Chapter 164 Lumbar Spinal Arthroplasty
Clinical Experiences of Motion Preservation
Lumbar spinal arthroplasty was first reported in clinical settings in 1994 by Griffith and colleagues.1 This early experience was acquired with the first lumbar artificial disc, the Charité I, in patients with degenerative disc disease. Since that time, a randomized, controlled trial comparing arthroplasty with the Charité Artificial Disc versus anterior lumbar interbody fusion with the BAK Cage and iliac crest bone was completed. Multiple other lumbar arthroplasty devices have been developed subsequent to the Charité and are undergoing or completing clinical trials.
Unlike other spinal medical devices, lumbar discs are required by the Food and Drug Administration (FDA) to complete randomized, controlled trials prior to market approval in the United States. As a result, lumbar arthroplasty devices have undergone more scrutiny and clinical evaluation than any other spinal medical devices. Specifically, a new device, the ProDisc-L, was granted FDA approval in 2006 and was described in a peer-review publication.2 In addition, the Maverick Total Disc Arthroplasty System (Medtronic Sofamor Danek), Kineflex Lumbar Disc (SpinalMotion), and FlexiCore Intervertebral Disc (SpineCore/Stryker) lumbar discs have both completed their randomized enrollments and are currently in continued access (nonrandomized) mode. All these ongoing and completed randomized clinical trials have generated a large body of evidence on the safety and efficacy of arthroplasty for lumbar spine in clinical applications and, in many cases, in level-1 publications.
Results
General Clinical Outcome Results
The Charité Artificial Disc manuscripts described clinical outcomes as early as 1 year3 and up to 13 years after surgery.4 The short-term and medium-term papers included herein typically disclosed early analyses from single sites involved in the Charité Artificial Disc IDE study comparing arthroplasty with Charité Artificial Disc versus anterior interbody fusion with BAK Cage and autograft.3,5,6 The complete randomized, controlled trial at 2-year follow-up included 205 arthroplasty and 99 fusion patients and was thoroughly described in two manuscripts, one focused on clinical outcomes7 the other on radiographic outcomes.8 Three additional medium- and long-term studies were also found: two papers with 10-year follow-up4,9 and one with an average of 6.6-year follow-up.10
Safety and efficacy of arthroplasty was demonstrated in all short- and medium-term studies. Specifically, at 2 years after surgery, Blumenthal and McAfee reported no device-related complications and a reoperation rate of 5.4% (vs 9.1% in the control arm). Efficacy was also demonstrated using validated disability (Oswestry Disability Index [ODI]) and pain (Visual Analogue Score [VAS]) clinical outcomes tools. At 2 years, the reduction in ODI reached 48.5% (vs. 42.4% in the control group) and the absolute reduction in VAS reached 40.6 points (vs. 34.1 points in the control group).7,8
Two of the three long-term studies confirmed these findings. Lemaire and colleagues reported 10-year follow-up results in 100 patients.4 This study included 54 patients operated on at one level, 45 patients operated on at two levels, and one patient operated on at three levels. Overall, the authors reported excellent or good clinical outcomes in 90% of cases. In a second long-term study, David and colleagues presented 10-year data on 106 patients.9 Only one-level surgeries were performed in this study. Excellent or good clinical outcome was obtained in 82.1% patients. Both papers thus concluded that arthroplasty was a viable option for disc degeneration.
A medium-term paper was published by Ross and colleagues, describing the long-term effect of arthroplasty in 160 patients (226 Charité Artificial Discs). This paper reported a cumulative survival rate at 156 months of 35% and a mean ODI score improvement of 14%. Implant removal was also described in 12 patients.10 These relatively poor findings were further discussed in two Letters to the Editor, which pointed out mathematical inconsistencies and overall clinical flaws in the manuscript and further highlighted the need for proper surgical technique and patient selection for optimal clinical outcome.11,12
The ProDisc-L manuscripts described clinical outcomes as early as 3 months13 and up to 8.7 years after surgery.14 As described earlier for the Charité Artificial Disc, The short-term papers typically disclosed early findings from one or two of the sites that participated in the randomized, controlled trial comparing ProDisc-L against a 360-degree fusion.13,15–21 The complete randomized, controlled trial at 2-year follow-up included 161 arthroplasty and 75 fusion patients.2 The long-term data included 64 patients operated on at one site, of whom 55 were available between 7 and 11 years after surgery for clinical and radiographic follow-up.14 All these studies concluded similarly that disc arthroplasty, at all evaluated time points, was safe and resulted in complication and/or reoperation rates comparable to those generally accepted for spinal surgery (complication rate of 9% at 8.7 years14; there were no major complications, but a reoperation rate of 3.7% at 2 years was reported2). In addition to safety, efficacy of spinal arthroplasty was also shown because all cases presented significant improvements in pain and disability. The final randomized, controlled trial data reported improvements in the arthroplasty group in VAS for pain by an average of 39 points and in disability, as determined by the ODI, by 28 points. It is worth noting, however, that the ODI tool used in this trial was not the validated and widely accepted ODI methodology Version 1.0 as defined by Fairbanks and colleagues.22 In a Letter to the Editor, Fairbanks denounced the use of the Oswestry Disability Index in the ProDisc-L study and thus cast doubt on the validity of the disability improvement outcomes observed herein.23
The two clinical data publications on the Maverick device were both based on the same data set of 64 patients, collected at one site.24,25 The clinical outcomes were described using the ODI Version 1.0 and VAS scores. The efficacy of arthroplasty was once again demonstrated using these tools, as ODI scores decreased by an average of 20.7 points and VAS scores by 4.4 points. As for the FlexiCore Intervertebral Disc, only one paper has been published (2008).26 This manuscript describes the clinical outcomes of 44 patients, of whom only 6 were available for 2-year follow-up. Although the clinical relevance of these data may therefore be questionable, the authors still concluded that the device may be safe and efficacious but that the data were not representative of the entire patient cohort.
Radiographic Analyses: Range of Motion, Heterotopic Ossification, and Sagittal Balance Analyses
Unlike other clinical and radiographic outcomes, accurate measurement of ROM was shown to be challenging and, to some extent, subjective, as patient positioning, imaging staff training, and other factors unrelated to the actual motion potential of the spine were shown to affect final readings.27 Using ProDisc-L cases, Lim and colleagues evaluated different methodologies and associated error margins for the measurement of ROM from radiographic images. Specifically, Lim and colleagues concluded that a ROM of at least 4.6 degrees must be observed to be 95% certain that a given device had any sagittal motion at all. Similarly, changes needed to be at least 9.6 degrees in ROM to confirm at 95% that change in motion really happened.28,29 These technical limitations might explain the inconsistent ROM data, particularly for ProDisc-L cases, found in the published literature. The flexion-extension ROM results from the 2-year randomized, controlled trial were determined at 7.7 degrees and 4.67 degrees and characterized as a normal ROM.2 However, at 8.7 years after surgery, Huang and colleagues reported a ROM less than that reported in asymptomatic normal person, with an average motion of 3.8 degrees.30 In a 2006 prospective study on 41 patients with 2-year follow-up, Leivseth and colleagues also reported that the device fails to restore normal segmental motion, and another retrospective study on 26 patients concluded that sagittal balance and ROM significantly improved after lumbar arthroplasty.31
Less controversy was observed when reviewing ROM data for the Charité Artificial Disc. A complete manuscript was dedicated to the radiographic data of the 2-year randomized, controlled trial of the Charité Artificial Disc.8 In this study, arthroplasty patients had a 13.6% increase in motion from preoperative to the 2-year postoperative time point. The ROM also correlated to device placement, as poor device placement resulted in a statistically significant reduction in motion. At 10-year follow-up, David reported an average 10.1-degree ROM, a value very similar to the 10.3 degrees reported by Lemaire and colleagues in their 10-year follow-up study.4,9
Le Huec and colleagues published the only data available on radiographic findings after arthroplasty with the Maverick Total Disc Arthroplasty System. In their study, Le Huec and colleagues broadened their analysis to include sagittal alignment and pelvic tilt.24,32 Using data related to 35 patients at an average of 14 months after surgery, authors showed maintenance of overall lordosis and unchanged sacral and pelvic tilts following arthroplasty.
More recently, a study by Tournier comparing all three—Charité Artificial Disc, ProDisc-L, and Maverick Total Disc Arthroplasty System—further refined the analyses from Le Huec on pelvic and sagittal tilt. In this study, authors found no difference in ROM among prostheses and observed maintenance of sagittal balance before and after surgery with all devices. However, modifications of the lumbar curvature were observed.33
The issue of heterotopic ossification in clinical cases of lumbar arthroplasty has been presented by McAfee and colleagues and, more recently, by Tortolani and colleagues.34,35 In his 2003 paper, McAfee introduced a novel method to characterize spinal heterotopic ossification. This method was applied by Tortolani and colleagues in reviewing the 276 arthroplasty patients from the Charité Artificial Disc trial (randomized and nonrandomized cases). From this analysis, 4.3% of cases of heterotopic ossification were noted. However, heterotopic ossification was not related to ROM, as the authors concluded that no difference in the ROM at 24 months after surgery was found between the patients who had and those who did not have heterotopic ossification.
Facets and Adjacent-Level Degeneration
Three long-term analyses evaluated adjacent-level degeneration, one with ProDisc-L at 8.7 years, and the other two with the Charité Artificial Disc at 10 years of follow-up. In the ProDisc-L study, 24% of patients developed adjacent level degeneration by the latest follow-up time point. A correlation was also found between a low ROM and the prevalence of adjacent level degeneration: All patients with adjacent-level degeneration had a ROM less than 5 degrees, whereas only 59% of patients without adjacent-level degeneration had a ROM less than 5 degrees.36 Lemaire and colleagues and David reported 2 (2%) and 3 (2.8%) cases of adjacent-level degeneration at the latest time point, respectively.4,9 Lemaire and colleagues and David also disclosed 11 cases (11%) and 5 cases (4.7%) with facet arthrosis at the latest time point, respectively.
The issue of facet degeneration was also recently discussed in a short-term study. From a 13-patient case series with 12 months of follow-up, Trouillier and colleagues alluded to possible maintenance of facet joint integrity following arthroplasty with Charité Artificial Disc based on the favorable results from their series.37 At the other end of the spectrum, Shim and colleagues, at the 3-year time point, observed 36.4% and 32.0% increase in index-level facet degeneration and 19.4% and 28.6% adjacent level disc degeneration with the Charité Artificial Disc and the ProDisc-L, respectively.
Revision and Revision Strategies
The first description of appropriate revision for a Charité Artificial Disc was presented by David.38 In this single-case example, a Charité Artificial Disc was replaced at 9.5 years after surgery with another Charité Artificial Disc. The author concluded that revision of the disc with another disc could be safely and adequately performed and was thus an alternative to a revision fusion procedure. David also noted that because of the inherent difficulty of an anterior approach, only experienced surgeons should undertake this operation.
Further revision and explantation of the disc were also described by McAfee and colleagues and Leary and colleagues.39,40 McAfee and colleagues confirmed David’s experience and concluded that arthroplasty with the Charité Artificial Disc did not preclude any further procedures at the index level during primary insertion, with nearly one third being revisable to a new motion-preserving prosthesis and just over two thirds being successfully converted to anterior lumbar interbody fusion (ALIF) and/or posterior pedicle screw arthrodesis, the original alternative procedure. Leary further implied that technical errors in position and sizing of the implant were largely to blame for further revision surgery. Finally, Punt and colleagues reviewed 75 revision cases from the Dutch experience (estimated by the authors at more than 1000 Dutch patients). In this series, patients received posterior fusion either with removal of the disc or without removal of the disc. No statistically significant difference was observed between these two groups.41 This paper included patients previously described by Van Ooij and colleagues.42
