Fusion Is a Simpler and More Predictable Operation than Total Disc Replacement

If a patient has failed nonsurgical treatment and a diagnostic workup warrants surgery, lumbar fusion has a single goal: to eliminate motion at the affected segment. If the fusion is solid and the patient continues to have pain, the pain is obviously not originating from the fused motion segment. In other words, the surgeon’s rationale for performing the operation was not correct. Recent evidence suggests that motion-sparing procedures are more challenging than fusion.^5–10 Implant design, size, precise positioning, status of facet joints, and other variables play significant roles in the outcome of the procedure. In fact, the current literature suggests that the majority of implant-related complications are due to a poor choice of size or position of the implant.^11,12 Although current implants are better than their predecessors, they are far from optimal. Mikeal et al. showed that current implants do not fit 98% of the patients. Huang et al. studied the prevalence of contraindications to TDR in a cohort of lumbar surgical patients.¹³ The authors reviewed 100 consecutive patients who had lumbar surgery carried out by one surgeon in 2002. Contraindications to TDR included central or lateral recess stenosis, facet joint arthrosis, spondylolysis or spondylolisthesis, herniated nucleus pulposus with radiculopathy, scoliosis, osteoporosis, postsurgical pseudarthrosis, or deficiency of the dorsal elements. Patients were divided into fusion or nonfusion groups, and the percentage of patients without contraindications to TDR was calculated. Out of 100 patients, 56 underwent spine fusion and 44 underwent nonfusion surgery. In the fusion group, 56 of 56 (100%) patients had contraindications to TDR. In the nonfusion group, 11% (5 of 44) were considered candidates for TDR. Overall, 5% of patients were considered candidates for TDR. The average number of contraindications to TDR was 2.48 (range, 0–5).

Total Disc Replacement Does Not Provide Greater Improvement of Symptoms and Function

Blumenthal et al. reported the outcomes of the first randomized control trial comparing Charité TDR and ALIF with the BAK cage filled with autograft from iliac crest.¹⁴ The study consisted of 304 patients with single-level L4-5 or L5-S1 degenerative disc disease and chronic low back pain. Clinical success was defined as at least 15 points (25%) improvement in Oswestry Disability Index (ODI) scores versus baseline data, no evidence of implant failure, absence of major complications, and maintenance or improvement of neurologic status. There were no differences in operative time, blood loss, or level of surgery between the two treatment groups. Length of hospitalization was significantly shorter in patients with TDR (3.7 vs. 4.2 days, P = .0039). TDR patients also enjoyed statistically greater improvement in visual analogue scale (VAS) and ODI scores at 12 months. However, this difference disappeared at 24 months. The VAS score was 40.6 in the TDR group and 34.1 in the fusion group. The ODI score change at 24 months was 48.5% for the TDR group and 42.4% for the fusion group. Zigler et al. reported the results of a similarly designed randomized clinical trial evaluating Prodisc-L TDR versus ALIF combined with dorsolateral fusion and instrumentation.¹⁵ The TDR group experienced statistically significant less intraoperative blood loss, operative time, and length of stay. These results are not surprising, considering that the control group underwent additional dorsal surgery. In contrast to Blumenthal et al.’s study,¹⁴ clinical success in Zigler et al.’s study was defined as at least 15% improvement in ODI score. At 24 months, ODI scores were similar in the two treatment groups. The VAS improvements for TDR patients and fusion patients were 39 and 32, respectively. Guyer et al reported on 5-year outcomes of Charité TDR RCT.¹⁶ The results showed a 57.8% success rate in the TDR group versus 51.2% in the fusion patients (P = .0359). No statistically significant differences were detected in ODI, VAS pain scores, or 36-Item Short Form Health Survey functional outcomes. However, 65.6% of TDR patients were employed at 5 years after surgery compared to 46.5% of fusion patients (P = .0403). Zindrick et al. performed an extensive review of the TDR literature from 1990 to 2007.¹⁷ The majority of studies were excluded, owing to lack of relevant data. The remaining clinical reports (level 3 or 4 evidence) often lacked consistency in outcomes or did not use a control group, precluding any meaningful conclusion. On the basis of available level I and II evidence, TDR does not result in better improvement of symptoms and function even against suboptimal controls.

Total Disc Replacement Maintains or Restores Motion at Affected Segments

McAfee et al. analyzed flexion-extension and translation angles in patients from the Charité IDE study.¹⁰ At 24 months, the TDR group demonstrated a 13.6% increase in range of motion compared to baseline. Interestingly, 40% of Charité TDR patients showed 5 degrees of motion less than baseline. These patients were potentially fused at 2 years using Food and Drug Administration criteria (adequate motion = 5 degrees or more; fusion = 5 degrees or less). Zigler et al. reported flexion-extension range of motion of 7.7 ± 4.67 degrees in a Prodisc-L IDE trial.¹⁵ Guyer el al. reported no change in range of motion at 5-year follow up.¹⁶ Despite these limitations, these results suggest early maintenance of range of motion with TDR.

Total Disc Replacement Maintains or Restores Motion to the Adjacent Segment—Does This Translate to a Decreased Incidence of Adjacent-Level Disease?

Adjacent-level disease is a well-recognized potential consequence of lumbar fusion. Harrop et al. performed a meta-analysis of available literature focusing on adjacent-level disease after fusion and TDR.¹⁸ The authors concluded that adjacent-level disease developed in 34% of patients after lumbar fusion and 9% after TDR. Fourteen percent of fusion patients developed clinically significant symptoms related to adjacent-level disease compared to 1% of TDR patients. Using odds ratio analysis, the authors concluded that fusion patients are 4.67 times more likely to develop adjacent-segment disease than TDR patients are. Huang et al. evaluated the relationship between range of motion and adjacent-level disease in patients with TDR. Lateral lumbar radiographs of 42 patients were analyzed.¹⁹ The authors defined adjacent-level disease as a loss of motion, disc height narrowing, and osteophyte formation. Ten patients (24%) developed radiographic evidence of adjacent-level disease. These findings correlated with loss of motion at the index TDR level. Putzier et al. reported similar findings in their series.²⁰ These data suggest that the development of adjacent-level disease is greater in patients with lumbar fusion. No level 1 or 2 data are available to address the issue of adjacent-level disease.

Conclusion

While TDR is certainly a very appealing technology, to date, no study has shown TDR to be superior to lumbar fusion in terms of clinical outcome. Currently available data suggest that a well-performed and appropriately positioned TDR may provide pain relief and functional improvement similar to those of lumbar fusion.

It seems that TDR may have positive effect on incidence of adjacent-level disease. Many other clinical questions remained unanswered. Therefore, TDR is not currently a replacement for lumbar fusion.

Lumbar Artificial Disc

Mechanical low back pain continues to be a major cause of morbidity. The diagnosis and management of mechanical low back pain are responsible for a significant portion of annual health-care costs. The etiology of mechanical low back pain (LBP) is multifactorial but may, in part, be related to segmental degenerative changes. Mechanical LBP may be difficult to accurately diagnose and effectively treat. Patients with mechanical LBP are first managed conservatively with therapies that include epidural steroid injections, nonsteroidal anti-inflammatory agents, oral narcotic medications, and physical therapy to strengthen core muscles. If conservative measures are exhausted without relief, surgical intervention may be considered. The surgical management of mechanical LBP has historically centered on fusion of the abnormal or dysfunctional motion segment, with subsequent loss of motion.

It has been opined that the fusion of a spine motion segment leads to increased adjacent-segment degenerative changes. The incidence of such changes has been reported to range from 10% to 30%,^1–10 which has led to the development and evolution of other treatment modalities for mechanical LBP, including artificial disc replacement. Lumbar artificial disc replacement facilitates the relief of pain by eliminating presumed pain generators (including the annulus fibrosus and nucleus pulposus) while preserving motion, thereby allegedly minimizing the development of adjacent-segment disease.

Lumbar artificial discs can be divided into total disc replacement and partial disc replacement, which is also known as nucleus replacement. The recent explosion in the number of lumbar artificial disc devices was spurred by the development of the first-generation Charité artificial disc by Butner-Janz and Schellnack and the Prodisc-L by Marnay. Lumbar artificial disc devices are categorized in several ways. These devices can be categorized on the basis of their biomechanical properties or on the basis of design of composite materials. Alternatively, they may be categorized on the basis of the number of components that make up the fully assembled device. In general, they are broadly classified as either metal-on-metal or metal-on-polymer (based on composite materials) or as unconstrained, semiconstrained, or constrained (based on biomechanics). An example of an unconstrained device is the Charité. Examples of the unconstrained devices with a fixed center of rotation include the ProDisc, Maverick, and FlexiCore discs. Thus far, the Food and Drug Administration (FDA) has approved several lumbar artificial discs for implantation in the United States. These include the Charité artificial disc (2004), Prodisc-L (2006), and the next generation Charité artificial disc (2007). Other devices are in various stages of development or have received an investigational device exemption from the FDA.

Lumbar Fusion: Advantages and Disadvantages

Prior to the development of lumbar artificial discs, the surgical treatment for mechanical LBP due to degenerative disc disease consisted of arthrodesis of the painful motion segment. Fusion is intended to eliminate painful, nonphysiologic motion across the dysfunctional motion segment. Spine fusion alters the biomechanics of the spine and results in the transmission forces to adjacent vertebral levels, which may lead to accelerated disc degeneration, facet arthropathy, and osteophyte formation. Recurrent back pain and symptomatic spinal stenosis at levels adjacent to the fusion site may result.^11–14

Advantages of Lumbar Artificial Disc

Lumbar artificial disc technology, to one degree or another, restores the normal disc space height and preserves physiologic motion rather than eliminating it as fusion does.^15–18 A degenerated disc leads to an abnormal joint that is unable to completely support the weight of the body, thereby causing pain with motion. Fusion of the joint eliminates the abnormal motion, thereby theoretically relieving the pain. Artificial lumbar discs preserve motion, restore disc height and lumbar lordosis, and attempt to replicate the normal instantaneous axis of rotation without exerting undue stress on the adjacent segments. Artificial joints are not a new concept in orthopedics as demonstrated by the various widely available artificial hip and knee joints. As technology continually improves, similar success may be encountered in lumbar artificial discs.¹⁹

To place an artificial disc, the degenerated disc is completely removed. This eliminates the risk of further disc herniation, thus ideally retarding spondylosis, stenosis, and instability. Placement of the artificial disc restores anatomic disc height and prevents exiting nerve root compression at the dynamically restored level.

Indications and Contraindications for Lumbar Artificial Discs

Current candidates for artificial lumbar discs are young patients with single-level disc disease and nonosteoporotic bone, without instability and facet arthropathy. Since disc replacements provide only anterior column reconstruction, patients with significant dorsal disease are not good candidates. Indications for disc replacement include the following:

The other indications for the implantation of the disc include the following:

Absolute contraindications for implanting a lumbar artificial disc include the following:

Relative contraindications to lumbar artificial disc implantation include the following:

Lumbar Artificial Disc Design

Lumbar artificial discs are designed to withstand a significant amount of wear and tear. The materials that are used are designed to be biocompatible and not cause an inflammatory reaction when implanted. The artificial disc is designed to replicate the biomechanics of the normal disc and should be able to withstand normal motion, which includes translation and rotation in the x-, y-, and z-axes.

Fernstrom, in the 1960s, is credited with the first attempt at implanting an artificial disc. He placed spherical metal balls in the disc space, but this treatment was fraught with migration of the spherical ball into the vertebral body, leading to disc space collapse. Contemporary devices are described as unconstrained (e.g., Charité) or semiconstrained with a fixed center of rotation (e.g., ProDisc, Maverick, and FlexiCore). It is important to note that these devices do not eliminate or even treat posterior column degenerative changes. The biomaterials that are used in the design of lumbar artificial discs are chosen carefully, since they must bear significant loads for a prolonged period of time. The biomaterials fall into three types: metal-metal design, metal-ceramic design, and metal-polymer design. Each of these designs may have potential problems that currently do not seem to have clinical significance. The metal-metal design may lead to the deposition of metal or metal-ionic debris. In the metal-ceramic designs, the ceramic may eventually be prone to shatter. Finally, the metal-polymer design may be prone to polymer wear. The potential wear of the metal-polymer design is extrapolated from the available data on artificial hips and knees, which require replacement as often as every 10 years. This information does not necessarily apply to a lumbar artificial disc because the hip and knee are exposed to more motion than the lumbar spine is.

Biomechanics of Lumbar Artificial Discs

Wear Testing

Lumbar artificial discs are relatively new devices, and as such, when compared to artificial hips and knees, little is known about in vivo degradation or the contribution of wear to implant failure. Retrieval studies have shown that clinically relevant wear and polyethylene degradation (metal-polyethylene implants) occur. Chronic inflammatory reactions and wear debris have also been observed in tissue surrounding failed artificial discs. The wear potential of the ProDisc L was determined by using an adaptive finite element technique.²³ The testing scheme was validated by using a model of a total hip replacement and was then used to model the ProDisc L. The degree of flexion-extension, lateral bending, axial twist, and axial loading were applied through 10 million simulation cycles. The polyethylene wear rate of the ProDisc L was 9.8 mg/million cycles. The comparable rate for a total hip replacement is 16.1 mg/million cycles.²³ Metal-on-metal wear for the Maverick artificial disc has been tested extensively, and there is essentially no metal wear at 31.5 years equivalent of simulated loading.

Surgical Implantation

The majority of the lumbar artificial discs are implanted through a ventral approach, with the notable exception of the Prosthetic Disc Nucleus (RayMedica, Bloomington, MN). Most are implanted via an approach similar to that utilized in performing anterior lumbar interbody fusions (ALIF).^24,25 Obviously, there are obligatory device-specific nuances. The implantation technique for the SB Charité III artificial disc (Depuy Spine, Johnson and Johnson, Raynham, MA) is as follows^24,25: The lumbar spine is approached via a retroperitoneal approach, usually on the right side. A standard anterior discectomy is performed, with care taken to avoid violating the end plates. The disc space is then distracted with a lamina spreader, followed by end-plate sizing and the actual implantation of the end plate of the prosthesis. It is crucial for the implant to be placed centrally in the disc space. This can be confirmed with anteroposterior fluoroscopy. Next, the polyethylene core is trialed and then implanted. The distracting device is released, and a final confirmatory radiograph is obtained.

ProDisc-L

ProDisc-L (Synthes, Paoli, PA; Fig. 234-1) is an artificial disc that has three implant components. It has two cobalt-chrome and molybdenum alloy end plates that are plasma sprayed with titanium. Each has a vertical keel for acute fixation to the vertebral bodies. The rostral component is composed of a solid one-piece design available in two lordotic angles: 6 and 11 degrees. There are two implant sizes, a medium and a large. The medium size is 27 mm in the anteroposterior dimension and 34.5 mm in the lateral-lateral dimension. The large size is 30 mm in the anteroposterior dimension and 39 mm in the medial-lateral dimension. The disc also has a high-molecular-weight polyethylene inlay. The inlay is available in 10-, 12-, and 14-mm sizes. The inlay locks into the caudal end plate. In terms of anatomic sizing, the anteroposterior dimension of each end plate must be at least 28 mm, and the lateral-lateral dimension of the end plate must be at least 35 mm to permit implantation of the ProDisc-L. This can be determined on preoperative axial CT scans.

FIGURE 234-1 ProDisc-L.

(Courtesy of Synthes, Paoli, PA.)

Charité III Disc

The SB Charité III disc (Fig. 234-2) consists of two cobalt chromium end plates with an intervening ultra-high-molecular-weight polyethylene sliding core. This free-floating core facilitates spacer dynamic shifting within the disc space during normal spinal motion. It moves dorsally in flexion and ventrally in extension. The end plates have anchoring teeth and a titanium/hydroxyapatite coating to enhance fixation to the vertebral body. The dynamic stabilizer of the Charité artificial disc comes in a variety sizes. During implantation, the largest size that the disc space will accommodate is typically inserted. There are various end-plate angles to match the distracted disc space anatomy. The polyethylene core is inserted between the two end plates and comes in a variety of heights. The appropriate size and angles of the end plates are determined intraoperatively, as is the height of the polyethylene core. The Charité lumbar artificial disc is implanted with the metal end plates on the rostral and caudal end plates, with the ultra-high-molecular-weight polyethylene core placed between the two metal end plates. The fact that the core is free floating facilitates replication of normal and natural translation. The concept behind the mobile sliding core of the Charité disc is similar to the mobile knee bearing in many of today’s artificial knee joints.

FIGURE 234-2 Charité lumbar artificial disc.

(Courtesy of Depuy Spine, Johnson and Johnson, Raynham, MA.)

Maverick Disc

The Maverick (Medtronic Sofamor-Danek, Memphis, TN) total disc replacement is associated with a relatively dorsal center of rotation that mimics that of the normal lumbar motion segment. It has a metal-on-metal cobalt chrome molybdenum design that allows for 16 degrees of motion in all planes. It has a semiconstrained design. This device simulates the intact functional spinal unit in flexion-extension and lateral bending and is stiffer than the intact spinal unit in compression, bending, and rotation. The implant has a keel and has a hydroxyapatite coating on both components that promotes bone growth on the implant. It is designed as a two-piece ball-and-socket joint replacement, with three overall sizes for end plate coverage: 25-, 27-, and 30-mm anteroposterior diameter coverage and 32-, 35-, and 39-mm lateral end-plate coverage. It comes in three dorsal disc heights: 10, 12, and 14 mm. There are three lordosis configurations: 6, 9, and 12 degrees (Fig. 234-3).

FIGURE 234-3 Maverick lumbar artificial disc.

(Courtesy of Medtronic Sofamor-Danek, Memphis, TN.)

FlexiCore Intervertebral Disc

The FlexiCore intervertebral disc (Stryker Spine, Allendale, NJ) is an all-metal total disc replacement device that consists of superior and inferior baseplates that articulate about a centrally located stationary center of rotation. Each baseplate has a central dome that is shaped to approximate the concavity of an adjacent vertebral body end plate. The domes have spikes that facilitate fixation. They are coated with a titanium plasma spray to aid fixation. The baseplates are joined by a central ball-and-socket joint that is set during the manufacturing process to prevent the device from separating under tension loads. The ball-and-socket joint provides ±15 degrees of flexion-extension and lateral bending. This exceeds the natural range of motion without allowing the baseplates to slide. The bearing surfaces of the ball-and-socket components are highly polished cobalt chromium, which allows for a metal-on-metal surface contact that virtually eliminates the creep that is observed with metal-on-polyethylene devices (Fig. 234-4). Sasso et al.²⁶ recently published their experience with the FlexiCore disc.

FIGURE 234-4 FlexiCore artificial disc. (Courtesy of Stryker Spine, Allendate, NJ.)

Kineflex Artificial Disc

The Kineflex artificial disc (SpinalMotion, Inc., Mountain View, CA) has a metal-on-metal design with a semiconstrained core. It is made of cobalt chrome, and the mobile core is seated within a retention ring. The rostral and caudal end plates have multiple serrations in addition to a central fin that allows for keel fixation and initial stability (Fig. 234-5).

FIGURE 234-5 Kineflex lumbar artificial disc.

(Courtesy of SpinalMotion, Mountain View, CA.)

Complications of Lumbar Artificial Disc Implants

As one might imagine, the complications associated with artificial discs are similar to those that occur with anterior lumbar interbody fusions. These include facet joint degeneration, unexplained radiculopathy and back pain, extrusion of the implant, facet fracture, and acquired spondylolysis.^27,28 Some of these complications have been ascribed to suboptimal device placement, sizing errors, surgical approach errors, inappropriate indications, and problems with patient compliance.^29,30 Implant extrusion may be encountered as a result of absence of the anterior longitudinal ligament, which is incised in order to place the implant. Under normal circumstances, this ligament serves to limit excessive extension in the disc space, and when it is excised, excessive lumbar extension may occur. This may increase the risk of implant extrusion.

Revision Strategies for Lumbar Artificial Disc Surgery

McAfee et al.³¹ report an 8.8% revision rate in a group of 688 patients. As one can imagine, the risks of the anterior surgical approach to the spine are higher in revision surgery. These include visceral injuries and vascular injuries that may occur as a result of increased fragility of the blood vessels due to adhesions. The natural anatomic planes are also ill defined and may lead to difficulty identifying normal anatomy. Most of the visceral injuries may be repaired by primarily suturing the damaged organs. The great vessels lie anterior to the discs above the L5-S1 level, and the bifurcation of the aorta and vena cava lie superior to the L5-S1 disc space, so vascular complications could be devastating and life threatening.^32–34 McAfee et al.³¹ report a 3.6% incidence of vessel injury in primary lumbar artificial disc placement and 16.7% in anterior revision surgery.

Other complications that may be associated with both primary and revision surgery include ureteral injuries and injury to the sympathetic chain, which could lead to both retrograde ejaculation and impotence.³⁵ It is advisable that a vascular surgeon assist in the reoperation approach with the spine surgeon because of the increased risk of complications associated with the exposure. Angiography and venography may also be helpful in identifying any areas of compression of the great vessels. Catheterization of the iliac vessels with inflatable balloon catheters provides the option for intraoperative tamponade of vessels to help prevent significant blood loss. Some surgeons have advocated a transperitoneal or contralateral approach for revision surgery, as it may help to avoid adhesions that result from the initial operation. There has been a lot of recent interest in adhesion prevention and a number of commercially available barriers have been developed to help limit or eliminate adhesions. Some of these materials are nonresorbable; others are resorbable. The nonresorbale barriers function as a permanent mechanical barrier between tissue and are placed at the time of the initial surgery. They can serve as a marker if reoperation is required. An example of this is Gore Preclude (W.L. Gore and Associates, Inc., Flagstaff, AZ). Resorbable barriers may be solid or liquid and can function as both physical and biologic barriers to adhesion formation.³⁶ Examples of the solid resorbable barriers include Interceed R (Ethicon, Somerville, NJ) and Seprafilm (Genzyme, Cambridge, MA). The liquid-based resorbable barriers offer increased ease of application and include Sepracoat (Genzyme), Adcon gels (Wright Medical, Arlington TN), and Intergel (Lifecore Biomedical Inc., Chaska, MN).

Conclusion

Lumbar artificial discs are used to minimize adjacent-segment disease by preserving segmental motion. They are also used to alleviate neuroforaminal stenosis by restoring disc height. Finally, they are used to manage mechanical low back pain. It has proved difficult to confirm efficacy to date. However, in the short term and in experienced hands, it appears that the results with lumbar artificial discs appear equivalent to fusion. Further research is required to determine the life span of these devices, the outcome compared to fusion, and the results of revision surgery. Also, further research will most certainly improve the design of the artificial disc. Unfortunately, there are currently many more unknowns than knowns. It remains to be seen where these devices will best fit into clinical practice.