Utility of cervical dynamic magnetic resonance imaging for evaluating patients with cervical myelopathy: a retrospective study

Article information

Asian Spine J. 2024;18(5):647-653

Publication date (electronic) : 2024 October 22

doi : https://doi.org/10.31616/asj.2024.0176

Mrugank Narvekar, Bharatkumar Rajendraprasad Dave, Ajay Krishnan, Devanand Degulmadi, Shivanand Mayi, Ravi Ranjan Rai, Mirant Dave, Charde Pranav, Abhijith Anil, Rohan Killekar, Panthackel Mikeson, Kishor Murkute

Department of Spine, Stavya Spine Hospital and Research Institute, Ahmedabad, India

Corresponding author: Bharatkumar Rajendraprasad Dave, Department of Spine, Stavya Spine Hospital and Research Institute, Rasala Marg, Mithakhali, Ellisbridge, Ahmedabad, Gujarat 380007 India, Tel: +91-07926483040, E-mail: brd_172@yahoo.com

Received 2024 May 22; Revised 2024 July 26; Accepted 2024 August 12.

Abstract

Introduction

Cervical spondylotic myelopathy (CSM), a common and debilitating condition characterized by spinal cord dysfunction in the cervical region, is considerably challenging to diagnose and manage. The pathophysiology of CSM still remains controversial and complex [1]. CSM has been intricately linked to both static and dynamic narrowing of the cervical spinal canal [1–4]. Static compressive factors, such as ligamentum flavum hypertrophy, disc protrusion, facet degeneration, and osteophytes, diminish the spinal canal volume, contributing to its compromise [5,6]. Moreover, dynamic factors, such as the infolding of the ligamentum flavum, exacerbate compression within an already compromised canal [4,7–9].

Traditionally, patients with cervical spine pathologies undergo evaluation using cervical spine magnetic resonance imaging (MRI) with the neck in the neutral position to establish a clinical diagnosis and guide surgical decision making. However, this approach often fails to identify the dynamic changes occurring during neck movement, which are crucial for comprehending the pathophysiology and guiding treatment strategies [1,6,10,11].

Various studies have reported on the dynamic changes in the cervical spine of healthy volunteers, describing the physiological morphologic features of the cervical spine on MRI at different positions [12,13]. The dynamic changes are reflected in the spinal cord diameter, transverse cross-sectional area of the spinal cord, and subarachnoid space [7,12,14]. The higher flexibility of the cervical canal likely accounts for the greater magnitude of changes observed therein than in other spinal regions [12–14].

Cervical dynamic MRI (dMRI) serves as a promising tool for evaluating cervical myelopathy, enabling clinicians to assess spinal cord compression, segmental instability, and alterations in range of motion, often missed on conventional imaging [6,15,16]. The utility of dMRI in cervical myelopathy extends beyond diagnosis, influencing treatment planning and outcome prediction [6,16]. Maximizing recovery following cervical decompression hinges on accurately determining the number of cervical levels requiring decompression before surgery. dMRI exhibits the potential to refine the surgical plan based on neutral MRI findings, potentially necessitating treatment of additional levels [6,15,16].

The current study aimed to evaluate the utility of cervical dMRI in the assessment of cervical canal stenosis.

Materials and Methods

This retrospective study was conducted at Stavya Spine Hospital and Research Institute, Ahmedabad. Data for all patients who had undergone cervical dMRI from May 2017 to August 2022 were retrieved from their electronic records on the hospital management system and picture archiving and communication system. This study was approved by the Stavya Spine Hospital and Research Institute and was registered under the Clinical Trial Registry India (CTRI/2022/09/045279). The ethics committee waived the need for informed consent for data collection.

Imaging protocol

All patients underwent MRI on a Philips 1.5-T (Multiva; Philips India Ltd., Gurugram, India) 16-channel MRI unit in the supine position with their neck in the neutral, flexion, and extension positions sequentially. Images for three sequences required approximately 8 minutes to acquire. The cervical flexion and extension positions were achieved using a foam pillow suitably sized based on the position at which the patients were most comfortable (Fig. 1A–C). Thus, the flexion and extension angles were not uniform for all patients. Patient tolerability during MRI was assessed intermittently by the technician/radiologist through a microphone. In case a patient experienced any discomfort during the procedure, such as increased tingling sensations or pain in the extremities, the scanning was stopped and resumed after readjusting the patient to a more comfortable position. Vital signs (pulse and oxygen saturation) were monitored throughout the procedure. These patients were excluded from the study and only underwent neutral MRI.

Fig. 1

Images showing the positioning of the patient during magnetic resonance imaging. (A) Neutral. (B) Flexion. (C) Extension. The flexion extension position was given using foam pillows of different height as per patient’s comfort. Written informed consent for the publication of this image was obtained from the patient.

Results

A total of 1,845 disc spaces from 369 patients were examined in this study. Table 1 summarizes the patient’s demographic data.

Table 1

Demographic data of the patients included in the study

Stenosis

Fig. 2 shows the number of disc levels showing grade 2/3 stenosis in all sequences. Accordingly, 653/1,845 (35.39%) disc levels showed an increase in the grade of stenosis on extension MRI when compared to neutral MRI. Table 2 summarizes the changes in the grade of stenosis on extension MRI. No disc levels showed an increase in the grade of stenosis on flexion MRI when compared to neutral MRI.

Fig. 2

The number of disc levels showing grade 2/3 compression in neutral, flexion and extension position. The number showed an increase in the extension position as compared to the neutral position.

Table 2

Change in the grade of stenosis at each disc level in the extension MRI as compared to the neutral MRI

Moreover, 168/653 (25.72%) and 180/653 (27.56%) disc levels showed a change in the grade of stenosis from grades 0/1 to grades 2 and 3, respectively on extension MRI. Table 3 and Fig. 3 describe the ARC at each level.

Table 3

The ARC at each disc level in extension MRI

Fig. 3

The appearance ratio of cord compression (ARC) calculated at each level. Maximum change was appreciated at the C4–5 level.

Overall, grade 3 stenosis was observed in 360/1,845 (19.51%), 675/1,845 (36.58%), and 198/1,845 (10.73%) disc levels on neutral, extension, and flexion MRI, respectively. This increase in the grade of stenosis was statistically significant (p<0.001) on extension MRI when compared to neutral and flexion MRI.

In total, 42/369 patients (11.38%) showed no compression on neutral MRI and dMRI, whereas 49/369 patients (13.27%) with grades 0/1 stenosis on neutral MRI converted to grades 2/3 at one or more levels in the extension MRI. A mean increase of 2.23±1.2 levels of compression was observed in these patients. Among the 90/369 patients with single-level compression on neutral MRI, 63/369 (17.07%) showed multiple levels of stenosis on extension MRI. Among the 84/369 patients with double-level stenosis on neutral MRI, 57/369 patients (15.44%) showed an increase in the levels of stenosis on extension MRI. The aforementioned patients showed a mean increase of 1.55±0.75 levels of compression on extension MRI when compared to neutral MRI. Moreover, 104/369 patients had multilevel stenosis on neutral MRI and dMRI.

Hyperintense lesion

The HILs at various positions are as described in Table 4. Notably, 79 disc levels with grade 0 HIL on neutral MRI showed type 1/2 on flexion MRI, whereas 62 disc levels with grade 1/2 HIL on neutral position changed to grade 0 on extension MRI.

Table 4

Number of disc levels according to the grade of hyperintense lesion seen in different neck positions and the segmental involvement

Discussion

In 1968, Penning [19] first described the adaptive changes of the spinal cord and subarachnoid space during the flexion and extension of the cervical spine. The aforementioned in vitro study involving functional myelogram showed a decrease in the subarachnoid space and increase in spinal cord diameter with neck flexion based on comparisons between neutral, flexion, and extension images. Muhle et al. [17] in 1998 had been the first to classify cord compression based on static and dynamic factors that may play a role in the pathogenesis of CSM.

Given its noninvasive nature and absence of radiation exposure, MRI has long been established as the diagnostic modality of choice for various cervical pathologies, such as cervical disc prolapse, cervical stenosis, congenitally narrow spinal canal, and ligamentum flavum hypertrophy [20]. Traditionally, MRI is obtained with the neck in the neutral position. The influence of the dynamic factors on cervical pathologies, as reported in the literature, sometimes complicates thew identic aition of the exact pathology on neutral MRI, thereby providing false negative results or downplaying the severity of dynamic pathology that occurs as a part of daily life [1,6,10,11,15]. Cases clinically suggestive of myelopathy but with no evidence of significant cervical spinal cord compression on neutral MRI often present as a challenge to the surgeon in clinical practice.

Flexion–extension radiographs and computed tomography myelography have remained the standard methods for obtaining positional images of the spine. MRI yields an image superior to that obtained via radiography while being less invasive than myelography. Thus, to understand the pathophysiology of the spine, further developments in functional clinical imaging seems to be needed. Different modalities and techniques, such as dynamic MRI, standing MRI, and dynamic MRI, have been published in the literature [6,15,21]. Physicians have been experimenting with methods for using dMRI in clinical practice. dMRI may prove to be efficacious as a part of the clinical diagnosis and treatment paradigm for patients with spinal, radicular, and referred pain syndromes originating from spinal pathology [21].

Muhle et al. [3] reported that spinal cord compression was aggravated in extension magnetic resonance (MR) images due to increased cervical cord sagittal diameter and decreased available space in the cervical spinal cord. Kim et al. [22] in their study noted that 72 % (23/32) of the patients had increased levels of compression on extension MR images when compared to neutral MRI. Lao et al. [23] observed increased disc bulge in flexion and extension MRI in 91 discs (3.03%) and 479 discs (15.97%), respectively. Zhang et al. [15] concluded similar findings that extension MRI showed far greater segmental canal stenosis than did neutral or flexion MRI. Zeitoun et al. [16], in their study on dynamic MRI changes in 51 CSM patients, reported that 22.5% of grade 1 compression on neutral imaging changed to grade 3 on extension imaging.

In our study, 348/1,865 levels (18.65%) showed an increase in stenosis from grades 0/1 to grades 2/3. The calculated ARC showed a maximum increase at the C4–5 level. This result is consistent with the findings obtained in previous studies [15,16,22,23]. Moreover, our findings revealed that 49/369 patients (13.27%) exhibited stenosis on extension MRI but not on neutral MRI. This finding may suggest the need to modify the treatment modality for such patients from conservative to surgical management. Furthermore, 120 patients (32.52%) showed a mean increase of 1.55±0.75 levels of compression on extension MRI when compared to neutral MRI. A sample case presented in Fig. 4 shows an increase in the grade and levels of compression in on extension MRI when compared to neutral MRI. This finding challenges the appropriacy of mandating dMRI as a protocol before surgery in cases with indicated cervical pathologies, such as CSM. Surgical planning based on neutral MRI may overlook levels of compression, causing inadequate decompression.

Fig. 4

Case example: magnetic resonance imaging (MRI) images in the neutral (A), flexion (B), and extension (C) positions (arrow). The extension image (C) shows increase in the grade and levels of stenosis. The hyperintense lesion is visualized better in the flexion MRI.

In their study, Zeitoun et al. [16] noted that 10% of HILs observed on flexion MRI were absent on neutral MRI. During flexion, the diameter of the cervical canal increases, allowing better visualization of the spinal cord [15,16]. In the present study, the grade of HILs increased on flexion MRI when compared to neutral MRI. Moreover, 79 (4.28%) disc levels showed a fresh-appearing HIL on flexion MRI, which was absent on neutral MRI. However, the number of disc levels showing the presence of HIL decreased in the extension position. Also, the extent of HIL was found to increase in the flexion images when compared to neutral MRI. This finding substantiates the claim of previous studies that flexion imaging provides better visualization of the presence and extent of HILs [15,16].

Our study noted that neutral MRI may be less effective than extension MRI in detecting, although additional flexion imaging helps. The presence of HILs may often be missed in the presence of stenosis and is only noted after adequate decompression due to better visualization of the cord margins (Fig. 4) [16]. The de-novo appearance of HILs, which have been identified as a prognostic factor, after decompression surgery may be considered as intraoperative mishandling of the cord and often becomes a factor for litigation [24,25]. dMRI with flexion overcomes this problem in the identification of cord signals preoperatively. Hence, cervical dMRI can help treating surgeons discuss the prognosis with their patients and safeguard themselves from medico-legal issues [16].

The strength of our study lies in the large study population, with all subaxial disc levels studied on neutral, flexion, and extension imaging. We have noted the changes in cervical stenosis, as well as the presence, absence, and extent of HILs in all three positions, thereby providing a better understanding of the dynamics of the spine. However, a few limitations of our study are worth mentioning. First, the included patients were not followed up, which would have allowed us to more precisely determining the utility of dMRI in terms of surgical decision making. Second, MRI was performed in the supine position to eliminate head weight. However, a standing MRI may potentially reveal the influence of the dynamic factors more accurately.

Conclusions

Cervical dMRI helps surgeons plan their surgery, discuss the prognosis with the patient, and safeguard themselves from medico-legal issues arising from improper or missed diagnosis and treatment. Extension imaging allows better identification of spinal canal stenosis than neutral imaging. Meanwhile, flexion imaging provides better visualization of the presence and extent of HILs.

Article information Continued