Abstracts

This is a Late Breaking abstract

Rationale: Dravet syndrome is a developmental and epileptic encephalopathy, most often due to a haploinsufficient mutation in the SCN1A sodium channel gene. It is the most common genetic epilepsy, associated with early onset, often refractory, epilepsy and intellectual disability. Traditionally, there has been strong agreement that the brain magnetic resonance images (MRIs) of patients with Dravet syndrome are initially normal, with a minority of patient develop cortical atrophy or hippocampal sclerosis over time._x000D_
Methods: We analyzed the radiologist reads and neurologist summaries of brain MRIs of children with Dravet syndrome in this large, single-center retrospective cohort study. This was one of the largest cohorts of brain MRIs analyzed in children with Dravet Syndrome, included 137 patients and a total of 258 MRI scans performed at either Ann and Robert H. Lurie Children’s Hospital (LCH) or at outside hospitals (OSH). The abnormalities detailed in the radiologists reads were categorized on a scale of 0-3: 0 (normal), 1 (incidental finding), 2 (moderate abnormality), 3 (significant, pathologic abnormality). These findings were then compared with clinical outcomes including history and type of status epilepticus, severity of intellectual disability, number of anti-seizure medications, and type of SCN1A mutation

Results: For all imaging available, there was a significantly higher percentage of abnormalities than previously reported, with 39.8% of MRIs identifying an abnormality of some kind. The radiologists reads from both LCH or OSH were more likely to have a higher abnormality rating than the neurologist summaries, but there was no difference between the ratings at LCH and other hospitals (Figure 1B-E). These abnormalities do not just develop late in the disease course, as only 58% of radiologists reads of the first available MRI at < 2 years of age were normal (Figure 1B’-E’). There was no one unifying abnormality in either infancy or later childhood, with the most common diagnosis being white matter or T₂ hyperintensities (10.4% at first scan < 2 yrs of age, 21.6% of all scans). There was no correlation between severity of MRI findings and any clinical outcome (Figure 2)._x000D_
Conclusions: This study revealed that there are significantly more MRI abnormalities described by radiologists and neurologists that previously reported. These findings are much more common than the rate of incidental findings in a general pediatric population, revealing that the consensus understanding of MRI findings in Dravet syndrome is underestimating abnormalities. When cohorted together by severity of MRI abnormality, there is no correlation that is useful to be able to predict or better understand the clinical outcomes of children with Dravet syndrome. This study is an important first step to understand that there are more MRI abnormalities than previously understood. A future, more detailed quantitative study of MRIs may be able to better elucidate how these changes relate to individualized outcomes.

Funding: Not applicable