Secondary Logo

Journal Logo

Focus Issue

Purified Bovine BMP Extract and Collagen for Spine Arthrodesis: Preclinical Safety and Efficacy

Damien, Christopher J., PhD*; Grob, Dieter, MD; Boden, Scott D., MD; Benedict, James J., PhD*

Author Information


Bone morphogenetic proteins (BMPs) have been in preclinical evaluations since their discovery by Dr. Marshall Urist and colleagues. 23–27 These proteins demonstrate osteoinductive properties that suggest their usefulness where bone formation is desired. In general, three BMP products are being developed and have demonstrated promising preclinical efficacy results. These products include two human recombinant BMPs and one mixture of BMPs extracted from bovine bone tissue. The recombinant forms, rhBMP-2 and OP-1 (rhBMP-7), are single factors produced using mammalian cell culture techniques and then purified. 17,28 bovine BMP extract (bBMPx) is a highly purified protein extract from bovine bone that contains BMP-2, BMP-3, BMP-4, BMP-5, BMP-6, and BMP-7; TGFβ1, TGFβ2, and TGFβ3; FGF-1; and other noncollagenous proteins. 1

bBMPx product, a composite of bBMPx and collagen, has demonstrated good spinal fusion results using a posterolateral intertransverse process fusion model in the rabbit. 3 Boden et al 4,5 have demonstrated a dose response (to 100% fusion) with bBMPx product in the rabbit, as illustrated in Table 1. The biologic power of this process also has been demonstrated in an impaired healing environment induced by the inhibitory effect of nicotine. 22 bBMPx product was able to overcome this adverse effect and enhance bone formation in a rabbit model. 21

Table 1
Table 1:
Rabbit Spinal Fusion Results With bBMPx product

In humans, semipurified BMP extracts has been used to heal segmental bone defects 12,29 or to accelerate fracture healing. 13 Reliable bone formation has been observed. In the past, BMP has been used only in combination with anterior interbody cages implanted into the lumbar spine. 19 Osteoinduction in humans was proven in these previous experiments. However, the use of BMP with the cages has provided an optimal biologic and mechanical environment that does correspond exactly to that of the animal models with posterolateral intertransverse process (PLITP) fusion.

The current studies were undertaken to demonstrate the safety of bBMPx product when placed near the spinal cord in the rabbit spine after laminectomy and the dose response efficacy of bBMPx in bBMPx product in a nonhuman primate lumbar PLITP spine fusion model. The results from these studies were used to justify a pilot study with humans. Preliminary human spine fusion results are described.

Materials and Methods

Part A: Rabbit Spine Safety Study.

For this part of the study, 54 New Zealand white rabbits were used to examine the safety of placing demineralized bone matrix (DBM) with bBMPx product near the spinal cord (Table 2). bBMPx product is composed of bBMPx and Type 1 collagen (Ne-Osteo; Sulzer Biologics, Austin, TX). Bovine BMP extract is a combination of proteins purified from bovine bone. Type 1 collagen, purified from bovine tendon, acts as carrier matrix in bBMPx product. This study was conducted under Good Laboratory Practices (GLP) at Bio-Research Laboratories (now ClinTrials Bioresearch—Inveresk, Senneville PQ, Canada).

Table 2
Table 2:
Distribution of Animals in the Rabbit Spine Safety Study

Preparation of Autograft.

Iliac crest bone was harvested from both crests of each rabbit as described previously. 3–5 Approximately 2 g of autogenous bone occupying a volume of 3 mL was implanted on each side of the spine.

Preparation of Demineralized Bone Matrix.

The DBM was prepared from the long bones of rabbits after they had been cleaned of soft tissue and marrow. The bone was crushed in a Wiley mill in liquid nitrogen and meshed to a particle size range of 125 to 500 μm. The particles were demineralized in 0.6 N of hydrochloric acid, washed with deionized water and ethanol, and lyophilized. Samples of these materials were weighed (1.2 g) and placed in sterilized 15-mL centrifuge tubes in a laminar flow hood under aseptic conditions. The samples were maintained at refrigerated temperatures until surgery.

Preparation of Bovine BMP Extract.

The extract 5 of bBMPx was supplied by Sulzer Biologics. To prepare the bBMPx, cortical diaphyses of bovine long bones were cleaned of soft tissue and marrow. The cleaned bones were pulverized and then demineralized in 1 N of hydrochloric acid for 8 hours at 25 C. The resulting particles were washed in deionized water followed by extraction using a solution of 4 mol/L guanidine hydrochloride buffered with 0.1 N Tris, pH 7.6, for 48 hours at 15 C. The extracted proteins were purified using a 100-K nominal molecular weight cutoff ultrafilter followed by a 10-K nominal molecular weight cutoff ultrafilter. Additional purification was completed using ion exchange chromatography and reverse-phase high-performance liquid chromatography (HPLC).

Preparation of bBMPx product (Collagen With bBMPx).

bBMPx product consists of bovine Type 1 tendon collagen (Regen Biologics, Redwood City, CA) combined with bBMPx in a 3% collagen dispersion with 5 mmol/L sodium phosphate buffer, pH 3.7. After bBMPx at a dose of 1.5 mg per 45 mg dry product was added, the mixture was adjusted to pH 3.7, quick frozen in liquid nitrogen, and lyophilized overnight. The resulting sponge was shredded in isopropyl alcohol using a Waring blender and vacuum dried, after which 45 mg was placed in a 3-mL syringe. The bBMPx product was packaged and sterilized with ethylene oxide (EtO).

Implant Materials (DBM With bBMPx product).

At surgery, the bBMPx product was hydrated in the syringe using 1.5 mL of sterile water for injection (WFI). The resulting gel was mixed with 1.2 g of rabbit DBM that had been hydrated with 3.2 mL of sterile WFI. The rabbits received the total 1.5-mg dose of bBMPx in a 6-mL volume divided equally (0.75 mg/3 mL) to both sides of the spine. This represented five times the effective dose of bBMPx, which is 0.15 mg per side (Table 1).

Surgical Procedure.

The study animals were anesthetized with ketamine (50 mg/kg) and xylazine (5 mg/kg) before the lower half of the back was shaved. The shaved areas were washed with Hibitane followed by a liberal application of Betadine. A bone densitometer unit (DEXA) was used to locate L4–L5. The animals were returned to the surgical suite and placed on isofluorane anesthetic for the duration of the surgical procedure.

After a 0.25% Marcaine injection, a longitudinal midline excision was made in the lumbar region to expose the lamina of L4 and L5 down to the lumbosacral fascia. The fascia and paraspinous muscles were dissected from the spinous process, and a laminectomy was performed using an air-driven drill with a burr bit. The laminectomy defect measured approximately 4 mm × 8 mm. Hemostasis was obtained using Gelfoam and a bipolar cauterizer when necessary. The ligamentum flavum and epidermal fat were removed, leaving the spinal cord exposed for the full extent of the laminectomy. After lavage, the bBMPx product or autograft was placed in the gutter on either side of the spinal cord. The lumbosacral fascia was sutured and the skin closed. Nine animals served as untreated control subjects, and 15 animals had laminectomies but no implant material, as shown in Table 2.


Clinical Examinations.

After randomization, all the animals were observed once daily before surgery for clinical signs of ill health, behavior changes, or reaction to treatment. A mortality check was performed at least twice daily throughout the study. More frequent examinations were undertaken when deemed necessary. A detailed physical examination was performed on all animals once before surgery and once weekly starting 1 week after surgery, except for a few animals on isolated occasions.

Cage-side neurobehavioral observations included body position, urination, locomotor activity, defecation, bizarre or abnormal behavior, respiration, tremors, nasal discharge, twitches, salivation, convulsions, lacrimation, and piloerection. Once before initiation of treatment and for the first 4 weeks after surgery, the neurobehavioral observations were performed daily on all the animals, and the individual observations recorded. After 4 weeks, these examinations were performed once weekly for the remainder of the study.

A detailed neurobehavioral examination in the arena was performed on all the animals on day 7 (week 1), day 28 (week 4), and day 56 (week 8). Neurobehavioral examinations in the arena included locomotor activity, bizarre behavior, skin sensitivity test, ataxic gait, hind paw reflex, urination, and defecation.

Tissue Preservation and Processing.

Five animals per group from Groups 2, 3, and 4 and three untreated control animals (Group 1) were killed on days 8, 29, and 57 (18 animals per occasion). One animal from Group 3 (#3113) and one animal from Group 4 (#4072) were killed before the scheduled sacrifice (on days 13 and 28, respectively) because of complications.

All the animals were perfused for histopathologic evaluation. They were deeply anesthetized by an intravenous injection of sodium pentobarbital. While the animals were under anesthesia, the thorax was opened, a 14-gauge needle was inserted into the left ventricle, and the right atrium was opened. Perfusion with lactated Ringer’s solution containing heparin (1000 IU/L) and sodium nitrite (0.02 g/L) was initiated and continued until the auricular effluent was essentially free of blood. The perfusion fluid then was changed to 10% neutral buffered formalin. Partial gross examination was performed on all the animals. After perfusion, the lumbar site was removed en bloc from L3 through S1 so that the laminectomy site was entirely retained. The cranial portion was marked with indelible ink, and the specimens were immersed in 10% neutral buffered formalin for a minimum of 48 hours. Radiographs of the explants were performed on all the animals.

For each specimen from the laminectomy site, two complete transverse sections (bone and spinal cord included in the same section) were taken before decalcification. In addition, one extra transverse section of the spinal cord from the laminectomy site was submitted separately from the bone.

After decalcification, the samples were processed, embedded in paraffin, and sectioned. The sections were stained with hematoxylin and eosin, Masson trichrome stain, and Klüver-Barrera. Where radiographs showed extensive bone proliferation, two extra sections were taken 1 cm apart from each extremity of the laminectomy site.


The laminectomy site of each animal was evaluated histopathologically, with particular attention given to the nervous tissue changes. Examination of the sections included identification of changes resulting from bone formation or ossification of the neural elements or others nearby. Any narrowing of the spinal canal was recorded. In addition, an evaluation of inflammation at the defect site was performed.

Part B: Efficacy Study of bBMPx product in Nonhuman Primates.

For this part of the study, 54 adult rhesus macaques, age, 6–15 years (Table 3), were obtained from the Yerkes Regional Primate Center, Atlanta, Georgia. Animals that had been involved with previous studies of bone-related hormone or studies involving growth factors or cytokines, and those that had any radiographic evidence of spinal disease in the lumbar spine were excluded. The animals were housed in an AAALAC-accredited facility, one per cage, and fed ad libitum. They were inspected daily for general health and neurologic examinations and weighed weekly.

Table 3
Table 3:
Summary of Rhesus Implantations

Preparation of Monkey Demineralized Bone Matrix.

Femurs, tibias, and humeri from the rhesus monkeys were shipped on ice from the Yerkes Primate Center to Sulzer Biologics. The bones were cleaned of soft tissue, and the ends were removed. Only the midshafts were used. Segments were crushed and cleaned of marrow. Chips were placed in ethanol for approximately 1 hour, lyophilized overnight, frozen in liquid nitrogen, and crushed to a size of 125 to 850 μm using a cooled Wiley mill. Particles (100 g) were added to 0.6 mol/L of HCl (2 L) at 4 C. After overnight demineralization, material was rinsed once with fresh acid, three times with deionized water (DI water), once with 0.1 mol/L of sodium phosphate buffer (100 g per 2 L of DI water at pH 7.4), and finally three times with DI water. Demineralized matrix was soaked in ethanol for 1 hour, air dried in a laminar flow hood, and vacuum dried overnight. Thereafter, 3 g of DBM was weighed and placed into stoppered vials. The samples were sent to AlloSource, Denver, Colorado, for EtO sterilization.

Preparation of bBMPx product.

Various experimental lots of bBMPx product were made, which consisted of bovine Type 1 tendon collagen combined with bBMPx in a 3% collagen dispersion with 5 mmol/L sodium phosphate buffer. To these was added bBMPx at doses of 0, 0.05, 0.15, 0.5, 1.5, 3, or 5 mg per side of the spine. The dispersion was frozen in liquid nitrogen and vacuum dried. The resulting sponge was shredded in isopropyl alcohol with a Waring blender and vacuum dried. Then 135 mg of dried product was placed in a 10-mL syringe. The bBMPx product was EtO sterilized.

At surgery, bBMPx product was hydrated with 4.5 mL of sterile WFI. The resulting gel was mixed with 3 g of monkey DBM that had been hydrated with 3 to 4 mL of sterile WFI to form a moldable paste. This 10- to 11-mL volume of material was split into two equal parts of ∼5 mL and implanted.

Intertransverse Process Fusion Surgical Technique.

Adult rhesus macaques weighing 10 to 16 kg were obtained. A single-level posterolateral intertransverse process fusion was performed bilaterally at L4–L5. The monkeys were anesthetized by an intravenous injection of Telazol (4 mg/kg) supplemented with pentobarbital (2 mg/kg). They then were shaved and prepared with Betadine. During surgery, anesthesia was maintained with Halothane up to 2%. A dorsal midline skin incision was made followed by two paramedian fascial incisions. The intermuscular plane between the multifidus and longissimus muscles was developed to expose the transverse processes of L4 and L5 as well as the intertransverse membrane. 11 Two separate fascial incisions were made to harvest 2 to 2.5 mL of corticocancellous bone graft from each iliac crest. The harvested autograft then was morselized with a rongeur. An electric burr (Stryker, Kalamazoo, MI) was used to decorticate the transverse processes. The iliac bone graft was placed between the transverse processes in the paraspinal bed. The fascial incisions were closed with 3.0 absorbable suture, and the skin was approximated using 4.0 prolene suture. For animals receiving DBM with or without bBMPx instead of autograft, the bone from their iliac crests was discarded after harvest (i.e., sham graft harvests were performed). Postoperative analgesia was controlled with buprenorphine hydrochloride for 48 to 72 hours. No perioperative antibiotics were used.


Spine fusions were evaluated via several methods: manual palpation, posteroanterior radiographs, computed tomography (CT), and undecalcified histology.

Manual Palpation.

At the time of harvest and after removal, the lumbar spines were palpated manually at the level of the fused motion segment, and at the levels of the adjacent proximal and distal motion segments. This simulated fusion exploration and palpation in humans, the gold standard for distinguishing nonunions from solid fusions in humans. Each motion segment was graded as solid or not solid. Only the levels graded as solid were considered for fusion.

Radiographic Analysis.

Posteroanterior radiographs were made of all the specimens after sacrifice. The tube-to-plate distance was 90 cm. The radiographs were read in a blinded fashion, and the fusions were graded as solid or not solid on the basis of the trabecular pattern in the intertransverse fusion mass.

Computed Tomography.

After excision, selected specimens from the 6-mg or higher-dose and autograft animals were stored frozen at −70 C until scanning. Computed tomography scans of the lumbar spine fusion mass were obtained on a high-speed CT scanner (General Electric, Milwaukee, WI) using the following parameters: 10-cm DFOV, 100 kV, 150 mA, 1-mm gap, and 1-mm slice thickness. The images then were qualitatively assessed.

Light Microscopic Analysis.

Lumbar spines were harvested from all the monkeys and fixed for at least 24 hours in 10% neutral buffered formalin. The specimens then were cut in half longitudinally in the sagittal plane and fixed for 1 week in 70% ethanol. After fixation, the specimens were dehydrated sequentially in 95% and 100% ethanol, then cleared in xylene. The undecalcified specimens were embedded in methylmethacrylate and sectioned on a Jung Model K Polycut microtome at a thickness of 5 μm. The left side of each fusion level was sectioned in the coronal plane, and the right side was sectioned in the sagittal plane. Contiguous sections were stained with Goldner trichrome.


Part A: Rabbit Safety Study

Two animals, one each from Groups (laminectomy and bBMPx product) and Group 4 (laminectomy and autograft), were killed before the study was completed because of automutilation.

Clinical Evaluation.

Daily examinations showed swelling at the surgical sites in 16 of 45 animals. As illustrated in Table 4, neurobehavioral assessments performed in the arena on days 7, 28, and 56 (before the animals were killed) showed impairment or slight impairment of hind paw reflex, hindlimb paralysis, and reduced skin sensitivity in the hindlimbs of some animals. No abnormal findings were noted in the untreated controls.

Table 4
Table 4:
Clinical Results of the Rabbit Spine Safety Study

Histopathologic Evaluation.

The defect site was present and intact in all the animals on day 8. Slight to marked repair of the defect site was observed in two of five animals in Group 2, five of five animals in Group 3, and four of five animals in Group 4 (Table 2) on day 28, and in none of four animals in Group 2, three of four animals in Group 3, and three of four animals in Group 4 on day 56. Variable degrees and patterns of inflammation were seen in the defect site of almost all the animals of Groups 2, 3, and 4 killed on days 8, 29, and 57. The nonrepair of the defect site observed in one Group 4 animal on day 28, and in four of four animals in Group 2, one of four animals in Group 3, and one of four animals in Group 4 on day 56 was attributed to an inflammatory reaction or a broken piece of bone within the defect site. Moreover, the inflammation was considered to be the major cause of spinal cord compression seen in the animals of Groups 2, 3, and 4 killed on days 8 and 29, and in one animal of Group 2 killed on day 57.

There was a decrease in the severity of the inflammatory reaction over time, which paralleled the decrease in the incidence of spinal cord compression, from 10 of 15 animals on day 9 to only one animal of Group 2 on day 57. In three animals of Group 3 killed on day 29, the compression of the spinal cord was related to bone proliferation in the vertebral canal. Degeneration of the spinal cord was seen mainly in the animals of Groups 3 and 4 on days 8 and 29, and generally was associated with spinal cord compression. This degeneration was not considered to be related directly to the test material, but was regarded as secondary to the inflammatory reaction from the surgical procedures alone or surgery combined with bone proliferation at the defect site. Degeneration alone was seen in one animal from Group 4 on day 57.

Part B: Nonhuman Primates

The efficacy of bBMPx product in nonhuman primates was evaluated in two separate studies. In the first study, 48 animals were tested for an initial dose response to 5 mg of bBMPx per side of the spine. In the second study, and six animals were used to test the effect of placing bBMPx product after a laminectomy. The results were pooled to demonstrate the dose response. Fusions were not noted before 18 weeks. Therefore, only the animals killed at 18 or 24 weeks were included in the calculation of the fusion rate.

The fusions were evaluated by manual palpation and radiography. It appears that autograft has a lower fusion rate in rhesus than in humans. Only 3 of 14 animals fused (21%) at 18 weeks or longer.

Animals receiving monkey DBM alone or with bBMPx doses of 0.05, 0.15, 0.5 or 1.5 mg per side did not fuse in any case (Table 5). Increasing the dose to 3 or 5 mg increased the rate of fusion from 0% to 54%.

Table 5
Table 5:
Fusion Results in the Rhesus at 18–24 Weeks Postsurgery

Histologic Examination.

Autograft Response.

The results of the autograft control samples at 12 weeks showed only minimal new bone at the transverse processes (TPs) and surrounding autograft particles. By 18 weeks, most of the autograft particles had been resorbed, with a small amount of reactive bone at the TPs.

The 24-week results were mixed. Two samples had some bone at the TPs, but there was a large gap between the TPs. Three samples had near bridging, but a mass of soft tissue still was evident (Figure 1). One sample had large bilateral fusion masses. In all cases, resorption of autograft particles had progressed.

Figure 1
Figure 1:
Implantation of autogenous iliac crest bone in the monkey lumbar spine resulted in a mass of bone formation after 24 weeks. Islands of bone are seen between the transverse processes (TP). There remains a layer of soft tissue between the islands, suggesting a nonunion. (Goldner’s trichrome, ×1)
Dose Response.

In the DBM samples with no bBMPx, there was very little new bone formation except for some reactive bone at the decorticated TPs. The DBM particles at the TPs were surrounded by bone, with little to no inflammatory response. Unresorbed DBM particles were observed even at 24 weeks (Figure 2). Similar results were seen in samples containing DBM and 0.05 mg of bBMPx, and those containing DBM and 0.15 mg of bBMPx. In some cases, small islands of bone were seen. In others, only bone at the TPs was visible. Unresorbed DBM particles remained at the defect site even at 24 weeks.

Figure 2
Figure 2:
Photomicrograph of a demineralized allograft bone matrix (DBM) explant with no bovine BMP extract (bBMPx) placed as a bone graft substitute for a lumbar spine fusion after 24 weeks. Some reactive bone (blue) is seen near the edge of the implant. Unresorbed DBM particles (red) remain. (Goldner’s trichrome, ×4.3)

Histologic specimens of the DBM and 0.5 mg of bBMPx showed larger islands of bone formation at 12 and 24 weeks. In one case, one third of the treated area between the TPs was bone. Although these samples did not demonstrate a continuous fusion mass, they did provide evidence of an osteoinductive effect of the bBMPx exceeding that of DBM alone. Results similar to those of the DBM and 0.5 mg of bBMPx were seen with DBM and 1.5 mg of bBMPx. There were more extensive islands of bone formation, and in one case, there was only a 5-mm gap between the bone forming from the TPs. In this case, a corticalized rim of bone encircled bony trabeculae and unresorbed DBM particles.

At 12 weeks, the samples of DBM and 3 mg of bBMPx showed a fair amount of bone formation surrounding unresorbed DBM particles. There was near bridging of the TPs with corticocancellous bone and marrow. By 24 weeks, a large mature fusion mass was evident with a cortical ring of bone, trabeculae, and marrow at the interior in one sample. Only a few DBM particles remained, and they were surrounded by new bone (Figure 3).

Figure 3
Figure 3:
After 24 weeks, a large fusion mass (blue) was observed bilaterally after implantation of bBMPx product with 6 mg of bovine BMP extract (bBMPx). A thick cortical rim of bone surrounds cancellous bone and marrow and attaches the transverse processes (TP). Only a few unresorbed demineralized allograft bone matrix (DBM) particles (red) remain. (Goldner’s trichrome, ×1)

Two 3-mg samples had a large mass of DBM particles, but did not show bone formation except at the TPs. Multinucleated giant cells (MNGC) in the process of resorbing residual DBM particles were observed. Two samples had little to no DBM remaining.

At 4 weeks, the bBMPx product with 5 mg of bBMPx demonstrated a large mass of DBM particles, with those at the edge surrounded by MNGC. By 8 weeks, bone induction was evident around various particles. An aggressive inflammatory response also was observed.

At 12 weeks, the 5-mg samples showed near bridging in one animal and extensive bilateral fusions with large masses in the second animal. In this animal, there was a mature rim of cortical bone and trabeculae, marrow, and unresorbed DBM particles.

The 18-week results showed that when some DBM remained, there was a robust bone-forming response. In one case, a classic osteoinductive response resulting in a solid fusion mass was observed. The samples with pseudarthroses, according to radiograph, had large masses of induced bone and a cartilaginous pseudarthrosis. The one sample without bone formation exhibited a total resorption of DBM particles.

By 24 weeks, results varied from little to no bone and an absence of DBM particles to a classic fusion response with mature bone (Figure 4). In general, when there was no bone formation, there also were no DBM particles present. When bone formed, it often surrounded the DBM particulate that remained in the site. These fusion masses demonstrated mature bone throughout the mass with a cortical edge and trabeculae in the central regions. What appeared to be a neoperiosteum forming at the cortical edge was suggestive of a maturating fusion mass. There were large numbers of osteoclast-like cells, further evidence of maturation. There did not appear to be any residual inflammatory response at 24 weeks because most of the DBM particles had been resorbed or remodeled into the new bone.

Figure 4
Figure 4:
Sagittal view photomicrograph of a fusion mass from a monkey 24 weeks after implantation using bBMPx product with 10 mg of bovine BMP extract (bBMPx). Mature bone (blue) can be seen throughout and attaches the transverse processes (TP). Only a few demineralized allograft bone matrix (DBM) particles remain. (Goldner’s trichrome, ×1)

Laminectomy Study.

Six animals were used in this part of the study. Two received laminectomy alone; two had laminectomies with autograft; and two had laminectomies with bBMPx product at 5 mg of bBMPx. Histologically, the laminectomy-only group showed no evidence of significant lamina regrowth in either animal. There was no compression of the neural elements inside the spinal canal. The laminectomy site was filled mostly with fibrous tissue.

Similarly, the autograft group demonstrated no evidence of significant regrowth. There was no compression of the neural elements inside the canal.

In the animals that received bBMPx product with 5 mg of bBMPx, there also was no evidence of regrowth of the lamina in either specimen. There was formation of bone (2 of 2 fusions) along the remaining lamina and spinous process away from the opening of the spinal canal, but it did not result in any impingement of bone into the canal or neurologic deficits.


Since the studies of Urist 24 in 1965, numerous reports on experimental work with BMP have been published, the vast majority showing a positive effect of bone formation in animal models. 2,4,5,7–10,12–15,17–20 For clinical application, both safety and efficacy must be demonstrated. Determination of safety involves a variety of preclinical tests, some of which are defined by the International Standards Organization (ISO). Document ISO 10993 entitled “Biologic Evaluation of Medical Devices”11 outlines various tests required for medical devices as well as additional tests for consideration that are product or indication specific. As bBMPx product was being used for lumbar spinal fusion, research was conducted to place a bBMPx product near the spinal cord to demonstrate its safety for these applications.

A rabbit model 3 was adapted using techniques from Saito et al. 16 In their study, a chronic compression myelopathy was induced in the rabbit by implanting extracted rabbit BMP with autograft onto a thinned ligamentum flavum. This study demonstrated early pathologic changes caused by the compression.

The current study used a similar technique, but placed material as for a spinal fusion and not directly on the ligamentum flavum. Initial compression was observed in all the surgically treated groups: laminectomy alone, laminectomy with autograft, and laminectomy with bBMPx product. The compression resolved itself over time as the inflammation decreased and was evident only in one animal in Group 2 (laminectomy-alone group) on day 57. Degeneration also was seen as transient, with only one animal in Group 4 (autograft group) demonstrating degeneration on day 57. These results indicate that whereas bone forms with bBMPx product, there is no difference with its use between the laminectomy procedure itself and autograft in terms of spinal cord compression or degeneration, and that it is safe to use bBMPx product near a laminectomy defect.

Preclinical efficacy also is important for medical products. Rabbits have demonstrated good bone formation with bBMPx product in the spine. 4,5 Dogs also have been used to demonstrate good healing of bone with BMP-containing products in spinal fusion models. 9,19 Higher animal data, however, remains the benchmark for dose evaluations of osteoinductive bone growth factors. In the human spinal fusion procedures, pseudarthroses are seen in 5% to 35% of the patients. 2 The current study demonstrates a 69% pseudarthrosis rate with autograft in the rhesus, suggesting that this model is very difficult to fuse. Boden et al, 2 in a study that tested BMP-2 with a ceramic composite in four animals, demonstrated that none of them fused with autograft. With bBMPx product and DBM, a dose response was seen that demonstrated twice the fusion rate of autograft with 3 mg of bBMPx per side, and higher rates (2.5 times higher) with 5 mg per side. Lower doses induced bone formation, but did not result in fusion. In comparison, the fusion rate with 3 mg of recombinant human BMP-2 (rhBMP-2) on ceramic was 100%. However, the carrier in this study also demonstrated a 100% (2/2) fusion rate, suggesting that the carrier did not require BMP to cause fusion.

The results of the rhesus suggest that efficacy can be achieved in humans with bBMPx product. This study is the first to report preliminary results of posterior intertransverse process fusion of the lumbar spine using bBMPx in humans. To minimize the clinical risk of nonunion, bBMPx product mixed with demineralized allograft was implanted on one side of the spine in 22 patients, and iliac crest autograft was placed on the contralateral side. This configuration allowed each patient to serve as his or her own control. The use of additional instrumentation was possible depending on the clinical situation.

As demonstrated in the rhesus, the response to bBMPx in humans is expected to be dose dependent. A clinical trial currently is underway after approval from the Ethics Committee of the Schulthess Klinik in Zurich, Switzerland, with all the patients providing informed consent. Different dosages were used to define the appropriate concentration of bBMPx necessary to stimulate bone formation at least equal to that of autologous bone. A 12.5-mg dose of bBMPx is a direct scale-up from the 10 mg (5 mg per side) dose in the rhesus on an implant volume basis. A dose of 25 mg per side is equivalent on a weight-to-weight scale-up of rhesus (∼13 kg) to humans (∼65 kg). One additional dose of 50 mg per side may be evaluated. An example from one patient shows that bBMPx product is capable of inducing bone in a human lumbar spine (Figure 5). The fusion on the side with bBMPx product was similar to that observed on the side that received autograft harvested from the iliac crest in terms of morphology, quality, and fusion success.

Figure 5
Figure 5:
Pilot clinical study patient that received bBMPx product. Good fusion masses with both autograft and bBMPx product can be seen at 12 months.

Radiographic evaluation with standard radiographs of the lumbar spine and CT scanning will be performed at 2 to 3 months, then 6, 12, and 24 months after surgery. Evaluations will include fusion rate as well as fusion morphology and quality. Initial results demonstrate good fusion with bBMPx product.

In previous animal studies, Boden et al 6 performed histologic analysis of the bone formation. They concluded that the fusion process using autologous bone depends primarily on the migration of osteogenic cells from the decorticated transverse process. The fusion achieved with bBMPx product may result from migration and differentiation of mesenchymal cells from the surrounding soft tissue into the open structure of the composite implant that allows mesenchymal cells to contact bBMPx.

In conclusion, the results of these studies demonstrate the species variation and the increased doses of bBMPx necessary for bone induction. In rabbits, 0.15 mg of bBMPx is sufficient for fusion rates greater than those from autograft. In the rhesus, 3 to 5 mg of bBMPx is required, and in humans, doses of 12.5 to 50 mg of bBMPx in bBMPx product appear to be needed for consistent spinal fusion.

The preclinical safety and efficacy studies demonstrate the usefulness of bBMPx product as an iliac crest autograft substitute for posterolateral intertransverse process spinal fusion. Preliminary clinical results are promising.

Key Points

  • A rabbit lumbar spine safety model demonstrated the safety of bBMPx product (BMPs and collagen) when placed near the spinal cord.
  • A primate lumbar spine efficacy model demonstrated a dose response of bBMPx product and consistent fusion results.
  • Evidence from one patient receiving bBMPx product suggests that this bovine BMP extract (bBMPx) can induce bone in a human spine fusion.


The authors thank the staffs at the Spine Center of the Schulthess Klinik, the Emory Spine Center, the Yerkes Primate Center, and Sulzer Biologics for their support. Thanks also are extended to those who reviewed the document including Drs. Michael Lewis, Veronique Bernhardt, Dunja Frey, and Hassan Achakri. These studies were supported by Sulzer Biologics.


1. Barnes TS, Lewis MJ, Benedict JJ. The Identification of Proteins in BP, an Osteoinductive Fraction Derived from Bovine Bone. Trans First European Conf BMP, Zagreb, Croatia, 1998.
2. Boden SD, Martin GJ, Morone MA, et al. Posterolateral lumbar intertransverse process spine arthrodesis with recombinant human bone morphogenetic protein 2/hydroxyapatite-tricalcium phosphate after laminectomy in the nonhuman primate. Spine 1999; 24: 1179–85.
3. Boden SD, Schimandle JH, Hutton WC. An experimental lumbar intertransverse process spinal fusion model: Radiographic, histologic, and biomechanical healing characteristics. Spine 1995; 20: 412–20.
4. Boden SD, Schimandle JH, Hutton WC. Lumbar intertransverse process spinal arthrodesis with use of a bovine bone-derived osteoinductive protein. J Bone Joint Surg [Am] 1995; 77: 1404–17.
5. Boden SD, Schimandle JH, Hutton WC. The use of an osteoinductive growth factor for lumbar spinal fusion: Part II. Study of dose, carrier, and species. Spine 1995; 20: 2633–44.
6. Boden SD, Schimandle JH, Hutton WC, et al. The use of an osteoinductive growth factor for lumbar spinal fusion: Part I. Biology of spinal fusion. Spine 1995; 20: 2626–32.
7. Cook, SD, Dalton JE, Tan EH, et al. In vivo evaluation of recombinant human osteogenic protein (rhOP-1) implants as a bone graft substitute for spinal fusions. Spine 1994; 19: 1655–63.
8. Damien CJ, Parsons JR. Bone graft and bone graft substitutes: A review of current technology and applications. J Appl Biomater 1991; 2: 187–208.
9. David SM, Gruber HE, Meyer RA. Lumbar spinal fusion using recombinant human bone morphogenetic protein in the canine. Spine 1999; 24: 1973–9.
10. Hecht BP, Fischgrund JS, Herkowitz HN, et al. The use of recombinant human bone morphogenetic protein 2 (rhBMP2) to promote spinal fusion in a nonhuman primate anterior interbody fusion model. Spine 1999; 24: 629–36.
11. International Standards Organization (ISO) Document 10993. Arlington, VA: Biological Evaluation of Medical Devices, 1994.
12. Johnson EE, Urist MR. Human bone morphogenetic protein allografting for reconstruction of femoral nonunion. Clin Orthop Rel Res 2000; 371: 61–74.
13. Johnson EE, Urist MR, Finerman GAM. Resistant nonunions and partial or complete segmental defects of long bones. Clin Orthop Rel Res 1992; 277: 229–37.
14. Meyer RA, Gruber HE, Howard BA, et al. Safety of recombinant human bone morphogenetic protein-2 after spinal laminectomy in the dog. Spine 1999; 24: 747–54.
15. Minamide A, Tamaki T, Kawakami M, et al. Experimental spinal fusion using sintered bovine bone coated with type I collagen and recombinant human bone morphogenetic protein-2. Spine 1999; 24: 1863–72.
16. Saito H, Mimatsu K, Sato K, et al. Histopathologic and morphometric study of spinal cord lesion in a chronic cord compression model using bone morphogenetic protein in rabbit. Spine 1992; 17: 1368–74.
17. Sampath TK, Maliakal JC, Hauschka PV, et al. Recombinant human osteogenic protein-1 (hOP-1) induces new bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro. J Biol Chem 1992; 267: 20352–62.
18. Sandhu HS. Anterior interbody fusion with osteoinductive growth factors. Clin Orthop Rel Res 2000; 371: 56–60.
19. Sandhu HS, Kanim LEA, Kabo JM, et al. Evaluation of rhBMP-2 with an OPLA carrier in a canine posterolateral (transverse process) spinal fusion model. Spine 1995; 20: 2669–82.
20. Schimandle JH, Boden SD, Hutton WC. Experimental spinal fusion with recombinant human bone morphogenetic protein-2. Spine 1995; 20: 1326–37.
21. Silcox DH III, Boden SD, Schimandle JH, et al. Reversing the inhibitory effect of nicotine on spinal fusion using an osteoinductive protein extract. Spine 1998; 23: 291–7.
22. Silcox DH III, Daftari T, Boden SD, et al. The effect of nicotine on spinal fusion. Spine 1995; 20: 1549–53.
23. Strates, BS, Urist MR. Origin of the inductive signal in implants of normal and lathyritic bone matrix. Clin Orthop Rel Res 1969; 66: 226–40.
24. Urist MR. Bone: Formation by autoinduction. Scand J Plast Reconstr Surg 1965; 150: 893–9.
25. Urist MR, Hay PH, Dubuc F, et al. Osteogenetic competence. Clin Orthop Rel Res 1969; 64: 194–220.
26. Urist MR, Silverman BF, Buring K, et al. The bone induction principle. Clin Orthop Rel Res 1967; 53: 243–83.
27. van de Putte KA, Urist MR. Osteogenesis in the interior of intramuscular implants of decalcified bone matrix. Clin Orthop Rel Res 1965; 43: 257–70.
28. Wang EA, Rosen V, Cordes P, et al. Purification and characterization of other distinct bone-inducing factors. Proc Natl Acad Sci U S A 1988; 85: 9484–8.
29. Yasko AW, Lane JM, Fellinger EJ, et al. The healing of segmental bone defects induced by recombinant human bone morphogenetic protein (rhBMP-2). J Bone Joint Surg [Am] 1992; 74: 659–70.

bone graft substitute; bone morphogenetic protein; growth factors; lumbar spine; primate model; rabbit model; spine fusion]Spine 2002;27:S50–S58

© 2002 Lippincott Williams & Wilkins, Inc.