Lumbar spinal fusion with β-TCP granules and variable Escherichia coli-derived rhBMP-2 dose

Background context The ideal tissue-engineered solution for any bone graft substitute is to assist in the rapid formation of bone and facilitate fusion. Purpose The present study aims to evaluate this E-BMP-2 (Escherichia coli-derived human bone morphogenetic protein-2) in ovine posterolateral lumbar fusion (PLF) to examine the influence of dose and overall performance in a model with similar graft size and diffusive challenges to the human. Study design/setting In vivo large animal model study. Methods An adult ovine PLF was performed in 30 animals with groups of E-BMP-2 with a beta-tricalcium phosphate (β-TCP) carrier at three different dosages, β-TCP alone, and autograft from the iliac crest. The fusions were assessed by radiography (X-ray and microcomputed tomography), mechanical testing, and hard-tissue histology with bone labels at 6, 8, and 10 weeks along with routine paraffin histology at 12 weeks. Results Results showed increasing new bone and fusion rate with E-BMP-2 dose, whereas β-TCP alone was largely resorbed and did not achieve fusion in this model at 12 weeks. Autograft showed similar grading for the amount of bone between the transverse processes but a lower fusion rate than β-TCP/E-BMP-2 groups. Bone labels revealed new bone formation at all time points for the E-BMP2 groups, whereas the autograft group showed active bone formation at 10 weeks. Beta-tricalcium phosphate displayed reliable incorporation into the decorticated host bone, whereas limited new bone was found between the transverse processes. At the center of the fusion mass, increased E-BMP-2 dose led to increased incorporation of β-TCP by new bone. Conclusions These results suggest that E-BMP-2 was capable of producing posterolateral fusion in the ovine model that is equal to or superior to autologous graft in terms of fusion rate and mechanical strength. E-BMP-2 dose had considerable influence on β-TCP granule resorption.