mail2

Spectrum Library

Can spinal MRI be used to predict neuromyelitis optica spectrum disorder ormultiple sclerosis at the onset of transverse myelitis?

Background: Acute transverse myelitis (ATM) in patients with no history of central nervous system (CNS) demyelinating disease may be idiopathic or herald the development of neuromyelitis optica spectrum disorder (NMOSD) or multiple sclerosis (MS). NMOSD may differ from MS in pathogenesis, prognosis and response to treatment, and radiological features at the first presentation of ATM that distinguish between NMOSD and MS would be desirable. Objective: To identify characteristic spinal radiological features of NMOSD and MS at first attack of ATM. Methods: We retrospectively reviewed all patients from the National Neuroscience Institute (NNI) Neuroimmunology Database and Tissue Repository who had i) presented with ATM from 2000 to 2008 in the absence of a history of CNS demyelinating disease, ii) spinal MRI within 30 days of symptom onset, and iii) definitive diagnosis available subsequently. Demographic, clinical, radiological and serological data were analysed. The spinal MRI images were reviewed by 2 neuroradiologists blinded to the final diagnosis. Results: Sixteen out of 158 patients fulfilled inclusion criteria (7 NMOSD, 8 MS, 1 recurrent myelitis). Of the 7 NMOSD patients, 3 fulfilled 2006 Wingerchuk NMO diagnostic criteria; 4 were NMO-IgG seropositive. There was no difference in the mean duration of follow-up (NMOSD 73.1 months vs MS 74.0 months, p=0.97). NMOSD patients were significantly older at disease onset (50 vs 29 years) and had higher Expanded Disability Status Score at last review (4.5 vs 1.5). There was no difference in relapse rates (NMOSD 0.75 vs MS 0.55 attacks per year, p=0.36). Spinal MRI in NMOSD patients showed significantly more cord swelling (NMOSD=7/9, MS=3/16, p<0.01) and T1 hypointensity (NMOSD=4/9, MS=0/16, p=0.01). Individual lesions in NMOSD were more longitudinally extensive (5.89 vs 2.13 segments, p<0.01), less likely to show axial peripheral T2 hyperintensity (NMO=0/7, MS=8/16, p=0.03) and tended to show more central grey-matter T2 hyperintensity (NMO=6/7, MS=8/16, p=0.12). Total lesion length was longer in NMOSD (7.57 vs 4.25 segments, p=0.13) and there were fewer non-contiguous foci (1.29 vs 2, p=0.2). Conclusions: NMOSD patients were older at the initial episode of ATM. Spinal MRI features of longitudinally extensive cord swelling, T1 hypointensity, and the absence of peripheral T2 hyperintensity may be helpful to distinguish them from patients with MS.

Read More: Can spinal MRI be used to predict neuromyelitis optica spectrum disorder ormultiple sclerosis at the onset of transverse myelitis?