Guest Discussants: Yu-Po Lee, MD, Tim Yoon, MD, PhD
SpineLine Section Editor: Jeffrey C. Wang, MD

Case Presentation
A 52-year-old gentleman presents with complaints of progressive unsteadiness of gait and difficulty with his handwriting and buttoning his shirt. He has significant neck pain and radicular pain in the C5 and C7 distributions with some weakness of his deltoids and triceps bilaterally. Approximately ten years previously, he had a C5-6 anterior cervical discectomy with fusion (ACDF) without cervical plating. He has failed conservative treatments and feels that his condition has worsened.

The patient has primarily radicular subjective complaints and does not wish to have a posterior surgery when offered the option of a posterior versus an anterior approach. For the sake of this Curve debate, we will consider only anterior options. What would you recommend?  

Yu-Po Lee, MD, Responds
The issue of this debate is the use of growth factors, such as bone morphogenetic protein (BMP), for cervical fusion. This is a controversial topic given the cost and the potential risks of BMP when used in the cervical spine. However, the case chosen is a very complex one and may benefit from the use of a BMP despite it being off-label when used in the cervical spine.

In the case described, the gentleman has four disc/osteophyte complexes that are compressing the spinal cord and nerve roots both above and below a previous anterior discectomy and fusion at C5-6. (Figure 2A) In addition to symptoms of myelopathy, he has radicular pain in the C5 and C7 nerve root distributions with associated muscular weakness. Given his progressive funtional loss, it is reasonable to consider surgery in this case. Although a good argument could be made for treating this case posteriorly or with a combined anterior and posterior approach, for the sake of this debate, let us assume that we intend to treat this patient with an anterior-only approach.

Once an anterior approach has been selected, various decompression and reconstruction options can be considered. The merits of individual discectomies with fusion at C3-4, 4-5, 6-7, and C7-T1 are considered against those of two level, C4 and C7, corpectomies with strut graft reconstructions. Each of these approaches has their associated risks and benefits. Both options allow direct removal of the sources of spinal cord and nerve root compression. The multi-level discectomy approach improves segmental fixation by retaining vertebral bodies for screw placement. Moreover, multilevel discectomy procedures are associated with better preservation or reconstruction of lordosis than multi-level corpectomy procedures.

The major disadvantage of multilevel anterior discectomies and fusions, however, is the increased pseudarthrosis rate caused by the number of bony surfaces that require healing. Fusion rates could be improved by performing corpectomies of C4 and C7 and placing strut grafts between C3-5 and C6-T1. But the instability cause by the two– level corpectomy would probably necessitate posterior instrumentation as well. This may be especially true when the bottom of your construct ends at the cervico-thoracic junction (C7-T1). Thus, if you wanted to treat this patient anteriorly, your best strategy would be with multilevel anterior discectomies and fusions.

Both anterior options risk pseudarthrosis and late term destabilization. Means to improve the rate and speed of bone healing are needed. Autogenous iliac crest grafts or BMP containing allograft struts or polyetheretherketone (PEEK) cages can be considered.

Autograft’s benefits of improved fusion must be weighed against the added morbidity of anterior iliac crest harvest. Anterior iliac crest harvests are well documented and potential complications include chronic pain, nerve injury, arterial injury, vascular injury, hematoma, gait abnormalities, stress fracture, peritoneal perforation and herniation of abdominal contents.1 With these potential complications, if a reasonable autogenous iliac crest alternative exists,
why not consider it?

When considering BMP use, consider necessity, efficacy, safety and cost.

Necessity. In this case, a multilevel anterior discectomy and fusion would have a high nonunion rate if it was not augmented with BMP. In one study by Hilibrand et al,2 the fusion rate of multilevel anterior fusions was only 63% despite the use of autograft. The poor fusion rate associated with multilevel cervical decompressions and fusions is due to the number of surfaces that need to heal. This challenging biologic environment has a high likelihood for fusion failure even with autograft. Thus, multilevel anterior cervical fusions need some means of improving local biology.

Efficacy. BMPs powerfully induce bony fusion in the cervical spine. Tumialán et al3 studied 200 patients who had multilevel anterior cervical discectomies and fusions with BMP-2 in a PEEK spacer. Included were six four-level anterior discectomy and fusion patients. Dynamic flexion/extension films and CT scans revealed a 100% fusion rate. Buttermann4 reviewed 66 consecutive multilevel anterior discectomy and fusion patients in whom either autograft or allograft struts augmented with BMP-2 were used. The autograft nonunion rate was twice as high as that of the allograft and BMP-2 group.

Safety. The use of BMPs in the cervical spine has been associated with postoperative swelling and dysphagia. However, many of those reports occurred before the correct dosage and effects of the BMPs were truly known. In the study by Tumialán et al,3 the authors noted that they took care to pack the BMP-2 soaked sponges entirely within the cage and made sure that none of the sponge migrated out from the cage. Consequently, they only observed significant dysphagia in 7% of the patients. This compares favorably with other studies evaluating postoperative dysphagia in anterior cervical fusions where a BMP is not used.5 To further decrease the risk of postoperative swelling and dysphagia, a barrier can also be used to limit the spread of BMP. In a study by Patel et al,6 the authors were able to limit the diffusion of BMP-2 with fibrin glue both in vitro and in vivo in a rat model. Thus, safety can be improved by tucking the BMP-2 sponges entirely within the PEEK cages or allograft struts and then injecting fibrin glue over the cages or allograft struts to contain the BMP-2. Although the risk of postoperative swelling and dysphagia cannot be completely eliminated with these techniques, using these techniques decreases the risk of postoperative swelling and dysphagia to a rate that is comparable to anterior cervical fusions where BMP is not used.3,5

Cost. With the rising cost of health care, our treatment plans must be cost sensitive. A small package of BMP typically costs about $2,500. In the cervical spine, only a small amount of BMP is necessary at each level. Thus, a small package of BMP is probably all that would be necessary for a four-level cervical discectomy and fusion. When compared to the cost associated with the added surgical time necessary to harvest anterior iliac crest graft and then cutting and fashioning four interbody grafts, the costs are probably equivalent. Further, consider the indirect costs resulting from slower mobilization and increased pain medication usage associated with harvesting iliac crest graft. In this case, BMP usage may even be cost–saving.

Summary
Many treatment options are available to this patient. Each has associated advantages and disadvantages. For an anterior-only approach, the best results would probably be obtained with BMP. BMP utilization increases fusion rates and avoids iliac crest harvest complications. Newer techniques limit the risks of postoperative swelling. Lastly, using BMP may be cost-saving when the increased operative time and postoperative pain and immobility associated with harvesting anterior iliac crest graft are factored into the equation. Thus, use of BMP in this case is definitely warranted.

References
1. Silber JS, Anderson DG, Daffner SD, et al. Donor site morbidity after anterior iliac crest bone harvest for single-level anterior cervical discectomy and fusion. Spine. 2003;28(2):134-139.

2. Hilibrand AS, Yoo JU, Carlson GD, et al. The success of anterior cervical arthrodesis adjacent to a previous fusion. Spine. 1997;22(14):1574-1579.

3. Tumialán LM, Pan J, Rodts GE, et al. The safety and efficacy of anterior cervical discectomy and fusion with polyetheretherketone spacer and recombinant human bone morphogenetic protein-2: a review of 200 patients. J Neurosurg Spine. 2008;8(6):529-535.

4. Buttermann GR. Prospective nonrandomized comparison of an allograft with bone morphogenic protein versus an iliac-crest autograft in anterior cervical discectomy and fusion. Spine J. 2008;8(3):426-435.

5. Fountas KN, Kapsalaki EZ, Nikolakakos LG, et al. Anterior cervical discectomy and fusion associated complications. Spine. 2007;32(21):2310-2317.

6. Patel VV, Zhao L, Wong P, et al. Controlling bone morphogenetic protein diffusion and bone morphogenetic protein-stimulated bone growth using fibrin glue. Spine. 2006;31(11):1201-1206.

Tim Yoon, MD, PhD, Responds
This 52-year-old with a history of C5-6 uninstrumented ACDF, complains of progressive myelopathy, radicular and axial neck pain. Furthermore, he has C5 and C7 weakness. The lateral X-ray image shows diffuse congenital stenosis with Torg ratio of less than 0.8, fused C5-6, and diffuse multilevel spondylosis. (Figure 1B) The sagittal MRI confirms the diffuse stenosis and also shows disc herniations at C3-4, C4-5, C6-7, and C7-T1 with a previous fusion at C5-6. (Figure 2) The worst stenosis appears to be at C3-4 and C4-5 with some retrovertebral involvement at C4. Myelomalacia is evident at C4.

Given the failure of nonoperative treatment and progressive myelopathy with cord compression, surgical intervention is the most appropriate treatment. Surgical treatment options include anterior only, posterior only, or combined anterior/posterior. For the sake of debate, we will consider anterior only as the desired surgical treatment option and focus on the issue of what type of bone grafting is the most appropriate under the circumstances. I will make the case for anything but BMP.

Based on images available, at least the disc spaces of C4-T1 (except for C5-6) are involved in the cord compression. This extreme scenario was clearly chosen in an attempt to legitimize consideration of BMP. BMP is not only unnecessary, but expensive, and increases the risk to both the patient and the surgeon.

Contrary to older literature, multilevel anterior cervical fusion with instrumentation has a reasonably high fusion rate, even without BMP. In a study by Samartis et al,1 a fusion rate of 97.5% (61 two–level, 19 three–level) was reported with a Smith Robinson ACDF, structural bone graft without BMP and anterior cervical plate. In that series, 56% of the patients had autogenous tricortical iliac crest bone and 44% had allograft cortical bone. Of note, the study population included a significant percentage of smokers and worker’s compensation patients. Despite this challenging situation, only two patients had pseudarthrosis (one two–level and one three–level). In both pseudarthrosis cases, allograft bone was used for the intervertebral graft, suggesting that autograft bone may be superior. Fraser and Härtl2 performed a meta-analysis of published literature of anterior cervical fusion. This study indicated that three–level ACDF without plate resulted in a fusion rate of 65%. This is expected based on the old literature which reported on cases without modern plating techniques. In contrast, the addition of anterior cervical plating increased the union rate to 82.5% and as high as 96.2% with corpectomy and plate.

Another option that is available is the hybrid technique that allows for judicious use of corpectomy and ACDF with plate. With this technique, local autograft bone from the corpectomy can be used to provide osteoinducitive, osteogenic, and osteoconductive graft material to supplement the biological environment to work together with the structural support provided by the cortical allograft bone. The corpectomy reduces the number of end plate/graft interfaces that have to fuse, allowing for higher fusion rates. However, by not performing corpectomies at every level, the vertebral bodies in between the corpectomies present fixation sites for the anterior cervical plate. By using these techniques, anterior cervical surgery encompassing even four levels can achieve relatively high fusion rates.

The powerful osteoinductive potential of BMP may entice some surgeons to use it in extreme cervical reconstruction challenges, as that presented here. However, consider the many recent reports of serious complications of BMP use in the anterior cervical spine. On July 1, 2008, the FDA issued a Public Health Notification entitled “Life-threatening Complications Associated with Recombinant Human Bone Morphogenetic Protein in Cervical Spine Fusion.” NASS has deemed this FDA warning to be “reasonable in content and an important message to emphasize given the extreme nature of possible complications.” The nature of the complication revolves around the proximity of the anterior cervical spine to the trachea and the esophagus. The literature indicates that, with BMP use, a higher percentage of patients developed airway and swallowing difficulty due to either soft tissue swelling and edema or hematoma formation.3-6 Debate continues as to whether safe use of BMP can be achieved by using a smaller dose of BMP, confining the BMP inside a cage or cortical graft, or using postoperative steroids.7,8 However, the safe and effective use of BMP in the anterior cervical spine has not been established.

Extensive studies exploring the appropriate dose and carrier were necessary for FDA approval of maxillofacial and anterior lumbar spine BMP. Similar, indepth studies will be required for cervical spine use. In the meantime, surgeons have to recognize that BMP is not FDA-approved for use in the cervical spine. With this recognition comes increased medicolegal risk. This demands great caution by the surgeon when considering the use of BMP in the anterior cervical spine. Additionally, BMP is an expensive product. In the absence of good literature supporti, it is difficult to justify BMP’s additional cost.

Returning to this specific case, I recommend a corpectomy of C4 with discectomy and fusion at C6-7 and C7-T1 and careful foraminotomies to decompress the symptomatic C5 and C7 roots. For bone graft, I would use structural allograft bone combined with local autogenous corpectomy defect bone that is placed at the ends of the allograft and anterior and lateral to the allograft. After bone-grafting, anterior plating can be performed either with two plates (C3-C5 and C6-C7-T1) or with one long plate with screws at C3, C5, C6, C7, and T1. The ACDFs at the C6-7 and C7-T1 level forms a solid biomechanical base for the construct. If the patient has a higher risk of nonunion (eg, heavy smoker) or very low tolerance for additional surgery (eg, elderly), then I would consider using autogenous tricortical iliac crest for structural bone instead of allograft bone.

Summary
In summary, sufficiently high fusion rates can be achieved for a four–level anterior cervical surgery, even without BMP. The use of plating and the use of hybrid constructs of corpectomy/ACDF makes this possible. At this point, the use of BMP for anterior cervical spine has serious life-threatening risks that demand great caution. In the extreme condition, it may be preferable to supplement the anterior construct with posterior instrumentation, rather than risk using BMP.

References
1. Samartzis D, Shen FH, Matthews DK, Yoon ST, Goldberg EJ, An HS. Comparison of allograft to autograft in multilevel anterior cervical discectomy and fusion with rigid plate fixation. Spine J. 2003;3(6):451-459.

2. Fraser JF, Härtl R. Anterior approaches to fusion of the cervical spine: a metaanalysis of fusion rates. J Neurosurg Spine. 2007;6(4):298-303.

3. Smucker JD, Rhee JM, Singh K, Yoon ST, Heller JG. Increased swelling complications associated with of-label usage of rhBMP-2 in the Anterior Cervical Spine. Spine. 2006;31(24):2813-9;

4. Shield LBE, Raque GH, Glassman SD, et al. Adverse effects associated with high-dose recombinant human bone morphogenetic protein-2 use in anterior cervical spine fusion. Spine. 2006;31(5):542-547.

5. Perri B, Cooper M, Lauryssen C, Anand M. Adverse swelling associated with use of rh-BMP-2 in anterior cervical discectomy and fusion: a case study. Spine J. 2007;7(2):235-239

6. Butterman GR. Prospective nonrandomized comparison of an allograft with bone morphogenic protein versus an iliac-crest autograft in anterior cervical discectomy and fusion. Spine J. 2008;8:426-435. (Though there are two rebuttal letters reporting the possibility of a safe dose)

7. Tumialán L, Rodts GE. rh-BMP-2 can be used safely in the cervical spine: dose and containment are the keys! Letter to the editor. Spine J. 2007;7(4):509-510.

8. Dickerman RD, Reynolds AS, Morgan BC, Tompkins J, Cattorini J, Bennett M. rh-BMP-2 can be used safely in the cervical spine: dose and containment are the keys! Letter to the editor. Spine J.2007;7(4):508-509. 

Author Disclosures

• Y Lee: c-3, Stryker; c-2, Medtronic.
• T Yoon: disclosure not received at press time.

Dr. Wang Comments
This Curve/Countercurve illustrates a nice debate on the offlabel use of BMP for the cervical spine. Recent reports have described several complications associated with the clinical use of BMP in the cervical spine. This lead the FDA to issue a public health statment July 1, 2008 notifying health care professionals of potential life-threatening complications that may occur when rhBMP is used in cervical spine fusion procedures. The FDA alert applied to both rhBMP-2 and rhBMP-7.

In certain clinical situations where the healing of a spinal fusion is severely compromised due to potential risk factors for nonunion, surgeons have used BMP for cervical fusion. Whether this FDA notification has an effect on the use of off-label use of BMP will be determined, but certainly, the debate illustrated in this case discussion is a current controversy in which both authors argue excellent clinical points supporting their views. This will be an interesting and continuing debate over the coming years.