Bone morphogenetic proteins (BMPs) are the most potent osteoinductive agents discovered to date. BMPs possess many times the osteoinductive activity of currently demineralized bone matrix (DBM) and blood-derived gel products. Two proteins, recombinant human (rh)BMP-2 and rhBMP-7 (also known as osteogenic protein-1 [OP-1) have recently received US Food and Drug Administration (FDA) approval for use to promote interbody spine fusions and to treat long-bone nonunion fractures, respectively. Although these approvals are an important milestone in orthopedic science, it is also a time to pause and examine unresolved issues so as not to misuse a clinical treatment modality in its infancy despite over three decades of research.
Preclinical studies of rhBMP-2 and rhBMP-7 using an array of animal species and surgical models have consistently yielded results as good as, or in most cases, superior to that obtained with autogenous bone graft. The use of rhBMP-2 to promote interbody and posterolateral spine fusions was demonstrated to be superior to autogenous bone graft in a variety of animals and human models 1, 2. Similarly, rhBMP-7 was shown to be capable of healing segmental bone defects in primates, which would not heal when treated with autogenous bone graft [3]. The need for autogenous graft and its associated donor site morbidity and quantity limitations appeared ended. The ability of BMPs to induce new bone and treat a variety of orthopedic indications appeared limitless.
Unfortunately, clinical results with BMPs have not been as conclusive as those in preclinical studies. The prospective clinical trial of rhBMP-7 did not achieve equivalent healing of tibia nonunion fractures compared with autogenous iliac crest bone graft. In a second trial of rhBMP-7 to treat other long-bone nonunions, successful healing was obtained in only 1 of 10 cases [4]. FDA approval for rhBMP-7 has been limited by the FDA to a humanitarian device exception. In its clinical trial rhBMP-2 was able to show equivalency, but not superiority, to autogenous bone graft in achieving anterior interbody spine fusions when placed in a titanium cage [5]. The clinical results with rhBMP-2 and rhBMP-7 were obtained in carefully designed and well-controlled trials, by investigators experienced in use of BMPs. It is unlikely that equivalent or superior clinical results will be obtained when BMPs are used routinely in the general orthopedic community.
The discrepancy in results observed between preclinical and clinical studies of BMPs may not be as simple as differences in the healing potential of young, healthy laboratory animals compared with human subjects. Limited and, in most cases, no information exists on the proper or optimal dose of BMPs in humans. No dosing data are available from an actual clinical study. Some limited dose–effect studies in animals have been reported for rhBMP-2 and rhBMP-7. However, Boden et al. and Martin et al. 6, 7 have warned that significant increases in dose are necessary in scale up from animals to humans, although no known formula for this increase exists. Although the relative activity of rhBMP-2 compared with rhBMP-7 is unknown, the two FDA-approved products contain significantly different amounts of active protein. It is possible that the lack of efficacy in some clinical trials may be related to suboptimal dose rather than the inability of the BMP to promote healing. It is likely that different indications will require not only different doses of proteins but potentially different carrier materials to obtain clinical efficacy. The rapid-protein-release kinetics of current carrier materials necessitate not only extremely high doses of the proteins but also limit potential indications because of compression of the matrix. It has been suggested that the current rhBMP-2 product indicated for interbody fusion may not be appropriate or as efficacious in posterolateral fusions. The use of the approved BMP products outside their very limited and specific indications is unsupported and may result in clinical failure.
The use of BMPs are not without potential significant complications. A great deal is unknown about these powerful therapeutic agents. Because of the powerful osteoinductive potential of BMPs the possibility of ectopic bone formation exists, particularly if the site of implantation is not well contained. Exuberant bone formation is often observed in the soft tissues in animal studies. The clinical trial of rhBMP-2 placed in a cage posteriorly was discontinued after bone formation within the canal was observed requiring reoperation in some patients. Whereas too much bone or misplaced bone formation may result in complications, it appears that too little bone formation with a BMP may have additional problems. Laursen et al. [8] reported significant bone resorption resulting from increased osteoclast activity with the use of rhBMP-7 when sufficient early new bone support was not obtained in clinically unstable thoracolumbar burst fractures.
Of unknown risk is the role of antibodies that develop to the BMPs in some patients after implantation. These antibodies may affect protein activity and occurred in 38% of patients receiving rhBMP-7 in its clinical trial. Antibody formation was also reported in the clinical trial of rhBMP-2 to a lesser degree. Little is known of the effect on bone induction of these antibodies and may be significant in future bone healing, particularly upon repeat applications of a BMP, Interestingly, Urist et al. [9], credited with the initial discovery of BMPs, observed increased levels of antibodies to BMPs in patients with osteoperosis. Urist et al. speculated osteoperosis may be a BMP autoimmune disorder. In addition, BMPs play a vital role in fetal development, and the loss of rhBMP-2 and rhBMP-7 activity has been shown to result in fetal death. The current BMP products are contraindicated in pregnant women, and women are advised to use contraception for 1 year after implantation of rhBMP-2 and rhBMP-7.
Finally, the use of BMPs in tumor patients is unclear. BMPs have been identified and isolated in osteosarcomas. Because BMPs are powerful differentiation factors, extreme caution should be exercised when contemplating BMP use in patients with tumors. The current BMP products are contraindicated and should not be used at or near the site of resected tumor or in patients with a history malignancy. Much research is needed before the clinical use of BMPs in these patients.
Although much is known about BMPs and their role in bone formation, much is unknown or unclear. The FDA-approved rhBMP-2 and rhBMP-7 products should be used only for the specific clinical indication studied and the manner described for use. The warnings, contraindications and possible future side effects of these products must also be carefully considered. As with any new technology, the cost of the BMP devices must be addressed in light of clinical benefit. Recently, Medicare denied add-on payment for rhBMP-2, citing a failure to demonstrate substantial added clinical improvement. BMPs offer the potential to dramatically change and improve a variety orthopedic treatments. However, their limitations and potential complications must be clearly understood and, until further research and clinical study is available, the BMP products should be used with caution.
Finally, there will be many surgeons who will assume that BMP will allow them to ignore basic principles of the biology of bone healing. Disregard for these principles, such as good fusion bed preparation, internal fixation, correction of metabolic abnormalties and so forth, will lead to failed surgery.
BMP is not a product that allows surgeons to take short cuts, and those patients who are subjected to this thinking will do poorly.
FULL TEXT