Cervical Total Disc Arthroplasty

Rationale for Cervical Arthroplasty

The excellent clinical success rate and long-term experience with ACDF for the treatment of cervical DDD raises the question as to the need for the development of alternate procedures. Although fusion is beneficial at the symptomatic level, it may be detrimental to the remaining motion segments because it induces increased motion and abnormal intradiscal pressure recordings in adjacent disc segments, which translate into increased stress on the adjacent nonoperated discs. ^{, ,} Several studies have reported the radiographic development of adjacent segment degeneration (ASD) after ACDF procedures. Baba and associates assessed more than 100 patients who underwent anterior cervical fusion for cervical myelopathy and observed new spinal canal stenosis above the previously fused segments in 25% of patients at an average follow-up of 8.5 years. Similarly, Gore and Sepic reported a 25% risk of developing new radiographic adjacent segment spondylosis 5 years following ACDF. In a comparative radiographic study, Herkowitz and colleagues evaluated 44 patients with 4.5 years of follow-up, randomized to either ACDF or posterior foraminotomy without fusion for the treatment of cervical radiculopathy, and found that the chances of developing ASD were similar between the two groups (41% for ACDF vs. 50% for foraminotomy group), with no correlation between the development of ASD and the onset of new clinical symptoms referable to those radiographic changes. In a series of more than 800 patients who underwent posterior foraminotomy without fusion evaluated by Henderson and coworkers, 9% of patients developed adjacent segment disease requiring additional surgery over an average follow-up time of 2.8 years, demonstrating that the development of ASD may not be unique to anterior cervical fusion, and in at least some cases simply reflects the natural history of the disease or failure to treat all of the symptomatic segments during the primary procedure.

Irrespective of the lack of clear correlation between radiographic changes following anterior cervical fusion and clinical symptoms, a subgroup of patients who undergo ACDF do develop symptomatic adjacent segment disease requiring treatment. ^, There has been an attempt to distinguish clinical versus radiographic adjacent segment problems after cervical fusion. The term adjacent segment degeneration has been used to describe radiographic changes adjacent to a previous spinal fusion procedure that do not necessarily correlate with any clinical findings, as compared with adjacent segment disease, which refers to the development of new clinical symptoms that correspond to segmental radiographic changes next to the level of a previous spinal fusion. Bohlman and colleagues, Gore and colleagues, and Williams and colleagues reported 9%, 14%, and 17% risks, respectively, of developing symptomatic adjacent segment disease following anterior fusion that required surgical treatment. ^{, ,} Hilibrand and coworkers retrospectively reported the long-term follow-up of 409 anterior cervical decompression and stabilization procedures performed for radiculopathy or myelopathy. They found symptomatic ASD occurred at a relatively constant rate of 2.9% per year in the 10 years following surgery. The results of this study have been interpreted by some as providing robust evidence for the development of ASD following ACDF. Patients with preexisting degenerative changes and who were over 60 years old had a more rapid onset of symptomatic adjacent segment disease. It is intriguing that patients who had multilevel arthrodesis were significantly less likely to have symptomatic adjacent segment disease than were those who had had a single-level fusion ( P < .001). Although still conjectural, with progression of natural history being an important factor in the development of ASD, the study by Goffin and colleagues showed that the rate of radiographic ASD following arthrodesis for traumatic cervical injuries was 60% over 5 years, indicating that the development of adjacent segment disease is at least partly related to the arthrodesis itself and may not be blamed totally on the natural history of the disease.

There is biomechanical evidence demonstrating that CDR may allow for a more normal restoration of load transfer and kinematics at adjacent levels when compared with fusion. Wigfield and associates recorded and compared intradiscal pressures in adjacent levels of normal/untreated cadaveric cervical spines, spines treated with a simulated fusion, and spines treated with a CDR and found that the motion and pressures of adjacent segments in specimens treated with a CDR did not significantly differ from those in the normal/untreated spines. Similarly, Puttlitz and colleagues demonstrated that spines treated with CDR approximated intact motion in all three rotational planes at the affected level compared with normal cadaveric spines. In a clinical prospective study, Robertson and coworkers compared the incidence of radiological documented changes and symptomatic adjacent level cervical disc disease after single-level discectomy and cervical fusion or arthroplasty using the Bryan disc and showed the appearance of new radiographic changes in 34.6% of the fusion-treated patients as compared with 17.5% of the arthroplasty-treated patients at 24 months ( P = .009). New symptomatic adjacent DDD occurred in 7% of the fusion group and in 0% of the arthroplasty group ( P = .018). This study strengthens the argument that maintaining motion with arthroplasty after single-level disc disease may delay or prevent, to a significant degree, the associated radiological disc degeneration.

Another rationale for the use of CDR as opposed to ACDF relates to the complications with pseudarthrosis and bone graft procurement for arthrodesis. It is generally believed that less soft tissue dissection, decreased esophageal retraction pressure, and the minimal profile of the current-generation CDRs may result in a lower incidence of postoperative dysphagia following cervical arthroplasty as compared with instrumented ACDF.

Nomenclature, Biomaterials, and Biomechanics

Fundamental to the principle of arthroplasty are the corollary concepts of repetitive stress and the generation of wear-related debris. Various factors are involved in the design of a cervical artificial disc, such as (1) implant kinematics, (2) implant materials, (3) device subsidence, and (4) fixation to bone. The choice of material used in creating the device must be governed by three principles: articular surface wear, generation of wear debris, and host inflammatory response. Many designs incorporate two different materials articulating with each other. In a metal-on-polymer pairing the polymer wears preferentially, whereas in a metal-on-ceramic pair the metal wears to a greater extent. ^, Metal wear debris generates a lesser inflammatory response compared with polymer debris.

An ideal artificial disc should resist corrosion and wear, reproduce the movements of a normal cervical disc without overloading adjacent biomechanical structures, be easy to insert, and be x-ray– and magnetic resonance imaging–compatible. ^, Implant design characteristics are important for the function and longevity of CDR. The articulating surfaces must be able to tolerate anticipated load without fatigue or failure while minimizing friction, should have superior wear characteristics with minimal debris generation, and should exceed patient life expectancy. The presence of wear debris and the associated inflammatory response can be detrimental to the stability of the device and varies depending on the design of the disc (metal-on-metal vs. metal-on-polymer). In addition, the implants must remain permanently affixed to the adjacent vertebral bodies.

With the development of numerous artificial discs since the early 2000s, the Cervical Spine Study Group developed a nomenclature system for cervical arthroplasty. Artificial discs are categorized by material, articulation, fixation, design, and kinematics. They can be classified as nonarticulating, uniarticulating, or biarticulating. Discs are either modular with replaceable parts or nonmodular. Some have points for vertebral body fixation, and some have surfaces that promote bone ingrowth at the disc–end plate interface. With regard to motion, they may be constrained, semiconstrained, or unconstrained. Constrained devices restrict motion to less than that seen physiologically, semiconstrained devices allow physiological motion, and unconstrained devices rely on soft tissue and the inherent compression across the disc space to limit motion. Each of these devices is available in a range of heights and depths to accommodate individual variances in anatomy. The currently approved cervical artificial discs are listed in Box 113.1 .

There are currently eight cervical arthroplasty devices with FDA approval, of which six are currently marketed in the United States. They each have significant differences in their construction. The Prestige LP (Medtronic Sofamor Danek, Memphis, TN) is a uniarticulating system made of a titanium alloy/titanium carbide composite (titanium ceramic composite) with a ball-in-trough design. Device components are placed through impaction of dual rails on the intervertebral contact surface at surgery and maintained via bony ingrowth over time, which is facilitated by a porous titanium plasma spray coating on the end plate surface. It is semiconstrained in motion. The Prestige LP design is further unique in that the ball-in-trough design creates a mobile center of rotation that provides the theoretical advantage of simulating normal kinematics over a range of device positions. The design also allows the postoperative center of rotation to mimic the preoperative state as closely as possible.

The ProDisc-C cervical disc (Centinel Spine, West Chester, PA) is a uniarticulating system constructed of cobalt–chromium alloy end plates with a locking core of ultra-high molecular weight polyethylene (UHMWPE) as a central polymer that provides a ball-and socket articulation. The device includes a central keel for anchorage to the vertebral bodies, and the end plates are coated with a titanium plasmapore for tissue compatibility and bony ingrowth. It is semiconstrained in motion. ^, The PCM cervical disc (NuVasive, San Diego, CA) is a uni-articulating system with cobalt chromium molybdenum endplates and a large radius UHMWPE core. The endplates are coated with titanium and electrochemically coated with calcium phosphate that encourages bone integration. PCM is unconstrained in motion. ^,

Secure-C (Globus Medical, Audubon, PA) is a biarticulating system consisting of two titanium plasma-sprayed cobalt–chromium–molybdenum alloy end plates and a central UHMWPE core. It is a semiconstrained device. ^, The Mobi-C cervical disc (Zimmer Biomet, Warsaw, PA) is a biarticulating system with two titanium plates with titanium plasma and hydroxyapatite coating and a central UHMWPE core. It is also a semiconstrained device.

M6-C (Orthofix, Lewisville, TX) is the most recently FDA-approved device. It is a nonarticulating system with titanium alloy end plates. To replicate a normal disc, it features a unique design, including both annular (polyethylene weave) and nuclear (viscoelastic polyurethane core) components that allow for axial compression with independent angular motions in flexion and extension, lateral bending, and axial rotation. It is unconstrained in motion. ^,