Portal de Periódicos da CAPES

Surgery vs Conservative Care for Cervical Spondylotic Myelopathy

2015; Lippincott Williams & Wilkins; Volume: 62; Issue: Supplement 1 Linguagem: Inglês

10.1227/neu.0000000000000816

1524-4040

George M. Ghobrial, James S. Harrop,

Spinal Hematomas and Complications

Throughout history, there has been a perception that with increasing age comes a tendency in progression to frailty and loss of neurological function. In 1928, Stookey1 illustrated that compression of the spinal cord by cartilaginous nodes from degenerative disk material may be a cause of neurological dysfunction known as cervical myelopathy. The idea of cervical myelopathy and neurological dysfunction resulting from cervical cord compression was further proclaimed in the landmark article by Brain et al2 in 1952. However, the natural history and progressive disorder of this disease entity were poorly understood; even today, confusion about those topics remains. In 1963, Lees and Turner3 prospectively reviewed 22 patients for >10 years, observing periods of exacerbation, recovery, and quiescence in symptomology. They noted, "Long periods of nonprogressive disability are the rule, and [a] progressive deteriorating course is an exception." They then concluded that in treatment a "conservative approach should be the rule."3 The benefits of nonoperative treatment in patients without progressive deterioration have been illustrated throughout the years in numerous publications (Table 1).4-12TABLE: Selected Studies Evaluating Nonsurgical Treatment for Cervical Spondylotic MyelopathyaHere, we review the literature highlighting the benefits of nonoperative treatment for cervical spondylotic myelopathy (CSM). METHODS Inclusion Criteria An electronic database search (PubMed; National Library of Medicine) was performed for English-language manuscripts published between January 1, 1990 and January 24, 2015, and containing the headings "cervical spondylotic myelopathy" and "cervical myelopathy." The results were then evaluated for significance in terms of whether they addressed the nonsurgical treatment of CSM or compared nonsurgical and surgical management of CSM in patients ≥19 years of age. Only studies for which both abstracts and full text were available were reviewed for relevance. Relevant abstracts were reviewed first, followed by selected full manuscripts. Bibliographies were then cross-referenced for additional studies of relevance. Exclusion Criteria Case reports and commentaries were excluded. Studies that did not include any objective outcomes measures were excluded. Studies that included 6 months using conservative treatment algorithms consisting of bracing and activity restriction. They measured patients' Japanese Orthopedic Association (JOA) scores, patient satisfaction, and magnetic resonance imaging results. Of their 27 patients, 10 developed some degree of neurological deterioration and were treated surgically. The conservative treatment group had a score of 14.9 at 3-month time compared to the surgical group which worsened at 3 months with a score of 12.9.8 Therefore, 3-month JOA scores were superior for patients who received conservative treatment compared with operative treatment. The authors concluded that conservative treatment was an effective treatment option for mild cervical myelopathy caused by cervical disk herniation.8 In an additional study, Nakamura et al4 followed up 66 patients for 3 months, retrospectively reviewing traction and collar treatment. Their outcome measure was also based on the JOA grading system; they noted an overall improvement rate to "no disability" in approximately 30% of patients.4 Matsumoto et al6 reported results from 52 patients with cervical myelopathy who had a JOA score >9. The authors examined radiographic characteristics of cord injury such as T2 signal hyperintensity, with the underlying hypothesis that patients with T2 signal abnormality would have the greatest degree of compromise with resultant neurological deficits. These patients were treated with a conservative algorithm consisting of bracing and monthly outpatient evaluations. Of the 52 patients, 10 deteriorated and underwent surgical treatment. The authors noted that the average JOA score increase with nonoperative management was 0.4 points. The groups were then segregated by underlying compressive pathology: spondylosis, disk herniation, and ossification of the posterior longitudinal ligament, with 62, 75, and 64 patients in each group, respectively. Interestingly, they noted that both patients with spondylosis and patients with disk herniation showed increases in their JOA score and had the highest degree of satisfaction with their outcome. This was in contrast to patients with underlying posterior longitudinal ligament, who demonstrated an average deterioration in their JOA score of 0.5 and the lowest average satisfaction score of only 45%.6 Thus, the authors concluded, "Increased signal intensity was not related to poor outcome of conservative treatment or severity of myelopathy in patients with mild cervical myelopathy." In 2007, Shimomura and colleagues10 prospectively enrolled 70 patients with mild CSM treated nonsurgically and defined as having a JOA score of ≥13.10 Patients were followed up from 1990 to 2003, with an average follow-up rate of 80% (56 patients). Overall, patients' JOA scores showed no statistically significant decline with nonsurgical care. Interestingly, 10 of 33 patients with magnetic resonance imaging demonstrating circumferential spinal cord encroachment resulting from ligamentous hypertrophy, cervical spondylosis, and degenerative disk arthropathy showed a significant radiologic deterioration but not clinical. The authors concluded that although it is reasonable and effective to treat patients with mild CSM nonsurgically (except in the case of circumferential spinal cord compression), the higher risk of deterioration—and the potential for surgical intervention if such deterioration should occur—should be discussed with the patient. In 2010, Nikolaidis et al11 published an evidence-based review and evaluation of the effects of operative treatment and timing for the symptomatic management of CSM through the Cochrane collaboration. This review specifically focused on Level 1 evidence comparing operative and nonoperative treatment for CSM. The article highlighted the paucity of high-quality studies in the literature comparing the 2 treatment options; only 2 prospective randomized studies were included in the report. Of those 2 studies, only 1 report adequately assessed surgical vs nonsurgical treatment for CSM.5 This study, by Kadanka et al,5 was a 2-year prospective randomized study comparing conservative and operative treatment of mild and moderate CSM. The study included 48 patients with a modified JOA score >12. The authors analyzed clinical outcomes, including modified JOA score, recovery rate, and results from a timed 10-m walk test. They analyzed qualitative results with a daily and subjective assessment score of function. They found no significant differences in modified JOA scores, the recovery rates, or the timed 10-m walk tests in their 2-year follow-up.5 There was a slight decline in scores of daily activities and subjective evaluation in the surgical treatment group. In summary, the authors concluded that there were significant differences favoring nonoperative treatment based on the score of daily activities recorded by video at 24 months and thus concluded, "Surgical treatment of mild and moderate forms of SCM in the present study design, comprising the patients with no or very slow, insidious progression and a relatively long duration of symptoms, did not show better results than conservative treatment over the 2-year follow-up."5 Kadanka and colleagues9 then continued to follow up this similar group in a 3-year follow-up to their prospective randomized study. Their objective was to compare conservative and operative treatment of mild to moderate nonprogressive and slowly progressive forms of spondylotic myelopathies. The authors powered the study appropriately, and 68 patients with modified JOA scores >12 were randomized for analysis. The nonoperative treatment group consisted of 35 patients, and the operative treatment group consisted of 33 patients. Again, the outcome measures were modified JOA score, timed 10-m walk, scores of daily activities recorded by video, and subjective assessments of patients at 6, 12, 24, and 36 months. There were no significant demographic differences in the groups. There was no significant deterioration in numerous outcomes measures, including no significant difference in modified JOA scores. However, there were a decrease in the self-evaluation scores of the surgical patients and a decreased daily activity score in the conservative treatment group. However, the timed 10-m walk test results favored the conservative group. Despite mixed results, the authors concluded by noting that their 3-year follow-up study did not show, on average, that surgery was superior to conservative treatment for mild or moderate forms of CSM.9 It is interesting to note in this cohort that 2 patients in the conservative treatment group went on to have surgical treatment. The patients had stable examinations, and the decision to undergo surgery was attributed by the authors as being due to a psychological reason. Therefore, in an intention-to-treat analysis, these patients were considered to be in the conservative group and deteriorated after surgical treatment. As a consequence of the mixed evidence and statistically similar results in both groups, the conclusion of the Cochrane evidence-based analysis was as follows: "The available small randomized trials do not provide reliable evidence on the effects of surgery on cervical spondylotic radiculopathy or myelopathy. It is not clear whether the short-term risks of surgery are all set by any long-term benefits."11 Kadaňka et al12 then followed up this group even further in a 10-year follow-up prospective, randomized study. As a result of loss of patients in the follow-up study (17 patients of the original population of 64 died), there were 25 patients in the nonoperative group and 22 patients in the surgical cohort. Again, the clinical outcomes were the same (modified JOA, timed 10-m walk, scores of daily activities, and subjective assessment of patients). At the 10-year-follow-up, there was again no significant difference between groups in modified JOA score, subjective evaluation of the patients based on video recordings, or ability to walk in a timed 10-m test, thus showing no benefit of surgical treatment in patients who were not clinically deteriorating. Providing support in favor of surgical intervention, Sampath et al13 reviewed outcomes of patients treated for cervical myelopathy based on a prospective, nonrandomized cohort group through the Cervical Spine Research Society. The reporting group included a large cohort of surgeons (n = 41); unfortunately, there was some degree of selection bias in enrollment in that only 62 of 503 patients were enrolled (approximately 12%). Of these, only 69% of patients returned to follow-up; at 11.2 months of follow-up, there were 20 surgically treated patients and 23 patients receiving medical treatment. Surgical patients showed significant improvement in functional status and overall outcome, with improvement in neurological symptoms. However, the main drawback to this study was the high degree of selection bias, which most likely resulted from the decision to operate urgently on patients with a precipitous deterioration in function. Consequently, the surgical treatment cohort comprised a different population than the nonoperative cohort, which was mainly nondeteriorating patients with a gradual clinical course. Surgical treatment is unfortunately an invasive treatment option, and although numerous patients with cervical myelopathy have an extreme benefit with surgical treatment, the treatment is not without potential morbidity. Yadla et al14 reviewed 75 patients with degenerative cervical spine and prospectively collected data on their morbidities, noting that 44% patients had early morbidities, 37.3% had minor complications, and 16.0% had major complications. In addition, there is the potential with the surgical treatment of cervical myelopathy that patients develop degeneration at mobile segments of the spine rostral and caudal to the fused cervical segments. Hilibrand et al15 calculated an annual incidence of 2.9% for symptomatic adjacent-level degeneration, thus potentially resulting in the need for further treatment. Yoshimatsu et al7 retrospectively reviewed 69 cases and categorized patients into 3 main groups: improvement, no change, and exacerbation. They noted that the multivariate analysis showed a significant correlation between clinical outcome, disease duration, and the presence of rigorous conservative treatment. They also noted that conservative treatment was considered to be effective if performed in selected groups of patients. However, they concluded, "Timely surgical intervention is considered to be important if the symptoms show no change or exacerbation with conservative treatment."7 CONCLUSION Studies suggest that consideration for surgical or nonsurgical management of cervical myelopathy should be preceded by an understanding of the patient's disease course, which is the greatest predictor of response to therapy. CSM is a very heterogeneous disease that unfortunately we still do not fully understand. Cervical stenosis itself may be the cause in a large proportion of patients with myelopathy but does not represent the entire pathogenesis of clinical deteriorations. Therefore, in selected individuals, serial observation and management is considered a viable and potential treatment option. Disclosure The authors have no personal, financial, or institutional interest in any of the drugs, materials, or devices described in this article.