Background: Brain MRI is a useful tool for diagnosing inflammatory demyelinating disorders in children. However, it remains unclear which are the most reliable criteria for distinguishing multiple sclerosis (MS) from monophasic disorders such as acute disseminated encephalomyelitis (ADEM). We therefore compared the 4 current sets of MRI criteria in our Dutch pediatric cohort and determined which are the most useful in clinical practice for distinguishing ADEM from MS. Methods: We included 49 children who had had a demyelinating event and an MRI scan within 2 months of their first clinical attack. Twenty-one patients had ADEM and remained relapse-free after at least 2 years of follow-up. Twenty-eight patients had a definitive diagnosis of MS. We assessed the sensitivity and specificity of the following MRI criteria: Barkhof criteria, KIDMUS criteria, Callen MS-ADEM criteria, and Callen diagnostic MS criteria. Results: The Callen MS-ADEM criteria had the best combination of sensitivity (75%) and specificity (95%). The KIDMUS criteria had higher specificity (100%), but much lower sensitivity (11%). The Barkhof criteria had a sensitivity of 61% and a specificity of 91%. The Callen diagnostic MS criteria were the most sensitive (82%), but were only 52% specific for distinguishing a first attack of MS from ADEM. Conclusions: The results in our cohort demonstrate that the new Callen criteria for multiple sclerosis-acute disseminated encephalomyelitis (MS-ADEM) are the most useful for differentiating a first attack of MS from monophasic ADEM. Although the Callen diagnostic MS criteria are more sensitive, they lack the specificity necessary to differentiate MS from ADEM.

, , , , , , , , , , , , , , , , , , ,
doi.org/10.1212/WNL.0b013e3181dc138b, hdl.handle.net/1765/19762
Neurology
Erasmus MC: University Medical Center Rotterdam

Ketelslegers, I., Neuteboom, R., Boon, M., Catsman-Berrevoets, C., & Hintzen, R. (2010). A comparison of MRI criteria for diagnosing pediatric ADEM and MS. Neurology, 74(18), 1412–1415. doi:10.1212/WNL.0b013e3181dc138b