Introduction
Cervical spondylotic amyotrophy (CSA) was first reported by Sobue et al. [1] in 1975 as a subtype of cervical spondylotic myelopathy (CSM). Crandall et al. [2] have indicated that approximately ≤7% of patients with CSM present with minimal sensation loss. In a study by Imajo et al. [3], preoperative factors predicting fair outcomes following surgery in 54 patients with proximal-type CSA (PCSA) using electrophysiological examination were reported [3]. Eight patients (14.8%) had C5 palsy. Moreover, Iizuka et al. [4] have reported that the prevalence of poor outcomes was 7.7% (1/13). Approximately 70% of patients with C5 palsy recover completely without treatment [5–7]. The mean time to full recovery is 4–5 months [5]. In a study by Imajo et al. [3], 12.5% of patients with C5 palsy could fully recover after 2 years. Postoperative C5 palsy was a significant issue following surgery for patients with PCSA because the prognosis for these patients was not good compared with that for patients with other compressive cervical myelopathies (CCMs) [3]. Therefore, preventing postoperative C5 palsy is important. The mean ratio of the compound muscle action potential (CMAP) amplitudes in the affected side to those in the normal side recorded in the deltoid and biceps brachii muscles could be used as a reliable predictive factor for a fair surgical outcome [3]. However, preoperative factors associated with poor outcomes remained unclear. Patients with poor outcomes should have achieved excellent outcomes as a result of the electrophysiological findings [3]. Ossification of the posterior longitudinal ligament (OPLL), cervical open-door laminoplasty together with posterior instrumented fusion, the width of the intervertebral foramen (WIF), a larger superior articular process (SAP), and a posterior shift of the spinal cord after laminoplasty were significant risk factors for postoperative C5 palsy for patients with other CCMs [6–10]. Factors associated with poor outcomes include long symptom duration, older age, and greater preoperative muscle weakness for patients with CSA [10–12]. This study was designed to investigate the factors associated with poor outcomes using radiological and neurological findings, instead of electrophysiological findings.
Materials and Methods
In this study, 68 patients with PCSA who underwent surgical treatment of the cervical spine between April 1997 and September 2017 at Yamaguchi University Hospital were enrolled. Four patients underwent cervical surgery and multi-muscle transfers simultaneously, two had severe muscle atrophy bilaterally, and two had coexisting rotator cuff tears. Therefore, we excluded these eight patients. Finally, this retrospective study enrolled 60 patients. We collected data on age, hypertension, diabetes mellitus (DM), dialysis, corticosteroid use, Parkinson’s disease, surgical approach, the duration of clinical history (DCH), the duration of postoperative follow-up (DPF), and radiological findings. The mean age of the patients was 61.3±9.8 years. An anterior approach was the preferred method of treatment for multilevel cervical myelopathies when less than three surgical segments were involved, regardless of the alignment of the cervical spine. A posterior approach was the preferred method of treatment for multilevel cervical myelopathies when three or more surgical segments were involved. When the alignment of the cervical spine was kyphotic (see evaluation of radiological findings: α <−5°), since 2014, we performed fixation with instrumentation in addition to laminoplasty. All 60 patients showed severe unilateral atrophy of the shoulder girdle muscles; however, the intrinsic muscles remained intact. When patients had muscle atrophy of the forearm and/or hand, in addition to the shoulder girdle muscles, we suspected amyotrophic lateral sclerosis (ALS) and other motor neuron diseases. To differentiate motor neuron diseases, such as ALS and PCSA, we performed preoperative electromyography on 35 patients. When ALS and other motor neuron diseases were suspected, we consulted neurologists preoperatively. Along with the neurologists, we determined that patients with PCSA did not have these diseases preoperatively.
This study was approved by the Institutional Review Board of Yamaguchi University Hospital (IRB approval no., H30–170). Informed consent was obtained from all individual participants included in the study.
1. Evaluation of radiological findings (Table 1)
The radiological findings on plain radiographs, computed tomography (CT), and magnetic resonance imaging (MRI) were evaluated. The cervical lordotic angles (cervical curvature, α; Cobb angle, β), K line, C2–C7 sagittal vertical axis (SVA), T1 slope, and diameter of the canal at the C5 vertebral level (C5 canal) were assessed on a lateral radiograph in the neutral position (Fig. 1A, B), and the intervertebral range at the C4–C5 (γ) and C5–C6 (δ) intervertebral levels was assessed using lateral radiographs in the flexion and extension positions [13]. The alignment of the cervical spine was classified as lordosis (>5°), straight (−5°≤ α ≤5°), sigmoid, or kyphosis (α <−5°). CT was used to assess the WIF at the C4–C5 intervertebral levels in the axial view and the anterior protrusion of the C5 SAP at the C5 vertebral level (Fig. 2A, B). MRI was used to assess the number of levels of compression (NLC), the presence of a high-intensity area (HIA) in the spinal cord, and the distance between the posterior edge of the C4–C5 disks and the anterior aspect of the spinal cord (D) in the T2-weighted midsagittal view (Fig. 3). With D, we evaluated 27 patients who underwent surgery using the posterior approach. The difference between the preoperative and follow-up values for C2–C7 SVA and T1 slope was also investigated, and this result was defined as the Δvalue for each parameter.
2. Assessment of muscle strength
In this study, C5 palsy was defined as a motor decline of the deltoid muscle function by at least one level in a standard manual muscle testing (MMT) within a 6-week postoperative period [8]. Preoperative and postoperative strengths of the most atrophic muscles were evaluated using MMT. Improvements in strength were classified as excellent (five grades recovered), good (more than one grade recovered), fair (no improvement by MMT), or poor (worsened by MMT), which were assessed using MMT. We classified surgical outcomes into two groups. Patients with poor outcomes were classified as group 1 and those with excellent, good, and fair outcomes were classified as group 2. Four surgeons who were certified by the Japanese Society for Spine Surgery and Related Research evaluated the MMT results at the first medical examination and the final follow-up. Two of the four surgeons evaluated one patient.
3. Evaluation of neurological findings
We examined the strength of the deltoid and biceps brachii muscles, forearm supination, and grip strength on the affected and healthy sides and the presence of sensory disturbance on the radial side in the affected upper extremity.
4. Statistical analysis
Statistical analyses were performed using StatFlex ver. 7.0 (Artech Co. Ltd., Osaka, Japan; http://www.statflex.net/), a general-purpose statistical software.
Parameters examined as possibly associated with the surgical outcomes, included surgical approaches (anterior approach versus posterior approach), K line, age, DCH, DPF, alignment, α, β, preoperative and postoperative C2–C7 SVA, ΔC2–C7 SVA, preoperative and postoperative T1 slope, ΔT1 slope, range of motion at C4–C5 and C5–C6 intervertebral levels, C5 canal, SAP and WIF at C4–C5 intervertebral levels on the bilateral side on axial CT, HIA, NLC, and D on T2-weighted sagittal MRI. Univariate statistical analyses were performed using the Mann-Whitney U-test to compare numerical parameters, and the chi-square test was used to compare categorical parameters. To identify the factors associated with poor and fair/good/excellent outcomes, multivariate logistic regression analysis (MLRA) was performed using a forward stepwise parameter selection procedure by setting p<0.1 as a criterion for inclusion in the regression model. The probability of patients belonging to group 1 (p-value) was computed from the following formula using the partial regression coefficients (β0, β1, β2, … βk) for explanatory variables (X1, X2, ... Xk), where k represents the number of explanatory parameters included in the regression model.
In this study, a dummy variable indicating the occurrence of poor outcomes was set as the objective variable, and all aforementioned parameters were set as the explanatory variables.
Results
The prevalence of C5 palsy was 17% (10/60). Postoperative C5 palsy was ipsilateral to the preoperative PCSA symptoms in nine of 10 patients. Paralysis was noted after a mean duration of 5.1 days (range, 0–18 days) postoperatively. Six patients underwent anterior decompression with spinal fusion (C4–C6), and four patients underwent laminoplasty (C3–C7) with foraminotomy (two patients: C5–C6 and two patients: C4–C5 and C5–C6). No patient underwent fixation with instrumentation. One patient had full recovery after 2 years and was categorized into group 2. However, nine patients with C5 palsy did not show adequate improvements even after ≥2 years postoperatively. These patients were categorized into group 1. The characteristics and neurological findings of the patients in groups 1 and 2 are shown in Table 2. No significant difference in preoperative neurological findings was observed between groups 1 and 2. However, significant differences in MMT of the deltoid and biceps brachii muscles, forearm supination on the affected side, and sensory disturbance in postoperative neurological findings (p<0.05) were noted between groups 1 and 2. No significant difference in grip strength on the healthy and affected sides was observed between groups 1 and 2. Furthermore, 28 patients (47%) presented with at least one comorbidity, the most common being hypertension (10 patients, 17%) and DM (six patients, 10%). In group 1, one patient had hypertension and one patient had DM.
1. Association between poor outcome and factors
We used univariate analyses to compare the aforementioned parameters between the two groups. Postoperative C2–C7 SVA, ΔC2–C7 SVA, and NLC differed significantly between groups 1 and 2 (p<0.05) (Table 3). ΔC2–C7 SVA and NLC differed significantly between groups 1 and 2 in the anterior approach (p<0.05) (Table 3). Ten patients had the K line (−) in this study. Moreover, seven patients had the K line (−) in the anterior approach and three patients in the posterior approach (Tables 3, 4). Only one patient in group 1 who had the K line (−). No significant difference in the K line (−) was observed between groups 1 and 2 (Table 3).
Postoperative C2–C7 SVA and ΔC2–C7 SVA significantly differed between groups 1 and 2 in the posterior approach (p<0.05) (Table 4). Regarding NLC, one patient had spinal cord compression from C3–C4 to C6–C7 at four levels, four patients from C3–C4 to C5–C6 at three levels, one patient from C3–C4 to C4–C5 at two levels, two patients from C4–C5 to C6–C7 at three levels, and one patient from C4–C5 to C5–C6 at two levels in group 1.
2. Factors in group 1
In the MLRA, the dummy variable representing group 1 was set as the objective variable, and all factors were set as the explanatory variables. Using the forward stepwise method, group 1 had higher NLC and ΔC2–C7 SVA than group 2 (p=0.015; odds ratio [OR], 5.758; 95% confidence interval [CI], 1.397–23.726 for a change of 10% and p=0.048; OR, 1.068; 95% CI, 0.992–1.141 for a change of 10%, respectively) (Table 5).
Discussion
The prevalence of C5 palsy was 17% (10/60). It was 18% and 15% in the anterior and posterior approaches, respectively. The average reported prevalence of C5 palsy following surgery for CCM for the anterior approach is 6.7% (range, 1.6%–9.1%) [7]. For the posterior approach, it is 5.8% [14]. Imajo et al. [15] have reported that the prevalence of C5 palsy was 4.6% in 219 patients who underwent anterior decompression with fusion (ADF) for cervical degenerative disease. Kalisvaart et al. [16] have reported that the prevalence of C5 palsy was 2.8%–12.1% in patients with CSM, 2.1%–14% in those with OPLL, and 0%–3% in patients with cervical disk herniation. The prevalence of C5 palsy in patients with PCSA was higher than that in patients with other diseases. Imajo et al. [15] have reported that the prognosis was good for patients with a CMAP amplitude of the deltoid muscle on the C5 palsy side that exceeded 1.5 mV. Even though patients with PCSA had a CMAP amplitude that exceeded 1.5 mV for the deltoid on the C5 palsy side, their deltoid function had not recovered. However, no difference in postoperative neurological results between patients with PCSA and those with other CCMs [15].
Factors associated with group 1 were ΔC2–C7 SVA and NLC. NLC was a preoperative factor. Thus, avoiding an increase in ΔC2–C7 SVA is important for surgeons. Rao et al. [17] have reported that the preoperative T1 slope–β had significant correlations with T1 slope, β, and preoperative C2–C7 SVA. ΔC2–C7 SVA, Δβ, and ΔT1 slope–β were correlated with surgical outcomes [17]. Sakai et al. [18] have reported the following important findings: (1) the postoperative α and cervical sagittal balance after ADF were maintained; however, those after laminoplasty worsened, especially in patients with a preoperative cervical sagittal imbalance; (2) the postoperative decrease in α was correlated with the preoperative center of gravity of the head—C7 SVA in laminoplasty, but not in ADF. Postoperative α was maintained; however, C2–C7 SVA was not maintained in this study (Table 3). Sakaura et al. [19] have reported that an increase in postoperative C1–C7 SVA was associated with older patients with CSM (≥75 years old) and operative levels. The patients were divided into two groups according to the most caudal level of laminoplasty (C6 group [C3–C6] and C7 group [C3–C7]). The C6 group had significantly shorter postoperative C1–C7 SVA than the C7 group. In this study, no significant age differences were observed between groups 1 and 2 (Table 4). Recently, various minimally invasive posterior decompression laminoplasty procedures, such as skip laminoplasty and selective laminoplasty, have been reported [20]. Shiraishi et al. [21] have shown a cervical curvature index after skip laminectomy that was superior to that after the conventional procedure. Therefore, we considered that selective laminoplasty was better than laminoplasty of the C6 and C7 groups. Some authors have reported that the surgical outcomes of selective laminoplasty (one above and one below the affected intervertebral level) for patients with CSM were almost the same as those of laminoplasty from C3 to C7 [22,23]. The ΔC2–C7 SVA after selective laminoplasty was significantly smaller than that after laminoplasty from C3 to C7 [24]. Selective laminoplasty for patients with PCSA may be important to prevent C5 palsy.
In this study, a larger SAP, WIF narrowing, and a posterior shift of the spinal cord were not associated with C5 palsy (Tables 3, 4). NLC was associated with poor outcomes in this study (Table 5). Hashimoto et al. [7] have reported that preexisting asymptomatic damage of the anterior horn cells at the C3–C4 and C4–C5 levels may contribute to developing motor weakness along with nerve root lesions that occur after ASF. Kaneko et al. [25] have reported the pathology of C5 palsy. The C5 nerve root was innervated predominantly for the deltoid compared with those at other cervical intervertebral lesions, probably due to the segmental gray matter lesions of the C6 spinal cord in patients with cervical myelopathy at the C4–C5 level. The weakness of the deltoid muscle must be clinically obvious if postoperative C5 radiculopathy has occurred in patients whose C5 nerve root predominantly innervates the deltoid muscles [25]. Similar to patients with cervical myelopathy at the C4–C5 level, the weakness of the deltoid muscle must be clinically obvious if postoperative C5 radiculopathy has occurred in patients with cervical myelopathy at the C3–C4 level. We regarded this pathology as a sure indication of PCSA. One patient with spinal cord compression at only the C4–C5 level had complete recovery 2 years postoperatively. All nine patients in group 1 had spinal cord compression from the anterior side at the C4–C5 intervertebral level. Therefore, they had compression of the C6 anterior horn and C5 root. In group 1, eight patients had spinal cord compression at the C5–C6 intervertebral level and six patients had spinal cord compression at the C3–C4 intervertebral level along with spinal cord compression at the C4–C5 intervertebral level. The former had compression of the C6 root, and the latter had compression of the C5 anterior horn in addition to compression of the C6 anterior horn and C5 root. The deltoid and biceps brachii muscles are innervated by the C5 and C6 anterior horns. Patients with lesions in the C5 anterior horn and/or C6 root in addition to compression of the C6 anterior horn and C5 root may have C5 palsy.
This study has few limitations that should be kept in mind when interpreting the results. The sample size was relatively small. The study may have been underpowered to identify any significant difference in surgical outcomes. However, to the best of our knowledge, this is the first study that identified preoperative factors related to poor outcomes using radiological examinations.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
