Use of bone morphogenetic protein-2 in the rabbit ulnar nonunion model.

The ability of the osteoinductive protein and recombinant human bone morphogenetic protein-2, combined with polylactic glycolic acid porous microspheres and autologous blood clot to heal a large segmental defect was tested in a rabbit diaphyseal defect model. Two centimeter nonuniting defects were surgically created in the bilateral ulnae of 50 male New Zealand white rabbits. Each defect was then implanted with a pastelike polylactic glycolic acid/blood clot combination that was mixed with 5 different concentrations of recombinant human bone morphogenetic protein-2. The forearms were radiographically assessed on a biweekly schedule for 8 weeks. At 8 weeks, all animals were sacrificed and forearms radiographed. Radiographs were then scored by 3 independent observers for bone formation and union rates. United limbs were tested in torsion for mechanical strength using a Burstein torsion tester. All nonunited limbs were analyzed histologically as were 2 united limbs from each dosage group. Radiographic evaluation revealed that there was a dose dependent response in healing of the ulnar defect with a higher bone formation rate in the 2 higher dose limbs than in the lower dose limbs. Union was achieved in 100% of the highest dose limbs, whereas only 50% of the lowest dose limbs achieved bony union. No defects implanted with carrier alone achieved union. Biomechanical studies revealed significantly stiffer bone than age matched controls. Histologic analysis demonstrated normal bone formation with abundant normal appearing osteoid. These dose response data further support the role of recombinant human bone morphogenetic protein-2 as a potent morphogen in bone regeneration.