Abstracts

Scant information exists regarding the effect of EEG or MRI findings on seizure recurrence in neurologically normal children. Our goal was to refine patient classification into high and low risk of seizure recurrence for normal children by defining risk factors based on EEG and MRI findings., Children were prospectively recruited at the parent institutions as part of a study of child and family adaptation to epilepsy and followed to 27 months after enrollment. Inclusion criteria: children 6-14 years old who had their first recognized afebrile seizure within 3 months of recruitment and who had both EEG and MRI data. Exclusion criteria: presence of a chronic neurological illness affecting activities of living, abnormal neurologic or developmental exam, or scoring [lt]80 on the Kaufman Brief Intelligence Test. Absence and atonic seizure patients were excluded. These criteria left 85 patients for the study.
Each child was assigned a predominant seizure type based on the ILAE classifications. EEG data was classified by a) overall interpretation, b) epileptiform discharges, c) slowing. Epileptiform activity and slowing were defined as focal or diffuse. MRI images were classified for the presence of cortical or white matter lesions, encephalomalacia, enlarged lateral ventricles, atrophy, hemorrhage, vascular lesion, prominence of extra-axial fluid spaces [gt] 1 cm, or [quot]other[quot] findings.
The children were followed as clinically appropriate and had an interview every 9 months to update seizure semiology and frequency. Recurrence was defined as a 2^nd unprovoked seizure occurring [gt]24 hours after the first seizure.
X² analysis was used to compare recurrence with EEG and MRI findings., A recurrent unprovoked seizure occurred in 65.9% of patients. EEG and MRI were abnormal in 64.7% and 32.9% of patients, respectively. A significant abnormality occurred in 10.6% of MRI[apos]s.
No association was found between recurrence of seizures and the baseline EEG (X² p=0.71), between recurrence of seizures and a normal MRI (X² p=0.23), or between recurrence of seizures and an abnormal MRI with significant abnormalities (X² p=0.12)., Our EEG and MRI findings in neurologically normal children lend support to the current practice of using the clinical course to dictate the treatment plan.
Although it appears that MRI abnormalities are not predictive of seizure recurrence, a trend toward significance is noted for significant MRI abnormalities., (Supported by Grant PHS R01 NS22416 from the National Institute of Neurological Disorders and Stroke to JKA.)