Long-Term Sustainability of Functional Improvement Following Central

CERVICAL SPINE

Long-Term Sustainability of Functional Improvement Following Central Corpectomy for Cervical Spondylotic Myelopathy and Ossification of Posterior Longitudinal Ligament

Sauradeep Sarkar, MBBS and Vedantam Rajshekhar, MCh

more years (P < 0.001). Early improvement in functional status Study Design. Retrospective study. was maintained in 90.5% and 76.3% of patients at 5 and Objective. To examine predictors of long-term outcome and 10 years follow-up, respectively. On survival analysis, patients sustainability of initial functional improvement in patients with shorter preoperative symptom duration (<1 yr) were more undergoing corpectomy for cervical spondylotic myelopathy likely to demonstrate sustained improvement in functional status (CSM) or ossification of the posterior longitudinal ligament after surgery (P ¼ 0.022). (OPLL). Conclusion. Initial gains in functional status after central Summary of Background Data. There are limited data on corpectomy for CSM and OPLL are maintained in more than the predictors of outcome and sustainability of initial functional 75% of patients at 10 years after surgery. Overall, the most improvement on long-term follow-up after cervical corpectomy. Methods. We studied the functional outcome at more than 1- favorable long-term outcomes are achieved in younger patients year follow-up after central corpectomy in 352 patients with who present early and with good preoperative functional status. Key words: cervical spine, cervical spondylotic myelopathy, CSM or OPLL. Functional status was evaluated with the Nurick corpectomy, myelopathy, ossification of the posterior grading system. Analysis was directed at identifying factors longitudinal ligament, outcomes, prognostic factors, spondylosis, associated with both improvement in functional status and the sustainability. achievement of a ‘‘cure’’ (improvement to a follow-up Nurick Level of Evidence: 4 grade of 0 or 1). A survival analysis was performed to identify Spine 2018;43:E703–E711 factors associated with sustained functional improvement in patients with serial follow-up evaluations. Results. Nurick grade improved from 3.2 0.1 to 1.9 0.1 over a mean follow-up period of 57.1 months (range 12–228 ervical cord compression due to cervical spondy- mo). On multivariate analysis, age 50 years (P ¼ 0.008) and lotic myelopathy (CSM) or ossification of the pos- symptom duration 1 year (P < 0.001) were negatively associ- C terior longitudinal ligament (OPLL) frequently ated with functional improvement by 1 Nurick grade. Indepen- results in debilitating symptoms that significantly impair 1 dent factors negatively associated with ‘‘cure’’ after surgery both neurological function and quality of life. The patho- included age 50 years or older (P ¼ 0.005), preoperative Nurick physiology of degenerative cervical myelopathy is complex, grade of 4 or higher (P < 0.001) and symptom duration of 1 or involving both static and dynamic processes that usually result in progressive neurological dysfunction.2–5 Thus surgical decompression is recommended for the vast majority From the Department of Neurological Sciences, Christian Medical College, of patients with CSM and OPLL who present with progres- Vellore, India. sive functional impairment. Surgical decompression may be Acknowledgment date: August 2, 2017. First revision date: October 5, 2017. Acceptance date: October 16, 2017. accomplished by using a number of approaches, both ante- The manuscript submitted does not contain information about medical rior and posterior. Central corpectomy is an anterior decom- device(s)/drug(s). pressive procedure, which is ideally suited for multilevel No funds were received in support of this work. CSM or OPLL. It directly addresses the ventral compression No relevant financial activities outside the submitted work. that is seen in more than 75% of patients with CSM and all Address correspondence and reprint requests to Vedantam Rajshekhar, patients with OPLL. MCh, Department of Neurological Sciences, Christian Medical College, Although most contemporary reports testify to excellent Vellore 632004, India; E-mail: [email protected] outcomes after surgery, a small proportion of patients fail to 6–14 DOI: 10.1097/BRS.0000000000002468 achieve a complete functional recovery. Most published Spine www.spinejournal.com E703 Copyright © 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. CERVICAL SPINE Long-Term Outcomes After Cervical Corpectomy Sarkar and Rajshekhar studies on the prognostic predictors of surgical outcome in each follow-up visit. Fusion status was assessed at the upper patients with CSM and OPLL suffer from small patient and lower ends of the graft, and stability was defined by lack cohorts, multiple surgeon involvement, short follow-up of sagittal translation on flexion-extension lateral cervical periods, or a heterogeneous choice of surgical spine radiographs. Patients unable to review in person were approaches.14–21 Even studies which report long-term out- periodically requested to send us postoperative radiographs comes often do not clarify whether there was a change in along with the follow-up questionnaire. function from the early postoperative period to late follow- up. Thus the proportion of patients who show a sustained Functional Evaluation improvement of function is unclear. Functional status before surgery and at all follow-up visits The objective of this study was to determine the sustain- was evaluated with the Nurick grading system. Follow-up ability of initially favorable postoperative outcome over details in the majority of cases had been obtained from long-term follow-up. We also planned to study the impact evaluation in the outpatient clinic and in the remainder by a of various preoperative clinical, imaging, and surgical fac- self-administered postal questionnaire. tors on long-term functional outcomes in a large single- surgeon cohort of patients undergoing the same surgical Outcome Measures and Prognostic Variables procedure, namely cervical corpectomy for CSM or OPLL. The primary outcome measures for this study were: (1) Functional improvement by at least one Nurick grade at MATERIALS AND METHODS last follow-up and (2) ‘‘Cure’’ (defined as improvement to a postoperative Nurick grade of 0 or 1). Patients with a Inclusion Criteria preoperative Nurick grade of 0 or 1 who retained their We consider cervical corpectomy for all patients with CSM Nurick grade after surgery were not considered ‘‘cured.’’ or OPLL, regardless of cervical alignment, unless the com- The prognostic variables studied were age at presentation, pression extends across more than three vertebral bodies or preoperative Nurick grade, symptom duration, the presence behind C2. We have been practicing uninstrumented fusion of diabetes mellitus, preoperative bladder dysfunction, pre- with a bone graft for nearly 3 decades and have achieved operative and postoperative sagittal alignment, a primary satisfactory neurological outcomes with acceptable graft diagnosis of OPLL, number of levels decompressed, and the migration and donor site morbidity rates, encouraging us occurrence of a perioperative complication. to maintain this practice.22,23 Between 1991 and 2008, 482 We sought to study whether improvement in functional patients underwent uninstrumented central corpectomy, status was sustained at long-term follow-up after CC for with all procedures performed by the senior author CSM. We identified 218 patients in whom serial evaluations (V.R.). For this study, we included only patients with a of postoperative functional status (i.e., two or more assess- primary diagnosis of CSM or OPLL and with >1 year ments performed at least 12 months apart). We then follow-up. Data were extracted from a prospectively excluded patients who had initially failed to improve or maintained database. worsened after surgery (n ¼ 24) or in whom the first follow- up evaluation was performed >2 years after surgery Surgical Technique (n ¼ 19). In this subcohort of 175 patients, we defined The surgical technique has been described in detail in several deterioration as a Nurick grade equal to or worse than that prior publications,24,25 and involved reconstruction of the before surgery or Nurick grade of 5 regardless of preopera- corpectomy defect with a tricorticate iliac bone or fibular tive status, on serial follow-up. For example, a patient with a strut graft, without instrumentation or anterior cervical preoperative Nurick grade of 3 who improved to Nurick plating. After confirming proper graft placement with an grade 1 at the first follow-up visit but later worsened to a immediate postoperative lateral cervical spine radiograph, Nurick grade of 3 at a subsequent visit was considered to patients were mobilized and discharged on a cervical collar have failed to sustain improvement in his/her functional that was maintained for 3 to 6 months. status.

Radiographic Evaluation Statistical Analysis Cervical spine curvature was classified by using a line drawn Continuous and categorical variables are presented as mean from the posterior end of the inferior endplate of the C2 standard error and frequencies, respectively. The Student t vertebral body to a similar point on C7. The spine was test or ANOVA and the chi-squared test were used to study described as lordotic if all the intervening vertebral bodies differences between continuous and categorical variables, were anterior to this line and straight if the intervening respectively. A univariate analysis was performed to esti- vertebral bodies touched the line without crossing it. If the mate the odds ratio and 95% confidence interval for each vertebral bodies intersected the line, the spine was classified variable. The logistic regression model was created by as kyphotic.26 Preoperative cervical alignment was classified selecting covariates that attained a P value less than before and after surgery in all patients in whom imaging was 0.150 on univariate analysis. A P value less than 0.05 available for review. Postoperative lateral neutral, flexion was considered statistically significant. The log-rank test and extension cervical spine radiographs were obtained at was performed to compare survival curves.

E704 www.spinejournal.com June 2018 Copyright © 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. CERVICAL SPINE Long-Term Outcomes After Cervical Corpectomy Sarkar and Rajshekhar

TABLE 1. Preoperative Characteristics of the Complications Cohort (n ¼ 352) Forty-five patients (12.8%) suffered at least one complication after surgery, with two patients suffering more than one Variable Value complication. Perioperative morbidity comprised dural Mean age at presentation 47.9 0.5 years tears (n ¼ 13), graft migrations (n ¼ 8), C5 radiculopathies Sex (n ¼ 6), recurrent laryngeal nerve palsy (n ¼ 1), severe dys- Male 330 (93.8%) phagia (n ¼ 3), superficial surgical site infections (n ¼ 5), Female 22 (6.2%) donor site morbidity (n ¼ 7), pulmonary complications Diabetes mellitus 16 of 333 (4.8%) (n ¼ 3), and cardiac failure (n ¼ 1). Bony nonunion was seen Median duration of symptoms 8 (1–240) months in 17.2% of patients and is discussed in more detail below. (range) Mean preoperative Nurick grade 3.2 0.1 Functional Outcomes at Last Follow-up Primary pathology The mean follow-up period was 57.1 months (range 12–228 CSM 243 (69.0%) mo). Nurick grades at baseline and at last follow-up are OPLL 109 (31.0%) summarized in Table 2. Overall, the mean preoperative Preoperative sagittal alignment Nurick grade improved significantly from 3.2 0.1 to Lordotic 21 (6.4%) 1.9 0.1 at last follow-up after surgery (P < 0.001) and Straight 260 (79.0%) 255 patients (72.4%) improved by at least one Nurick grade. Kyphotic 48 (14.6%) Postoperative improvement to a Nurick grade of 0 or 1 was Number of operated segments achieved in 126 patients (35.8%). Functional outcomes strat- One 112 (31.8%) ified by duration of follow-up are described in Table 3. The Two 214 (60.8%) y mean change in Nurick grade and the proportion of patients Three 26 (7.4%) who achieved cure was similar in patients followed up for Perioperative morbidity 45 (12.8%) different periods of time (P ¼ 0.054 and P ¼ 0.236, respec- Includes only patients with appropriate data. tively). However, a slightly lower proportion of patients y Includes patients who underwent skip corpectomies. followed up for between 5 and 10 years had improved by CSM indicates cervical spondylotic myelopathy; OPLL, ossification of posterior longitudinal ligament. one Nurick grade or more. By last follow-up, 14 patients (3.9%) had symptomatic adjacent segment disease. All these patients underwent reoperation, with the exception of three RESULTS patients, two of whom refused surgery, and one in whom cardiac comorbidities rendered him unfit for general anes- Preoperative Clinical Features of the Cohort thesia. Reoperations consisted of ventral decompression in A total of 352 patients fulfilled the inclusion criteria for this six patients, and laminectomy in five patients. study. Table 1 summarizes their preoperative clinical and imaging characteristics. The vast majority of patients Sustainability of Improved Functional Outcome (93.8%) were men and 45.7% were 50 years or older at After Surgery the time of presentation. Axial pain was a significant symp- Sustainability of early postoperative improvement in func- tom at presentation in 91 patients (25.8%), whereas 169 tional status was assessed in 175 patients over a mean patients (48.0%) had features of bladder dysfunction. OPLL follow-up period of 71.8 months (range 24–228 mo). was the primary diagnosis in almost a third of patients. Nine The proportions of patients who showed sustained improve- patients had undergone prior surgery elsewhere for degen- ment in functional status were 90.5% and 76.3% at 5 and erative cervical spine disease. Seven of these patients had 10 years’ follow-up, respectively (Figure 1A–D). Of all the undergone prior anterior approaches, whereas two patients prognostic factors studied, only symptom duration was had undergone cervical laminectomy without fusion. associated with the likelihood of sustained improvement

TABLE 2. Preoperative Versus Last Follow-up Nurick grade for the Cohort (n ¼ 352) Nurick Grade at Last Follow-up Preoperative Nurick Grade 0 12345Total 01000001 1 57 100013 218213622079 3 13 39 54 8 6 0 120 4 7 14 44 7 33 1 106 5168315033 Total 45 87 143 20 56 1 352

Spine www.spinejournal.com E705 Copyright © 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. CERVICAL SPINE Long-Term Outcomes After Cervical Corpectomy Sarkar and Rajshekhar

TABLE 3. Outcomes Stratified by Duration of Follow-up Total Follow-up Duration 1 and <5yr 5 and <10 yr 10 yr Number of patients 211 110 31 Mean preoperative Nurick grade 3.3 0.1 3.0 0.1 3.0 0.2 Mean postoperative Nurick grade 1.9 0.1 2.0 0.1 1.5 0.2 Change in Nurick grade 1.4 0.1 1.1 0.1 1.5 0.2 Improved postoperative Nurick gradey 163 (77.3%) 68 (61.8%) 24 (77.4%) Nurick grade 0 or 1 at last follow-upz 76 (36.0%) 35 (31.8%) 15 (48.4%) 0.054. yP ¼ 0.011. zP ¼ 0.236. in function (Figures 1 and 2A–D). Patients with shorter significantly differ between patients in the curvature was lost preoperative symptom duration were more likely to dem- or preserved after surgery (P ¼ 0.238). Similarly, complete or onstrate sustained improvements in functional status after partial restoration of lordosis in patients with a preopera- surgery (P ¼ 0.022; log-rank test). tively straight or kyphotic spine was not associated with mean change in Nurick grade at last follow-up (P ¼ 0.953) Cervical Alignment and Fusion Data on preoperative and postoperative cervical alignment Predictors of Functional Improvement were present in 329 patients. At last follow-up, 110 patients On univariate analysis, there were statistically significant (34.0%) had lordotic spines, 149 patients (45.0%) had straight negative associations between age 50 years or older spines, and 65 patients (20.0%) had kyphotic spines. Data on (P ¼ 0.005) and symptom duration of 1 year or more fusion were available in 314 patients. Bony fusion was (P < 0.001) and functional improvement by at least 1 Nur- achieved at both ends of the graft in 260 (82.8%) of patients ick grade (Table 4). On multivariate analysis, both factors and stable fibrous union at either end of the graft was seen in 51 retained their statistical significance. patients (16.2%) Three patients (1.0%) demonstrated a pseudoarthrosis at the upper or lower end. None of the patients Predictors of ‘‘Cure’’ with pseudoarthrosis were symptomatic for the same. On univariate analysis, age 50 years or older (P < 0.001), In a subset analysis of patients with a preoperatively preoperative Nurick grade of 4 or higher (P < 0.001), symp- lordotic spine, the mean change in Nurick grade did not tom duration of 1 year or more (P ¼ 0.001), preoperative

Figure 1. Kaplan-Meier plots showing proportion of patients with sustained functional outcome overall (A) and stratified by age (B), duration of symptoms (C), and preoperative Nurick grade (D). The log-rank test was used to determine statistical significance.

E706 www.spinejournal.com June 2018 Copyright © 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. CERVICAL SPINE Long-Term Outcomes After Cervical Corpectomy Sarkar and Rajshekhar

Figure 2. Kaplan-Meier plots showing proportion of patients with sustained functional outcome stratified by number of operated levels (A), primary diagnosis (B), preoperative kyphosis (C), and postoperative kyphosis (D). The log-rank test was used to determine statistical significance. CSM indicates cervical spondylotic myelopathy; OPLL, ossification of posterior longitudinal ligament.

bladder dysfunction (P ¼ 0.011), and the occurrence of a recovery rates tend to decline over time, most patients complication after surgery (P ¼ 0.046) were negatively asso- continue to demonstrate the benefits of surgical decompres- ciated with outcome. On multivariate analysis, the only sion at long-term follow-up.12,19,28 The results of our study independent factors negatively associated with ‘‘cure’’ after are largely in agreement with these findings, with more than surgery included age 50 years or older (P ¼ 0.005), baseline 70% of patients continuing to demonstrate quantifiable Nurick grade of 4 or higher (P < 0.001), and symptom gains in functional status over a mean follow-up period duration 1 year or more (P < 0.001) (Table 5). of almost 5 years.

DISCUSSION Is Initial Improvement in Functional Status Most contemporary reports describe favorable outcomes in Sustained at Long-Term Follow-up? patients undergoing anterior or posterior approaches for A number of studies have described long-term outcomes CSM and OPLL.6–11,27 Although functional outcomes and after anterior or posterior decompression for degenerative

TABLE 4. Univariate and Multivariate Analysis of Factors Associated With Improvement by At least 1 Nurick Grade at Last Follow-up (n ¼ 338) Univariate Analysis Multivariate Analysis Odds Ratio; 95% Odds Ratio; 95% Variable Confidence Interval P Confidence Interval P Age 50 years 0.493; 0.301–0.808 0.005y 0.502; 0.301–0.838 0.008 Preoperative Nurick grade 4/5 1.150; 0.700–1.891 0.581 Symptom duration 1 yr 0.286; 0.171–0.478 <0.001y 0.293; 0.174–0.493 <0.001 Diabetes mellitus 1.548; 0.430–5.572 0.504 Preoperative bladder dysfunction 1.000; 0.615–1.626 1.000 Preoperative kyphotic alignment 0.959; 0.334–2.750 0.937 Presence of OPLL 0.674; 0.404–1.124 0.131y 0.723; 0.422–1.238 0.237 Single versus multilevel corpectomy 1.042; 0.612–1.772 0.881 Perioperative morbidity 0.664; 0.338–1.302 0.233 Postoperative kyphotic alignment 0.701; 0.383–1.283 0.701 Excludes 14 patients with a preoperative Nurick grade of 0 or 1. ySelected for multivariate analysis.

Spine www.spinejournal.com E707 Copyright © 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. CERVICAL SPINE Long-Term Outcomes After Cervical Corpectomy Sarkar and Rajshekhar

TABLE 5. Univariate and Multivariate Analysis of Factors Associated With Nurick Grade 0 or 1 at Last Follow-up (n ¼ 352) Univariate Analysis Multivariate Analysis Odds Ratio; 95% Odds Ratio; 95% Variable Confidence Interval P Confidence Interval P Age 50 yr 0.382; 0.241–0.605 <0.001 0.492; 0.301–0.804 0.005 Preoperative Nurick grade 4/5 0. 296; 0.181–0.485 <0.001 0.331; 0.192–0.573 <0.001 Symptom duration 1 year 0.449; 0.285–0.708 0.001 0.408; 0.251–0.665 <0.001 Diabetes mellitus 0.400; 0.112–1.432 0.159 Preoperative bladder dysfunction 0.562; 0.361–0.876 0.011 0.777; 0.473–1.279 0.321 Preoperative kyphotic alignment 0.700; 0.264–1.855 0.473 Presence of OPLL 0.744; 0.460–1.204 0.228 Single versus multilevel corpectomy 1.394; 0.878–2.213 0.159 Perioperative morbidity 0.470; 0.225–0.986 0.046 0.524; 0.238–1.152 0.108 Postoperative kyphotic alignment 0.784; 0.440–1.397 0.409 Selected for multivariate analysis. OPLL indicates ossification of posterior longitudinal ligament. cervical myelopathy. Japanese Orthopedic Association deterioration is important in counseling patients regarding (JOA) scores at 1 month and at final follow-up did not their expectations before surgery. change significantly in a group of 91 patients undergoing laminectomy for long-segment cervical canal stenosis fol- Predictors of Outcome lowed over a decade.15 In contrast, Kato et al report signifi- Advanced age, long symptom duration and poor preopera- cant late deterioration in postoperative (JOA) scores and tive myelopathy scores have been established by a number of recovery rates over serial follow-up in patients who under- studies as predictors of suboptimal postoperative recovery went laminectomy for OPLL followed for a mean period of in patients with CSM and OPLL.10,17,18,36–45 Age-related 14.1 years.29 Similar reports of decline in functional status loss of corticospinal tract, dorsal column, and gamma motor of patients over long-term follow-up after anterior decom- neurons coupled with significantly higher comorbidity pression have also been noted.10,17,30 scores may account for the association between age and Most of the patients in our study demonstrated sustained poor outcome.46–48 Furthermore, elderly patients often improvement in functional status on serial postoperative have more severe myelopathy at presentation.49–51 Notably, evaluations performed more than 2 years after surgery. patients in our study were on average at least a decade However, in the 175 patients followed serially, 31 younger that those in most published reports. This suggests (17.7%) deteriorated to their preoperative functional status that even middle-aged or ‘‘nonelderly’’ patients achieve less or worse. This is similar to the late deterioration rates favorable outcomes. Suri et al52 also made a similar obser- reported by other studies.16,17,29 Long preoperative symp- vation. Advanced age also remains a risk factor for devel- tom duration was associated with failure to maintain initial oping perioperative complications.21,53 Nevertheless, gains in functional status in our study, and may reflect a Nagashima et al54 report favorable outcomes in patients phenomenon of transient improvement in a background of ages 80 years or older and recommend early surgery in view chronic irreversible spinal cord dysfunction that finally of rapid disease progression in these patients. results in a late deterioration in neurological status. The poorer outcome seen in patients with a long duration Symptomatic adjacent segment disease remains a concern of symptoms and worse preoperative functional scores has in patients undergoing cervical fusion procedures, although been attributed to chronic and irreversible compressive distinguishing the consequences of this phenomenon from injury to the spinal cord. Most studies have reported signifi- natural age-related degenerative changes remains diffi- cantly worse outcomes in patients who present more than cult.10,31–33 A recent publication from our institution showed 1 year after symptom onset, and our results also indicate a 3.9% incidence of clinical adjacent segment pathology at that beyond this time period, functional recovery is limited. long-term follow-up.34 Other causes of late deterioration in The design of our study does not allow us to comment on the patients with CSM or OPLL are myriad, and include lumbar role of expectant or nonoperative treatment of patients with or thoracic canal stenosis, cerebrovascular accidents and hip CSM or OPLL. However, combined with the view that the or knee osteoarthritis.13,19,29,35 The failure of our study to underlying disease process usually progresses over time, this identify additional factors associated with late deterioration argues in favor of early and aggressive surgical decompres- may therefore be attributed to the etiology of the deteriora- sion in patients with significant myelopathy.14 tion itself, which in most cases appears to be unrelated to the Preoperative functional status scores may also act as a cervical spine. Nevertheless, information regarding late-onset surrogate marker for disease severity. Patients with a poor

E708 www.spinejournal.com June 2018 Copyright © 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. CERVICAL SPINE Long-Term Outcomes After Cervical Corpectomy Sarkar and Rajshekhar

preoperative functional status are also at a significantly and correction of the kyphosis might lead to better func- higher risk for perioperative morbidity.55 Nevertheless, this tional outcome after surgery. In a recent study employing should not dissuade surgeons from offering surgery to poor- anterior, posterior, and combined approaches, Shamji et al grade patients, because at least 75% of patients demonstrate reported poorer outcomes in patients with preoperative some degree of improvement and 60% achieve return-to- kyphosis, although correction of this deformity did not employment status.25 impact outcomes.62 Ferch et al have recommended correc- The presence of diabetes mellitus and the longitudinal tion of preoperative kyphosis to a target local sagittal angle extent of cord compression did not influence outcomes in of 0 to 4 degrees to enhance neurological recovery.63 our study. Some authors have reported poorer outcomes in patients with diabetes, but with strict perioperative glycemic Limitations of This Study control, diabetic patients may also achieve good functional Aside from its retrospective nature and the fact that data outcomes after surgery.56–58 Nakashima et al59 report simi- were lacking in a number of patients, this study suffers from lar 2-year outcomes in patients with OPLL versus those with other limitations. We were unable to assess the impact of other forms of degenerative cervical myelopathy. A primary cigarette smoking on outcomes because of difficulty in diagnosis of OPLL was also not associated with long-term quantifying the various forms of indigenous tobacco prod- outcomes in our study. Similarly, the number of levels ucts used by our population. Furthermore, patients in our decompressed was unrelated to outcome, as has also been study were overall younger and hence serious medical reported by previous authors.7,17 comorbidities were less common. This did not allow us to study the impact of coexisting illnesses on outcomes. Nev- Predictors of ‘‘Cure’’ ertheless, this study describes long-term outcomes and Although a minimum clinically important difference for the adverse prognostic factors in the largest single-surgeon Nurick grading system has not been documented, unlike the cohort of patients undergoing cervical corpectomy for JOA score, a 1-point increase in grade in the former would CSM and OPLL described in literature. correspond to a substantial improvement in walking ability, as per the design of this scale.21 In this regard, the best CONCLUSION possible outcome after surgery for degenerative cervical Uninstrumented cervical corpectomy achieves excellent out- spine disease would be a postoperative Nurick grade of 0 comes in patients with CSM and OPLL. Poor prognostic or 1 or its equivalent on other scales, which essentially factors include age 50 years or older, symptom duration of corresponds to complete resolution of symptoms related 1 year or more, and a baseline Nurick grade of 4 or higher. to myelopathy with no residual difficulty in walking. This Initial gains in functional status after surgery are maintained was achieved in approximately 36% of our patients. This in the majority of patients on long-term follow-up; however, outcome measure differs from simple assessment of change longer preoperative symptom duration is associated with a in functional status, as has been highlighted by prior pub- failure to sustain these early improvements. lications from our institution,25,60 and more recently by the AOSpine CSM studies.14,61 One third of patients in our study achieved this outcome. On multivariate analysis, age Key Points 50 years or older, symptom duration 12 months or more, Long-term outcome following corpectomy for CSM and poor preoperative Nurick grades were each individually or OPLL is good in more than 70% of patients. associated with failure to achieve a Nurick grade of 0 or 1 at Predictors of poor outcome included advanced last follow-up. These prognostic factors might be associated age, poor preoperative functional status, and long with irreversible or only partially reversible neural injury symptom duration. that accounts for persistent postoperative deficits. This Initial gains in functional status are maintained in group of patients will most likely be unable to return to a more than 75% patients at 10 years after surgery. ‘‘normal’’ functional status postoperatively, despite improv- ing significantly after surgery. Although this does not under- Longer preoperative symptom duration (>1 year) is associated with a failure to sustain early mine the benefits of surgical decompression, it is essential improvement. for patients to moderate their expectations before surgery.14

Significance of Sagittal Alignment Preoperative and postoperative cervical alignment was largely unrelated to outcomes in our cohort. A previous study from References 1. Thakar S, Christopher S, Rajshekhar V. Quality of life assessment our institution also found no impact of postoperative kyphotic after central corpectomy for cervical spondylotic myelopathy: change on short-term outcomes after one-level or two-level comparative evaluation of the 36-Item Short Form Health Survey corpectomy.24 Preservation or restoration of cervical lordosis and the World Health Organization Quality of Life–Bref: Clinical article. J Neurosurg Spine 2009;11:402–12. was also unrelated to long-term outcome in this study. 2. Baron EM, Young WF. Cervical spondylotic myelopathy: a brief However, other authors suggest that preoperative kypho- review of its pathophysiology, clinical course, and diagnosis. sis might be associated with worse postoperative outcome Neurosurgery 2007;60 (1 supp1 1):S35–41. Spine www.spinejournal.com E709 Copyright © 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. CERVICAL SPINE Long-Term Outcomes After Cervical Corpectomy Sarkar and Rajshekhar

3. Zhang L, Zeitoun D, Rangel A, et al. Preoperative evaluation of the 25. Rajshekhar V, Kumar GSS. Functional outcome after central cervical spondylotic myelopathy with flexion-extension magnetic corpectomy in poor-grade patients with cervical spondylotic mye- resonance imaging: about a prospective study of fifty patients. lopathy or ossified posterior longitudinal ligament. Neurosurgery Spine (Phila Pa 1976) 2011;36:E1134–9. 2005;56:1279–85. 4. Sampath P, Bendebba M, Davis JD, et al. Outcome of patients treated 26. Batzdorf U, Batzdorff A. Analysis of cervical spine curvature for cervical myelopathy: a prospective, multicenter study with inde- in patients with cervical spondylosis. Neurosurgery 1988;22: pendent clinical review. Spine (Phila Pa 1976) 2000;25:670–6. 827–36. 5. Matz PG, Anderson PA, Holly LT, et al. The natural history of 27. Kumar VG, Rea GL, Mervis LJ, et al. Cervical spondylotic cervical spondylotic myelopathy. J Neurosurg Spine 2009;11: myelopathy: functional and radiographic long-term outcome 104–11. after laminectomy and posterior fusion. Neurosurgery 1999;44: 6. Wang MY, Shah S, Green BA. Clinical outcomes following cervical 771–7. laminoplasty for 204 patients with cervical spondylotic myelopa- 28. Matsuoka T, Yamaura I, Kurosa Y, et al. Long-term results of the thy. Surg Neurol 2004;62:487–92. anterior floating method for cervical myelopathy caused by ossifi- 7. Fessler RG, Steck JC, Giovanini MA. Anterior cervical corpectomy cation of the posterior longitudinal ligament. Spine (Phila Pa 1976) for cervical spondylotic myelopathy. Neurosurgery 1998;43: 2001;26:241–8. 257–65. 29. Kato Y, Iwasaki M, Fuji T, et al. Long-term follow-up results of 8. Chagas H, Domingues F, Aversa A, et al. Cervical spondylotic laminectomy for cervical myelopathy caused by ossification of the myelopathy: 10 years of prospective outcome analysis of anterior posterior longitudinal ligament. J Neurosurg 1998;89:217–23. decompression and fusion. Surg Neurol 2005;64:S30–5. 30. Gao R, Yang L, Chen H, et al. Long term results of anterior 9. Chiles BW III, Leonard MA, Choudhri HF, et al. Cervical spon- corpectomy and fusion for cervical spondylotic myelopathy. PLoS dylotic myelopathy: patterns of neurological deficit and recovery One 2012;7:e34811. after anterior cervical decompression. Neurosurgery 1999;44: 31. Hilibrand AS, Carlson GD, Palumbo MA, et al. Radiculopathy and 762–9. myelopathy at segments adjacent to the site of a previous anterior 10. Emery SE, Bohlman HH, Bolesta MJ, et al. Anterior cervical cervical arthrodesis. J Bone Joint Surg Am 1999;81:519–28. decompression and arthrodesis for the treatment of cervical spon- 32. Hilibrand AS, Robbins M. Adjacent segment degeneration and dylotic myelopathy. Two to seventeen-year follow-up. J Bone Joint adjacent segment disease: the consequences of spinal fusion? Spine Surg Am 1998;80:941–51. J 2004;4:S190–4. 11. Gok B, Sciubba DM, McLoughlin GS, et al. Surgical treatment of 33. Kulkarni V, Rajshekhar V, Raghuram L. Accelerated spondylotic cervical spondylotic myelopathy with anterior compression: a changes adjacent to the fused segment following central cervical review of 67 cases. J Neurosurg Spine 2008;9:152–7. corpectomy: magnetic resonance imaging study evidence. J Neuro- 12. Lee SE, Chung CK, Jahng T-A, et al. Long-term outcome of surg 2004;100:2–6. laminectomy for cervical ossification of the posterior longitudinal 34. Vedantam A, Rajshekhar V. Clinical adjacent-segment pathology ligament: Clinical article. J Neurosurg Spine 2013;18:465–71. after central corpectomy for cervical spondylotic myelopathy: 13. Sakaura H, Hosono N, Mukai Y, et al. Medium-term outcomes of incidence and risk factors. Neurosurg Focus 2016;40:E12. C3-6 laminoplasty for cervical myelopathy: a prospective study 35. Chiba K, Ogawa Y, Ishii K, et al. Long-term results of expansive with a minimum 5-year follow-up. Eur Spine J 2011;20:928–33. open-door laminoplasty for cervical myelopathy—average 14-year 14. Tetreault L, Kopjar B, Coˆ te´ P, et al. A clinical prediction rule for follow-up study. Spine (Phila Pa 1976) 2006;31:2998–3005. functional outcomes in patients undergoing surgery for degenera- 36. King JT Jr, Moossy JJ, Tsevat J, et al. Multimodal assessment after tive cervical myelopathy. J Bone Joint Surg Am 2015;97:2038–46. surgery for cervical spondylotic myelopathy. J Neurosurg Spine 15. Li Z, Xue Y, He D, et al. Extensive laminectomy for multilevel 2005;2:526–34. cervical stenosis with ligamentum flavum hypertrophy: more than 37. Holly LT, Matz PG, Anderson PA, et al. Clinical prognostic 10 years follow-up. Eur Spine J 2015;24:1605–12. indicators of surgical outcome in cervical spondylotic myelopathy. 16. Al-Tamimi YZ, Guilfoyle M, Seeley H, et al. Measurement of long- J Neurosurg Spine 2009;11:112–8. term outcome in patients with cervical spondylotic myelopathy 38. Morio Y, Teshima R, Nagashima H, et al. Correlation between treated surgically. Eur Spine J 2013;22:2552–7. operative outcomes of cervical compression myelopathy and MRI 17. Ebersold MJ, Pare MC, Quast LM. Surgical treatment for cervical of the spinal cord. Spine (Phila Pa 1976) 2001;26:1238–45. spondylitic myelopathy. J Neurosurg 1995;82:745–51. 39. Chibbaro S, Benvenuti L, Carnesecchi S, et al. Anterior cervical 18. Kiris T, Kilinc¸er C. Cervical spondylotic myelopathy treated by oblique corpectomy for cervical spondylotic myelopathy: experience and corpectomy: a prospective study. Neurosurgery 2008;62:674–82. surgical results in a series of 70 consecutive patients. J Clin Neuro- 19. Iwasaki M, Kawaguchi Y, Kimura T, et al. Long-term results of sci 2006;13:233–8. expansive laminoplasty for ossification of the posterior longitudi- 40. Suzuki A, Misawa H, Simogata M, et al. Recovery process follow- nal ligament of the cervical spine: more than 10 years follow up. J ing cervical laminoplasty in patients with cervical compression Neurosurg 2002;96:180–9. myelopathy: prospective cohort study. Spine (Phila Pa 1976) 20. Machino M, Yukawa Y, Hida T, et al. Modified double-door 2009;34:2874–9. laminoplasty in managing multilevel cervical spondylotic myelop- 41. Alafifi T, Kern R, Fehlings M. Clinical and MRI predictors of athy: surgical outcome in 520 patients and technique description. J outcome after surgical intervention for cervical spondylotic mye- Spinal Disord Tech 2013;26:135–40. lopathy. J Neuroimaging 2007;17:315–22. 21. Furlan JC, Kalsi-Ryan S, Kailaya-Vasan A, et al. Functional and 42. Zhang YZ, Shen Y, Wang LF, et al. Magnetic resonance T2 image clinical outcomes following surgical treatment in patients with signal intensity ratio and clinical manifestation predict prognosis cervical spondylotic myelopathy: a prospective study of 81 cases: after surgical intervention for cervical spondylotic myelopathy. clinical article. J Neurosurg Spine 2011;14:348–55. Spine (Phila Pa 1976) 2010;35:E396–9. 22. Kumar GSS, Rajshekhar V. Acute graft extrusion following central 43. Koyanagi T, Hirabayashi K, Satomi K, et al. Predictability of corpectomy in patients with cervical spondylotic myelopathy and operative results of cervical compression myelopathy based on ossified posterior longitudinal ligament. J Clin Neurosc 2009;16: preoperative computed tomographic myelography. Spine (Phila Pa 373–7. 1976) 1993;18:1958–63. 23. Sarkar S, Nair BR, Rajshekhar V. Complications following central 44. Hamanishi C, Tanaka S. Bilateral multilevel laminectomy with or corpectomy in 468 consecutive patients with degenerative cervical without posterolateral fusion for cervical spondylotic myelopathy: spine disease. Neurosurg Focus 2016;40:E10. relationship to type of onset and time until operation. J Neurosurg 24. Rajshekhar V, Arunkumar MJ, Kumar SS. Changes in cervical spine 1996;85:447–51. curvature after uninstrumented one-and two-level corpectomy in 45. Satomi K, Ogawa J, Ishii Y, et al. Short-term complications and patients with spondylotic myelopathy. Neurosurgery 2003;52: long-term results of expansive open-door laminoplasty for cervical 799–805. stenotic myelopathy. Spine J 2001;1:26–30.

E710 www.spinejournal.com June 2018 Copyright © 2018 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited. CERVICAL SPINE Long-Term Outcomes After Cervical Corpectomy Sarkar and Rajshekhar

46. Terao S, Sobue G, Hashizume Y, et al. Age-related changes in 55. Boakye M, Patil CG, Ho C, et al. Cervical corpectomy: compli- human spinal ventral horn cells with special reference to the loss of cations and outcomes. Neurosurgery 2008;63:295–302. small neurons in the intermediate zone: a quantitative analysis. 56. Arnold PM, Fehlings MG, Kopjar B, et al. Mild diabetes is not a Acta Neuropathol 1996;92:109–14. contraindication for surgical decompression in cervical spondy- 47. Terao S, Sobue G, Hashizume Y, et al. Age-related changes of the lotic myelopathy: results of the AOSpine North America multi- myelinated fibers in the human corticospinal tract: a quantitative center prospective study (CSM). Spine J 2014;14:65–72. analysis. Acta Neuropathol 1994;88:137–42. 57. Machino M, Yukawa Y, Ito K, et al. Impact of diabetes on the 48. Nakashima H, Tetreault LA, Nagoshi N, et al. Does age affect outcomes of cervical laminoplasty: a prospective cohort study of surgical outcomes in patients with degenerative cervical myelopa- more than 500 patients with cervical spondylotic myelopathy. thy? Results from the prospective multicenter AOSpine Interna- Spine (Phila Pa 1976) 2014;39:220–7. tional study on 479 patients. J Neurol Neurosurg Psychiatry 58. Kusin DJ, Ahn UM, Ahn NU. The influence of diabetes on surgical 2016;87:734–40. outcomes in cervical myelopathy. Spine (Phila Pa 1976) 2016;41: 49. Machino M, Yukawa Y, Hida T, et al. Can elderly patients recover 1436–40. adequately after laminoplasty? A comparative study of 520 59. Nakashima H, Tetreault L, Kato S, et al. Prediction of outcome patients with cervical spondylotic myelopathy. Spine (Phila Pa following surgical treatment of cervical myelopathy based on 1976) 2012;37:667–71. features of ossification of the posterior longitudinal ligament: a 50. Matsuda Y, Shibata T, Oki S, et al. Outcomes of surgical treatment systematic review. JBJS Rev 2017;5. for cervical myelopathy in patients more than 75 years of age. 60. Vedantam A, Jonathan A, Rajshekhar V. Association of magnetic Spine (Phila Pa 1976) 1999;24:529–34. resonance imaging signal changes and outcome prediction after 51. Yamazaki T, Yanaka K, Sato H, et al. Cervical spondylotic surgery for cervical spondylotic myelopathy: clinical article. myelopathy: surgical results and factors affecting outcome with J Neurosurg Spine 2011;15:660–6. special reference to age differences. Neurosurgery 2003;52: 61. Tetreault LA, Kopjar B, Vaccaro A, et al. A clinical prediction 122–6. model to determine outcomes in patients with cervical spondylotic 52. Suri A, Chabbra RPS, Mehta VS, et al. Effect of intramedullary myelopathy undergoing surgical treatment: data from the prospec- signal changes on the surgical outcome of patients with cervical tive, multi-center AOSpine North America study. J Bone Joint Surg spondylotic myelopathy. Spine J 2003;3:33–45. Am 2013;95:1659–66. 53. Holly LT, Moftakhar P, Khoo LT, et al. Surgical outcomes of 62. Shamji MF, Mohanty C, Massicotte EM, et al. The association of elderly patients with cervical spondylotic myelopathy. Surg Neurol cervical spine alignment with neurologic recovery in a prospective 2008;69:233–40. cohort of patients with surgical myelopathy: analysis of a series of 54. Nagashima H, Dokai T, Hashiguchi H, et al. Clinical features and 124 cases. World Neurosurg 2016;86:112–9. surgical outcomes of cervical spondylotic myelopathy in patients 63. Ferch RD, Shad A, Cadoux-Hudson TA, et al. Anterior correction aged 80 years or older: a multi-center retrospective study. Eur of cervical kyphotic deformity: effects on myelopathy, neck pain, Spine J 2011;20:240–6. and sagittal alignment. J Neurosurg Spine 2004;100:13–9.