Abstracts

To evaluate the need for magnetic resonance imaging (MRI) in children with newly diagnosed epilepsy. All children with newly diagnosed epilepsy who came from the primary catchment area of our hospital were prospectively recruited from 1998-2002. The epilepsy was defined as at least 2 unprovoked seizures occurring = 1 day apart. MRI was performed in all children at the onset of epilepsy (n=81) with the exception of idiopathic epilepsies [childhood and juvenile absence epilepsy (n=5), Rolandic epilepsy (n=2), juvenile myoclonic epilepsy (n=1), Panayiotopoulos epilepsy (n=1)], in one child with focal epilepsy and Down syndrome, and in one child with focal epilepsy who had been examined with MRI just prior to the onset of epilepsy in the routine follow-up of his leukemia. MRI findings were compared to EEG findings and to the clinical presentation. The etiologically significant MRI finding was defined as an abnormality that is associated with an increased risk of epilepsy and which is presumed to be relevant to the child s epilepsy. Of the 70 MRIs, 52 (74.3 %) were normal, 7 (10.0 %) had etiologically insignificant findings (2 Chiari I malformations, 2 arachnoideal cysts, one pineal cyst, one hypophyseal cyst, and one enlarged cisterna magna). In 5 children (7.1 %) MRI had been performed earlier for other reasons (2 hydrocephalus, one intrauterine insult, one encephalitis, and one achondroplasia) and the repeated MRI at the onset of epilepsy showed no new findings. Six children (8.6 %) had new etiologically significant findings: one frontal glioma (1,4 %), 3 large cortical dysplasias, one hippocampal dysplasia, and one hemimegalencephaly. The child with frontal glioma had repeated daily myoclonic jerks of the head, trunk, and the extremities, but his neurological examination and the routine EEG were normal. Video EEG revealed ictal spike-and-slow wave discharges arising from the right frontal area. The other 5 children with etiologically significant MRI findings had abnormalities in the neurological examination and/or in the routine EEG. It seems not reasonable to perform an MRI in all children at the onset of epilepsy. However, if the neurological examination is abnormal or if the routine EEG or video EEG show focal features, an MRI should be performed. One frontal brain tumor was found in this series (1,4 %). The rest of the etiologically significant MRI findings did not influence the primary treatment of epilepsy but may be important in planning surgical procedures if the epilepsy turns out to be medically intractable.