Abstracts

Rationale: It has long been reported that up to 1/3 of individuals with autism have epilepsy. However, the reported rates of EEG abnormalities in the absence of clinical seizures have been more variable (15-60%). Furthermore, the role of these abnormalities in the autism phenotype is unknown and there is substantial controversy regarding treatment. This study aims to characterize epilepsy and/or EEG abnormalities in young children with autism enrolled in a large ongoing study investigating clinical subtypes of autism. Methods: Behavioral and EEG data were analyzed for the first 51 children with autism (age 1-6 yrs) enrolled. Both routine (unsedated) and overnight (natural sleep) EEG were performed during inpatient admission using a standard 21 channel EEG (simultaneous EKG and time locked video). Children were not sedated for lead placement, rather distraction (videos and books) and positive behavioral techniques were employed. All studies were read by EEG experts (SS & RF) and a child neurologist (SJS) collected seizure histories. Records were analyzed for presence of both interictal epileptiform discharges (IEDs)(sharp waves, spikes and spike wave complexes) and non-epileptiform abnormalities (slowing, asymmetry). IEDs were then categorized regarding frequency and location. An analysis of the relationship between the presence of EEG abnormalities and basic cognitive profiles was performed. Results: Two subjects (4%) had a history of clinical epilepsy, one of whom would not even tolerate lead placement. Further analysis was limited to the 49 subjects with no history of clinical seizures. Routine EEG showed clear abnormalities in 10 (20%). Abnormalities were epileptiform in 9 and non-epileptiform in 3. Overnight EEGs were abnormal in 26 (59%). Of these 25 showed IEDs. Frequency and localization were variable: 15 frequent and 9 infrequent; 11 diffuse, 4 multifocal, and 11 focal (mostly left temporal). Only 6 (12%) showed non-epileptiform abnormalities (5 with slowing, 1 with asynchrony). Routine and overnight studies were discordant in a high percentage (43%). No significant cognitive differences were seen between those with and without IEDs on overnight studies, but this result is considered preliminary due to small sample sizes limiting power. Conclusions: Data from this young autistic cohort show a lower than expected prevalence of epilepsy and higher than expected rate of EEG abnormalities (predominantly epileptiform) in children without clinical seizure history. The low rate of clinical epilepsy may be explained by the young age of the sample, since the onset of epilepsy in ASD is known to have a bimodal distribution. Age may also explain the high rate of IEDs if this is a developmental phenomenon. The left temporal predominance in cases of focal IEDs is intriguing in light of the severe language difficulties seen in this cohort, but a larger sample is needed to detect subtle behavioral, cognitive or language differences associated with the presence of IEDs. The high discordance rate between routine and overnight EEG demonstrates the value of prolonged studies in these children.