Background

The cervical spine consists of seven vertebral bodies with intervening discs. These discs function in load bearing and motion transfer. In addition to its biomechanical functions in motion, the cervical spine serves as the protective passage for the spinal cord and vertebral arteries. Much is known about the macrobiology of the intervertebral disc. Disc degeneration and the subsequent processes that ensue in the cervical spine are also well documented as the transition from mild degenerative disc disease to cervical spondylosis progresses. For many years, the surgical treatment for pathology in the cervical intervertebral disc has been limited to procedures that remove pathologic disc material and address the bony and neurologic pathology in the region of the excised disc.

Anterior cervical discectomy and fusion (ACDF) is a proven intervention for patients with radiculopathy and myelopathy.¹ It has served as the standard by which other cervical and spinal disorders may be judged as the result of its high rate of success. The success of this technique is often judged based on its consistent ability to relieve symptoms related to neurologic dysfunction. In this sense, the clinical results with regard to the patient’s index complaint are outstanding. The radiographic results of this technique are also initially predictable with a high rate of fusion. Plating techniques have diminished the need for postoperative immobilization or eliminated them entirely.² Because of limitations specific to this procedure, investigators have developed surgical alternatives to fusion that attempt to address the kinematic and biomechanical issues inherent in it.

A major concern related to the treatment of cervical degenerative disc disease and spondylosis with ACDF is the issue of adjacent segment degeneration. This event is manifest as the radiographic appearance of degenerative change at a level directly above or below a level treated with a surgical intervention—typically being associated with degeneration of a level adjacent to a fused level. The incidence of this phenomenon has been reported to be 92% by Goffin and colleagues,³ who wrote a long-term follow-up on patients after treatment with anterior interbody fusion.

Although some debate remains regarding the causation of adjacent segment degeneration—with a mix of postsurgical and naturally determined aging cited as root causes—there is little debate regarding the existence of this phenomenon. It is also relevant to note the clinical distinction between adjacent segment degeneration and adjacent segment disease. Adjacent segment disease is defined as adjacent segment degeneration that causes clinical symptoms (pain or neurologic disorders or both) severe enough to lead to patient complaint or require operative intervention.⁴ Although this distinction has not remained consistent in published literature, it is an important consideration with regard to the phenomenon that occurs in discs adjacent to discs that have undergone a surgical intervention. Numerous studies have made a consistent point of distinguishing between radiographic degeneration and symptomatic disease.^3,5

There is clinical evidence to support the postsurgical nature of adjacent segment disease. In patients previously treated with fusion, adjacent segment disease has been documented at a rate of 2.9% of patients per annum by Hilibrand and colleagues,⁴ and 25% of patients undergoing cervical fusion have new onset of symptoms within 10 years of fusion. This study has received a great deal of attention and has led to further investigations into biomechanical causation. Other reports have focused on the recurrence of neurologic symptoms and degenerative changes adjacent to fused cervical levels.^3,6 The concept that adjacent levels need to compensate for loss of motion in the fused segment may also be valid. Segments adjacent to a fusion have an increased range of motion and increased intradiscal pressures.^7,8

Bone graft materials used in traditional ACDF procedures have also been a source of controversy. Current ACDF techniques make use of allograft bone, premanufactured allograft bone, and autologous iliac crest. Complications associated with autologous iliac crest harvest used as a fusion graft in ACDF are well documented. Sandhu and colleagues⁹ reported a complication rate of 1% to 25% with such procedures. Complications such as acute and chronic pain, infection, meralgia paresthetica, and pelvic fracture are known to occur at harvest donor sites.^10,11

Although allograft removes the risks associated with the harvest of autograft, it has the detriment of having the theoretical risk of disease transmission. In practice, this risk is believed to be extremely minute, although the U.S. Food and Drug Administration (FDA) has taken this issue seriously. The issues of disease transmission and contaminated graft materials have been highlighted by allograft tissue recalls by the FDA in recent years.¹² Although bone graft substitutes may play a role in the future practice of ACDF, this continues to be a minority stake in the overall graft selection of modern surgeons.

Pseudarthrosis is another complication encountered with anterior cervical fusion procedures. Pseudarthrosis is the failure of bony bridging or nonunion of a segment that has previously been treated with a bone graft or a bone graft substitute—an attempt at fusion has been made. In multilevel ACDF procedures, there is a relationship between the rate of pseudarthrosis and the number of levels fused. Brodke and Zdeblick¹³ reported a 97% fusion rate in single-level ACDF, which decreased to 83% with fusion at three levels. Bohlman and colleagues¹ reported an 11% pseudarthrosis rate in single-level fusions that increased to 27% with multilevel fusions.

In recent years, the use of bone morphogenetic proteins has been proposed as an alternative or adjunct to traditional bone grafting techniques to combat the pseudarthrosis issue in patients deemed to be at higher risk for this complication.^14,15 This off-label use has been associated with an increased incidence of swelling complications and concerns for graft resorption and migration of interbody implants.^16–19 These issues serve to strengthen the argument for fusion alternatives in the treatment of discogenic pathology in the anterior cervical spine.

Total intervertebral disc replacement (TDR) is designed to preserve motion, avoid limitations of fusion, and allow patients to return quickly to routine activities. The primary goals of the procedure in the cervical spine are to restore disc height and segmental motion after removing local pathology that is deemed to be the source of a patient’s index complaint. A secondary intention is to preserve normal motion at adjacent cervical levels, which may be theorized to prevent later adjacent level degeneration. Cervical TDR avoids the morbidity of bone graft harvest.^20,21 It also avoids complications such as pseudarthrosis, issues caused by anterior cervical plating, and cervical immobilization side effects.

History of Disc Arthroplasty and Device Design

An understanding of the evolution of cervical TDR serves as an important lesson in the concepts of device design, TDR bearing and wear characteristics, and articular constraint. In the late 1980s, Cummins and colleagues²² developed a metal-on-metal ball-and-socket cervical disc replacement composed of 316L stainless steel. With the acquisition of this technology and the later development of new metal-on-metal devices came a rapid transition from this device to the most recent device, the PRESTIGE LP (Medtronic Sofamor Danek, Memphis, TN). A predecessor of this device, the PRESTIGE ST (Medtronic Sofamor Danek, Memphis, TN) is currently approved by the FDA for human use in the United States (Figs. 43-1 and 43-2). More recent additions to the metal-on-metal category of arthroplasty devices include the Kineflex-C disc (Spinal Motion, Mountain View, CA) and the CerviCore intervertebral disc (Stryker Spine, Allendale, NJ) (Figs. 43-3 and 43-4), which are in the process of U.S. FDA investigational device exemption (IDE) trials.

FIGURE 43–1 PRESTIGE ST cervical disc prosthesis is currently approved by the FDA for use in the United States. This stainless steel uniarticulating device attains primary fixation to the vertebral bodies via use of locked screws.

(Implant representations courtesy Medtronic Sofamor Danek, Memphis, TN; with permission.)

FIGURE 43–2 PRESTIGE ST prosthesis in C5-6 arthroplasty. Lateral flexion and extension radiographs show motion through arthroplasty device in this postoperative patient.

(Courtesy Medtronic Sofamor Danek, Memphis, TN; with permission.)

FIGURE 43–3 CerviCore cervical disc prosthesis on lateral and expanded anteroposterior and lateral views. Initial fixation is obtained via vertical rails on the endplates.

(Courtesy Stryker Spine, Allendale, NJ; with permission.)

FIGURE 43–4 CerviCore cervical disc prosthesis is shown ex vivo and on T2-weighted MRI. MRI may show some artifact in the region of an arthroplasty device.

(Courtesy Stryker Spine, Allendale, NJ; with permission.)

Numerous devices have evolved in parallel to the metal-on-metal implants, including the BRYAN cervical disc (Medtronic Sofamor Danek, Memphis, TN) (Figs. 43-5 through 43-7), the PCM (CerviTech, Rockaway, NJ), the DISCOVER (DePuy Spine, Raynham, MA), and the MOBI-C (LDR, Austin, TX). Each of these devices is in the process of limited human trials or U.S. FDA IDE submission and represents an alternative to metal-on-metal bearing surfaces, which have the potential for metal debris and systemic concentration of metal ions. To date, one such device, the Prodisc-C (Synthes Spine, Paoli, PA) (Figs. 43-8 and 43-9), has obtained approval for use in the United States.

FIGURE 43–5 BRYAN cervical disc prosthesis is shown ex vivo and in unassembled form. The endplates of this device are unique in their design and promote ingrowth of bone into metallic surface. At the time of this publication, the BRYAN device remains under FDA review.

(Courtesy Medtronic Sofamor Danek, Memphis, TN; with permission.)

FIGURE 43–6 BRYAN cervical disc prosthesis is visualized on postoperative MRI. Titanium alloy devices such as the BRYAN device may have less MRI artifact than similar devices constructed with cobalt-chromium or stainless steel. These images show the imaging characteristics of this device at index and adjacent surgical levels.

(Courtesy Rick Sasso, Indianapolis, IN.)

FIGURE 43–7 Upright lateral view of a patient who underwent successful cervical arthroplasty at C5-6 with a BRYAN device.

(Courtesy Medtronic Sofamor Danek, Memphis, TN; with permission.)

FIGURE 43–8 Ex vivo image of Prodisc-C. This device is approved by the FDA in the United States for cervical arthroplasty and obtains initial fixation via a central keel. Bone ingrowth is promoted via the surface alterations of the superior and inferior endplates of this device.

(Courtesy Synthes Spine, Paoli, PA; with permission.)

FIGURE 43–9 Prodisc-C is visualized on lateral flexion-extension radiographs. The device retains motion at index surgical level in this patient successfully treated with arthroplasty at C5-6.

(Courtesy Synthes Spine, Paoli, PA; with permission.)

A summary of the design characteristics of each of these devices is presented in Table 43–1. Although the ideas of bearing surface, wear debris, and constraint are not new to discussions with regard to arthroplasty in general, they are relatively new in regard to the spine. A full understanding of the term constraint has not been agreed on because constraint may arise within the device or as a result of the local anatomy (e.g., facets, posterior longitudinal ligament). As the knowledge base in spine TDR increases, intelligent investigations and discussions are sure to include many of these concepts and may redefine understanding of them.

It is relevant to understand the fact that the load borne by devices in the cervical spine is dissimilar to the load borne in the lumbar spine. The biomechanical environment of the cervical spine has been taken into account in the design of the current generation of these devices. As intermediate-term and long-term studies on individual devices become available, the design concepts of these initial devices will have the opportunity for continued examination in their in vitro environment.

Indications for Use, Contraindications, and Complications

Cervical disc arthroplasty trials have included patients refractory to nonoperative treatment modalities with and without radiculopathy or myelopathy or both and with one-level and two-level degenerative disc disease or spondylosis.^23–25 These indications have been retained throughout the FDA approval process. At the time of this writing, two devices, the PRESTIGE ST and the Prodisc-C, have achieved FDA approval for single-level use in the United States. Other devices are in various stages of the IDE and approval process (Table 43–2). ACDF may be discussed as part of the indication process for an arthroplasty procedure. The historical challenges associated with ACDF presented in this chapter may be weighed against the early nature of data with respect to cervical arthroplasty in a patient’s informed consent discussion.

TABLE 43–2 Current Status of Cervical Arthroplasty Devices in the United States*

IDE, investigational device exemption.

* Table current as of April 1, 2009.

In determining indications for cervical arthroplasty of any type in a patient, it is relevant and appropriate to discuss verbally and obtain written consent for an intraoperative alternative to arthroplasty. In current practice, ACDF with plating and anterior corpectomy and fusion with plating remain options when it becomes clear to the operating surgeon that placement of an arthroplasty device may be compromised. This judgment to proceed with a fusion may occur as the result of endplate defects; arthroplasty sizing and fixation issues; or other bony, vascular, or neurologic issues that would prevent the appropriate placement of the device.

The appropriate time to discuss complications related to cervical arthroplasty is at the time of informed consent. The approach-related risks of cervical arthroplasty are similar to ACDF and should be discussed as such. These risks have been adequately covered in other portions of this text. A unique risk of arthroplasty is the concern of heterotopic ossification in the region of the arthroplasty device.^26–30 Heterotopic ossification may result in loss of motion or frank fusion of the index level. Heterotopic ossification may be associated with the physiologic response to the implantation process, the amount of bleeding or hemostasis necessary in a particular procedure, or the amount of bone debris created at the time of preparation for implantation. Postoperative use of nonsteroidal anti-inflammatory drugs has been suggested to moderate the prevalence of this risk.

Other complications may occur with cervical arthroplasty that are independent of the anterior cervical approach, including infection, implant migration, subsidence, continued or new neurologic findings, vascular injury, dural injury or cerebrospinal fluid leak, hematoma, and reoperation for adjacent level disease. Many complications associated with placement of arthroplasty devices have come to light as the result of reporting and analysis of the prospective randomized multicenter U.S. FDA IDEs and large studies performed outside the United States.^{23–25,31,32}

Cervical arthroplasty is contraindicated in patients with active or prior infection, osteoporosis or poor host bone, segmental cervical instability or segmental kyphosis, trauma, tumor, primary axial neck pain, significant facet arthropathy, posterior neurologic compression, anterior soft tissue abnormality (e.g., tracheal or esophageal abnormality, prior radiation), allergy to any of the device materials, severe spondylosis, pediatric patients, compromised vertebral body morphology, or (presently) disease involving more than one level. Other relative contraindications are similar to ACDF with the exception of nicotine use.

Preoperative Imaging

At a baseline, preoperative imaging for cervical arthroplasty should include plain radiographs with anteroposterior, neutral lateral, odontoid, and lateral flexion-extension films. The flexion-extension lateral views serve as a preoperative assessment of normal and abnormal mobility at the index and adjacent surgical levels. An advanced imaging modality such as magnetic resonance imaging (MRI) or computed tomography (CT) with the possible addition of myelography can be crucial to evaluating the index surgical level. These studies combine to allow assessment of spondylosis, preexisting facet arthropathy, neurologic compression, and cause and may provide insight into any preoperative contraindications in candidates for cervical arthroplasty.

Technique of Implantation

The technique of anterior cervical discectomy and the anterior approach to the spine are beyond the scope of this chapter. The following description assumes a surgeon’s comfort with this technique and suggests only specific modifications to the current approach relevant to anterior cervical arthroplasty compared with ACDF.

Intraoperatively, patient position is important. A “physiologic” or slightly lordotic cervical spine position is preferred.³³ Assessment via fluoroscopy (C-arm) is crucial to patient positioning and implant insertion and fixation portions of these procedures. It is important to keep the head, neck, and shoulders in a stable and neutral position throughout this surgical procedure. A small towel roll may be placed under the neck to assist with appropriate positioning of the neck and shoulders and to keep a physiologic lordosis without creating a hyperlordosis (Fig. 43–10). This positioning technique differs from the typical placement of a roll under the shoulders or thoracic spine, which could place the cervical spine in hyperlordosis. The head is placed on a doughnut-type pillow or a folded towel to keep it from rolling during the procedure. The careful positioning of shoulders with a taping technique can also allow for less motion during this procedure and must be carefully weighed against the risk of traction to the shoulders. Taping of the shoulders differs from the commonly used wrist restraints in a typical ACDF procedure that are used to obtain additional longitudinal traction via a temporary “pull” on the arms.

FIGURE 43–10 Positioning for cervical arthroplasty is as crucial to the technique as any portion of the procedure. Correct positioning of a patient maintains physiologic lordosis without creating hyperlordosis in the cervical spine and may be facilitated through use of a towel roll placed under the cervical spine. Techniques have moved away from traction through spine (as shown in this picture). Preoperative and intraoperative use of fluoroscopy allows for confirmation of patient positioning and device alignment.

(Courtesy Rick Sasso, Indianapolis, IN.)

A standard right- or left-sided Smith-Robinson approach may proceed with appropriate localization and exposure of the index surgical level being the intent of this exposure. It is crucial to obtain a surgical exposure that allows for identification of the center of the index disc and vertebral bodies for later placement of the arthroplasty device. Disc arthroplasty is performed only after adequate decompression of the affected cervical level. At the surgeon’s discretion for treatment of the index neurologic complaint, this may involve a complete discectomy from ventral to dorsal that also allows for placement of a device of appropriate width and adequate decompression, symmetrical resection of uncovertebral osteophytes and spurs, resection of all or part of the posterior longitudinal ligament, and any resection of central spondylotic osteophytes associated with degenerative disc disease. Meticulous hemostasis is recommended throughout this procedure to diminish the blood loss and minimize the risk of heterotopic ossification. It may become clear at any point during the neurologic decompression, endplate preparation, or device trialing process that arthroplasty is contraindicated. Should this occur, the surgeon must adjust the surgical plan intraoperatively and proceed with a fusion-based alternative.

After neurologic decompression, assessment for placement of an appropriately sized disc and planning for proper orientation of the implant are crucial to successful arthroplasty. To this end, it should be the surgeon’s goal to place as large a device (with respect to diameter) as possible in the prepared space.³⁵ Device-specific tools may aid in this assessment.

Before any intervention that prepares the endplates, it is important to ensure the exact sagittal position of the vertebrae with lateral fluoroscopic imaging. Anteroposterior views are important to place the spinous processes at the target disc level between the pedicles to ensure perfect alignment and centering in the coronal plane. Sizing of a cervical arthroplasty device may be determined with a combination of preoperative templates and preoperative radiographic studies including CT. The use of intraoperative trials and fluoroscopic imaging allows for additional assessment of proper device sizing and placement in the coronal and sagittal planes.

Endplates are prepared in a manner consistent with the device to be implanted. This preparation may include milling of the endplate (as in the BRYAN technique) or creation of a bony trough to accommodate an endplate keel (as in the Prodisc-C technique). Preservation of subchondral bone is otherwise crucial to the prevention of implant subsidence. Instrumentation specific to each arthroplasty device may be of great assistance in endplate preparation and may include special endplate distracters, keel preparation mills, rasps, and endplate mills. After the endplate preparation has been completed, it is appropriate to reassess the centering of the preparation and recheck the neurologic decompression.

Insertion of the artificial disc device may proceed and is implant-specific. Common to all devices is the principle of implantation to an appropriate depth based on implant design, with a repeat assessment of implant centering and endplate coverage. After an assessment of the implant position in the coronal and sagittal planes has been done, the implant may be fixed to the spine with any implant-specific instrumentation such as screws.

Final imaging of the device implantation is performed before wound closure. Hemostasis is rechecked, and the surgical wound is closed in a standard fashion. Postoperative immobilization is not required. Upright flexion-extension radiographs may be obtained before discharge from the hospital and serve as a comparison to postdischarge radiographs for the purposes of follow-up.

Postoperative Imaging

Follow-up imaging of arthroplasty devices is crucial to the assessment of motion retention, adjacent segment disease, device wear and settling, device fixation, and neurologic decompression and status. Because the current generation of disc arthroplasty devices retains metallic components either in the endplates or in the bearing mechanism, radiation-based technologies have predominated as a mechanism of assessment. The workhorse studies remain flexion-extension and lateral bending plain radiographs because they are easily accessible, require less technician and physician technique in acquisition, maintain the ability to assess motion, eliminate the concerns of claustrophobia with MRI and CT, and are associated with moderate patient risk from radiation exposure. CT and CT myelography require an increased dose of radiation and are technique driven with respect to myelography.

To moderate the risks of technique and radiation associated with CT myelography, MRI has been proposed as an alternative that allows for postoperative assessment of neurologic status adjacent to and at the level of a prior cervical arthroplasty procedure. Success with this technique has been described by Sekhon and colleagues³⁴ in several devices, including the BRYAN and the PRESTIGE LP arthroplasty devices. Both of these devices use titanium alloy in their endplates, which was shown to produce less MRI artifact at the index and adjacent surgical levels than the metals associated with the manufacture of the Prodisc-C and PCM arthroplasty devices.³⁴ The BRYAN prosthesis has a polyurethane core, and the PRESTIGE LP had a titanium carbide alloy (as tested in the study). Investigators had difficulty assessing the neural structures at the index and adjacent levels with devices manufactured with nontitanium metals (cobalt-chromium-molybdenum) used in the Prodisc-C and PCM devices.