenetic protein-2 showed that 73% ± 6% (mean and standard deviation) of the administered dose was initially retained at the fracture site. Approximately 37% ± 10% of the initial dose remained at the site one week after surgery, and 8% ± 7% remained after two weeks. The mineralized callus area was similar in all groups at two weeks, but it was 20% to 60% greater in the ulnae treated with recombinant human bone morphogenetic protein-2 than in either the ulnae treated with buffer or the untreated ulnae at three, four, and six weeks (p < 0.05). Biomechanical properties were similar in all groups at two weeks, but they were at least 80% greater in the ulnae treated with recombinant human bone morphogenetic protein-2 at three and four weeks than in either the ulnae treated with buffer (p < 0.005) or the untreated ulnae (p < 0.01). By four weeks, the biomechanical properties of the ulnae treated with recombinant human bone morphogenetic protein-2 were equivalent to those of the intact ulnae, whereas the biomechanical properties of both the ulnae treated with buffer and the untreated ulnae had reached only approximately 45% of those of the intact ulnae. At six weeks, the biomechanical properties were similar in all groups and were equivalent to those of the intact ulnae. The callus geometry and biomechanical properties of the ulnae treated with buffer were equivalent to those of the untreated ulnae at all time-points. Conclusions and Clinical Relevance: These findings indicate that treatment with an absorbable collagen sponge containing recombinant human bone morphogenetic protein-2 enhances healing of a long-bone osteotomy that heals spontaneously. Specifically, osteotomies treated with recombinant human bone morphogenetic protein-2 healed 33% faster than osteotomies left untreated. The results of this study provide a rationale for testing the ability of recombinant human bone morphogenetic protein-2 to accelerate healing in patients with fractures requiring open surgical management.