Objective

We evaluated the performance of the newly proposed 2017 McDonald criteria for the diagnosis of multiple sclerosis (MS) in children at onset and compared this to the 2010 criteria.

Methods

To identify eligible patients for this study we retrospectively examined a cohort of 279 children with acquired demyelinated syndromes with either a follow-up of ³2 years or ³2 recorded clinical attacks. We then excluded all children who presented with ADEM(N=73) and patients who were MOG-Ab(N=79) or AQP4-Ab(N=14) positive. The remaining 158 were included. A neuroradiologist scored the MRI scans acquired at baseline (within 3months from clinical onset).

Results

At baseline, 83/158(53%) and 86/158(54%) children met the 2010 and 2017 dissemination in space (DIS) criteria, respectively (this modest increase in number was due to the inclusion of lesions within the symptomatic region). 43/59(73%) of the children who underwent MRI scan with gadolinium fulfilled the 2010 dissemination in time (DIT) criteria. When the presence of oligoclonal bands (OCBs) was used to substitute DIT, an additional 35 children, were diagnosed with MS.

The 2017 criteria had higher accuracy [93.6% vs 66.0%], higher specificity [79.2%(95% CI 68.4-87.0) vs. 54.1%(44.8-63.2)], and similar sensitivity [94.1%(86.8-97.4) vs. 93.6%(82.8-97.8) when compared to 2010 criteria. No differences were observed when comparing children younger and older than 11yrs.

Conclusion

The improved performance of the 2017 compared to 2010 McDonald criteria in children was predominantly due to the inclusion of intrathecal OCBs. Our finding suggests that the same criteria could be applied routinely to children of any age.