Bone morphogenetic protein-2 induced posterolateral lumbar fusion in a mouse model

Objective: The goal of this study was to establish a rhBMP-2 induced posterolateral intertransverse lumbar fusion model in the mouse to serve as an acceptable substitute for models using larger animals, such as the rat, by dening the pertinent anatomy and establishing an effective dosing regimen in the mouse. Methods: Forty 10-week-old male Institute of Cancer Research (ICR) mice were randomized into ve groups and were subjected to posterolateral lumbar fusion surgery at the L5-L6 level and received the following assigned treatment regimens: Group A: carrier collagen sponge only; Group B: 0.05μg rhBMP-2 on a collagen sponge; Group C: 0.15μg rhBMP-2 on a collagen sponge; Group D: 0.5μg rhBMP-2 on a collagen sponge; and Group E: 1.5μg rhBMP-2 on a collagen sponge. The study was ended at eight weeks post-operation and standard x-ray and micro-computed tomography imaging were performed in addition to histological examination of dissected spine specimens. Results: BMP-2 induced new bone at all dose levels in a dose-related manner. The mice implanted with rhBMP-2 at the dose of 0.5μg demonstrated more bone formation and a higher fusion rate compared to mice receiving lower doses. More abundant bone formation could be induced by higher dose regimen, which could also result in poorer bone quality. Conclusions: Posterolateral lumbar fusion at the L5-L6 level can be successfully accomplished in mice. The dose of 0.5μg of rhBMP-2 delivered on a collagen sponge appears to be the optimal regimen to produce the most satisfactory fusion results.


Introduction
Since its approval by the United States Food and Drug Administration (FDA) for clinical use, recombinant human bone morphogenetic protein-2 (rhBMP-2) has been widely used in the clinical practice of orthopaedic, including a substantial number of "off label" applications 1; 2 . In addition to the approved anterior lumbar interbody fusion, rhBMP-2 is used extensively in other lumbar spinal fusion techniques and at doses that seem to exceed the physiological need 3 . Safety issues pertaining rhBMP-2 have been of concern since the inception of its clinical use. Although many animal and clinical studies suggested that the use of rhBMP-2 in clinical settings should be generally safe 4 , some studies found reasons for concern with respect to uncontrolled bone formation, soft tissue in ammation, and carcinogenesis 5 . In 2011 the FDA denied the approval for a high dose of rhBMP-2 (40 mg) on a compression-resistant hydroxyapatite/β-tricalcium phosphate carrier (Amplify, Medtronic, Memphis, TN) for use in posterolateral lumbar fusion because of concerns for an increased cancer risk associated with rhBMP-2 6 . A recent publication demonstrated a signi cant increase of cancer risk in patients treated with high dose rhBMP-2 7 , however, two other studies reported different results 6; 8 . Controversy continues to the present and no consensus has been reached.
In addition to clinical studies, a large number of basic science studies, both in vitro and in vivo, have been undertaken to investigated the effects of BMP-2 on tumor growth 9 . However, integrating the results of the clinical studies and the basic science studies into a coherent understanding of the relevant processes remains very problematic. Most basic research focused on the direct effects of BMP-2 on different tumor cell types, however, in clinical settings, the systematic carcinogenic effect of BMP-2 is certainly more complex. Animal studies that are more similar to the clinical situation would be helpful to reveal the systematic effect of BMP-2 use. While most in vivo studies for tumor research are performed in mouse models 10 , pre-clinical studies of BMP-2 in orthopaedic applications are mostly done in larger animals, such as rats 11 , rabbits 12 , or canines 13 . A BMP-2 induced spinal fusion model in the mouse would, therefore, provide a much-needed bridge to ll the gap between these two research areas. Although several attempts have been made to establish a mouse spinal fusion model, no standard procedure has been established with detailed protocols and appropriate dosing regimens 14; 15 . This study established a posterolateral intertransverse lumbar fusion model in mice using rhBMP-2 by delineating the pertinent anatomy and establishing the optimal dosing regimen for rhBMP-2 in order to validate it as a suitable Group D: 0.5μg rhBMP-2; and Group E: 1.5μg rhBMP-2.

Surgical Procedures
Posterolateral lumbar fusion surgery was performed in a manner similar to the protocol used for the rat spinal fusion model with which we have considerable experience 16 . Mice were pre-medicated with sustained release buprenorphine thirty minutes prior to surgery. Sodium pentobarbital solution was administered intraperitoneally at the dose of 50 mg/kg. The surgical site, including dorsal aspect of the lumbar spine and iliac crest, was shaved and disinfected with Betadine and 70% ethanol. The iliac crest was used as a landmark to locate the body of the L6 vertebra. A 2-cm longitudinal midline incision was made through the skin and subcutaneous tissue over L5-L6 down to the lumbodorsal fascia and the paraspinal muscles were dissected bilaterally to expose the transverse processes of L5 and L6, which were then decorticated with a high-speed burr. The surgical site was irrigated with sterile saline and 1.5mm×1.5mm×2mm pieces of absorbable collagen sponge (ACS) (Helistat, Integra Life Sciences, Plainsboro, NJ) containing the assigned doses of rhBMP-2 were placed bilaterally, with each implant spanning the transverse processes. The ACS, which is made of collagen derived from bovine Achilles tendon, was prepared to deliver the test material by soaking the ACS in a PBS solution of the rhBMP-2 before applying the sponge to the intertransverse process space. The implants were then covered with the overlying paraspinal muscles and the lumbodorsal fascia and skin were closed with 5-0 Prolene sutures (Ethicon, Inc., Somerville, NJ) (Fig 1).

Radiographic Analysis
Fusion between L5 and L6 was examined by means of radiographs obtained at eight weeks after the surgical procedure. In each mouse, the fusion between the L5 and L6 transverse processes was evaluated by two independent observers, who were blinded with respect to the treatment group assignments, employing the following standardized scale: 0, no fusion; 1, unilateral fusion; and 2, complete bilateral fusion. The scores from the observers were added together and only a score of 4 was considered as complete fusion.

Microcomputed Tomography (Micro CT) Analysis
Mice were euthanized using a standard CO 2 chamber eight weeks after the spinal fusion procedure, and their spines were excised and stored in 40% ethyl alcohol. Each excised spine specimen was analyzed by high resolution microcomputed tomography (micro-CT) using a SkyScan 1172 scanner (SkyScan/Bruker; Kontich, Belgium) as previously described in detail 17 . Virtual image slices were reconstructed using the cone-beam reconstruction software version 2.6 based on the Feldkamp algorithm (SkyScan/Bruker; Kontich, Belgium). Sample reorientation and 2D visualization were performed using Data Viewer (SkyScan/Bruker; Kontich, Belgium), and 3D visualization was performed using Dolphin Imaging version 11 (Dolphin Imaging & Management Solutions; Chatsworth, CA)

Histology
The spines were dissected out of the animal, xed in paraformaldehyde, decalci ed in a standard decalcifying solution (Richard-Allan Scienti c, Kalamazoo, MI), washed with running tap water, and then transferred to 75% ethanol 18 . Serial sagittal sections near the transverse processes were cut with a blade carefully at the level of the transverse process. The specimens were embedded in wax and then sectioned. The sections were stained with hematoxylin and eosin, alcian blue, and also alizarin red. The slides were scanned by a Pannoramic SCAN digital slide scanner (3DHISTECH, Budapest, Hungary) using a Zeiss plan-apochromatic objective (magni cation: 20x, numerical aperture: 0.8) and a Hitachi (HV-F22CL) 3CCD progressive scan color camera (resolution: 0.2325 μm/pixel). JPEG image encoding with a quality factor of 80 and an interpolated focus distance of 15 with stitching in the scan options was chosen 19 .

Statistical Methods
PASW software (version 18; SPSS, Chicago, Illinois) was used to perform the statistical analyses. The Mann-Whitney U test was used for comparisons between different groups. Data are expressed as mean ±

Results
The average weight of the mice was 30.39 ± 1.89 g, and there were no signi cant differences among the groups (P = 0.394). The mice were allowed to eat and move about ad libitum after surgery. No wound infections or edema were observed in any animal subject. At eight weeks, radiographs of the spines of the mice in the control group showed no new bone formation, while in all of the other treatment groups evidence of new bone formation was observed, although the amount of new bone formation observed in the 0.05 µg rhBMP-2 group (Group B) was minimal (Fig. 2). Osseous bridging was detected in Groups C, D and E, and complete fusion was seen in Groups D and E. Group E showed the most abundant bone formation. The radiographic scores for the degree of fusion increased in relation to the dose of rhBMP-2 and both Groups D and E were given the same score as complete fusion was achieved in both groups (Table 1). In general, the results of the micro-CT analyses were in agreement with those of the plain X-rays. The micro-CT images, however, are more clear and also allow for the amount of new bone formation to be quanti ed (Fig. 3). The bone volume (BV) in the area of interest (yellow box in Fig. 4A) was measured, and the ratio of bone volume/ total volume (BV/TV) was calculated. Although the 1.5 µg of rhBMP-2 group (Group E) resulted in the highest BV compare with other groups, the 0.5 µg rhBMP-2 group (Group D) showed better bone quality as re ected in the BV/TV ratio (Fig. 4B,C).
Histology demonstrated that, although the contour of new bone is largest in Group E, the inside of the bone shell was occupied mostly by adipocytes. In Group D, on the other hand, the dose of 0.5 µg rhBMP-2 induced better bone structure with more trabecular bone inside the tissue bridging the transverse processes (Fig. 5).

Discussion
BMP-2 belongs to the TGF-beta superfamily of proteins and it can stimulate proliferation in some cell types. Although utilization of BMP-2 can eliminate the associated comorbidities of harvesting allograft, use of high-dose rhBMP-2 (> 40 mg) in orthopaedic surgery should be undertaken cautiously considering its potential adverse effects. A recent publication demonstrated a signi cant 6.75-fold increase in the incidence of new cancer two years after single level anterior spinal arthrodesis in patients treated with a high dose of BMP-2 as compared to patients who received autogenous bone grafts 20 . Thus, there is continued concern with respect to carcinogenesis in situations in which exogenous BMP-2 is employed.
Nevertheless, in a retrospective study of 502 adult patients who underwent spinal surgery with high-dose BMP-2 the authors reported no correlation between increasing BMP-2 dosage and cancer with an average follow-up period of 42 months 21 . There was, however, a cancer prevalence of 3.4% in the study patients 21 . Kelly et al. 22  In addition to the con icting results of these basic science studies, the consideration remains that most of these studies focused on the direct effect of BMP-2 on tumor growth, whereas, in clinical settings, BMP-2 is usually administrated in a normal area and the cancer risk is associated with an effect at a distant site. Animal studies that more closely resemble the actual clinical situation would be helpful to address this short-coming. Most in vivo studies for tumor research are conducted in mouse models and, therefore, a mouse spinal fusion model involving BMP-2 will eliminate a substantial barrier to progress in this research. To date, larger animals such as rats, rabbits, and canines have been used for spinal fusion models as it was thought that mice are too small for such surgical procedures. We have now demonstrated that this surgical procedure can effectively and reliably be accomplished. In this study we determined the optimal dose of rhBMP-2. A dose of 0.5 µg provided for the highest fusion rate with the highest bone quality and, therefore, it is our recommendation that this dose be employed in future studies.  Surgical procedure of the posterolateral intertransverse lumbar fusion model using mice. A, a 2-cm longitudinal midline skin incision was made over L5-L6 ; B, soft tissue was dissected down to the lumbodorsal fascia; C, the paraspinal muscles were dissected bilaterally to expose the transverse processes of L5-L6; D, the transverse processes were decorticated with a high-speed burr; E, absorbable collagen sponge (ACS) (Helistat, Integra Life Sciences) containing different doses of rhBMP-2 were placed bilaterally to the intertransverse process space; F, the incision was closed with 5-0 Prolene sutures (Ethicon, Inc., Somerville, NJ).

Figure 2
Radiographs of the spines of the mice at eight weeks. Control group showed no new bone formation; minimal bone formation in the 0.05μg rhBMP-2 group; osseous bridging was detected in Groups C; Groups D, E appeared complete fusion.

Figure 3
Images of the fusion segments reconstructed from micro-CT data.

Figure 4
Micro-CT data analysis of the fusion segments. A, The bone volume (BV) was measured in the area of interest (yellow box); B, BV was compared between groups; C, the ratio of bone volume/ total volume (BV/TV) was calculated and compared among groups.