A Possible Specific MRI Appearance Found in Spondylotic Myelopathy

S. Andrew Josephson


AccessMedicine from McGraw-Hill 

The most common cause of nontraumatic spinal cord dysfunction is cervical spondylosis. While the treatment of choice is surgical decompression, items on the differential diagnosis such as infection, autoimmune disorders (e.g., multiple sclerosis), and sarcoidosis are treated differently, with a variety of medical therapies. Distinguishing these conditions from spondylosis is particularly difficult given the large number of patients with cervical degenerative disease that is asymptomatic and unrelated to cord dysfunction at nearby levels. In a recent case series, Flanagan and colleagues (2014) attempted to identify a specific enhancement pattern on MRI in spondylotic myelopathy that would allow more rapid and accurate diagnosis.

The authors retrospectively examined patients at a single institution over a 17-year period. Patients with suspected spondylotic myelopathy who underwent preoperative MRI imaging showing gadolinium enhancement and then underwent surgery were included. Patients were selected based on the authors’ recollection of patients, a search of surgical records, and review of those patients with progressive myelopathy of unknown origin who were assessed for sarcoidosis.

A total of 56 patients were included; they had subacute (≤8 weeks) onset (37%) or insidious onset (65%) of symptoms. Preoperatively, 11% were wheelchair bound, 73% were ambulatory, and 16% used some assistive device for gait such as a cane or walker. On MRI, all patients were found to have a T2-weighted MRI lesion in the spinal cord that was typically spindle shaped on sagittal images. All but 1 of the patients had a single enhancing lesion that was ≤1 spinal segment in length; the other patient had 3 focal lesions of enhancement at 3 distinct sites of narrowing. The pattern of enhancement was in a transverse band with a "pancake-like" appearance on the sagittal images in 73%. This band of enhancement was most commonly located just inferior to the site of maximal stenosis. On the axial images, the enhancement was typically circumferential and spared grey matter in 88%.

The median time to surgery from symptom onset was 11 months (range, 1–64 months), often due to other conditions being considered. A total of 70% of the patients improved after surgery, and 14% remained stable. In all, 7 patients required another surgery at the same level, all of whom reported benefit. On follow-up MRI (median duration, 23 months; range, 1–131 months), persistent enhancement was present at 3 months in 34 of 34 patients and at 12 months in 12 of 16 patients. In 46% of patients, this persistent enhancement led the treating physician to question the diagnosis of spondylotic myelopathy.

Although the study certainly suffers from the biases introduced in any case series, the MRI characteristics described may prove helpful to clinicians and radiologists. The pattern and location of enhancement in this common condition as well as the high frequency of persistent enhancement post-surgery will be helpful in cases of diagnostic uncertainty. Of course, some of these patients actually may have had an alternative diagnosis of an inflammatory myelopathy, further complicating the interpretation of this study, but the authors went to great pains to ensure that this was unlikely.