Effect of recombinant human bone morphogenetic protein–2 on bone regeneration in large defects of the growing canine skull after dura mater replacement with a dura mater substitute

OBJECT: This study was designed to evaluate the bone regeneration potential of the dura mater and dura mater substitute (Durepair) in the presence of recombinant human bone morphogenetic protein-2 (rhBMP-2) delivered in a collagen sponge-collagen-ceramic matrix (CCM; MasterGraft Matrix) in a large skull defect in growing canines. METHODS: Forty immature male beagles were used to create two 2.5 x 4-cm cranial defects on each side of the sagittal suture. The dura mater on the left side was cut to make a 1 x 3-cm defect and replaced with bovine skin collagen (Durepair). The dura mater on the right side remained intact. Different doses of rhBMP-2 (none 8 animals, 0.11 mg/ml 4 animals, 0.21 mg/ml 4 animals, and 0.43 mg/ml 8 animals) were infused on 2 Type I bovine absorbable collagen sponge (ACS) strips. The strips were layered with the CCM (15% hydroxyapatite HA/85% tricalcium phosphate TCP) to reconstruct both cranial defects. In a fifth group (8 animals), 0.43 mg/ml rhBMP-2 was directly infused into the CCM. Demineralized canine cancellous freeze-dried demineralized bone matrix (DBM; 8 animals) was used as a control in a sixth group. All materials were fixed under 2 resorbable protective sheets (MacroPore). Skulls were resected 16 weeks after operation. Histological and histomorphometric analyses on the percentage of the defect spanned by bone, and the percentage of residual HA-TCP granules and collagen were analyzed. RESULTS: Calcified seroma was the only complication observed and only occurred in the 0.43-mg/ml rhBMP-2 groups (Groups 4 and 5). Dura mater repair appeared complete at 4 months in all animals. New bone was formed sporadically throughout the skull defect in the ACS+CCM and DBM groups without rhBMP-2. In all rhBMP-2 groups, mature new bone (compact and trabecular) was uniformly formed across the defect on both the repaired and intact dura mater sides. There was significant new compact bone formation on top of the repaired dura mater, which did not appear in the ACS+CCM and DBM groups lacking rhBMP-2. Greater HA-TCP and collagen scaffold resorption was noted in rhBMP-2 groups compared with non-rhBMP-2 groups. Statistical analysis showed there was a significantly lower percentage of bone spanning the defect in the ACS+CCM group compared with groups with rhBMP-2, with more residual HA-TCP and collagen on the repaired dura mater side than the intact dura mater side (p 0.05). CONCLUSIONS: The ACS+CCM combination had an effect similar to demineralized bone-on-bone regeneration in craniofacial reconstruction. The addition of rhBMP-2 to CCM directly or with ACS induces mature new bone formation in large cranial defects both in the presence of intact dura mater and repaired dura mater.