Neuropediatrics 2006; 37 - PS4_6_2
DOI: 10.1055/s-2006-945817

MAGNETIC RESONANCE IMAGING CRITERIA FOR THE DIAGNOSIS OF CHILDHOOD MULTIPLE SCLEROSIS

D Callen 1, M Shroff 1, D Li 1, D Stephens 1, B Banwell 1
  • 1University of Toronto, Toronto, ON, Canada

Objectives: Magnetic resonance imaging (MRI) is a well-established tool used to support the diagnosis of multiple sclerosis (MS) in adults, but imaging criteria have not yet been developed in pediatric MS. Previous work from our lab has shown that MRI criteria currently used to support the diagnosis of MS in adults have reduced sensitivity in the pediatric population. The goal of this study was to establish MRI criteria to support the diagnosis of MS in children.

Methods: The study population consisted of 39 children, aged 2–18 years, with clinically definite MS, and 50 age-matched children with other white matter related neurological disorders, including mitochondrial disease, migraine, and vasculitis. Scans were evaluated in a blinded manner. Axial T2 and FLAIR images were simultaneously viewed to identify hyperintense lesions. For all lesions, maximal expanse was noted in both transverse and longitudinal planes. Lesions were differentiated using pre-determined parameters. Lesions were categorized based on their location as follows: periventricular (white matter, confluent, other), juxtacortical white matter, deep white matter, internal capsule, corpus callosal, cortical grey, deep grey nuclei, brainstem, and cerebellar. For each patient, the total number of lesions, individual lesion size, and regional distribution was determined.

Results: Intra-rater reliability was >0.8 for all measures with the majority >0.95. Most location, descriptor, and size categories differed between MS patients and controls with a medium-large effect size. The presence of five or more parenchymal lesions plus one lesion spanning >7.5mm longitudinally plus one lesion in any of the corpus callosum, brainstem, or cortical grey matter, predicted MS group membership with a 90% sensitivity and 92% specificity.

Conclusion: Diagnostic criteria for MS in children are proposed, which differ from criteria currently in use for adult MS. Creation of MRI criteria for pediatric MS will facilitate prompt diagnosis, and reduce diagnostic uncertainty.