Mandibular reconstruction using bone morphogenetic protein 2: long-term follow-up in a canine model

To determine the degree of bone resorption and stability of 3-cm, full-thickness canine mandibular defects reconstructed with recombinant human bone morphogenetic protein 2 (rhBMP-2) and a bioerodible particle carrier followed for 30 months after reconstruction.

Nine dogs, divided into three groups, underwent reconstruction of surgically created 3-cm, full-thickness defects of the body of the mandible.

Mandibular reconstruction was performed via a combined intraoral and extraoral approach. Using standard plating techniques, a unilateral full-thickness, 3-cm defect was created in the body of the mandible. After stabilizing the defects with titanium reconstruction plates, test implants composed of rhBMP-2 and poly(lactide-co-glycolide) bioerodible particles were placed in the mandibular defects of six animals. Reconstruction plates were removed from test animals at 10 weeks. Three short-term test animals were sacrificed 3 months after reconstruction. Three long-term test animals were sacrificed 30 months after reconstruction to determine the degree of resorption and long-term stability of the rhBMP-2-induced bone. Control implants (carrier without rhBMP-2) were used in three animals and were sacrificed at 3 months. At 9 months, long-term animals were advanced to a solid diet. Masticatory function and body weight were monitored periodically to assess diet tolerance. Roentgenographic photodensitometry was performed on serial dental roentgenograms of the reconstructed segments to determine bone density and the degree of bone resorption over 30 months. After sacrifice, reconstructed segments were harvested and embedded in plastic for histological analysis and histomorphometry to determine the percentage of the defect replaced by mineralized bone (area density) and degree of resorption from 3 to 30 months after reconstruction. The main outcome measures were bone density and bone height determined from serial roentgenograms and percentage of the reconstructed segment replaced by mineralized bone (area density) determined from histomorphometry.

Control animals without rhBMP-2 showed no evidence of bone formation across the defect. Histological examination revealed good bone formation in two of three of the short-term test animals with a mean area density of 41.0%. The long-term test animals treated with rhBMP-2 demonstrated good bone formation that was comparable to that of normal host bone by 3 months. The roentgenographic photodensity measurements stabilized at 5 months without evidence of persistent bone resorption. The height of the reconstructed segment (rhBMP-2-induced bone) initially decreased, then stabilized by 11 months after reconstruction with no indication of resorption or failure. Histological examination of the long-term test animals revealed good bone formation across the mandibular defects. However, there were localized areas of thinning of the cortical bone as compared with the short-term test animals sacrificed at 3 months. Histological examination verified the loss of height of the bone in the reconstructed segments. The area density (mean) of the long-term test animals was 56.5%. Despite the decrease in height of the induced bone, there was an increase in area density of the bone over time.

This study demonstrated that rhBMP-2 in a bioerodible particle carrier induced new host bone formation across critical-size mandibular defects. The newly formed bone successfully integrated with existing host bone creating a stable union capable of withstanding the forces of masticatory function in a canine. There was some evidence of early bone resorption (thinning of the cortical bone and decrease in height) in the rhBMP-2-induced bone. The rhBMP-2-induced bone stabilized by 11 months after reconstruction and no further resorption was noted. The percentage of area of the defect replaced by rhBMP-2-induced bone (area density) increased over 30 months. (ABSTRACT TRUNCATED)