The objective of this study was to investigate the effects of different doses rhBMP-2 on bone healing in an ovine lumbar interbody fusion model. In this study 22 sheep underwent two level lumbar interbody fusion using a ventrolateral approach with secondary dorsal fixation at L1/2 and L3/4. After randomization in one level a PEEK-cage was implanted filled with one of three doses rhBMP-2 (0,5mg; 1mg; 2mg) delivered on an ACS. The other level received an empty PEEK-cage or ACS filled cage. Animals were sacrificed after 3 and 6 months and decalcified histology was performed. This included histomorphological analysis well as histomorphometry of the tissues within the cage.

At 3 months after surgery the groups treated with rhBMP-2 showed higher amounts of bone tissue within the cage. At 6 months the amounts of bone tissue increased in all groups, were still lower in the groups without growth factor. At 3 months there was only one active osteolysis in the cage/ACS. 7 of 8 segments of the rhBMP-2 groups had a compromised bone structure around the implant. These areas were filled with fibrous tissue and fibrocartilage. This finding was not detected in the groups without rhBMP-2 at 3 months. At 6 months most of the segments with an empty cage or cage/ACS showed a chronic inflammation. Predominant cells were macrophages and giant cells. The groups treated with rhBMP-2 showed only a few mild chronic inflammatory reactions. The well-known dose dependent effect of rhBMP-2 on bone healing could also be recognized in our study. Attention has to be payed to the proinflammatory properties of the growth factor. Consistent with other studies we found 2 strong inflammatory reactions, each one in the lowest and highest dose group. Also, the potential for causing transient bone resorptions, according to the results of others, was demonstrated. At 3 months 7 of 8 segments treated with rhBMP-2 showed compromised peri-implant bone. Osteoblasts, but not osteoclasts, were seen in the periphery of these areas. It can be concluded that there where bone resorptions which already merged into an increased osteoblastic activity. Usually resorptions occur between 2 and 12 weeks and are followed by a period of increased osteoblastic activity. This finding wasn't recognized at 6 months anymore. Striking is that at 6 months most of the segments without rhBMP-2 showed a compromised bone structure around the implant with a mild to mainly moderate chronic inflammatory reaction. This cannot be attributed to the growth factor. Also, the ACS is degraded at 6 months and is unlikely a possible explanation. Therefore, the cage as a reason must be considered and it has to be questioned whether PEEK is the optimal material for interbody cages.