Abstracts

Rationale: The correlation between autism spectrum disorder (ASD) and epilepsy is well described in the literature, with the frequency of epilepsy in ASD patients ranging from 5% to 40% in a number of studies (Tuchman & Rapin 2002). Abnormal electroencephalograms (EEGs) have been described in patients with ASD even in the absence of epilepsy, with reported rates as high as 60% (Spence & Schneider 2009); however, the prognostic implications of these abnormalities remain unclear, particularly with respect to risk of subsequent epilepsy. The purpose of this study was to determine the relationship between epilepsy risk and factors such as preceding EEG abnormalities, gender, febrile seizures, family history of epilepsy, family history of febrile seizures, and family history of autism.  Methods: Data were collected from medical records at the University of California-Los Angeles and at Pediatric Minds Medical Clinic in Torrance, California. We retrospectively identified patients with (1) a clinical diagnosis of ASD, (2) no history of epilepsy prior to overnight continuous EEG evaluation, (3) overnight video-EEG evaluation, and (4) at least one week of follow-up. We used the Kaplan-Meier method and Cox proportional hazards regression to evaluate whether time to onset of epilepsy is associated with an abnormal EEG (on the basis of epileptiform discharges or focal slowing), as well as family history of epilepsy, febrile seizures, or autism. Results: We identified 160 patients who met the inclusion criteria. Median age at initial EEG was 4.6 years (interquartile range [IQR] 3.6 - 7.0). Sixty-three (38.9%) had an abnormal EEG. Median follow-up after EEG was 2.4 years (IQR 1.0 - 4.4). Eighteen (11.3%) children subsequently developed epilepsy after a median of 1.9 years (IQR 0.7- 4.1). In multivariate Cox proportional hazards regression, time to epilepsy onset was independently associated with abnormal initial overnight continuous EEG (hazard ratio 8.0; 95%CI 1.7 - 37.9; P= 0.008) and family history of febrile seizures (hazard ratio 12.6, 95%CI 1.3 - 122.7; P= 0.03). Conclusions: This study suggests that an abnormal initial overnight EEG in patients with ASD is associated with an approximately 8-fold increased risk of subsequent epilepsy. These results warrant further study in a prospective observational cohort utilizing overnight EEG as a screening tool. Funding: No funding