Background: Thoracoscopically assisted anterior spinal arthrodesis and instrumentation is being used more widely to treat idiopathic scoliosis. However, harvesting autologous bone increases operative time and morbidity. The purpose of this study was to compare autologous iliac crest and rib graft with recombinant human bone morphogenetic protein-2 (rhBMP-2) in thoracoscopically assisted anterior spinal arthrodesis and instrumentation in an animal model.

Methods: Twenty-two pigs underwent thoracoscopically assisted anterior spinal arthrodesis. Each animal had five contiguous thoracic discectomies followed by anterior instrumentation. The animals were randomly assigned to five treatment groups. Group I consisted of control animals that received no graft material; group II, animals treated with autologous rib graft; group III, animals treated with autologous iliac crest graft; group IV, animals treated with an rhBMP-2-composite sponge (collagen-hydroxyapatite-tricalcium phosphate carrier); and group V, animals treated with a composite sponge carrier alone. The animals were killed four months after the procedure, and the spines were harvested. The fusion mass was assessed with use of axial and sagittal computed tomography scans. The spines were tested biomechanically with incremental loads applied in the frontal and axial planes to achieve bending moments of up to 6.0 N-m. Angular motion at each segment was recorded with use of a three-dimensional motion analysis system. Histomorphometric analysis of each undecalcified disc segment was also performed.

Results: The fusion grades, according to computed tomography analysis with use of a 4-point grading system in which scores of 3 and 4 indicated a solid fusion, were 0.6 point for group I, 2.1 points for group II, 2.3 points for group III, 3.8 points for group IV, and 0.4 point for group V. Group IV (the rhBMP-2-treated animals) had a higher grade than all of the other groups. Group II (rib graft) and group III (iliac crest) had similar grades, and both were greater than group I (the untreated controls) and group V (composite sponge alone) (p < 0.05). In axial rotation, lateral bending, and flexion-extension, the spines in group IV were stiffer than those in the four other groups (p < 0.05); the spines in groups II and III were similar, and the spines in both of those groups were stiffer than those in groups I and V (the control groups). Histologic analysis demonstrated that the total new-bone area, expressed as a percentage of the total disc space area, was 23.2% in group I, 37.1% in group II, 37.2% in group III, 48.5% in group IV, and 5.9% in group V. Group IV had significantly greater bone formation than all of the other groups (p < 0.001). The animals treated with rib graft (group II) and iliac crest (group III) had a similar amount of bone formation, and it was greater than that in both control groups (p < 0.001).

Conclusions: The rhBMP-2 significantly increased the prevalence and quality of the spinal fusion after thoracoscopically assisted anterior arthrodesis and instrumentation in an animal model compared with that in the other treatment groups and in the controls.

Clinical Relevance: The use of rhBMP-2-composite sponge in the thoracoscopic approach to idiopathic scoliosis may decrease the morbidity from autologous bone-graft harvesting and increase the fusion mass, adding to the advantages of this minimally invasive technique.