The development of adjacent level pathology, also termed adjacent segment degeneration or symptomatic adjacent segment disease (ASD) requiring treatment, is a significant problem in spine surgery. These phenomena are often seen following a spinal fusion where either the proximal or distal adjacent segment can degenerate and lead to further symptoms. There is a lack of precision regarding the terminology used to describe the pathologies of ASD. The term adjacent segment pathology (ASP) is proposed as an umbrella term to refer to the breadth of clinical and/or radiographical changes at adjacent motion segments that developed subsequent to a previous spinal intervention. Under this umbrella, radiological adjacent segment pathology and clinical adjacent segment pathology are then used to categorize radiographical features (e.g., degenerative changes on magnetic resonance imaging, MRI) and clinical manifestations (e.g., new radiculopathy), respectively. The development of ASP next to a previous spinal fusion is thought to be related to the increased stresses or alterations at the adjacent level brought forth by the spinal fusion. However, degenerative changes can also develop at levels that have not undergone spinal fusion. There is therefore considerable controversy over whether the development of ASP is a result of the spinal fusion and loss of spinal mobility, or is a result of the natural history of progressive degeneration at the adjacent level and would occur even without the presence of a fusion. Confounding factors to these 2 vastly differing viewpoints do exist, which include the alterations in adjacent segment anatomy at the time of surgery; a genetic predisposition to degeneration; implantation of spinal instrumentation that may directly alter the adjacent segments; or the condition being solely a biomechanical phenomenon. In reality, there may be a combination of factors influencing the development of ASP.
If the development of ASP is because of the fusion and loss of spinal mobility, then motion-preservation technologies may prevent or slow its development. Based on the theory that motion preservation may protect against ASP, several motion-preserving devices have been developed that allow surgical decompression to address the spinal pathology, but obviate the need for a spinal fusion.
One of these motion-preservation technologies is spinal arthroplasty, or total disc replacement (TDR) devices. These disc arthroplasty devices are often composed of a metallic endplate that is seated onto the adjacent vertebral endplates, with an articulating plastic or ceramic surface that allows for some movement. Considerable argument exists regarding the optimal biomechanical design of these devices and the optimal articulating materials. However, all TDR devices are designed to restore some movement and normal biomechanics to the spinal segment.
There are other classes of devices that have been developed to retain spinal mobility and obviate the need for a fusion. A few of these devices are FDA-approved for use in the lumbar and cervical spine, with multiple other devices currently being studied in clinical trials. In addition, there are many devices that are approved and being used outside the United States and have longer-term follow-up studies.
Yajun et al 1 published a meta-analysis of 5 randomized controlled trials (RCTs)2 – 6 involving 837 patients comparing TDR with fusion in patients with lumbar degenerative disc disease. They evaluated the following outcomes and pooled the data where appropriate: Oswestry Disability Index, pain visual analogue scale, patient satisfaction, complications, return to work, and reoperation rates. The authors concluded that TDR does not show significant superiority for the treatment of lumbar degenerative disc disease (DDD) compared with fusion; however, the evaluation of symptomatic adjacent segment pathology was not included in their analysis.
Harrop et al 7 published a review article evaluating the risk of ASP after fusion and TDR. Twenty-seven case series were included (20 fusions and 7 arthroplasties). The authors reported a mean radiographic ASP risk of 34% and 9% in the fusion and TDR series, respectively (P < 0.0001). The mean clinical ASP risk was 14% and 1%, respectively (P < 0.0001).
This study suggests that the rate of both radiographical and clinical ASP may be significantly greater after fusion than TDR; however, definitive conclusions cannot be made because these findings were established from individual case series evaluating separate patient populations. The literature is lacking a systematic review establishing the relative effect of fusion versus TDR on the development of ASP in the same population. Furthermore, it is unclear if other motion-sparing devices are associated with a lower risk of ASP compared with fusion. Comparative studies are required to establish the relative risk comparing motion-sparing devices with fusion.
The primary goal of this review is to perform an evidence-based synthesis of the literature comparing motion preservation devices with fusion to determine if the use of these devices decreases the development of ASP compared with fusion. The review will also attempt to compare different classes of motion preservation devices. To accomplish these goals, we sought to answer the following key questions regarding various lumbar conditions:
- Is there evidence that TDR is associated with a lower risk of radiographical or clinical symptomatic ASP compared with fusion?
- Is there evidence that other motion preservation devices are associated with a lower risk of radiographical or clinical ASP compared with fusion?
- Is one type of motion preservation device associated with a lower risk of radiographical or clinical ASP compared with other devices?
MATERIALS AND METHODS
Electronic Literature Search
A systematic search was conducted for articles published between January 1990 and February 2012 using PubMed and the Cochrane Library. For all key questions, we identified all cohort studies and RCTs, making the comparison of interest independent of the outcomes measured. We did not require that ASP be reported in the abstract, recognizing that this rare outcome may not have been the primary focus of the study, but reported as one of the possible complications in the results section of the article. Therefore, we searched each full-text article for a report of any type of structural or degenerative condition specifically occurring at an adjacent segment. We limited our results to humans and to articles published in the English language. We included articles with adult lumbar patients who had degenerative disc disease, disc herniation, radiculopathy, kyphosis, scoliosis, and spondylolisthesis, who were treated with TDR, other motion-sparing procedure, or fusion. Articles were excluded if patients were less than 18 years of age, treated for ASP, tumor, infection, had more than 20% trauma, neuromuscular scoliosis, or ankylosing spondylitis. Other exclusions included case series, case reports, studies with fewer than 10 subjects, and biomechanical studies and cadaver studies (Table 1). Full texts of articles meeting the inclusion criteria were reviewed by 2 independent investigators (D.C.N., J.T.H.) to obtain the final collection of included studies.
From the included articles, the following data were extracted: study design and study purpose, patient demographics, inclusion and exclusion criteria, follow-up duration and the rate of follow-up for each treatment group, treatment interventions, and definition of ASP. We reported the risks (% cumulative incidence) of 3 potential methods of reporting ASP by treatment group: radiographical, clinical requiring nonsurgical treatment, and clinical requiring surgical treatment, as these may have been reported separately.
Study Quality and Overall Strength of Body of Literature
Level of evidence ratings were assigned to each article independently by 2 reviewers (J.T.H, D.C.N.) using criteria set by The Journal of Bone & Joint Surgery, American Volume 8 for therapeutic studies, and modified to delineate criteria associated with methodological quality described elsewhere.9 (See the Supplemental Digital Content 1, available at: http://links.lww.com/BRS/A695, for individual study ratings.)
The overall body of evidence with respect to each key question was determined based on precepts outlined by the Grades of Recommendation Assessment, Development and Evaluation (GRADE) working group10 and recommendations made by the Agency for Healthcare Research and Quality (AHRQ).11 Risk of bias was evaluated during the individual study evaluation described above in the section “Study Quality.” This system, which derives a strength-of-evidence grade of “high,” “moderate,” “low,” or “insufficient” for each outcome or key question, is described in further detail in the methodology article in this Focus Issue.9 A detailed description of how we arrived at the strength of evidence for each key question can be found in the Supplemental Digital Content.
Where the data were available, we report the risk (cumulative incidence) of radiographical, clinical ASP requiring nonsurgical treatment, and clinical ASP requiring surgical treatment using the number of events (adjacent levels with disease) in the numerator and the number of adjacent levels at risk in the denominator. For studies only providing the patient as the denominator, the proportion of patients with one or more levels of ASP was considered a count of 1 in the numerator and the number of patients at risk was counted in the denominator.
These were recorded and noted in the text and appropriate tables. Data were pooled if there did not seem to be a significant amount of clinical heterogeneity with respect to the definition of ASP, indication for surgery, treatments, and patient populations. To evaluate the effect of a treatment type (e.g., fusion vs total disc arthroplasty) on the risk of ASP, we calculated the relative risk (RR) and 95% confidence interval (CI). We also calculated the attributable risk (AR) and its 95% CI. Attributable risk, or risk difference, is the difference between the risks in exposed and nonexposed groups—or, in this case, one treatment group versus the other. The AR is used to quantify the increased risk attributed to a specific exposure or treatment group (e.g., increased risk of receiving fusion versus TDR). Forest plots for RRs with their 95% CIs were constructed (for pooled comparisons only) to provide a graphical display illustrating the relative strength of the safety effects. A fixed-effect model was used if the I 2 value was less than 30%, suggesting minimal statistical heterogeneity. We also calculated the number needed to harm (NNH) when the results nearly achieved or achieved statistical significance (P ≤ 0.05) for symptomatic ASP. The NNH represents the number of patients one would need to treat with the intervention (e.g., fusion) to cause one case of ASP compared with the control (e.g., total disc arthroplasty). All calculations were performed using Stata 9.0.12 Forest plots were created using Review Manager (Rev-Man).
Clinical Recommendations and Consensus Statements
Clinical recommendations or consensus statements were made through a modified Delphi approach by applying the GRADE/AHRQ criteria that imparts a deliberate separation between the strength of the evidence (i.e., high, moderate, low, or insufficient) from the strength of the recommendation. When appropriate, recommendations or statements “for” or “against” were given “strong” or “weak” designations based on the quality of the evidence, the balance of benefits/harms, and values and patient preferences. In some instances, costs may have been considered. A thorough description of this process can be found in the methodology article in this Focus Issue.9
For study question 1 comparing TDR with fusion, we identified 30 total citations from our literature search. Of these, 19 were excluded by the title/abstract, and 11 full-text articles were evaluated to determine if they met the inclusion criteria. From these 11 studies, 9 were excluded because they did not report ASP as an outcome or they were evaluating risk factors for ASP rather than the treatments of interest. Details of the excluded articles can be found in the Supplemental Digital Content. The remaining 2 studies, RCTs comparing fusion with TDR, met our inclusion criteria and are summarized in this report2 , 13 (Figure 1A). Patient demographic information and 2-year follow-up data for the Guyer study can be found in 2 previously published studies.3 , 14
For study question 2 comparing other motion sparing devices with fusion, we identified 45 total citations from our literature search. Of these, 33 were excluded by the title/abstract, and 12 full-text articles were evaluated to determine whether they met the inclusion criteria. From these 12 studies, 7 were excluded because they did not report ASP. Details of the excluded articles can be found in the Supplemental Digital Content. The remaining 5 studies met our inclusion criteria and are summarized in this report15 – 19 (Figure 1A). We identified 1 prospective cohort study comparing lumbar decompression and dynamic stabilization with lumbar decompression and rigid stabilization,17 and 1 RCT comparing dynamic stabilization to semirigid stabilization and rigid stabilization.18 We identified 3 retrospective cohort studies.15 , 16 , 19 Kanayama et al 15 compared Graf ligamentoplasty with posterior lateral fusion in the 2001 publication, and compared Graf ligamentoplasty with posterior lateral fusion and posterior lumbar interbody fusion in the 2009 publication.
For study question 3 comparing different motion-sparing devices, we identified 54 total citations from our literature search. Of these, 41 were excluded by the title/abstract, and 13 full-text articles were evaluated to determine whether they met the inclusion criteria. From these 13 studies, 12 were excluded because they did not report ASP or were review studies. Details of the excluded articles can be found in the Supplemental Digital Content. The 1 remaining study, a prospective cohort study comparing dynamic stabilization and nucleotomy with nucleotomy alone20 met our inclusion criteria and is summarized in this report (Figure 1C).
Risk of ASP Comparing TDR With Fusion
Four RCTs comparing TDR with fusion and measuring some form of ASP were identified. Two of these studies3 , 14 were the precursors to the study published by Guyer et al 13 that measured ASP at 60 months.13 The other study, by Berg et al,2 measured ASP at 48 months.
The primary diagnosis in both the Berg and Guyer study populations was DDD. In the RCT by Berg, all patients were required to be 20 to 55 years of age, have confirmed DDD on MRI, have low back pain with or without leg pain for more than 1 year, and have had 3 months of failed conservative management.2 In the RCT by Guyer, all patients were required to be 18 to 60 years of age, have confirmed DDD by discography, have low back pain with or without leg pain, and have had at least 6 months of failed nonoperative care.13 Both studies had similar exclusion criteria, which included symptomatic multilevel degeneration, previous fusion, and a spondylolisthesis more than 3 mm. The mean age was 39.4 and 40 years in the Berg and Guyer studies, respectively, and males comprised 41% and 53% of the populations, respectively.
The Berg study included CHARITÉ, ProDisc, and Maverick implants compared with either posterior lateral fusion (PLF) or posterior lumbar interbody fusion (PLIF). The Guyer study included CHARITÉ discs compared with anterior lumbar interbody fusion with BAK cages. Both studies were designed to measure effectiveness and safety, with ASP being one safety measure. In the Berg trial, operation at an adjacent level was used as the diagnosis for clinical ASP. In the Guyer trial, there were 2 measurements of ASP: clinical ASP defined as requiring pain management treatment, and, as in the Berg trial, clinical ASP requiring surgery. Because the populations of these 2 studies were similar, their follow-up times were similar, and both studies reported the surgical treatment of clinical ASP, we felt justified in pooling these studies despite different fusion techniques and despite 1 trial including multiple disc replacement devices.
A total of 152 and 304 subjects were enrolled in the Berg and Guyer trials, respectively. In the Berg trial, the authors reported a follow-up rate of 100%. The risk of clinical ASP (treated with surgery) was 1.3% (1/80) and 8.3% (6/72) in the TDR and fusion group, respectively (Table 2). The relative risk of clinical ASP (treated surgically) comparing the fusion with the TDR group was 6.7 (95% CI: 0.82, 54.1; P = 0.05). The Berg trial was well executed; however, the diagnosis of clinical ASP was not made by an independent observer. In the Guyer trial, the authors reported a follow-up rate of 57% with a nearly equivalent loss of follow-up in each group. The risk of clinical ASP requiring pain management was 4.4% (4/90) and 14.0% (6/43) in the TDR and fusion group, respectively (Table 2). The relative risk of clinical ASP comparing the fusion with the TDR group was 3.1 (95% CI: 0.93, 10.5; P = 0.05). The risk of clinical ASP (treated surgically) was 1.1% (1/90) and 4.7% (2/43) in the TDR and fusion group, respectively (Table 2). The relative risk of clinical ASP (treated surgically) comparing the fusion with the TDR group was 4.2 (95% CI: 0.39, 44.9; P = 0.20). The Guyer trial also lacked independent assessment of clinical ASP, but, more importantly, it had a significantly low follow-up rate. The authors went to great details to rule out potential bias through a sensitivity analysis and comparison of those lost to follow-up with those who remained in the trial; however, with such a high loss to follow-up, the findings are at risk of bias. This bias is likely nondifferential because the loss to follow-up is nearly identical in each group; therefore, observed differences are more likely to be a conservative estimate.
After combining the data from the 2 trials, the risk of clinical ASP (treated surgically) was 1.2% (2/170) and 7.0% (8/115), respectively (Table 2). The pooled relative risk of clinical ASP (treated surgically) comparing the fusion with the TDR group was 5.9 (95% CI: 1.3, 27.3; P = 0.02) (Figure 2). This difference was statistically significant. The increased risk attributed to fusion was 5.8% (AR = 5.8; 95% CI: 0.01, 10.7). The number needed to harm was 17.3 (95% CI: 9.3, 116.8).
Risk of ASP Comparing Other Motion-Sparing Devices With Fusion
Two retrospective cohort studies, both published by Kanayama et al,15 , 16 compared Graf ligamentoplasty with fusion. Both study populations had degenerative conditions including spinal stenosis, herniated disc, and spondylolisthesis. Both studies measured clinical ASP requiring treatment confirmed by plain radiography. Radiographical evidence of ASP included disc-space narrowing (>2 mm loss of posterior disc height), spur formation, spondylolisthesis (anterior or posterior slip of the vertebra by >2 mm), and vacuum phenomenon, comparing preoperative and final follow-up radiographs. Sagittal T2-weighted MRI was also used for the assessment of radiographical ASP.
The first study by Kanayama et al 15 compared Graf ligamentoplasty (n = 18) with posterior lateral fusion (PLF; n = 27). Patients averaged 57 years of age and were 49% male; follow-up was 60 months and the follow-up rate was 64.3%. The risk of radiographical ASP was 37.0% (10/27) and 5.6% (1/18) in the PLF and ligamentoplasty groups, respectively (Table 3). The relative risk of radiographical ASP comparing the PLF to the ligamentoplasty group was 6.7 (95% CI: .93, 47.7; P = 0.03). The risk of clinical ASP requiring treatment was 18.5% (5/27) and 5.6% (1/18) in the PLF and ligamentoplasty groups, respectively (Table 3). The relative risk of clinical ASP requiring treatment comparing the PLF to the ligamentoplasty group was 3.3 (95% CI: 0.42, 26.2; P = 0.21). This study had several weaknesses, including the retrospective design, a low follow-up rate (a relatively equal loss to follow-up in each arm), a small sample size, and the inability to control for potential baseline imbalances between groups.
The second study by Kanayama et al 16 compared Graf ligamentoplasty (n = 65) with PLF (n = 75) or posterior lumbar interbody fusion (PLIF; n = 78). The authors reported data separately for the PLF and PLIF groups. We were able to combine the data into one fusion group because the rates of ASP in the PLF and PLIF were similar at 13.3% and 14.1%, respectively. Patients averaged 63 years of age and were 42% male; follow-up was 48 months and the follow-up rate was 97.3%. The risk of clinical ASP requiring treatment was 7.2% (11/153) and 1.5% (1/65) in the PLF/PLIF and ligamentoplasty groups, respectively (Table 3). The relative risk of clinical ASP requiring treatment comparing the PLF/PLIF with the ligamentoplasty group was 4.7 (95% CI: 0.62, 35.5; P = 0.09). We felt justified in pooling these 2 studies because they were drawn from similar patient populations, had similar treatments, and used similar measurements of ASP. Because our study is a safety review of a rare event rather than an effectiveness review, we were comfortable pooling the data of these retrospective cohort studies. After combining the data from the 2 studies, the risk of clinical ASP (requiring surgery) was 8.9% (16/180) and 2.4% (2/83), respectively (Table 3). The pooled relative risk of clinical ASP (requiring surgery) comparing the fusion group with the Graf ligamentoplasty group was 3.7 (95% CI: 0.87, 15.7; P = 0.05) (Figure 3). The increased risk attributed to fusion was 6.5% (AR = 6.5; 95% CI: 0.01, 11.8). The number needed to harm was 15.4 (95% CI: 8.5, 85.3).
One RCT18 and 1 prospective cohort study17 compared decompression and dynamic stabilization with decompression using some form of semirigid stabilization with fusion. In the RCT by Korovessis et al,18 patients were included if they had symptomatic degenerative spinal stenosis for at least 1 year. Patients were excluded if they had prior spine surgery, active infection, or congenital deformity. Patients averaged 62 years of age and were 76% male; mean follow-up was 47 months and the follow-up rate was 100%. This study had 3 arms: rigid (n = 15), semirigid (n = 15), and dynamic instrumentation (n = 15). There were no reported observations of ASP in any group (Table 3). This RCT did not report concealed treatment allocation, and the sample size was likely too small to detect an adequate number of ASP events to establish potential differences between groups.
In the prospective cohort study by Kaner et al,17 patients were included if they had single-level grade 1 or 2 degenerative spondylolisthesis causing central and/or lateral recess syndrome with previous failed medical treatment. Patients were excluded if they had isthmic spondylolisthesis, more than 1 level involved, or a history of prior fusion, infection, or congenital deformity. Patients in the group treated with decompression and posterior dynamic transpedicular stabilization (n = 26) averaged 64 years and were 23% male; the mean follow-up was 38 months and the follow-up rate was 100%. Patients in the group treated with decompression and rigid transpedicular stabilization with fusion (n = 20) averaged 62 years and were 35% male; the mean follow-up was 44 months and the follow-up rate was 100%. There was only 1 reported case of clinical ASP in the posterior rigid stabilization group and there were no reported cases in the dynamic stabilization group (Table 3). This study lacked an independent assessment of ASP, and the sample size was likely too small to detect differences.
One retrospective cohort study by Satoh et al 19 compared discectomy (n = 177) to PLIF (n = 174) in patients with massive herniation defined by a complete block on myelogram and segmental instability. Patients in the discectomy and PLIF groups averaged 39 and 42 years and were 67% and 78% male, respectively. Adjacent segment pathology was defined as adjacent level instability determined by anterior slip more than 3 mm and/or local kyphosis more than 5° at maximal flexion on lateral radiographs. They calculated separate rates for radiographical ASP and clinical ASP. The risk of radiographical ASP was 8.6% (15/174) and 1.7% (3/177) in the PLIF and discectomy groups, respectively (Table 3). The relative risk of radiographical ASP comparing the PLIF with discectomy was 5.1 (95% CI: 1.5, 12.3; P = 0.003). The risk of clinical ASP was 6.9% (12/174) and 3.4% (1/65) in the PLIF and discectomy groups, respectively. The relative risk of clinical ASP comparing PLIF with discectomy was 2.0 (95% CI: 0.78, 5.3; P = 0.14). The weaknesses of this study were its retrospective nature as well as some baseline treatment group differences of important variables that were not controlled for in the analysis that may have had an influence on the relative differences between treatment groups.
Risk of ASP Comparing Motion-Sparing Devices With Other Motion-Sparing Devices
We identified only 1 study comparing 2 different motion-sparing procedures that measured ASP as an outcome. Putzier et al 20 performed a retrospective cohort study comparing dynamic stabilization and nucleotomy (n = 35) with nucleotomy alone (n = 49) in patients with therapy-resistant lumbar radicular complaints due to disc prolapse via MRI, with stage 1 disc degeneration in a maximum of 2 segments. Patients in the dynamic stabilization with nucleotomy and nucleotomy alone groups averaged 39 and 36 years and were 63% and 59% male, respectively. Adjacent segment pathology was defined as the presence of deviation or osseous remodeling processes, stenosis, or spondyloarthroses assessed by MRI. Patients were followed for an average of 34 months and the follow-up rate was 97.6%. There were no reported observations of ASP in either group (Table 4). Despite being reported as a prospective study, one prospective arm was compared with a retrospective arm. Furthermore, this study did not include an independent assessment of ASP, and the sample size was likely too small to detect ASP events.
The final overall strength of the body of literature expresses our confidence in the estimate of effect and the impact that further research may have on the results. The overall strength of the evidence evaluating whether TDR is associated with a lower risk of radiographical or clinical ASP compared with fusion is “moderate,” meaning we have moderate confidence that the evidence reflects the true effect, and further research may change our confidence in the estimate of effect and may change the estimate. The overall strength of evidence evaluating whether other motion preservation devices are associated with a lower risk of radiographical or clinical ASP compared with fusion is “low” to “insufficient,” meaning we have low confidence that the evidence reflects the true effect, and further research is likely to change the confidence in the estimate of effect and likely to change the estimate, or that the evidence is either unavailable or does not permit a conclusion. The overall strength of the evidence evaluating whether 1 type of motion preservation device is associated with a lower risk of ASP compared with other types is “insufficient,” meaning the evidence is either unavailable or does not permit a conclusion (see Supplemental Digital 1, available at: http://links.lww.com/BRS/A695, Table 3).
This is the first article to produce an evidence synthesis of studies comparing lumbar motion preservation devices with lumbar fusion, with the development of ASP as the outcome. The primary justification for the development of motion-sparing technology is to prevent or minimize the development of adjacent segment pathology. Fusion is a commonly performed procedure. The development of adjacent segment degeneration and symptomatic disease is considered either a reflection of the natural history of progressive degeneration, or a result of altered biomechanics from the fusion procedure that induces or promotes disease at the remaining motion levels. There exists considerable controversy regarding the true etiology of ASP, and it may in fact result from a combination of natural history, genetics, biomechanical alterations from the fusion, and other factors not yet identified.
We performed a rigorous systematic search reviewing dozens of full-text articles to identify comparison studies that measured radiographical and symptomatic ASP. The literature is sparse, as with most novel technology, and time is needed to provide consistent, high-quality follow-up data that can provide more definitive evidence. However, it is important to periodically assess the available literature to glean as much information in these new areas and provide for the best evidence for treatment decisions for our patients with these specific spinal issues. Often in these situations, the quality of the evidence will increase with time, with early case reports leading to more valuable RCTs contributing to larger numbers of subjects and longer follow-up periods.
When pooling 2 similar RCTs with 4 to 5 years of follow-up comparing TDR with fusion, the risk of clinical ASP requiring surgery is very low in both groups (1.2% and 7%, respectively). Only 10 events were observed among 285 patients when pooling these studies. The majority of these events occurred in the fusion group, who were nearly 6 times more likely to receive surgery for ASP than those who underwent TDR. The increased risk of clinical ASP requiring surgery associated with fusion is 5.8%. From a clinical perspective calculating the number needed to harm, this increased risk suggests that for every 17 operations, one might expect a new surgically treated ASP event following fusion in those otherwise not harmed by TDR. Despite the risk being greater after fusion, it is important to remember that the risk is still low. In addition, the overall evidence for this risk was considered moderate rather than high, because the estimates from the meta-analysis were not precise because of the limited number of events observed in both groups.
Previous publications either did not evaluate ASP in comparison studies, or were case series and thus could not establish relative safety of one treatment over another. Case series data suggest a nearly 14 times increased risk of ASP comparing fusion with motion-sparing. Our meta-analysis of 2 RCTs suggest this increased risk of ASP is likely less than half this amount (relative risk = 5.9). This highlights the weaknesses of using case series to establish relative treatment differences and highlights the need for more comparative studies in which outcomes are collected longitudinally in the same populations.
The evidence that lumbar TDR is associated with a lower risk of clinical ASP requiring treatment comes essentially from 2 published RCTs with long-term follow-up. We rate the evidence as moderate because we pooled the data from these 2 controlled studies to obtain the relative risk, and the confidence interval around the statistically significant relative risk was wide. This is in part because ASP is a rare outcome and there were a limited number of ASP events. Certainly there are criticisms, such as these trials were likely industry funded, patients enrolled in the study may have wanted the study device rather than the control, and there may have been unintended institutional bias in the evaluations of the technology (neither study included a blinded or independent assessor). More studies are needed to support this important question. In addition, with the many lumbar TDR devices that exist and are being implanted worldwide, it is unclear whether we can generalize that these devices are all essentially equal and are a generic class of devices that can be grouped together with similar results. Based on these criticisms, the strength of our consensus statement is rated as “weak.”
This last point, that these devices may indeed not be a generic class of devices that can be grouped together with similar results. Certainly, total disc replacement should be considered separately from posterior dynamic stabilization. Further, there are considerable differences between other motion sparing devices; therefore, one must be cautious especially with our second question of whether there is evidence that other motion preservation devices are associated with a lower risk of radiographical or clinical ASP compared with fusion. We discuss them as a class of devices, but there are such obvious differences between the varying technologies that further studies are needed to support the individual devices.
We found insufficient evidence that the risk of ASP after discectomy was different from fusion, or that dynamic or rigid stabilization would reduce the risk of ASP compared with fusion. The Graf ligamentoplasty data did provide low evidence of a lesser risk of ASP than fusion. This product is certainly not in widespread use, and there are serious doubts that these studies can be generalized to all pedicle-based motion-sparing devices because of the wide variety of devices that currently exist worldwide. Further studies are necessary to understand the efficacy of these types of devices.
The third question we attempted to answer, that one type of motion preservation device might be associated with a lower risk of either radiographical or clinical ASP, had insufficient evidence to permit a conclusion. With the paucity of literature, there are even fewer comparison studies. It is difficult then to draw any conclusions or to compare the devices. Further study is necessary to compare different devices or even to generalize the results of single-device studies to the general class of similar devices.
Although we are making our recommendations based on the current available evidence, it is clear that more studies, longer follow-up, and additional publications are needed in this area. It is important to consider that the selection of one treatment instead of another cannot be made based on one outcome alone—in particular, a rare outcome such as ASP that occurs years after the operation. The information in this review should be used to inform patients of the potential increased risk during a discussion of the potential benefits and harms of all treatments.
- Evidence demonstrates that the risk of clinical ASP requiring surgery is likely greater after fusion but the risk is still quite rare. The increased risk compared with TDR could be as small as less than 1% or as great as 10%.Strength of Statement: Weak
- There is insufficient evidence to make a definitive statement regarding fusion versus other motion-sparing devices with respect to the risk of ASP.
- The risk of adjacent segment pathology is rare after both TDR and lumbar fusion.
- The increased risk of developing adjacent segment pathology after fusion versus after TDR could be as small as less than 1% or as great as 10%.
- There is limited evidence that fusion may increase the risk of developing adjacent segment pathology compared with motion-sparing procedures, but the risks are relatively low for both procedures.
Supplemental digital contents are available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.spinejournal.com).