Cervical and Lumbar Arthroplasty, Interspinous Devices, and Motion-Sparing Technology

John M. Rhee

Bradley Moatz

Cervical Arthroplasty

Background/Rationale for Arthroplasty

For over 50 years, anterior cervical discectomy and fusion (ACDF) has been a proven intervention for patients with cervical radiculopathy or myelopathy unresponsive to conservative management. In 1958, Smith and Robinson first described the procedure in which disk material is removed along with subchondral bone while the interspace is then filled with iliac crest autograft. Cloward subsequently reported using a cylindrical allograft bone dowel instead of autograft. Plating techniques were introduced in the 1990s and decreased the need for postoperative immobilization. This procedure has served as the gold standard by which other spinal disorders have been compared due to its high rate of success in relieving symptoms of neural compression. Nevertheless, ACDF eliminates motion at the operated segment and may be complicated by pseudarthrosis and/or donor-site morbidity. Another major concern with regard to ACDF is the potential issue of accelerated adjacent segment degeneration. Cervical disk arthroplasty offers potential, although as of yet not proven, advantages in terms of motion and adjacent segment degeneration while allowing early return to function.

In 1956, the first patent for disk replacement was issued to van Streenbrugghe in France. Six years later Fernström of Sweden implanted stainless steel spheres following discectomy in 133 patients, 125 lumbar disk levels and 8 cervical disk levels. In 1995, McKenzie reported 10- to 20-year follow-up data on 67 of Fernström’s patients. He noted a high success rate with 95% of patients reporting that the procedure was worthwhile. Only one patient required removal of the prosthesis in this study. Ultimately, however, the device caused hypermobility and was frequently found to have eroded into the vertebral end plates. Cervical arthroplasty subsequently fell out of favor for the next two decades.

The use of lumbar arthroplasty devices in the 1980s helped to renew interest in its cervical counterpart. Drs. Kurt Schellnack and Karin Büttner-Jans, working at the Charité hospital in East Germany, designed a three-piece implant with a sliding polyethylene core sandwiched between two metallic end plates. The first SB Charité lumbar disk replacement was implanted in Berlin in 1984. There were, however, issues with subsidence and implant fracture.

Current FDA-Approved Devices

The following cervical disk replacement products are FDA approved at the time of this writing:

PrestigeTM ST Cervical Disk (Medtronic Sofamor Danek, Memphis, TN)
PRODISC-C Total Disk Replacement (Synthes, Inc., New York, NY)
BRYAN Cervical Disk (Medtronic Sofamor Danek, Memphis, TN)
SECURE-C Artificial Cervical Disk (Globus Medical Inc., Audubon, PA)
PCM cervical disk system (NuVasive, Inc., San Diego, CA)
Mobi-C Cervical Disk (LDS Spine USA, Inc. Austin, TX)

The Bryan cervical prosthesis consists of two titanium alloy shells with a polyurethane nucleus. The device end plates have a porous coating to facilitate bony ingrowth and is therefore not secured into the disk space with any instrumentation. Its central portion is surrounded by a polyurethane sheath that is filled with saline in order to both contain potential wear debris and to act as an initial lubricant.

The ProDisc-C is composed of two cobalt-chromium- molybdenum end plates with a central keel and an ultrahigh–molecular-weight polyethylene (UHMWPE) inner core. This modular implant is similar in design to the lumbar arthroplasty device, the ProDisc-L.

The Porous Coated Motion Disk Prosthesis, or PCM device, is a two-piece metal device with an UHMWPE inner core. It has a broad radius of curvature which is thought to provide more end plate support laterally. The outer surface of the end plates has a calcium phosphate coating, as well as serrations, both of which help anchor it to the end plate.

There are also multiple other devices that are currently in development. They all seek to maintain segmental motion, alignment, and disk height. Furthermore, they should produce minimal wear debris that may incite an inflammatory reaction all the while having a risk-to-benefit profile at least equivalent to ACDF.

Range of Motion

Cervical disk arthroplasty may offer preservation of motion when compared to ACDF and, therefore, the possibility of averting adjacent degenerative changes. Several studies have shown that intersegmental motion after cervical arthroplasty more closely resembles physiologic values when compared to ACDF. Auerbach et al. reported that cervical disk arthroplasty increases cervical range of motion more than ACDF and this discrepancy is maintained at 2 years postoperatively. Goffin et al. measured intersegmental motion by cine fluoroscopy during flexion and extension in 40 patients. They found that the disk replacement cohort had the same range of motion as normal volunteers and they had nearly identical synchronous patterns of motion. The fused patient cohort demonstrated reduced overall range of motion, but increased motion at adjacent segments.

Adjacent Segment Degeneration

A major concern with ACDF is the potential risk of inducing accelerated degeneration at adjacent segments. It is important to note that adjacent segment degeneration is a radiographic phenomenon in which degenerative changes are seen directly above or below a surgical level. Goffin et al. studied the long-term follow-up of patients after treatment with anterior fusion and reported the incidence of radiographic degenerative changes to be 92%. Gore and Sepic followed 121 patients who had undergone an ACDF on average 5 years prior. They found that 25% of patients had progression of pre-existing spondylosis and 25% had new-onset spondylosis. Baba et al. studied 106 patients who underwent ACDF for cervical myelopathy with an average follow-up of 8.5 years. He reported that 25% of patients developed spinal canal stenosis at the level above the fused segment. However, in all of these studies there was no correlation between adjacent segment degeneration and clinical symptoms. Furthermore, it is also important to keep in mind that degeneration at adjacent levels occurs as a result of normal aging as well. The question is whether ACDF truly accelerates adjacent segment degeneration versus the natural history and whether arthroplasty could lead to less adjacent segment degeneration than ACDF.

Adjacent segment disease refers to segment degeneration that causes clinical symptoms (pain or neurologic deficit). Hillibrand et al. followed 374 patients for a maximum of 21 years. They found the adjacent segment disease occurred at a relatively constant rate of 2.9% per year for the first 10 years after surgery. Furthermore, 25% of patients undergoing cervical fusion had new onset of symptoms within 10 years of fusion. Symptomatic adjacent segment degeneration was more likely to occur at C5–C6 and C6–C7. Counterintuitively, patients who underwent multilevel fusion were significantly less likely to develop adjacent segment degeneration than those who had a single-level fusion. The authors concluded that adjacent segment degeneration is more likely attributable to the natural history of cervical spondylosis and not to the fusion itself.

The concept that cervical fusion forces adjacent levels to compensate for the loss of motion may be valid. Biomechanical studies strongly support altered adjacent level kinematics after fusion. DiAngelo et al. observed that after anterior cervical fusion, there was an increase in motion at adjacent segments in order to compensate for the loss of motion at the index level. However, artificial disk replacement did not alter motion at either the index or the adjacent levels. In a cadaveric model, Eck et al. measured disk pressures at C4–C5 and C6–C7 both before and after simulated fusion at C5–C6. They found a 73% and 45% increase in intradiscal pressure at levels cephalad and caudad to the simulated fusion. Gurvinder et al. performed finite element analysis in a fusion model
and found a 30% increase in motion in flexion, extension, and rotation when one level was fused. This number was found to double when two levels were fused. Critics of cadaveric studies note that there are inherent design limitations including: inability to simulate a true bony fusion, difficulty in accounting for the stabilizing effects of the paraspinal musculature, evaluation of only immediate rather than long-term changes, and inability to assess clinical relevance.

Potential Complications of ACDF

Several studies have found a direct relationship between the number of levels fused and the rate of pseudarthrosis. Brodke and Zdeblick reported a 97% fusion rate for single-level procedures and 83% fusion rate at three levels. Bohlman et al. reported an 11% pseudarthrosis rate in single-level fusions that increased to 27% in multilevel fusion. Approximately two-thirds of patients with a pseudarthrosis were symptomatic, and 17% required revision surgery. Modern fusion rates for a single-level ACDF with allograft and plate approximates 90%.

Cervical disk arthroplasty requires no interbody bone graft harvest from the iliac crest thus decreasing operative time and postoperative morbidity versus ACDF when performed with iliac autograft bone. Bone graft morbidity such as pain and infection has been reported to occur in 18% to 30% of cases. Arthroplasty also obviates the need for anterior plate fixation. Many patients who undergo cervical fusion are placed in an external orthosis postoperatively and are restricted in terms of activity for a variable duration of time. On the other hand, many patients who undergo cervical disk arthroplasty are allowed to return to normal activities, including driving and work, within 1 to 2 weeks following surgery because fusion is not intended. It should be noted that a randomized trial found no difference between patients using a collar after ACDF versus those who did not. Thus, postoperative immobilization is not a universal disadvantage of ACDF.

Proper Patient Selection

Similar to spinal fusion, success of cervical disk arthroplasty is contingent upon proper patient selection. Indications for cervical disk replacement include:

Symptomatic disk herniations or spondylosis causing radiculopathy or mild myelopathy at one or more levels.
As is the case for ACDF, those without significant neurologic deficits should undergo a trial of nonoperative management first.

Alternative indications, yet unproven, include treatment of three or more symptomatic levels or levels adjacent to a fusion.

Contraindications for cervical disk replacement include:

instability
osteomyelitis
infection
osteoporosis
significant facet arthrosis

Patients with pre-existing conditions that lead to loss of motion such as ossification of the posterior longitudinal ligament, ankylosing spondylitis, and diffuse idiopathic skeletal hyperostosis, may not benefit from arthroplasty. Use of cervical disk arthroplasty in the management of axial neck pain secondary to disk degeneration is unknown. The role of arthroplasty in patients with significant spondylosis is debatable, particularly those with spondylotic myelopathy. Our opinion is that the optimal indications for arthroplasty are refractory soft-disk herniations in younger patients with relatively little spondylosis.

Clinical Outcomes

To date there have been five randomized clinical trials that have been completed in the United States under the Food and Drug Administration (FDA) Investigational Device Exemption (IDE) protocols to study cervical arthroplasty. These five studies are large multicenter prospective randomized clinical trials comparing cervical arthroplasty with ACDF using interbody allograft and titanium plates. These studies were sponsored by industry with the goal of product approval. Individually, the studies were not statistically powered in a manner to demonstrate superiority of one procedure over another.

Mummaneni et al. performed a prospective randomized multicenter study to evaluate the Prestige ST with ACDF. In total, 541 patients were included with 276 in the investigational group who underwent anterior cervical arthroplasty with the Prestige II, and 265 patients in the control group underwent ACDF. Clinical outcomes included Neck Disability Index (NDI), the 36-Item Short Form Health Survey (SF-36), and neck and arm pain scores. Radiographs were also evaluated for angle of motion as well as for fusion. Patients were followed at regular intervals postoperatively with the 7-year follow-up data being recently published. Of the 541 patients treated, 395 patients (73%) (212 investigational and 183 control patients) completed 7 years of clinical follow-up. Significant improvements that were achieved in both groups by 1.5 months were maintained at 7 years. In the investigational group, mean NDI improvement from preoperative baseline was 37.5 at 7 years and 31.9 for the control group. This difference between groups achieved statistical significance (p = 0.002). Overall rate of maintenance or improvement in neurologic status was significantly (p = 0.011) higher at 7 years for the investigational group (88.2%) versus the control (79.7%). The implant was found to maintain an average angular motion of 6.75 degrees at 7 years. Cumulative rates for subsequent revision surgery or supplemental fixation were lower in the investigational group. Furthermore, rates for additional
surgical procedures that involved the adjacent level were found to be lower in the investigational group (4.6%) than in the control group (11.9%).

Zigler et al. recently reported on the 5-year data resulting from the FDA-regulated IDE study evaluating the safety and effectiveness of the ProDisc-C compared to fusion. Two hundred nine patients (103 ProDisc-C and 106 ACDF) from 13 sites were randomized and treated. Clinical outcome measures included NDI, visual analog scale (VAS) neck and arm pain, SF-36, neurologic examination, adverse event occurrence, and VAS patient satisfaction. Rates of follow-up at 2 years were 98.1% for the ProDisc-C group and 94.8% for the ACDF group. At 5 years the rate of follow-up fell markedly to 72.7% for ProDisc-C group and 63.5% for ACDF. The authors reported that for all clinical outcomes for both groups, there was a statistically and clinically significant improvement at 2 and 5 years compared to baseline. At 5 years, the investigational group had significantly less neck pain intensity and frequency compared with the controls. There were no reports of device failures or implant migration with ProDisc-C. The authors also noted that motion was maintained at 5 years at the index level and that the rate of reoperation (index or adjacent level) was 2.9% at 5 years for the investigational group compared to 11.3% for the control group.

Sasso et al. reported on the 4-year clinical outcomes data for the multicenter FDA IDE trial for the Bryan cervical disk. In total, 242 patients were randomized to the Bryan disk while 221 patients received ACDF. The primary outcome measure was overall success, which is a composite variable of safety and efficacy measures. Secondary measures were also assessed. Four-year follow-up data was collected on 181 patients who received the Bryan disk and 138 patients who underwent ACDF. The authors found a substantial reduction in NDI scores in both groups when compared to preoperative values. The greater improvement in NDI score in the Bryan cohort that was seen at 2-year follow-up persisted through the 4-year follow-up period. The 4-year overall success rates were 85.1% and 72.5% for the arthroplasty and fusion groups, respectively (p = 0.004). The average range of motion for the Bryan disk was 8.48 degrees at 48 months. Total and serious adverse events were similar between the cohorts as were secondary surgical procedures at both the index and adjacent levels.

Coric et al. performed a prospective, randomized multicenter study evaluating 2-year outcomes of the Kineflex C artificial disk with ACDF for single-level pathology. Primary outcome measures such as VAS pain scores and overall clinical success favored the arthroplasty group. NDI improved significantly in both groups but there was no statistically significant difference between the two cohorts at 24 months. Reoperation rates at the adjacent levels showed no significant difference between the groups, although radiographic evidence of severe adjacent segment degeneration was significantly lower in the arthroplasty group.

Phillips and colleagues reported on a prospective, randomized series of 342 patients (189 PCM, 153 ACDF) with 2-year follow-up. At 2 years postoperatively, neck and arm VAS scores, NDI, SF-36, and neurologic status were significantly improved from preoperative baselines in both groups. Average NDI score at 2 years was found to be lower in the PCM group. There were no statistical differences between the groups in terms of serious surgery-related adverse events or secondary surgical procedures. The PCM cohort was found to have lower dysphagia scores and higher patient satisfaction.

It is important to note that although these IDE studies are of high methodologic quality they are not without inherent biases that may substantially impact the findings. Critics of the FDA IDE studies note several biases inherent in the study design. For example, the lack of blinding in these studies potentially allows for the introduction of patient bias favoring arthroplasty in reported outcomes, which is perceived as the newer, cutting edge, and, therefore, “better” option. Furthermore, the decision to reoperate, which affects the rate of secondary surgical procedures or adverse outcomes, is not a completely objective decision in most cases and is certainly subject to biases on the part of both surgeons and patients. Patients who believe that arthroplasty has allowed them to avoid all the pitfalls of ACDF they have read about on the internet may be much more tolerant of any lingering symptoms they may have since they believe they received the better option. Time to return to work favors arthroplasty in most studies, but may be due to less restrictions being placed by treating surgeons on arthroplasty versus fusion patients, rather than being a reflection of any inherent superiority of arthroplasty. Another possible criticism is the low overall success rate of roughly 70% in the fusion controls. In clinical practice, and in other studies, overall success rate is believed to be much higher. One possible reason for a difference in outcomes for ACDF patients within and outside of the IDE trials could be related to “buyer’s remorse” in patients who seek inclusion in trials wanting arthroplasty but end up randomized to the “old fashioned” ACDF procedure. Finally, the studies with longer-term follow-up data are limited by the relatively low rate of patient follow-up.

Only gold members can continue reading. Log In or Register to continue