Bone Morphogenic Protein

Introduction
An arthrodesis (fusion) of adjacent vertebral segments of the spinal column is surgically created in nearly 250,000 patients per year in the United States. Despite recent technical advances, the incidence of failure of such intervention is still significant. Furthermore, substantial amounts of autogenous bone graft must be procured and transplanted from the ilium to the spine to enable consolidation of a fusion mass with elements of spinal column. The additional morbidity associated with this procedure is well known. Finally, secondary attempts at spinal fusion following initial failure have even poorer prognosis and pose greater challenges.

Advances in spinal fixation devices have addressed some mechanical considerations. However, osteogenesis and skeletal healing (fusion consolidation) are biologic processes which are facilitated but not replaced by instrumented fixation. Recently, considerable attention has been directed toward biologic considerations and, in particular, novel growth factors that regulate and promote these biologic processes. This renewed focus has been a consequence of the successful isolation, purification, and characterization of an entire family of bone morphogenetic proteins (BMPs), and the demonstration that a singular molecular species of recombinant human BMP (rhBMP) could induce the entire cascade of endochondral osteogenesis. Optimism for using such growth factors as effective bone-graft substitutes increased following a series of valuable pre-clinical animal experiments.

The concept of osteogenic bone proteins was introduced by Urist in 1965. His pioneering work demonstrated the osteogenic potential of demineralized bone segments after implantation at intramuscular sites in animals. Urist postulated that bone matrix must contain a substance of substances that induce the differentiation of pluripotential precursor cells along an osteogenic line.

Numerous bone-inducting proteins have been isolated from bone using a variety of purification procedures. As many as 14 different BMPs have been isolated. With the purification of these proteins in sufficient quantity and purity to provide amino acid sequence data, human complementary cDNAs were isolated and cloned.

Two bone morphogenic proteins have proven particularly capable of bone induction. These have experimentally been tested in spinal and long bone applications with dramatic success.

Preclinical Studies with rhBMPs
During the past four years several investigators have demonstrated that rhBMP-2 and rhBMP-7, combined with appropriate delivery vehicles, can induce successful posterolateral spinal fusion in rabbits, dogs, and rhesus monkeys without need for autogenous bone graft. The fusion rate and strength of fusion mass was equal or superior to the rate and strength of fusions produced with autogenous graft from the iliac crest. Cook and associates demonstrated that when using osteogenic protein-1 (OP-1 or rhBMP-7), posterior lumbar interlaminar fusions in dogs occurred in 12 weeks compared to 26 weeks when iliac crest bone was used. Sandhu and David, in separate studies, reported a 100% rate of lumbar intertransverse process fusion in beagles within six months of implantation of rhBMP-2 compared to 0% and 33% rate of fusion with autogenous iliac crest. Boden, using a rabbit lumbar intertransverse process fusion model with a four-week postoperative endpoint, reported a 100% fusion rate with rhBMP-2 and a 42% fusion rate with autogenous graft.

The efficacy of rhBMP-2 for producing anterior spinal column fusions was also explored. Sandhu, et, demonstrated that implantation of threaded titanium intervertebral "cages" containing rhBMP-2 without autograft in a sheep anterior lumbar interbody fusion (ALIF) mode3l consistently resulted in rapid intervertebral osseous fusion through the fixation device. Similarly, the same investigators showed that when rhBMP-2 was added to porous tantalum ALIF cages in the same model, significantly greater bone ingrowth into the porous macrostructure occurred.

Data obtained from these lower animal models prompted further investigation in the higher non-human primate model. Boden, et al recently demonstrated the efficacy of rhBMP-2 in a rhesus monkey ALIF model. Laparoscopic assisted L5-S1 ALIF procedures with threaded, tapered, titanium cages were performed in seven animals. A type I collagen sponge carrying high dose of rhBMP-2, low dose of rhBMP-2, or no growth factor was placed into the cage prior to implantation. Animals implanted with a cage containing rhBMP-2 (high or low dose) all had successful fusions. The control animals implanted with a cage containing collagen alone did not fuse. In another recently presented study, Fishgrund, et al, reported successful anterior cervical interbody fusions in a primate model using cadaveric allograft bone dowels filled with rhBMP-2 also carried by a type I collagen sponge. Histologic analysis of the rhBMP-2 associated cervical fusions revealed complete resorption of the allograft by six months, an observation not seen when the allograft bone dowels were filled with autogenous bone graft from the iliac crest.

Testing
Radiographic histologic and mechanical evaluation has been carried out in animal studies by numerous authors. Results indicate less fibrous ingrowth, greater bone formation and stronger fusions in torsion and stiffness at earlier time periods. The overall rate of fusion is higher.

Clinical Trials with rhBMP-2 Early Observations
Based upon the promising results obtained from the preclinical studies, carefully selected and controlled clinical trials with rhBMP-2 have begun. Each of the studies is still in progress and formal interpretation and presentation of the data has not been performed. Some preliminary observations have been made and are included in the summary below.

A clinical trial has been designed to examine the dose-response relationship of rhBMP-2 carried on a fibrillar collagen vehicle in a human skeletal defect site. A controlled defect is made in the anterior iliac crest during tricortical bone harvest for anterior cervical discectomy and fusion. The collagen sponge is implanted in the iliac crest site in each use. One of three specific doses of rhBMP-2 in buffered solution are combined with the collagen sponge prior to implantation. A soak time of 30 to 90 minutes is used during preparation of the composite device. A sham condition in which the sponge has not been combined with the rhBMP-2 is also used. CT examination of the iliac crest site along the rhBMP-2/collagen implant (Figure 1). Comparison of bone formation associated with the different doses of rhBMP-2 including the sham condition has not been completed.

A clinical trial examining the use of a rhBMP-2/collagen sponge composite combined with a threaded titanium interbody fusion cage for anterior lumbar interbody fusion has been completed. No autogenous bone graft was used in any patient. A total of fourteen patients were enrolled. CT images of the fusions were obtained three and six months following surgery. These images demonstrated de novo bone formation within the cages bridging adjacent vertebral segments. Bone formation anterior, posterior, and lateral to the cages were also noted (Figure 2 and Figure 3). Formal evaluation of fusion status and clinical outcome has not yet been completed.

A clinical trial examining the use of rhBMP-2 in posterior lumbar fusion surgery for the treatment of degenerative disease of the lumbar spine is underway. Evidence of de novo bone formation attributable to the growth factor is present in follow up plain radiographs of the lumbar spine (Figure 4). Formal evaluation of these cases and their outcomers is currently in progress.

Conclusion
At this time, evaluations of the clinical applications of the growth factor rhBMP-2 for the spinal column are in the earliest stages of study. Early observations suggest that this growth factor is capable of inducing de novo bone formation in the spinal environments. Formal examination of these clinical cases, however, is pending and necessary.