Clinical Correlation of a New Practical MRI Method for Assessing Cervical Spinal Canal Compression

Musculoskeletal Imaging Original Research Park et al. MRI Assessment of Cervical Spinal Canal Compression Musculoskeletal Imaging Original Research Hee-Jin Park 1,2 Sam Soo Kim 2 Eun-Chul Chung 1 So-Yeon Lee 1 Noh-Hyuck Park 3 Myung-Ho Rho 1 Sun-Hyung Choi 1 Park HJ, Kim SS, Chung EC, et al. Keywords: canal, cervical spine, MRI, spine DOI:1.2214/AJR.11.7599 Received July 18, 211; accepted after revision November 8, 211. 1 Department of Radiology, Kangbuk Samsung Hospital, Sungkyunkwan University School of Medicine, Seoul, Korea. 2 Department of Radiology, Kangwon National University School of Medicine, Gangwon-do, Korea. 3 Department of Radiology, Myongji Hospital, Kwandong University College of Medicine, Dugyang-ku, Hwajongdong 697-24, Koyang Kyunggido 412-27, Korea. Address correspondence to N. H. Park (radiology11@hanmail.net). WEB This is a Web exclusive article. AJR 212; 199:W197 W21 361 83X/12/1992 W197 American Roentgen Ray Society Clinical Correlation of a New Practical MRI Method for Assessing Cervical Spinal Canal Compression OBJECTIVE. The purpose of this study was to evaluate interobserver agreement and whether or not a new MRI grading system correlates with symptoms and neurologic signs for assessing spinal canal compression. MATERIALS AND METHODS. One hundred patients (52 men and 48 women; mean age, 5 years) underwent MRI of the cervical spine at our institution and were evaluated by two musculoskeletal radiologists. The presence and grade of cervical canal stenosis at the maximal narrowing point was assessed according to the new grading system suggested by Kang et al. (Kang system). The results correlated with the clinical manifestations and neurologic examination. Statistical analysis was performed using kappa statistics, categoric regression analysis, and nonparametric correlation analysis (Spearman correlation). RESULTS. Interobserver agreement in the grading of spinal stenosis between the two readers was almost perfect (k =.925). Most of the patients with grade cervical canal stenosis showed no neurologic manifestation, and patients with grades 2 and 3 cervical canal stenosis had positive neurologic The correlation coefficient (R) of reader 1 between MRI grade (, 1, 2, and 3) and neurologic manifestations (positive or negative) was.846. The R of reader 2 was.88. In the younger age group (< 5 years old), the R of reader 1 was.834 and the R of reader 2 was.745. In the older age group ( 5 years old), the R of reader 1 was.839 and the R of reader 2 was.839. CONCLUSION. The interobserver agreement of the Kang system was almost perfect and was higher than in the study by Kang et al. Grade cervical canal stenosis represents negative neurologic manifestations and grades 2 and 3 cervical canal stenosis represent positive neurologic The Kang system and clinical manifestations are significantly correlated, especially in the older age group ( 5 years). C ervical spinal stenosis is a common condition that results in considerable morbidity. MRI is used in the evaluation of patients with symptoms related to cervical spinal stenosis; however, there are no widely used diagnostic criteria or grading systems for cervical spinal stenosis on MRI. Harrop et al. [1] subdivided cervical spinal cord compression and the presence of hyperintense signal within the cord on T2-weighted imaging. They evaluated the correlation between the radiologic findings and cord myelopathy and suggested close correlation between those radiologic findings and cord myelopathy, but they did not grade the spinal cord compression. Takahashi et al. [2] reported the frequency of the high signal intensity of the cervical cord on T2-weighted imaging is directly proportional to the severity of clinical myelopathy and degree of spinal canal compression. Recently, Kang et al. [3] reported a new MRI grading system for cervical canal stenosis. They classified cervical canal stenosis into the following grades based on T2- weighted sagittal images: grade, absence of canal stenosis (subarachnoid space obliteration 5%); grade 1, subarachnoid space obliteration > 5%; grade 2, spinal cord deformity (compressed); and grade 3, spinal cord signal change. Kang et al. suggested that this new grading system provides a reliable assessment of cervical canal stenosis. The purpose of this study was to evaluate whether the new MRI grading system for cervical canal compression correlates with symptoms and neurologic signs and to evaluate whether each grade represents clinical significance. W197

Park et al. Materials and Methods Case Selection Of the patients who visited our hospital between January and September 21, 1 consecutive patients (52 men and 48 women) who underwent MRI of the cervical spine were included in this retrospective study. The age distribution was as follows: < 5 years, n = 5; and 5 years, n = 5. The mean age was 5 years (± SD, 13.3 years; median age, 5.5 years). The mean age of the older group was 61.7 years (± 8.13 years; median age, 59 years), and the mean age of the younger group was 4.5 years (± 7.94 years; median age, 42 years). The sex distribution was 29 men and 21 women in the older group and 23 men and 27 women in the younger group. Sixty-one patients were excluded from the study. The exclusion criteria were as follows: infections; tumors; acute trauma; surgical history; neural foraminal stenosis; lumbar spinal stenosis; combined brain infarction or other intracranial lesion; and peripheral neuropathy, such as carpal tunnel syndrome. The patient with symptoms at a different cord level was excluded from the study after a review of the medical records. This study was approved by our institutional ethics review board, and the requirement for informed consent was waived because of the retrospective design. Image Analysis MRI examinations were interpreted by two fellowship-trained academic musculoskeletal radiologists who had 12 and 1 years of experience. The radiologists were blinded to the clinical information and radiologic reports. A total of four sequential levels (C3 C4, C4 C5, C5 C6, and C6 C7) were qualitatively analyzed. The radiologists assessed the presence and grade of cervical spinal canal stenosis at the maximal narrowing point, in 6 5 4 3 2 1 accordance with the new MR grading system suggested by Kang et al. [3]; hereafter, we refer to this grading system as the Kang system. Cervical canal stenosis was classified into the following grades on the basis of T2-weighted midsagittal images: grade, subarachnoid space obliteration 5% (originally Kang et al. defined grade as the absence of stenosis, but we defined it as 5% stenosis); grade 1, subarachnoid space obliteration > 5% without cord compression; grade 2, spinal cord compression without cord signal change; and grade 3, spinal cord compression with cord signal change near the compressed level. When the radiologists were not confident about their findings, the axial T2-weighted images were used as a supplementary evaluation method. We excluded the possibility of a partial volume artifact from true spinal stenosis through the axial T2-weighted images. Clinical Correlation Neurologic examinations were performed and clinical manifestations were acquired by the same physician. We considered positive neurologic manifestations as observed paresthesias, extremity weakness, numbness, and funicular or radicular pain. The positive neurologic signs were positive Lhermitte sign, Spurling sign, increased response of deep tendon reflex, and positive denervation sign on electromyography. Deep tendon reflex evaluations were performed in all cases, and electromyography was performed in four cases. More than one positive neurologic sign combined with more than one neurologic clinical manifestation was considered a positive neurologic manifestation of cervical canal stenosis. MRI Parameters All MRI examinations were performed using the same protocol on a 1.5-T magnet (Intera, 6 5 4 3 2 1 Philips Healthcare) using a Syn-head coil (Philips Healthcare) and fast spin-echo imaging. T2- weighted images were obtained in the axial plane and T2-weighted images in the sagittal plane in the supine position with the following parameters: FOV, 27 cm; matrix, 512 512; slice thickness, 3 mm; interslice gap,.3 mm (sagittal image); and FOV, 17 cm; matrix, 512 32; and slice thickness, 3 mm; interslice gap,.3 mm (axial image). The MRI sequences were as follows: sagittal T2- weighted spin echo (TR/TE, 35/12) and axial T2-weighted turbo spin echo (TR/TE, 27/11). Statistical Analysis The interobserver agreement between the two radiologists was analyzed using kappa statistics. The interpretation of kappa values was as follows: poor, <.1; slight,.1 to.2; fair, >.2 to.4; moderate, >.4 to.6; substantial, >.6 to.8; and almost perfect, >.8 to 1. The correlation coefficients (R) were calculated with categoric regression analysis and nonparametric correlation analysis (Spearman correlation). For analysis of the relationship between the findings and patient characteristics, the association between MRI findings and clinical manifestations were evaluated with age (< 5 years and 5 years). An R between.7 and.9 indicated a relatively high correlation and R >.9 indicated a very high correlation. The level of correlation significance was.1. Statistical analyses were performed using SPSS statistical software (version 1.1). Results The results for each grade detected by readers 1 and 2 are shown in Figure 1. We found positive neurologic manifestations in 45 patients, with negative findings for 55 patients (Figs. 2 and 3). Reader 1 noted only one pa- Grade Grade Fig. 1 Chart shows prevalence of each MRI grade. Black bars indicate reader 1, and gray bars indicate reader 2. Fig. 2 Chart shows correlation between MRI grade and neurologic manifestations according to reader 1 (gray bars). Black bars indicate positive neurologic W198

MRI Assessment of Cervical Spinal Canal Compression tient with positive neurologic manifestations in 47 patients with grade ; eight patients had negative neurologic manifestations and 14 patients had positive neurologic manifestations in 22 patients with grade 1; and only one patient with negative neurologic manifestations was noted in 1 patients with grade 3 (Figs. 4 8). No negative neurologic manifestations were noted in 21 patients with grade 2 (Fig. 2). Reader 2 noted three patients with positive neurologic manifestations in 49 patients with grade ; eight patients with negative neurologic manifestations and 12 patients with positive neurologic manifestation in 2 patients with grade 1; and only one patient with negative neurologic manifestations was noted in 24 patients with grade 2 (Fig. 8). No negative neurologic manifestations were noted in seven patients with grade 3 (Fig. 3). The interobserver agreement in the grading of spinal stenosis between the two readers was near perfect (k =.925). Although the kappa value suggested strong overall interobserver agreement, the agreement was weaker for more severe stenosis (grades 2 and 3). The R of reader 1 between MRI grades (, 1, 2, or 3) and neurologic manifestations (negative or positive) was.846 (Table 1); the R of reader 2 was.88. In the younger age group (< 5 years), the R of reader 1 was.834 and the R of reader 2 was.745. In the older age group ( 5 years), the R of reader 1 was.839 and the R of reader 2 was.839. Fig. 3 Chart shows correlation between MRI grade and neurologic manifestations according to reader 2 (gray bars). Black bars indicate positive neurologic Discussion The MRI classification method of the cervical canal must be accurate, easy to apply, and highly reproducible between observers to facilitate a clinical trial. A labor-intense, precise quantitative analysis may not be practical in a busy clinical practice. A more practical semiquantitative measurement may be easily incorporated in a clinical setting and may help in eliminating some of the factors that lead to variability caused by internal subjective standards [4]. Sagittal T2-weighted MRI provides a simple objective method for detecting cervical spinal canal compression [5]. Larsson et al. [6] reported an assessment using a single dimension, in which mild narrowing was defined as 5% reduction in the width of the subarachnoid space, moderate narrowing involved > 5% reduction in the width of the subarachnoid space, and severe stenosis was defined as cord compression. Recently, Kang et al. [3] reported a new MRI grading system for cervical canal stenosis. They suggested that this new grading system provides a reliable assessment of cervical canal stenosis, with interobserver agreement for the four grades ranging from.6 to.62. The prevalence of the each grade showed a similar distribution between the two readers in this study, but the results were somewhat different from the prevalence reported by Kang et al. Interestingly, Kang et al. reported the incidence of grade to be minimal (six patients) and grade 1 to be more common (36 patients); however, in our study, the incidence of grade was more common (47 and 49 patients by readers 1 and 2, respectively) than the incidence of other grades. This difference may have resulted from case selection bias. The population in the Kang et al. study included patients > 6 years old, but we included patients with a wider range of ages (2 82 years). Thus, the proportion of grade patients might be greater. In the current study, the interobserver agreement for the new grading system was near perfect (k =.925) and much higher than that of Kang et al. [3]. This discrepancy cannot be explained satisfactorily. We may presume that the differences are because our results were extracted from two observers but previous results were from three or more observers and mild differences in the grading system (originally Kang et al. defined grade as the absence of stenosis, but we defined it as 5% stenosis). We also correlated the new grading system with clinical manifestations and neurologic signs. Only one of three grade patients had positive neurologic manifestations and only one of the grade 2 or 3 patients had negative neurologic Thus, grade represents negative neurologic manifestations and grades 2 and 3 represent positive neurologic 6 5 4 3 2 1 Grade Fig. 4 Grade 1 stenosis in 32-year-old man with radiating pain in both upper extremities. Sagittal T2- weighted turbo spin-echo image (TR/TE, 35/12) shows cervical canal stenosis with obliteration of CSF space > 5% at C3 C4 (arrow). Patient had positive neurologic W199

Park et al. Fig. 5 Grade 1 stenosis in 5-year-old man with neck discomfort. Sagittal T2-weighted turbo spinecho image (TR/TE, 35/12) shows cervical canal stenosis with obliteration of CSF space > 5% at C5 C6 (arrow). Patient had negative neurologic Fig. 8 Grade 2 stenosis in 47-year-old woman with bilateral hand pain. Sagittal T2-weighted turbo spinecho image (TR/TE, 35/12) shows cervical canal stenosis and cord compression without cord edema at C5 C6 (arrow). Patient had negative neurologic Fig. 6 Grade 2 stenosis in 35-year-old man with radiating pain in right upper extremity. Sagittal T2-weighted turbo spin-echo image (TR/TE, 35/12) shows cervical canal stenosis and cord compression without edematous changes of cord at C6 C7 (arrow). Patient had positive neurologic The clinical significance of grade 1 cervical canal stenosis is controversial. We suggest that surgical intervention in the case of grade cannot be justified and surgical intervention in the case of grade 1 must be performed when clinical and neurologic manifestations are evident. The R of readers 1 and 2 was high (.846 and.88, respectively). The R of the readers for the younger group of patients differed (.834 and.745, respectively) but was similar for the older group of patients (.839 and.839, respectively). These results suggest that the new grading system reflects clinical symptoms precisely, with very good agreement between readers. One of the limitations of this study was the single posture of the cervical spine MRI because cervical spine posture affects the di- Fig. 7 Grade 3 stenosis in 54-year-old man with paresthesias of both upper extremities. Sagittal T2-weighted turbo spin-echo image (TR/TE, 35/12) shows cervical canal stenosis and cord compression with cord edema at corresponding level of C5 C6 (arrow). Patient had positive neurologic TABLE 1: Correlation Coefficients of Cervical Spinal Stenosis Between MRI Grade and Neurologic Manifestations Age Observer Total < 5 y 5 y Reader 1.846 (<.1).834 (<.1).839 (<.1) Reader 2.88 (<.1).745 (<.1).839 (<.1) Note The level of correlation significance was.1. Data in parentheses are p values. mensions of the spinal canal. Muhle et al. [7] reported the prevalence of spinal stenosis and cervical cord impingement increase at flexion and extension. In the current study, the patients were in a neutral supine position, and no flexion or extension was applied. However, all examinations were in the same position; therefore, any differences should have been minimal. Another limitation was that the study was not based on a quantitative evaluation of the clinical We classified clinical manifestations as positive or negative, so the dependent variables were unordered qualitative variables. Nevertheless, the purpose of this study was to evaluate the clinical effectiveness of the new grading system because differentiation of spinal stenosis in symptomatic and asymptomatic conditions might be valuable. Another limitation W2

MRI Assessment of Cervical Spinal Canal Compression was that most of the patients underwent MRI for some reason or discomfort other than radiating pain or neurologic symptoms. A true healthy control group was not established. In conclusion, interobserver agreement for the new grading system of cervical canal compression was near perfect. Grade represents negative neurologic manifestations and grades 2 and 3 represent positive neurologic The clinical significance of grade 1 cervical canal stenosis is controversial. The correlation between the MRI grade and clinical manifestations was high, especially in the older group of patients. References 1. Harrop JS, Naroji S, Maltenfort M, et al. Cervical myelopathy: a clinical and radiographic evaluation and correlation to cervical spondylotic myelopathy. Spine 21; 35:62 624 2. Takahashi M, Yamashita Y, Sakamoto Y, Kojima R. Chronic cervical cord compression: clinical significance of increased signal intensity on MR images. Radiology 1989; 173:219 224 3. Kang Y, Lee JW, Koh YH, et al. New MRI grading system for the cervical canal stenosis. AJR 211; 197:193; [web]w134 W14 4. Stafira JS, Sonnad JR, Yuh WT, et al. Qualitative assessment of cervical spinal stenosis: observer variability on CT and MR images. AJNR 23; 24:766 769 5. Fehlings MG, Rao SC, Tator CH, et al. The optimal radiologic method for assessing spinal canal compromise and cord compression in patients with cervical spinal cord injury: results of a multicenter study. Spine 1999; 24:65 613 6. Larsson EM, Holtas S, Cronqvist S, Brandt L. Comparison of myelography, CT myelography and magnetic resonance imaging in cervical spondylosis and disk herniation: pre- and postoperative findings. Acta Radiol 1989; 3:233 239 7. Muhle C, Metzner J, Weinert D, et al. Classification system based on kinematic MR imaging in cervical spondylitic myelopathy. AJNR 1998; 19:1763 1771 W21