Lumbar Disc Replacement: The Artificial Disc

43 Lumbar Disc Replacement: The Artificial Disc

Fred H. Geisler

Abstract

Lumbar spinal arthroplasty was first reported in clinical settings nearly 30 years ago in patients with lumbar degenerative disc disease. The early device underwent several mechanical improvements, including changes in its sterilization procedure to ensure long-term viability, and evolved into its final configuration of the CHARITÉ III. The CHARITÉ III artificial lumbar disc was studied in a Food and Drug Administration (FDA) investigational device exemption (IDE) randomized controlled trial (RCT) that compared arthroplasty to anterior lumbar interbody fusion. This trial was reported in 2005 and, since then, multiple other lumbar arthroplasty devices have been developed and are undergoing or completing clinical trials both in the United States and in Europe. Because lumbar total disc replacements require successful completion of RCT prior to market approval in the United States, lumbar arthroplasty devices have undergone more scrutiny and clinical evaluation than any other spinal medical devices. Although the randomized studies were initially planned for 2-year follow-up, the total follow-up was extended to 5 years in all the studies. These prospective randomized clinical trials have generated a large body of evidence on the safety and efficacy of arthroplasty for lumbar spine. Additionally, significant insights were developed regarding the impact of lumbar arthroplasty on sagittal alignment and motion, possible adverse events and reoperation, and optimal patient selection and indication. Health economics papers have been presented to understand the societal impact of this new technology. This review chapter is aimed at providing an overview of the surgical technique and clinical data related to spinal arthroplasty.

Keywords: lumbar arthroplasty, total disc replacement, lumbar artificial disc, clinical trial, statistical analysis, lumbar spine, revisions, reoperations, lumbar fusion, randomized study, FDA IDE randomized controlled trial, arthrodesis, prior surgery, effect of age

43.1 Introduction

Multiple intervertebral disc procedures have been developed to deal with abnormalities in the intervertebral disc. These include herniation of the nucleus pulposus, degenerative disc disease (DDD), and segmental instability. In recent years, the diagnostic accuracy and description of these abnormalities have been aided by the development of water-soluble myelography, MRI, provocative discograms, diagnostic blocks, and high-resolution CT scanning techniques with both intravenous and intrathecal contrast. Over the past 15 years, multiple therapeutic advances have also occurred to aid in managing intervertebral disc disease. These have included rigid segmental pedicle screw fixation (which has been shown to enhance the fusion rate over a noninstrumented fusion), carbon fiber–reinforced polymer cages, and allograft spacers placed in the anterior column to promote anterior column fusion; demineralized bone matrix, platelet-derived autologous growth factors (AGF), bone morphogenetic proteins (BMP), and numerous bone graft extenders to eliminate or minimize iliac crest bone graft harvested during a lumbar fusion procedure. There has also been recognition over the past decade that interbody stabilization and arthrodesis, in addition to posterior instrumentation and arthrodesis, enhances the total lumbar joint fusion rate. Interbody fusion can be accomplished either anteriorly through a separate incision or posteriorly via a posterior lumbar interbody fusion or transforaminal lumbar interbody fusion approach. Laparoscopic surgery has also been used in spine surgery for anterior cage insertion, and minimally invasive techniques posteriorly and posterolaterally have been developed. There are also several intradiscal therapies with internal decompression of the disc center or heating of the posterior annulus to minimize the patient’s surgical discomfort while potentially relieving some symptoms of low back disorder.

43.2 Advantages of an Artificial Disc

All of the above-mentioned techniques, however, either mask the true disease process or eliminate the joint motion and its normal physiologic function.

With lumbar artificial disc technology, we now have the ability to fix the problem and restore normal anatomy and physiologic motion rather than simply fuse the back.¹^,²^,³^,⁴ Fusion works in many instances because the motion of the joint itself causes pain through its inability to comfortably support the weight of the body. Thus, when it is fused, it no longer moves and hence the motion cannot cause pain. The fusion does, however, cause stress and increased motion in the joints adjacent to the fused level as a direct effect of eliminating motion at the fused level. The theory behind an artificial disc in the lumbar area is not only to preserve the motion but also to correct the abnormal motion that would be present in the degenerative disc and restore the disc height, lordosis, and a normal instantaneous axis of rotation. By doing so, the joints adjacent to the dynamically stable segment would not be subject to abnormal loads and motions. It is hoped with this new technology of the artificial lumbar disc that the good results that have followed the introduction of artificial knees and hips will likewise be seen in the lumbar spine.⁵

The advantage of the artificial lumbar disc as compared with a lumbar fusion is that it reproduces the biomechanics of the normal disc. Additionally, it reduces the mechanical forces transmitted to the adjacent segments. It has the promise of slowing or halting degenerative changes at adjacent levels. Performing a total discectomy eliminates the chance of disc herniation and will hopefully retard spondylosis, stenosis, and instability at the dynamically stabilized segment. By restoring the anatomic disc height, the artificial disc increases the exiting foraminal height and prevents compression on the exiting nerve roots at the level stabilized.

The typical diseased lumbar segment considered for artificial lumbar disc treatment is often collapsed in vertical height and exhibits loss of normal lordosis, modic end plate changes in adjacent vertebral bodies, and little motion on flexion–extension. Modic end plate changes in the bodies adjacent to the affected disc space, and little motion on flexion–extension. Because of the mechanical changes in the degenerative condition in the disc space, this natural disease process is already placing more forces on the adjacent levels. The application of an artificial lumbar disc will restore normal motion, height, and lordosis, and the forces on the adjacent level will be decreased. Thus, an artificial lumbar disc may have beneficial effects compared to the natural history of the nonoperated degenerative state.

43.3 Artificial Lumbar Disc Design

The design of a lumbar artificial disc has many very strict requirements. These devices must have superb mechanical strength and endurance. They are designed to last several decades, as many of these devices will be implanted in young individuals. Mechanical testing of 5,000,000 motion cycles over a 40-year life span would be a typical design criterion. The base materials need to be biocompatible with no significant surrounding inflammatory reaction either due to the base material reaction or secondary to any debris. The devices need to induce no organotoxic or carcinogenic reaction from the base material or potential debris. The biomechanical functional movement requirements of an artificial lumbar disc are quite strict, as they need to replicate the full biomechanics of a normal disc. This normal motion includes translation and rotation in all three planes of motion—X, Y, and Z axes. The implant geometry and materials would determine the statistical configuration, dynamic motion, schematics, and any constrained nature of the motion. The exact placement of the lumbar artificial disc in the disc space is determined by its biomechanical design. Different designs will require different placement accuracy—the “sweet spot” for the implant. Fixed pivot devices may need a higher placement precision than devices that use a sliding core or an elastopolymer.

The history of the lumbar artificial disc goes back 35 years to Fernström, who first placed spherical metal balls in the disc space. It was noted that a majority of these patients had ball migration into the vertebral body with subsequent collapse of the disc space. Relatively recently, a nucleus pulposus replacement with a hygroscopic gel or fluid-filled cylindrical sacs has been developed for use after a standard discectomy in which the annulus is still holding the disc space to a normal height. These gels are currently under development and have not yet begun U.S. Food and Drug Administration (FDA) trials. Replacements of the entire disc after severe degenerative changes have included a variety of designs, such as mechanical bearing devices and a rubber/silicone/polymer nucleus between metal end plates made out of either cobalt-chromium or titanium, with the potential for bony ingrowth surfaces at the end plates.

Although many different spinal dynamic stabilization systems go under the category of “artificial disc,” these need to be separated because they have different indications and potentially different applicable disease states. The first group of devices for the lumbar disc is intended to prevent the collapse of a lumbar disc space after a standard free-fragment disc herniation surgery.⁶^,⁷^,⁸ These devices are designed to be placed in the center of the disc to halt the secondary changes that would occur over the subsequent years and would hopefully provide stability over many decades, eliminating the need for fusion or rebuilding of the disc space at a later date. The second class of devices is for patients with severe DDD with loss of disc height but normal lordosis, instability of the disc, and little to no significant bony pathology posteriorly. This set of devices requires good facets, posterior ligaments, and muscular structures, as the aim is to replace only the degenerative disc component of the entire lumbar joint. These are currently what will be termed “artificial lumbar discs” and will be the focus of the remainder of this chapter.

Four different designs have finished FDA investigational device exemption (IDE) trials (ProDisc, Maverick, FlexiCore, and Kineflex), and two designs have completed IDE trials and received marketing approval (CHARITÉ and ProDisc-L; Fig. 43.1). It is notable that these devices do not replace the posterior column degenerative changes; neither do they augment them. In fact, a contraindication to any of these devices would be spondylolysis or significant spondylosis with facet hypertrophy and potential or ongoing nerve root compression. The third category of devices increases posterior column stiffness,⁹^,¹⁰^,¹¹ with one currently in an IDE trial and others reported in European surgical series. The fourth class of devices is the total lumbar joint replacement, which would replace both anterior and posterior components. Currently, no such devices are available in any U.S. FDA trials, and none are being implanted elsewhere in the world. In the lumbar spine, in addition to the hard implants, which have metal ends that attach onto the bony end plates, there are also some soft implants made of either all elastic with potential laminations or a sack of fiber filled with some fluid or matrix.¹²^,¹³^,¹⁴^,¹⁵ Currently, none of the soft implants are in FDA trials.

Fig. 43.1 Artificial lumbar discs that have completed Food and Drug Administration clinical trials separated by unconstrained and semiconstrained (fixed core) designs.

There are potential base material problems with all current technology solutions to the bearing surface for the hard lumbar artificial disc replacement designs. Broadly, these fall into three separate classifications: a metal–metal design, a metal–ceramic design, and a metal–plastic design. The metal–metal designs have the potential problem of metal and/or metal ionic debris; with the metal–ceramic designs, there is a risk that the ceramic component may shatter; and with the metal–plastic designs, there is the problem of the plastic wearing out. At first thought, the wear associated with a metal–plastic bearing surface would seem to exclude it from use in the lumbar spine because of excessive long-term wear. This initial opinion is an extrapolation from the well-known fact that the plastic components in the current artificial hips and knees have a 10-year lifetime and then require revision. Because the lumbar artificial disc is made of these same base materials, cobalt-chromium and high-density polyethylene, it might be inferred that the lumbar artificial disc would also require replacement of the plastic cores every 10 years. There are three facts, however, that refute this seemingly common sense idea. First of all, with each step, the hip and the knee move approximately 50 degrees, whereas the lumbar spine will only tilt a few degrees. This greatly decreases the “sandpaper effect” by over an order of magnitude. Next, in the lumbar design the high-density polyethylene is not constrained but is open on the sides. This is in marked contrast to the hips, where the plastic is constrained in a ball-/socket-type joint. In the hip joints, the high-pressure points that arise at the constrained metal–plastic interface greatly accelerate the wearing of the plastic. Because of the nonconstrained nature of the plastic in the lumbar application, there are no wear-accelerated pressure points. Furthermore, there is good experience from Europe that there is no plastic wear in 10 years of implantation, verifying the estimation of the expected lifetime to be far greater than that of the hips and knees.

A separate class of dynamic stabilization devices is currently being studied and tested in the cervical spine. These devices, although they are also called “artificial discs,” vary greatly from the lumbar artificial disc. First of all, cervical discs are experiencing much lower loads than lumbar discs, and they have different biomechanical characteristics. But more importantly, in the cervical spine, bony pathology and osteophytes causing radiculopathy and/or myelopathy predominate as causes of intervention rather than pure axial neck pain, as is the case in the lumbar indication (axial back pain) for an artificial disc. The potential patient groups to be studied and the outcome variables are quite different between the cervical and lumbar artificial disc studies. Furthermore, the results achieved in the lumbar area are not necessarily directly transferable to the cervical spine.

In summary, all artificial discs are not the same. There will be major biomechanical differences between cervical and lumbar implants in the design, the disease treated, and the outcome expected. One needs to consider the pathology that is being treated, whether the disc is of normal disc height, and whether the patient has DDD, osteophytes, and/or facet disease before selecting the best treatment option.

43.3.1 Acroflex Artificial Lumbar Disc

The earliest lumbar motion–preserving implant was the Acroflex disc. It was composed of a rubber core vulcanized to two titanium end plates and was pioneered by Dr. A. Steffee in Cleveland. The spinal kinematics, histologic osseointegration, and particulate wear debris after total disc arthroplasty using in vitro and in vivo models were studied.¹⁶ The Acroflex disc was also investigated in nonhuman primate model for bony ingrowth to the metal end plates regarding fixation as well as preservation of lumbar motion.¹⁷ Three-year follow-up showed that in a small number of patients (n = 6) the results were sufficiently satisfactory to proceed on with a larger study.¹⁸ However, larger and longer-term pivotal studies were not done because of detection of mechanical failure of the elastomer on thin-cut CT scans, and examples of osteolysis were also discovered.¹⁹^,²⁰ Although this design was found to be flawed, it laid the foundation for the modern designs of lumbar total disc replacement (TDR).

43.3.2 ProDisc-L Artificial Lumbar Disc

The initial ProDisc-L product was designed in the late 1980s and used by Thierry Marnay, a French orthopedic spine surgeon. From March 1990 to February 1993, Dr. Marnay implanted this artificial disc in 64 patients. In 1999, he went back to examine these patients. He was able to locate 58 of the surviving 61 patients for a 95% follow-up rate at 7 to 10 years after the procedure. At that time, he found that all of the implants were intact and mechanically functioning. There had been no implant removals, revisions, or failures. Furthermore, there was no evidence of subsidence into the bony end plate on follow-up radiographs compared with the perioperative films. There was a highly significant reduction in patient-reported back pain and leg pain, and 92.7% of these patients were either satisfied or extremely satisfied with the results of this procedure. In this study, two-thirds of the patients had single-level implants and one-third had two-level implants. No differences were noted between oneand two-level diseases. Most importantly, at this long-term follow-up there were no device-related safety issues, no untoward effects, no complications, and no adverse events.²¹^,²²^,²³^,²⁴ This ProDisc-L was based on spherical articulation and had metal end plates. The current ProDisc-L (Synthes; Fig. 43.2), which was approved by the FDA in August 2006, consists of two cobalt-chromium end plates and a high-density polyethylene core, and is applied with an inserter no wider than the implant. It has a fin in the midline to help in stabilization and positioning. Because the high-density polyethylene is fixed to the inferior plate, it functions as a fixed pivot design, and the instantaneous axis of rotation is within the lower body rather than in the disc space.

Fig. 43.2 ProDisc-L artificial lumbar disc. (a) Expanded view showing the two metal end plates and polyethylene core that attach to the lower metal end plate. (b) Assembled ProDisc-L artificial lumbar disc construct. (c–e) Sequence of steps involved in the insertion of the ProDisc-L artificial disc into the L4–L5 disc space. (c–e; reproduced with permission of Synthes Spine Co., LP.) (f,g) L5–S1 one-level ProDisc stabilization. (h,i) L4–L5 and L5–S1 two-level ProDisc-L stabilization.

The ProDisc-L has been studied in a FDA IDE randomized trial. The initial 2-year results of this trial were published in 2007 ²⁵ and report that the ProDisc-L showed superior results to circumferential fusion by multiple clinical criteria. Zigler and Delamarter²⁶^,²⁷^,²⁸ and Zigler et al²⁸ published the 5-year follow-up data on this study and reported that the large clinical improvements that occurred immediately after surgery were maintained throughout the 5-year follow-up period. This data set was also used to report in detail the degenerative changes at the adjacent lumbar level and found that the ProDisc-L case that maintained range of motion (ROM) of the segment was associated with a significantly lower rate of adjacent-level disease compared with the circumferential group.²⁷ The use of two-level ProDisc-L has been shown to lead to improvement in visual analog scale (VAS) and Oswestry Disability Index scores similar to the one-level ProDisc-L.²⁹

43.3.3 Maverick, FlexiCore, Kineflex, Mobidisc, and Activ-L Artificial Lumbar Discs

Four other TDR implants, Maverick,³⁰^,³¹ FlexiCore,³² Kineflex, and Mobidisc, were in FDA IDE trials for which enrollment was initiated about 10 years ago, and have completed planned enrollment and 2-year follow-up ( Fig. 43.3). Notably, the data in the Kineflex and the Mobidisc study were so incompletely reported that they were not included at all in a recent review of the TDR literature.³³ All four of these prospective studies are incompletely reported in the literature as a result, in large part, of poor reimbursement insurance potential and perceived future FDA regulatory difficulties regarding a potential metal-on-metal bearing problem related to the real metal debris problem in some metal-on-metal hip prostheses. The lack of complete reporting in these studies has left several open questions, which may in fact be answerable by the studies.

These questions include the metal-on-metal bearing surface safety in TDR, and long-term safety information on the use of BMP-2 (Infuse) in conjunction with a metal anterior lumbar interbody fusion (ALIF) implant in the Maverick study. Despite the poor reporting of the entire series, some papers on these studies have provided valuable information such as the histological and retrieval findings of Activ-L and Mobidisc TDR in two patients,³⁴ and biomechanics with TDR by examining the mismatched center of rotation on the clinical outcomes and flexion–extension ROM with 5.5-year follow-up.³⁵

A newer lumbar TDR Activ-L (B. Braun; Aesculap, Tuttlingen, Germany) just completed the FDA IDE trial in 2015.³⁶ This new design was a result of wear simulation tests to predict the mid- and long-term clinical wear behavior for total disc arthroplasty and of biomechanical analysis in human cadaveric spines.³⁷^,³⁸

Zander et al³⁹ used a validated finite-element model of the lumbosacral spine to compare the results of total disc arthroplasty influence from different artificial disc (CHARITÉ, ProDisc, Activ-L) kinematics on spine biomechanics. The spinal kinematic alterations due to an artificial disc exceed by far the interimplant differences, while facet joint contact force alterations are strongly implant and load case dependent.

Additionally, Wiechert⁴⁰ commented on the modular design of the Activ-L TDR, which allows for a flexible anchoring concept with either spikes or one or two keels. The advantages of a semiconstrained design that allows for some movement of an ultrahigh-molecular-weight polyethylene (UHMWPE) inlay were also discussed. Lu et al⁴¹ reported on the early experience in China with both a prospective and a retrospective study series⁴² and found the clinical outcome comparable to other forms of treatment for DDD.

43.3.4 CHARITÉ Artificial Disc

The CHARITÉ artificial disc (DePuy Spine, Raynham, MA) was designed to restore disc space height and motion segment flexibility, and to duplicate the kinematics and dynamics of a normal motion segment.⁴³^,⁴⁴ It was designed to restore anatomic lordosis, which will result in normal facet joint motion loading and unloading ( Fig. 43.4). The CHARITÉ artificial disc uses two metal alloy end plates of chromium-cobalt and a high-density polyethylene free-floating core. The free-floating core offers the theoretical advantage of allowing the spacer to shift dynamically within the disc space during regular spinal motion, moving posteriorly in flexion and anteriorly in lumbar extension. This not only provides unloading of the posterior facet structures during this normal replication of motion, but also allows forgiveness for slight off-center positioning of the implant.

Several clinical studies have been published documenting the European experience with this disc since 1987. Worldwide experience with this unconstrained anatomic disc replacement now exceeds 11,000 cases. Several studies are historically notable. Cinotti et al⁴⁵ reported on 46 Italian patients in 1996 with 2- to 5-year follow-ups. They noted no implant failures but did report a reoperation rate of 19% for continued pain. Overall satisfaction was 63%. Lemaire,⁴⁶ who reported on his French series in 1997, followed 105 patients for a mean follow-up period of 51 months with 79% good outcomes and no device failures. Zeegers et al,⁴⁷ in 1999, reported on 50 patients in a Dutch series, which showed 70% good results with 2-year follow-up.

Recently, Lemaire et al⁴⁸ reported long-term follow-up of clinical and radiologic outcomes for the CHARITÉ artificial disc with a minimum follow-up of 10 years. Of the 107 patients treated with the CHARITÉ artificial disc, 100 were followed for a minimum of 10 years (range 10–13.4 years). A total of 147 prostheses were implanted with 54 one-level procedures, 45 two-level procedures, and 1 three-level procedure. Clinically, 62% had an excellent outcome, 28% had a good outcome, and 10% had a poor outcome. Of the 95 patients who were eligible to return to work, 88 (91.5%) either returned to the same job they had before surgery or took a different job. Mean flexion–extension motion was 10.3 degrees for all levels (12.0 degrees at L3–L4, 9.6 degrees at L4–L5, and 9.2 degrees at L5–S1). Slight subsidence was observed in two patients, but they did not require further surgery. No subluxation of the prostheses and no cases of spontaneous arthrodesis were identified. There was one case of disc height loss of 1 mm. Five patients required a secondary posterior arthrodesis. A good or excellent clinical outcome rate of 90% and a return-to-work rate of 91.5% compare favorably with results described in the literature for fusion for the treatment of lumbar DDD. Lemaire et al⁴⁸ concluded that with a minimum follow-up of 10 years, the CHARITÉ artificial disc demonstrated excellent flexion–extension and lateral ROM with no significant complications.

Fig. 43.3 FlexiCore artificial lumbar disc shown in anteroposterior (a) and lateral (b) views. This is a metal-on-metal device with some contour to the end plate attachment to match the normal anatomy and cleat attachment.

Fig. 43.4 CHARITÉ artificial disc. (a) Fully assembled construct and disassembled parts showing the two metal end plates (b) and the ultrahigh-density polyethylene core (c). Note the various-sized footprints that are available. The metal end plates are also available in different angulations.

The FDA IDE study of the CHARITÉ artificial disc initiated patient enrollment in May 2000, with the Texas Back Institute as the principal institution. Since that time, all patients have been enrolled in the FDA multicenter study with complete 2-year follow-up completed in December 2003. Following the randomized arm of the study, the centers had access to the CHARITÉ artificial disc, on a limited basis, as part of a continuing access study. The FDA granted approval for marketing the CHARITÉ artificial disc in the United States on October 26, 2004. The FDA label states ⁴⁹^,⁵⁰^,⁵¹^,⁵²:

The CHARITÉ artificial disc is indicated for spinal arthroplasty in skeletally mature patients with DDD at one level from L4–S1. DDD is defined as discogenic back pain with degeneration of the disc confirmed by patient history and radiographic studies. These patients with DDD should have no more than 3 mm of spondylolisthesis at the involved level. Patients receiving the CHARITÉ artificial disc should have failed at least 6 months of conservative treatment prior to implantation of the CHARITÉ artificial disc.

The CHARITÉ artificial disc comes in a variety of base metal sizes as far as the footplate to fit different-sized disc spaces.⁴³ In addition, there are various end plate angles to match the distracted disc space anatomy. The plastic core is inserted between the two metal end plates and also comes in a variety of heights. The sizing of the end plates, angles, and heights is done intraoperatively. The CHARITÉ artificial disc is implanted with metal end plates of chromium-cobalt on the superior and inferior bony end plates of the disc space and an UHMWPE core between the highly polished insert interfaces. There is a slight difference in the curvature between the polyethylene core and the metal end plates, which allows the core to slide. Spinal forces are transmitted down through the anterior column in a normal manner after the disc is inserted. The core translation allows duplication of anatomic translation ( Fig. 43.5, Fig. 43.6, Fig. 43.7). Under normal physiologic circumstances, there is a slight translation during the flexion–extension motion and lateral bending motion. A normal disc is able to handle this translation. In sagittal rotation (flexion–extension) under normal circumstances, the instantaneous axis of rotation, although generally in the center of the disc, moves in a pattern that duplicates the Greek letter alpha. The CHARITÉ artificial disc duplicates this motion. Coronal motion, likewise, has this slight translation in order to reproduce the normal biomechanics of the intact disc space. Axial rotation also requires a slight coupled rotation–translation to reduce the forces on the posterior facets. Pure axial rotation on a pivot point in the disc space will result in direct compression of one facet joint while releasing the pressure on the other. If one compares a fixed pivot design to a sliding core design, the facet pressure would be more in a fixed pivot design than in a sliding core design. The exact clinical benefit for those patients who are most helped by the sliding core design will be determined by the outcome of the clinical studies currently underway. It is evident from measurements of the center of intervertebral rotation in cadavers that the CHARITÉ artificial disc preserves normal motion ( Fig. 43.8) not only at the repaired disc space, but also at adjacent levels.⁵³ Fusion has been reported to greatly distort the instantaneous axis of rotation at adjacent levels.⁵³