Spinal Disc Replacement: The Development of Artificial Discs
Aging results in deterioration of the biological and mechanical integrity of the intervertebral discs. Disc degeneration may produce pain directly or perturb the functional spinal unit in such a way as to produce a number of painful entities. Whether through direct or indirect pathways, intervertebral disc degeneration is a leading cause of pain and disability in adults (1). Approximately 80% of Americans experience at least a single episode of significant back pain in their lifetime, and for many individuals, spinal disorders become a lifelong malady. The morbidity associated with disc degeneration and its spectrum of associated spinal disorders is responsible for significant economic and social costs. The treatment of this disease entity in the United States is estimated to exceed $60 billion annually in health care costs (2). The indirect economic losses associated with lost wages and decreased productivity are staggering.
Age-related disc changes occur early and are progressive. Almost all individuals experience diminished nuclear water content and increased collagen content by the 4th decade. This desiccation and fibrosis of the disc blur the nuclear/annular boundary (3). These senescent changes allow repeated minor rotational trauma to produce circumferential tears between annular layers. These defects, usually in the posterior or posterolateral portions of the annulus, may enlarge and combine to form one or more radial tears through which nuclear material may herniate (4). Pain and dysfunction due to compression of neural structures by herniated disc fragments are widely recognized phenomena. It should be noted, however, that annular injuries may be responsible for axial pain with or without the presence of a frank disc herniation (5,6).
Progression of the degenerative process alters intradiscal pressures, causing a relative shift of axial load-bearing to the peripheral regions of the endplates and facets. This transfer of biomechanical loads appears to be associated with the development of both facet and ligament hypertrophy (7,8). There is a direct relation between disc degeneration and osteophyte formation (9). In particular, deterioration of the intervertebral disc leads to increased traction on the attachment of the outermost annular fibers, thereby predisposing to the growth of laterally situated osteophytes (10). Disc degeneration also results in a significant shift of the instantaneous axis of rotation of the functional spinal unit (11). The exact long-term consequences of such a perturbation of spinal biomechanics are unknown, but it has been postulated that this change promotes abnormal loading of adjacent segments and an alteration in spinal balance.
Nonoperative therapeutic options for individuals with neck and back pain include rest, heat, analgesics, physical therapy, and manipulation. These treatments fail in a significant number of patients. Current surgical management options for spinal disease include decompressive surgery, decompression with fusion, and arthrodesis alone.
Greater than 200,000 discectomies are performed annually in the United States (12). Although discectomy is exceptionally effective in promptly relieving significant radicular pain, the overall success rates for these procedures range from 48% to 89% (13,14,15). In general, the return of pain increases with the length of time from surgery. Ten years following lumbar discectomy, 50-60% of patients will experience significant back pain and 20-30% will suffer from recurrent sciatica (16). In general, the reasons for these less than optimal results are probably related to continued degenerative processes, recurrent disc rupture, instability, and spinal stenosis (17,18).
There are several specific reasons for failure of surgical discectomy. The actual disc herniation may not have been the primary pain generator in some patients. A number of relapses are due to disc space collapse. Although the disc height is often decreased in the preoperative patient with a herniated nucleus pulposus, it is an exceedingly common occurrence following surgical discectomy (14). Disc space narrowing is very important in terms of decreasing the size of the neural foramina and altering facet loading and function. Disc space narrowing increases intra-articular pressure, and abnormal loading patterns have been shown to produce biochemical changes in the intra-articular cartilage at both the level of the affected disc and the adjacent level (19,20). The entire process predisposes to the development of hypertrophic changes of the articular processes (21). Disc space narrowing also allows for rostral and anterior displacement of the superior facet. This displacement of the superior facet becomes significant when it impinges upon the exiting nerve root which is traversing an already compromised foramen (4). Destabilization of the functional spinal unit is another potential source of continued pain. A partial disc excision is associated with significant increases in flexion, rotation, lateral bending, and extension across the affected segment. As the amount of nuclear material which is removed increases, stiffness across the level decreases accordingly (22). Disc excision has also been demonstrated to lead to instability at the level above the injured segment in cadaver studies. This situation has been documented to occur clinically as well (23,24,25).
Arthrodesis, with or without decompression, is another means of surgically treating symptomatic spondylosis in all regions of the mobile spine. Fusion has the capability of eliminating segmental instability, maintaining normal disc space height, preserving sagittal balance, and halting further degeneration at the operated level. Discectomy with fusion has been the major surgical treatment for symptomatic cervical spondylosis for over 40 years (26,27,28). A report in 1986 estimated that over 70,000 lumbar fusions were performed annually in the United States (29). Given the explosive development of the instrumentation and interbody device technology, the current annual number of patients treated with a lumbar fusion is even higher. The major rationale for spinal arthrodesis is that pain can be relieved by eliminating motion across a destabilized or degenerated segment (30). Good to excellent results have been reported in 52-100% of anterior lumbar interbody fusions and 50-95% of posterior lumbar interbody fusions (31,32,33,34,35).
Spinal fusion is not, however, a benign procedure. In numerous patients, recurrent symptoms develop years after the original procedure. Fusion perturbs the biomechanics of adjacent levels. Hypertrophic facet arthropathy, spinal stenosis, disc degeneration, and osteophyte formation have all been reported to occur at levels adjacent to a fusion, and these pathological processes are responsible for pain in many patients (17,18,36,37,38,39,40,41). The long-term results of lumbar fusions have been reported by Lehman et al. These investigators described a series of patients who were treated with uninstrumented fusions and followed for 21-33 years. Roughly half the patients had lumbar pain requiring medication at last follow-up, and about 15% had been treated with further surgery over the study period (38). Finally, there are a number of other drawbacks to fusion as a treatment for spinal pain, including loss of spinal mobility, graft collapse resulting in alterations of sagittal balance, autograft harvest site pain, and the possibility of alteration of muscular synergy.
Sir John Charnley revolutionized modern orthopedics with his development of total hip replacement (42). Today, hip and knee arthroplasties are two of the most highly rated surgical procedures in terms of patient satisfaction. It is possible that the development of an artificial disc may impact the treatment of degenerative disc disease in a similar fashion. Although the challenges associated with developing a prosthetic disc are great, the potential to improve the lives of many individuals suffering from symptoms of spinal spondylosis is tremendous.
The idea of spinal disc replacement is not new. One of the first attempts to perform disc arthroplasty was undertaken by Nachemson 40 years ago (43). Fernstrom attempted to reconstruct intervertebral discs by implanting stainless steel balls in the disc space (44). 1966 he published a report on 191 implanted prostheses in 125 patients. Subsidence occurred in 88% of patients over a 4- to 7-year period of follow-up. These pioneering efforts were followed by more than a decade of research on the degenerative processes of the spine, spinal biomechanics, and biomaterials before serious efforts to produce a prosthetic disc resumed.
Challenges of Design and Implantation
There are a number of factors which must be considered in the design and implantation of an effective disc prosthesis. The device must maintain the proper intervertebral spacing, allow for motion, and provide stability. Natural discs also act as shock absorbers, and this may be an important quality to incorporate into prosthetic disc design, particularly when considered for multilevel lumbar reconstruction. The artificial disc must not shift significant axial load to the facets. Placement of the artificial disc must be done in such a way as to avoid the destruction of important spinal elements such as the facets and ligaments. The importance of these structures cannot be overemphasized. Facets not only contribute strength and stability to the spine, but they could be a source of pain. This may be especially important to determine prior to disc arthroplasty because it is currently believed that disc replacement will probably be ineffective as a treatment for facet pain. Excessive ligamentous laxity may adversely affect disc prosthesis outcome by predisposing to implant migration or spinal instability.
An artificial disc must exhibit tremendous endurance. The average age of a patient needing a lumbar disc replacement has been estimated to be 35 years. This means that to avoid the need for revision surgery, the prosthesis must last 50 years. It has been estimated that an individual will take 2 million strides per year and perform 125,000 significant bends; therefore, over the 50-year life expectancy of the artificial disc, there would be over 106 million cycles. This estimate discounts the subtle disc motion which may occur with the 6 million breaths taken per year (45). A number of factors in addition to endurance must be considered when choosing the materials with which to construct an intervertebral disc prosthesis. The materials must be biocompatible and display no corrosion. They must not incite any significant inflammatory response. The fatigue strength must be high and the wear debris minimal. Finally, it would be ideal if the implant were imaging "friendly."
All currently proposed intervertebral disc prostheses are contained within the disc space; therefore, allowance must be made for variations in patient size, level, and height. There may be a need for instrumentation to restore collapsed disc space height prior to placement of the prosthesis.
The intervertebral disc prosthesis ideally would replicate normal range of motion in all planes. At the same time it must constrain motion. A disc prosthesis must reproduce physiologic stiffness in all planes of motion plus axial compression. Furthermore, it must accurately transmit physiologic stress. For example, if the global stiffness of a device is physiologic but a significant nonphysiologic mismatch is present at the bone-implant interface, there may be bone resorption, abnormal bone deposition, endplate or implant failure.
The disc prosthesis must have immediate and long-term fixation to bone. Immediate fixation may be accomplished with screws, staples, or "teeth" which are integral to the implant. While these techniques may offer long-term stability, other options include porous or macrotexture surfaces which allow for bone ingrowth. Regardless of how fixation is achieved, there must also be the capability for revision.
Finally, the implant must be designed and constructed such that failure of any individual component will not result in a catastrophic event. Furthermore, neural, vascular, and spinal structures must be protected and spinal stability maintained in the event of an accident or unexpected loading.
Current Prosthetic Devices
Prosthetic discs have been constructed based on the utilization of one of the following primary properties: hydraulic, elastic, mechanical, and composite.
PDN Prosthetic Disc Nucleus
Hydrogel disc replacements primarily have hydraulic properties. Hydrogel prostheses are used to replace the nucleus while retaining the annulus fibrosis. One potential advantage is that such a prosthesis may have the capability of percutaneous placement. The PDN implant is a nucleus replacement which consists of a hydrogel core constrained in a woven polyethylene jacket (Raymedica, Inc., Bloomington, MN) (Figure 1) (46,47)
PDN Prosthetic Disc Nucleus
The pellet-shaped hydrogel core is compressed and dehydrated to minimize its size prior to placement. Upon implantation, the hydrogel immediately begins to absorb fluid and expand. The tightly woven ultrahigh molecular weight polyethylene (UHMWPE) allows fluid to pass through to the hydrogel. This flexible but inelastic jacket permits the hydrogel core to deform and reform in response to changes in compressive forces yet constrains horizontal and vertical expansion upon hydration. Although most hydration takes place in the first 24 hours after implant, it takes approximately 4-5 days for the hydrogel to reach maximum expansion. Placement of two PDN implants within the disc space provides the lift that is necessary to restore and maintain disc space height. This device has been extensively assessed with mechanical and in vitro testing, and the results have been good (46,47). Schönmayr et al. reported on 10 patients treated with the PDN with a minimum of 2 years follow-up (47). Significant improvement was seen in both the Prolo and Oswestry scores, and segmental motion was preserved. Overall, 8 patients were considered to have an excellent result. Migration of the implant was noted in 3 patients, but only 1 required reoperation. One patient, a professional golfer, responded favorably for 4 months until his pain returned. He had marked degeneration of his facets, and his pain was relieved by facet injections. He underwent a fusion procedure and since has done well. The devices have been primarily inserted via a posterior route. Bertagnoli recently reported placing the PDN via an anterolateral transpsoatic route (48). The PDN is undergoing clinical evaluation in Europe, South Africa, and the United States.
Two elastic type disc prostheses are the Acroflex prosthesis proposed by Steffee and the thermoplastic composite of Lee (49,50). The first Acroflex disc consisted of a hexene-based polyolefin rubber core vulcanized to two titanium endplates. The endplates had 7 mm posts for immediate fixation and were coated with sintered 250 micron titanium beads on each surface to provide an increased surface area for bony ingrowth and adhesion of the rubber. The disc was manufactured in several sizes and underwent extensive fatigue testing prior to implantation. Only 6 patients were implanted before the clinical trial was stopped due to a report that 2-mercaptobenzothiazole, a chemical used in the vulcanization process of the rubber core, was possibly carcinogenic in rats (51). The 6 patients were evaluated after a minimum of 3 years, at which time the results were graded as follows: 2 excellent, 1 good, 1 fair, and 2 poor (49). One of the protheses in a patient with a poor result developed a tear in the rubber at the junction of vulcanization. The second generation Acroflex-100 consists of an HP-100 silicone elastomer core bonded to two titanium endplates (DePuy Acromed, Raynham, MA) (Figure 2).
In 1993 the FDA approved 13 additional patients for implantation (52). The results of this study have not yet been published.
Lee et al. have published a report on the development of two different disc prostheses created in a manner to simulate the anisotropic properties of the normal intervertebral disc (50). I am not aware of any publications describing the implantation of these devices in humans.
Several articulating pivot or ball type disc prostheses have been developed for the lumbar spine. Hedman and Kostuik developed a set of cobalt-chromium-molybdenum alloy hinged plates with an interposed spring (53). These devices have been tested in sheep. At 3 and 6 months post-implantation there was no inflammatory reaction noted and none of the prostheses migrated. Two of the three 6-month implants had significant bony ingrowth. It is not clear whether motion was preserved across the operated segments (45). I am not aware of any publications describing the implantation of these devices in humans.
Dr. Thierry Marnay of France developed an articulating disc prosthesis with a polyethylene core (Aesculap AG & Co. KG., Tuttlingen, Germany). The metal endplates have two vertical wings and the surfaces which contact the endplates are plasma-sprayed with titanium. Good to excellent results were reported in the majority of patients receiving this implant (54).
Link SB Charité Disc
The most widely implanted disc to date is the Link SB Charité disc (Waldemar Link GmbH & Co, Hamburg, Germany). Currently more than 2000 of these lumbar intervertebral prostheses have been implanted worldwide (55). The Charité III consists of a biconvex ultra high molecular weight polyethylene (UHMWPE) spacer. There is a radiopaque ring around the spacer for x-ray localization. The spacers are available in different sizes. This core spacer interfaces with two separate endplates. The endplates are made of casted cobalt-chromium-molybdenum alloy, each with three ventral and dorsal teeth. The endplates are coated with titanium and hydroxyapatite to promote bone bonding (Figure 3).
CHARITÉ Artificial Disc (DePuy Spine, Inc.)
Photograph Courtesy of DePuy Spine, Inc.
The Food and Drug Administration (FDA) has approved the CHARITÉ Artificial Disc (DePuy Spine, Inc. of Raynham, MA) for use in treating pain associated with degenerative disc disease. The device was approved for use at one level in the lumbar spine (from L4-S1) for patients who have had no relief from low back pain after at least six months of non-surgical treatment.
Although there is great concern regarding wear debris in hip prostheses in which UHMWPE articulates with metal, this does not appear to occur in the Charité III (55). This prosthesis has been implanted in over a thousand European patients with relatively good results. In 1994 Griffith et al. reported the results in 93 patients with 1-year follow-up (56). Significant improvements in pain, walking distance, and mobility were noted. 6.5% of patients experienced a device failure, dislocation, or migration. There were 3 ring deformations, and 3 patients required reoperation. Lemaire et al. described the results of implantation of the SB Charité III disc in 105 patients with a mean of 51 months of follow-up (57). There was no displacement of any of the implants, but 3 settled. The failures were felt to be secondary to facet pain. David described a cohort of 85 patients reviewed after a minimum of at least 5 years post-implantation of the Charité prosthesis (58). 97% of the patients were available for follow-up. 68% had good or better results. 14 patients reported the result as poor. Eleven of these patients underwent secondary arthrodesis at the prosthesis level. Despite the concern of many other investigators, it is interesting to note that David treated 20 patients with spondylolisthesis or retrolisthesis with an outcome identical to that of the entire group. Clinical trials using the Charité III prosthesis are ongoing in Europe, the United States, Argentina, China, Korea, and Australia.
The Bristol Disc
There have been several reports on results from a cervical disc prosthesis which was originally developed in Bristol, England. This device was designed by Cummins (59). The original design has been modified. The second generation of the Cummins disc is a ball and socket type device constructed of stainless steel. It is secured to the vertebral bodies with screws. Cummins et al. described 20 patients who were followed for an average of 2.4 years. Patients with radiculopathy improved, and those with myelopathy either improved or were stabilized. Of this group, only 3 experienced continued axial pain. Two screws broke, and there were two partial screw back-outs. These did not require removal of the implant. One joint was removed because it was "loose." The failure was due to a manufacturing error. At the time of removal, the joint was firmly imbedded in the bone and was covered by a smooth scar anteriorly. Detailed examination revealed that the ball and socket fit was asymmetric. It is important to note that the surrounding tissues did not contain any significant wear debris. Joint motion was preserved in all but 2 patients (Figure 4).
"The Bristol Disc; a. lateral cervical radiograph in extension; b. lateral cervical radiograph in flexion"
Both of these patients had implants at the C6-7 level which were so large that the facets were completely separated. This size mismatch was felt to be the reason motion was not maintained. Subsidence did not occur. This disc prosthesis is currently being evaluated in additional clinical studies in Europe and Australia.
Bryan Cervical Disc Prosthesis
The Bryan Cervical Disc System (Spinal Dynamics Corporation, Seattle) is designed based on a proprietary, low friction, wear resistant, elastic nucleus. This nucleus is located between and articulates with anatomically shaped titanium plates (shells) that are fitted to the vertebral body endplates (Figure 5).
"Bryan Cervical Disc Prosthesis"
The shells are covered with a rough porous coating. A flexible membrane that surrounds the articulation forms a sealed space containing a lubricant to reduce friction and prevent migration of any wear debris that may be generated. It also serves to prevent the intrusion of connective tissue. The implant allows for normal range of motion in flexion/extension, lateral bending, axial rotation, and translation. The implant is manufactured in five sizes ranging from 14 mm to 18 mm in diameter. The initial clinical experience with the Bryan Total Cervical Disc Prosthesis has been promising (Jan Goffin, personal communication, March 2000). 52 devices were implanted in 51 patients by 8 surgeons in 6 centers in Belgium, France, Sweden, Germany, and Italy. There were no serious operative or postoperative complications. Twenty-six of the patients have been followed for 6 months, and complete clinical and radiographic data is available on 23 patients. 92% of the patients were classified as excellent or good outcomes at last follow-up. Flexion/extension motion was preserved in all patients, and there was no significant subsidence or migration of the devices.
Spinal disc replacement is not only possible but is an exciting area of clinical investigation which has the potential of revolutionizing the treatment of spinal degeneration. The development of a prosthetic disc poses tremendous challenges, but the results from initial efforts have been promising. The future for this field, and our patients, is bright.