Vaccaro, Alexander R. MD, PhD*; Fisher, Charles G. MD, MHSc†; Whang, Peter G. MD‡; Patel, Alpesh A. MD§; Thomas, Ken C. MD, MHSc¶; Mulpuri, Kishore MBBS, MHSc‖; Chi, John MD, MPH**; Prasad, Srinivas K. MS, MD††
Transforaminal lumbar interbody fusion (TLIF) versus posterolateral instrumented fusion (PLF) in degenerative lumbar disorders: a randomized clinical trial with 2-year follow-up. Eur Spine J 2013;22:2022–9.
Lumbar fusion is a commonly used technique to address low back pain associated with degenerative lumbar disorders. A variety of surgical techniques and implants are currently available, yet there is a paucity of comparative data. The TLIF, a commonly performed adjunct to traditional PLF, has been demonstrated to lead to greater blood loss, longer surgical time, and greater cost. Nonetheless, it offers theoretical benefits of a greater fusion success rate. The authors address this issue by performing a prospective, randomized study comparing PLF with TLIF.
The authors define a study population of 100 patients at a single center randomized 1:1 to either TLIF or PLF procedures during a 5-year period (2003–2008). The authors used a number of reported outcome measures: the Dallas Pain Questionnaire, the pain index from the low back pain rating scale, Oswestry Disability Index, and Short-Form 36 version 2. All assessments were completed at 1 and 2 years postoperatively. The authors performed a power-analysis based on the Dallas Pain Questionnaire with a reported standard deviation of 25 and a 15-point difference considered clinically relevant.
The authors do not state if these were consecutive patients enrolled in the study; a complete description would provide more information about the study population. Though the randomization scheme seems effective, the groups remain heterogeneous as they include both primary and revision procedures, single- and multilevel procedures, and a variety of degenerative disorders. These factors limit the validity of the findings by introducing bias.
Furthermore, although a power analysis was presented by the authors, the predetermined standard deviation of 25 and clinical difference of 15 is quite large considering mean scores were 40. The authors should have specified 95% confidence intervals when presenting their data. In addition, the authors could have performed a power analysis based on the Oswestry Disability Index, a more commonly reported outcome measure with more reported studies upon which standard deviation and clinical differences could be based. Furthermore, a post hoc power analysis should have been performed. Given the findings of this study (standard deviation of 11) as well as the smaller differences observed between groups, a post hoc analysis would have found that this study required a much large study population to determine statistical significance.
The authors have addressed a very relevant clinical question. The addition of the TLIF procedure to our surgical options for lumbar degenerative conditions occurred in the 1990s and, yet, the strength of clinical evidence demonstrating an improvement in either fusion rate or patient outcomes is glaringly low. The increase in complex fusion has been suggested to increase both complications and costs in the surgical treatment of degenerative stenosis.1 Nonetheless, TLIF procedures have been commonly used and with increasing frequency.1
The authors address this question and suggest that there is no added benefit either to patient-reported outcomes or to radiographical fusion rate. The authors reported longer operative time and greater blood loss in the TLIF group but no differences in outcomes. Reoperation rates were, also, no different between PLF and TLIF.
The findings of this prospective randomized study would suggest that TLIF does not have an added benefit over PLF procedures for adult lumbar degenerative spine surgery. The limitations of the study, however, need to be accounted for. The study presents a heterogeneous study population both in terms of diagnosis and surgically treated levels. Grouping different diagnoses and surgical procedures creates too much uncertainty in the study populations' surgical indications to effectively apply the findings to a specific patient. The study seems to be underpowered and thus equivalence can be challenged. Finally, although the authors report equivalent fusion rates (not reported), the criteria for fusion and for pseudarthrosis are not reported. The authors report reoperation and, presumably, used this to report equivalent fusion rates. As such, the fusion rates between TLIF and PLF are not appropriately compared in the 2 arms and the authors statements are not supported by their published study.
RECOMMENDATION REGARDING IMPACT ON CLINICAL PRACTICE
This study suggests that the use of TLIF procedure across a broad array of surgical indications and posterior lumbar procedures does not provide any additional benefit to fusion rate or to surgical outcomes. Given the significant limitations of the study, there is insufficient information to apply these findings into clinical practice. However, the equivalent success rates in this study justify a “weak recommendation” that treating surgeons should individualize the care of the patient, taking into account the added risks of surgical time, blood loss, and cost, when considering the utility of the TLIF procedure.
Cancer risk after use of recombinant bone morphogenetic protein-2 for spinal arthrodesis. JBJS 2013;95:1537–45.
The use of recombinant human bone morphogenetic protein-2 (rhBMP-2) in lumbar spinal arthrodesis has increased significantly since its introduction in the US market.2 The FDA-approved indication for the use of rhBMP-2 (Infuse, Medtronic, Memphis, TN) represents only a small number of procedures, with the large majority of utilization in an off-label manner. With its utilization, concerns regarding the safety and complications associated with BMP-2 have been reported including radiculitis, heterotopic bone formation, and osteolysis.3,4
One of the most emotionally charged concerns is that of a risk of cancer associated with the use of rhBMP-2. Although there is the theoretical potential for carcinogenesis with rhBMP-2 use, the clinical evidence is not definitive that such a relationship exists. The authors address the potential for cancer risk after the utilization of one preparation of rhBMP-2, in particular (Amplify).5
This is a retrospective review of prospectively collected data that were submitted to the FDA as part of an investigational device exemptions (IDE) study. Patients were enrolled in a prospective study comparing Amplify (2.0 mg/cm3 rhBMP-2 plus a compression resistant matrix; Medtronic, Inc, Memphis, TN) to iliac crest autograft in a single-level posterolateral instrumented lumbar fusion.
The patients had previously been reported on by Dimar et al6 who reported no statistically significant differences in cancer risk at 2 years postoperatively. The authors in this study present the results of the same patient group at both 2 and 5 years postoperatively. Unlike the initial report, this study demonstrates a statistically significant increased cancer risk and incidence among patients who were administered Amplify compared with control patients. The authors suggest that this difference could be due to a larger number of patients and longer follow-up of those patients.
There are a number of significant limitations to this study. Most importantly, the authors do not provide clear diagnostic criteria for the diagnosis of cancer events (e.g., histopathology, staging, etc.), the primary reported outcome of this study. Furthermore, the authors do not address potential confounding factors that could influence cancer outcomes.
The authors have addressed a timely clinical question. The concerns about complication associated with rhBMP-2 have been well reported. The gravest of complications, the development of cancer after rhBMP-2 use remains highly controversial. This study investigated 1 specific preparation (Amplify) and reported a statistically significant association between new cancer diagnosis and the use of Amplify in posterolateral lumbar fusion compared with iliac crest autograft bone.
Although this study should further the significant concerns about the clinical use of Amplify, it is insufficient to generalize the findings to all patients and to all uses of rhBMP-2. Patients with a pre-existing history of cancer were excluded. Other concentrations and carriers of rhBMP-2 were not assessed. In addition, perhaps most importantly, the diagnostic criteria for new cancers are not reported. Given that this is the primary outcome measure of this study and that the absolute number of new cancers is low, the uncertainty of the accuracy of cancer reporting in this study creates significant concerns.
RECOMMENDATION REGARDING IMPACT ON CLINICAL PRACTICE
On the basis of the results of this study, there is low-quality evidence and thus a “weak recommendation” that surgeons not use Amplify in instrumented posterolateral lumbar fusions for patients with adult spine problems. However, the findings of this study cannot be generalized to all patients and to all clinical use of rhBMP-2.
Efficacy and safety of surgical decompression in patients with cervical spondylotic myelopathy: results of the AOSpine North America prospective multicenter study. J Bone Joint Surg Am 2013;95:1651–8
Cervical spondylotic myelopathy (CSM) is a degenerative condition that frequently gives rise to progressive spinal cord dysfunction. Radiographical evidence of cervical stenosis is an extremely prevalent finding among patients older than 50 years, the majority of whom will not require formal treatment.7,8 The natural history of CSM has yet to be definitively established with some patients experiencing unremitting neurological deterioration and worsening disability over time whereas others will remain stable with conservative measures alone.9,10 Although multiple observational studies have reported that decompression of the spinal cord may bring about improvements in clinical outcomes for moderate to severe CSM, 1 small prospective randomized trial failed to identify any significant benefits associated with surgery for milder disease.11–14 Thus, although operative intervention is routinely advocated for individuals with signs and symptoms attributable to neural element compression, the specific surgical indications for CSM are still a matter of some debate. The purpose of this investigation was to prospectively evaluate the effects of decompressive procedures on the function, quality of life, and disability of a large cohort of patients with CSM at 1 year after surgery.15
As part of this study's prospective multicenter observational design, a total of 278 subjects with symptomatic CSM and evidence of spinal cord compression on their magnetic resonance images were enrolled at 12 institutions across North America during 3 years. Each patient underwent decompression of the spinal cord in conjunction with an instrumented fusion although the specific technical details (e.g., anterior vs. posterior approach, number of levels addressed, etc.) as well as the rehabilitation protocol were left to the discretion of the treating surgeon. Several different general and CSM-specific outcome instruments were used to assess neurological status, functional disability, and quality of life including Nurick grade, modified Japanese Orthopaedic Association scale (mJOA), and SF-36v2, which were administered preoperatively and 1 year after surgery. In addition, investigators were asked to review a list of predetermined complications at the 6- and 12-month visits to characterize the number and types of adverse events that occurred during the postoperative time period.
Patients were stratified according to their preoperative mJOA scores and the percentages of patients with mild, moderate, and severe CSM were 30.6%, 39.6%, and 29.9%, respectively; not surprisingly, those with more advanced disease were significantly older and required more extensive decompressions. Of the 278 subjects who were enrolled and subsequently underwent surgery, adequate follow-up data were available for 222 (85.4%). Compared with baseline values, all of the clinical outcomes were significantly improved after surgical intervention except for the general health component of the SF-36v2. Patients with milder symptoms exhibited less improvement in their mJOA scores relative to those with severe CSM but there were no other significant differences in the degree of improvement observed among the 3 groups according to the other outcome instruments. At 1 year, 52 of the 278 subjects (18.7%) had experienced an adverse event but the incidence of complications did not seem to be dependent upon the severity of disease.
Although the various inclusion criteria employed for this investigation are documented, the authors do not provide an explicit operational definition of “symptomatic” CSM or report the data for individuals who were found to be eligible for the study but were not enrolled. Using preoperative mJOA scores, the subjects were essentially equally distributed among the mild, moderate, and severe CSM categories; the stratification of patients in this manner allowed for the secondary analysis of outcomes based upon the extent of the initial disease. A wide range of decompression and fusion procedures were performed, varying in terms of the surgical approach and number of levels treated.
Appropriate statistical testing was implemented including univariate as well as multivariate regression techniques and any missing scores were accounted for using imputation. One-way analysis of variance with the Bonferroni correction was employed to minimize the risk of a type I error that may occur secondary to multiple comparisons. Besides the general health domain of the SF-36v2, all of the measured outcomes had significantly improved at 1 year after surgery. The patients with most profound myelopathy experienced the largest postoperative improvement as defined by their mJOA scores, consistent with a “regression toward the mean” phenomenon. Otherwise, the remaining outcome measures were similar between the 3 disease severity cohorts.
One methodological limitation of this investigation is the lack of a control group, in keeping with its observational design. This inherent challenge cannot be easily overcome because the randomization of subjects to nonoperative treatment, especially those with moderate or severe symptoms, would likely be deemed unethical. Thus, this protocol may represent the best available option for elucidating the benefits of surgical intervention for CSM. Given the heterogeneity of procedures performed, it was not possible to determine the optimal operative approach for decompressing the spinal cord; in actuality, this clinical trial may be more aptly described as an “effectiveness” study rather than an “efficacy” study because it incorporated fairly broad enrollment criteria with no strict guidelines regarding surgical technique or postoperative rehabilitation.
RECOMMENDATION REGARDING IMPACT ON CLINICAL PRACTICE
Fehlings et al15 have provided relatively strong evidence that surgical decompression and stabilization may be expected to give rise to meaningful improvements in the functional status and quality of life of individuals with CSM, regardless of the severity of their symptoms at the time of presentation. Given that operative intervention is a generally well-accepted method for managing patients with this condition, “we do not recommend any changes in clinical practice on the basis of these results.”
Effects of bracing in adolescents with idiopathic scoliosis. New Engl J Med 2013;369:1512–1.
Although adolescent idiopathic scoliosis (AIS) is a relatively common condition, the majority of cases may simply be observed and will not require any formal treatment. However, based upon the natural history of this condition, curves measuring greater than 50° are likely to increase over time so this threshold is generally considered to be an indication for surgical stabilization.16 For this reason, the application of a rigid orthosis is routinely advocated as a method for limiting the progression of deformities with significant growth potential. Numerous authors have suggested that the implementation of bracing regimens may give rise to smaller curves and therefore reduce the need for operative intervention.17–20 Although these findings are largely accepted in clinical practice, many of the existing studies exhibit a range of methodological deficiencies and there continues to be a paucity of high-quality evidence establishing the utility of bracing for patients with AIS. To this end, Weinstein et al21 reported the results of a prospective multicenter randomized controlled trial comparing the effectiveness of bracing relative to simple observation for preventing curve progression to 50° or more.
The Bracing in Adolescent Idiopathic Scoliosis Trial involved 242 skeletally immature children (defined as Risser grade of 0, 1, or 2) between the ages of 10 and 15 years with Cobb angles between 20° and 40° who were enrolled across 25 institutions in North America during 4 years. Because recruitment was initially slower than anticipated secondary to subjects' concerns regarding randomization, a preference arm was subsequently added to the study with otherwise identical inclusion criteria, protocols, and outcomes assessments. Patients in the bracing group were provided with a rigid thoracolumbosacral orthosis that was intended to be worn for a minimum of 18 hours a day; the actual time in the orthosis was calculated using an embedded temperature sensor. Subjects assigned to the observation cohort did not receive any specific treatment for their curves. The primary outcomes were curve progression to 50° or greater versus reaching skeletal maturity with a curve of smaller magnitude (i.e., failure and success, respectively). Radiographical assessments were performed at 6-month intervals at which time clinical data such as adverse events and quality-of-life scores were also collected.
Of the 242 patients, 116 underwent randomization and 126 were in the preference cohort; 146 received a thoracolumbosacral orthosis whereas 96 were simply observed. Overall, the rates of treatment success were 72% in the bracing group and 48% for the observation cohort with an adjusted odds ratio for treatment success of 1.93 (95% confidence interval,10 1.08–3.46). According to the intention-to-treat analysis, the rates of treatment success were 75% and 42% for those randomly assigned to bracing and observation, respectively (95% CI, 1.85–9.16). There was a statistically significant positive association between duration of brace wear and treatment success (P < 0.001) but there were no clinically relevant differences between the quality-of-life scores and adverse events reported for the bracing and observation groups. Although these findings ostensibly demonstrated a clear benefit in favor of bracing, the decision was made to proceed with early termination of the trial.
The authors clearly document the numbers of children who were screened (1183), found to be eligible (1086), consented (383), and included in the randomized and preference cohorts (116 and 126, respectively). Subjects who declined to participate in the study were registered in a rejection log and they exhibited similar demographic characteristics in terms of age and sex. As another control, the randomization of subjects to the bracing and observation cohorts was stratified by curve type (i.e., single thoracic curve vs. all other curves). The majority of patients who were consented were not considered in the primary analysis because they did not reach the definition of treatment success or failure at the time the investigation was concluded.
For those in the bracing group, a novel temperature-sensitive monitor was incorporated into the orthosis to more accurately quantify compliance. As part of the primary endpoint, the criteria used to assess skeletal maturity were explicitly defined. The determination of radiographical success or failure was made by 2 independent raters who were blinded to the treatment designation of each subject.
Data were analyzed using both univariate and multivariate techniques to compare the efficacy of bracing with that with observation. Applying multiple statistical tests in this fashion inherently increases the likelihood of a type 1 error but this issue was addressed with appropriate safeguards. In addition to the superior rates of success associated with bracing, the authors estimated that only 3 patients would have to be treated in this fashion to avoid a single case of curve progression warranting surgery, corresponding to a reduction in relative risk of 56%. Moreover, there was a significantly positive dose-response relationship between the number of hours spent in the orthosis and the frequency of treatment success. Although the preliminary results so strongly supported the efficacy of bracing for AIS and met certain prespecified criteria, the investigators elected to suspend the study at the time of the first interim analysis.
The primary strengths of this clinical trial are its prospective multicenter design and blinded methods of outcome assessment. The slow accrual of subjects amenable to randomization was overcome by establishing a preference-based cohort as well; any potential bias that may have been introduced by considering subjects who selected their own treatment was mitigated by the use of multivariable regression.
RECOMMENDATION REGARDING IMPACT ON CLINICAL PRACTICE
This study provides sound evidence that the bracing of skeletally immature patients with AIS may be expected to decrease curve progression and possibly minimize the need for surgical intervention. Furthermore, it also seems as if the benefits of these regimens increase with longer times in the orthosis. This study provides the first high-quality evidence to validate the efficacy of bracing for AIS curves. Although there has historically been general acceptance of this treatment for AIS, we think there is now a “strong recommendation” to incorporate these results into clinical practice.
Protective effects of preserving the posterior complex on the development of adjacent-segment degeneration after lumbar fusion. J Neurosurg Spine 2013;19:201–6
Adjacent segment degeneration (ASDeg) after lumbar spinal fusion is of increasing interest as the rate of spinal fusions increases worldwide. Numerous motion-sparing and other technologies have been developed with the belief that ASDeg is a consequence of the biomechanical impact of spinal segmental immobilization. It is important to distinguish ASDeg from adjacent segment disease (ASDis): ASDeg represents new radiographical degeneration of a segment adjacent to a surgically treated level without symptomatology; ASDis represents the degeneration of an adjacent level accompanied by related symptoms like radiculopathy, mechanical back pain, or claudication.22 A number of factors have been suggested to predispose patients to ASDeg after lumbar spine surgery including age, sagittal imbalance, posterior element disruption, elevated body mass index, and pre-existing degeneration.23–29 The authors attempt to isolate a single factor—posterior element disruption—to determine if different posterior interbody fusion techniques confer different risks of ASDeg.
Liu et al30 present a randomized cohort study from a single hospital in China on patients having L4–L5 posterior interbody single-level fusion with instrumentation. From 1998 to 2006, 120 patients were randomized into the following 3 groups: group A, facet joint resection and fusion (midline structures preserved); group B, semilaminectomy and fusion (only the bottom half of L4 and the top one-third of L5 removed); and group C, complete L4 laminectomy and fusion (all of L4 lamina with midline structures removed). Fusion consisted of PEEK interbody cages with posterior pedicle screws at L4–L5 and local autograft. Indications for surgery included lumbar disk herniation with disk endplate degeneration, disk herniation with instability, and disk herniation with stenosis. Standard preoperative variables were collected as well as JOA scores. Radiographical variables focused on disk height and angulation at the L3–L4 level, given its historical propensity for adjacent level disease. All surgery was reportedly performed by the same team at 1 hospital. One-way and repeated measures analysis of variance and χ2 analysis was used to evaluate outcomes.
The method of randomization was not clear and seemed to only include age, sex, and diagnosis. Even with that, there seemed to be a sex imbalance with more females in group C than group A and more males in group A than in group C. This may hardly be a cause of bias, but does question the method of randomization employed. Because the main outcome of the study is dependent on radiographical evaluation, 3 independent radiologists were used to rate and measure the pre- and postoperative images. However, it is unclear whether these radiologists were blinded to surgery group and whether or not true blinding is even possible given the ability to see remaining or absent bone structures on radiographs. Although this should not play a strong role in bias so long as the radiologists were “blinded” to the goal of the study, it nonetheless could be an issue if they were aware.
The authors found that adjacent level degeneration was higher in group C than in the other 2 groups with regard to disk height and disk angulation at final follow-up (average, 5.9 yr). This suggests that disrupting the posterior midline bony/ligamentous structures between L3 and L4 (as in group C) may play a role in subsequent ASDeg that significantly exceeds the impact of less-aggressive techniques that preserve the intervening structures at L3–L4 (as in groups A and B). This conclusion is supported by this study based on its reported methods, but is highly dependent on the integrity of the selection and randomization process, which was somewhat obscure.
The authors present a series of 120 single-level L4–L5 posterior interbody fusions randomized to 3 different degrees of posterior element disruption, as outlined in the earlier text. They evaluated ASDeg only at the supra-adjacent L3–L4 level limiting the generalizability of their results. They were careful to exclude patients with pre-existing L3–L4 degeneration greater than grade III using the UCLA classification system. The authors report that preoperative demographic and clinical data were not significantly different between groups although body mass index and the presence of medical comorbidities were not reported. Cho et al3 identified body mass index, and pre-existing stenosis to be independent risk factors for ASDeg. The authors present a thorough analysis of plain radiograph and report equivalent height restoration and preservation at L4–L5. They do not report pseudarthrosis rates for these 3 techniques. At the L3–L4 level, they report significant loss of disk height only in group C at final follow-up. All groups showed a modest but statistically significant increase in angular ROM at L3–L4 though this was significantly greater in group C (3.5°) than in groups A (1.1°) and B (1.1°). Similarly, all groups had a modest but statistically significant increase in L3–L4 listhesis though, again, it was significantly greater in group C (2.19 mm) than groups A (0.68 mm) and B (0.56 mm). Finally, L1–S1 lordosis was comparable preoperatively in all 3 groups and increased by 1° in both groups A and B but decreased by 4° in group C by final follow-up. In aggregate, the authors report that 3 (7.5%) patients in group A, 4 (10%) of patients in group B, and 17 (42.5%) of patients in group C developed ASDeg at the L3–L4 level.
With respect to clinical outcome, the authors present comparable preoperative JOA scores and show statistically significant improvements in all 3 groups at 3 months postoperatively and final follow-up. Interestingly, there was no difference between the 3 groups at the 3-month time point and both groups A and B preserved this improvement at final follow-up, whereas there was loss of improvement in group C that rendered a statistically significant difference in final follow-up JOA score between group C and groups A and B. None of the patients in groups A and B underwent reoperation, whereas 7 (17.5%) of the patients in group C underwent reoperation for ASDis.
Although this is a well-performed study, a few other concerns bear reporting. As the authors indicate in their article, MRI data are not available to more thoroughly characterize ASDeg. Sagittal imbalance has been shown to significantly increase ASDeg but is not reported in this article. Although L1–S1 lordosis is presented, it is difficult to infer anything convincing about global sagittal balance or spinopelvic anatomy. By design, none of the patients had significant structural disease at the L3–L4 level yet patients in group C underwent removal of the L3–L4 interspinous and supraspinous ligaments and detachment of the L3–L4 ligamentum flavum. Numerous other studies have described degenerative disease at both cranial and caudal adjacent levels though the authors evaluate only the L3–L4 level. In short, this article suggests that disruption of the L3–L4 posterior ligamentous and osseous elements predisposes patients to accelerated degeneration at this level when it is above an L4–L5 interbody fusion.
RECOMMENDATION REGARDING IMPACT ON CLINICAL PRACTICE
Although structural differences limit the generalizability of the ASDeg and ASDis rates identified in this study, the authors have attempted to isolate extent of posterior element disruption as a contributor to ASDeg. Their findings suggest that unnecessary posterior element disruption should be minimized particularly when it undermines an unfused level adjacent to a fused segment. On the basis of these results, a “weak recommendation” to incorporate these findings into clinical practice can be made.
Positive predictive factors and subgroup analysis of clinically relevant improvement after anterior cervical decompression and fusion for cervical disk disease: a 10- to 13-year follow-up of a prospective randomized study. J Neurosurg Spine 2013;19:403–11.
Managing patient expectations is an increasingly important part of surgical practice and surgeons often cite generalized outcomes for procedures, adjusting them modestly on the basis of relevant risk factors. On a simple level, patients want to know if they will feel better after surgery and how long that “clinically relevant improvement” (CRI) may last. In an ideal world, we would have a prediction model that would allow individualized outcome prediction for each patient and circumstance. Hermansen et al31 have sought to take us one step closer to this ideal by identifying predictive factors for CRI in patients undergoing ACDF in Sweden with more than 10-year follow-up.
Hermansen et al31 used data and subjects from a prior single center, prospective randomized trial conducted at their hospital in 1995. The initial study ran from 1995 to 1998 with the objective of comparing ACDF using the Cloward procedure (iliac crest bone dowel) or cervical intervertebral fusion cage (carbon fiber cage) in a randomized fashion. Initial variables included standard clinical and surgical metrics such as age, sex, smoking status, duration of symptoms, number of levels operated on, and VAS and NDI scores as well as radiographical parameters such as fusion, disk height, and segmental kyphosis. Questionnaire surveys were then sent to patients at 6 and 10 years from surgery to provide data for this study, including VAS, NDI, EQ-5D, and CSQ. CRI was defined as a 3-point improvement in pain score and a 20% improvement in disability score. Patients were then dichotomized into binary outcome groups (presence or absence of CRI) and analyzed with preoperative variables using linear regression techniques.
Of the 95 patients from the initial 1995 study, 90 patients were sent questionnaires and 73 patients responded for an 88% response rate, which is very high. Of these patients, 46 had single-level surgery, 24 had 2-level surgery, and 3 had 3-level surgery. Because the subjects were randomized and because this study did not need to compare the 2 surgical procedures, patient variables remained balanced. Overall, the patient group represents a very relevant group of patients mainly having 1- to 2-level ACDF surgery at a single university hospital in Sweden with excellent preoperative data fidelity. Follow-up responses were also very high, and in all but 1 case, questionnaires were completely filled out. This however was not done in a proctored method to help control for confounding and bias and therefore introduces possible error in the fidelity of the long-term outcome data. Though the VAS and NDI questionnaires do not require particular recall of the original responses, they also do not identify potential new or unrelated causes of pain and disability such as natural progression of degenerative disk disease, history of injury, or other health changes nor do they capture other therapeutic treatments such as medications and alternative treatments currently employed for related or unrelated ailments. Although the presence or absence of a CRI can still be associated with preoperative variables, other data that could interact with the association are lacking. Whether or not there is another variable that interacts with the predictive association in a significant way is unlikely, but unclear as well.
This study demonstrates that high neck VAS and nonsmoking status were predictive of CRI in neck VAS score, and that male sex was predictive of CRI in NDI, which is generally consistent with prior studies. This conclusion is supported by the methods of this article but does not preclude other underpowered or unrecognized factors from being important or interacting with CRI. Also, interestingly, fusion at 2 years from surgery did not seem to be a positive predictor of CRI in long-term follow-up, which only adds to the growing debate of whether a bony fusion is necessary for ACDF to be considered successful.
RECOMMENDATION REGARDING IMPACT ON CLINICAL PRACTICE
There is a “strong recommendation” that the factors identified by Hermansen et al,31 which predict CRI in ACDF should be incorporated into the guidance offered patients and the management of patient expectations. However, these results cannot be extrapolated to influence candidacy for surgical intervention.
1. Deyo RA, Mirza SK, Martin BI, et al. Trends, major medical complications, and charges associated with surgery for lumbar spinal stenosis in older adults. JAMA 2010;303:1259–65.
2. Ong KL, Villarraga ML, Lau E, et al. Off-label use of bone morphogenetic proteins in the United States using administrative data. Spine (Phila Pa 1976) 2010;35:1794–800.
3. Cho TK, Lim JH, Kim SH, et al. Preoperative predictable factors for the occurrence of adjacent segment degeneration requiring second operation after spinal fusion at isolated L4–L5 level. J Neurol Surg A Cent Eur Neurosurg 2014;75:270–5.
4. Singh K, Ahmadinia K, Park DK, et al. Complications of spinal fusion with utilization of bone morphogenetic protein: a systematic review of the literature. Spine (Phila Pa 1976) 2014;39:91–101.
5. Carragee EJ, Chu G, Rohatgi R, et al. Cancer risk after use of recombinant bone morphogenetic protein-2 for spinal arthrodesis. J Bone Joint Surg Am 2013;95:1537–45.
6. Dimar JR II, Glassman SD, Burkus JK, et al. Clinical and radiographic analysis of an optimized rhBMP-2 formulation as an autograft replacement in posterolateral lumbar spine arthrodesis. J Bone Joint Surg Am 2009;91:1377–86.
7. Bednarik J, Kadanka Z, Dusek L, et al. Presymptomatic spondylotic cervical cord compression. Spine (Phila Pa 1976) 2004;29:2260–9.
8. Irvine DH, Foster JB, Newell DJ, et al. Prevalence of Cervical spondylosis in a general practice. Lancet 1965;1:1089–92.
9. Fehlings MG, Arvin B. Surgical management of cervical degenerative disease: the evidence related to indications, impact, and outcome. J Neurosurg Spine 2009;11:97–100.
10. Matz PG, Anderson PA, Holly LT, et al. The natural history of cervical spondylotic myelopathy. J Neurosurg Spine 2009;11:104–11.
11. Kadanka Z, Mares M, Bednaník J, et al. Approaches to spondylotic cervical myelopathy: conservative versus surgical results in a 3-year follow-up study. Spine (Phila Pa 1976) 2002;27:2205–10; discussion 2210–1.
12. Kaminsky SB, Clark CR, Traynelis VC. Operative treatment of cervical spondylotic myelopathy and radiculopathy. A comparison of laminectomy and laminoplasty at five-year average follow-up. Iowa Orthop J 2004;24:95–105.
13. Kiris T, Kilincer C. Cervical spondylotic myelopathy treated by oblique corpectomy: a prospective study. Neurosurgery 2008;62:674–82; discussion 674–82.
14. Papadopoulos CA, Katonis P, Papagelopoulos PJ, et al. Surgical decompression for cervical spondylotic myelopathy: correlation between operative outcomes and MRI of the spinal cord. Orthopedics 2004;27:1087–91.
15. Fehlings MG, Wilson JR, Kopjar B, et al. Efficacy and safety of surgical decompression in patients with cervical spondylotic myelopathy: results of the AOSpine North America prospective multi-center study. J Bone Joint Surg Am 2013;95:1651–8.
16. Weinstein SL, Ponseti IV. Curve progression in idiopathic scoliosis. J Bone Joint Surg Am 1983;65:447–55.
17. Dolan LA, Weinstein SL. Surgical rates after observation and bracing for adolescent idiopathic scoliosis: an evidence-based review. Spine (Phila Pa 1976) 2007;32(suppl 19):S91–100.
18. Focarile FA, Bonaldi A, Giarolo MA, et al. Effectiveness of nonsurgical treatment for idiopathic scoliosis. Overview of available evidence. Spine (Phila Pa 1976) 1991;16:395–401.
19. Negrini S, Minozzi S, Bettany-Saltikov J, et al. Braces for idiopathic scoliosis in adolescents. Cochrane Database Syst Rev 2010;1:CD006850.
20. Rowe DE, Bernstein SM, Riddick MF, et al. A meta-analysis of the efficacy of non-operative treatments for idiopathic scoliosis. J Bone Joint Surg Am 1997;79:664–74.
21. Weinstein SL, Dolan LA, Wright JG, et al. Effects of bracing in adolescents with idiopathic scoliosis. N Engl J Med 2013;369:1512–21.
22. Saavedra-Pozo FM, Deusdara RA, Benzel EC. Adjacent segment disease perspective and review of the literature. Ochsner J 2014;14:78–83.
23. Helgeson MD, Bevevino AJ, Hilibrand AS, Update on the evidence for adjacent segment degeneration and disease. Spine J 2013;13:342–51.
24. Keorochana G, Taghavi CE, Lee KB, et al. Effect of sagittal alignment on kinematic changes and degree of disc degeneration in the lumbar spine: an analysis using positional MRI. Spine (Phila Pa 1976) 2011;36:893–8.
25. Min JH, Jang JS, Lee SH. Comparison of anterior- and posterior-approach instrumented lumbar interbody fusion for spondylolisthesis. J Neurosurg Spine 2007;7:21–6.
26. Radcliff KE, Kepler CK, Jakoi A, et al. Adjacent segment disease in the lumbar spine following different treatment interventions. Spine J 2013;13:1339–49.
27. Ren C, Song Y, Liu L, et al. Adjacent segment degeneration and disease after lumbar fusion compared with motion-preserving procedures: a meta-analysis [published online ahead of print April 12, 2014]. Eur J Orthop Surg Traumatol 2014;(suppl 1):245–53. doi:10.1007/s00590-014-1445-9.
28. Soh J, Lee JC, Shin BJ. Analysis of risk factors for adjacent segment degeneration occurring more than 5 years after fusion with pedicle screw fixation for degenerative lumbar spine. Asian Spine J 2013;7:273–81.
29. Wai EK, Santos ER, Morcom RA, et al. Magnetic resonance imaging 20 years after anterior lumbar interbody fusion. Spine (Phila Pa 1976) 2006;31:1952–6.
30. Liu H, Wu W, Li Y, et al. Protective effects of preserving the posterior complex on the development of adjacent-segment degeneration after lumbar fusion: clinical article. J Neurosurg Spine 2013;19:201–6.
31. Hermansen A, Hedlund R, Vavruch L, et al. Positive predictive factors and subgroup analysis of clinically relevant improvement after anterior cervical decompression and fusion for cervical disc disease: a 10- to 13-year follow-up of a prospective randomized study: clinical article. J Neurosurg Spine 2013;19:403–11.