Abstracts

RATIONALE: To determine the diagnostic value of overnight video EEG in patients with suspected Landau-Kleffner Syndrome (LKS) or electrical status epilepticus of slow wave sleep (ESES), and to identify features in the clinical history predictive of overnight EEG results.
METHODS: Retrospective review of clinical database at Washington University Pediatric Epilepsy Center at St. Louis Childrens Hospital from 1995-2001. Children referred by a neurologist for 12-48 hours of video-EEG monitoring to evaluate for possible LKS or ESES were enrolled. All studies included an overnight sleep recording. Medical records were reviewed to identify diagnosis of seizures, autistic spectrum disorder, other cognitive or language disorder, and the presence or absence of regression in language or cognitive skills.
RESULTS: Of 70 children identified, 48 were boys. Age at first video EEG was 14 to 213 months (mean 85.7 months). 34 had autistic spectrum disorder (ASD), of whom 11 had a history of regression. 36 had other language/cognitive disorder (L/CD), 19 with regression. Overall, 28 children had normal studies (40%), 7 had non-epileptiform abnormalities (10%). Of the 35 children (50%) with epileptiform activity (EA), 18 showed significant sleep activation (SAEA) that could be compatible with the diagnosis of LKS and 6 revealed ESES. EA was found in 63% of 27 children with known seizures, and 44% of those without. 47% of children with ASD had EA, including 52% of the 25 ASD cases without known seizures, and 43% of the 23 ASD cases without regression.
Of 31 children with a history of regression (11 with ASD, 20 with L/CD), 58% had EA, and 45% had SAEA. Of 39 children without regression, 44% had EA, and 26% had SAEA. Children with regression after age 3 years were more likely to have SAEA compared with those who regressed at younger than age 3 (not significant in current sample size) and children who did not have a documented regression (p[lt]0.05). SAEA was present in 9/15 older children with regression (60%) versus 5/16 younger children (31%). 6 children were diagnosed with ESES. All had L/CD with regression after age 3 years, a history of seizures and a prior epileptiform EEG.
53 children had a routine EEG prior to the first video EEG. Of 15 with a normal routine EEG, 5 had EA in the video study and 2 of these were SAEA. Of the 35 children with epileptiform video-EEGs, 30 had prior routine EEGs, of which 25 were epileptiform.
CONCLUSIONS: In a highly selected group of children referred by pediatric neurologists for possible LKS or ESES, overnight EEG revealed epileptiform activity in 50%. In 34% this was significantly activated by sleep, supporting the suspected diagnosis. In 5 children the abnormality was not evident on routine EEG, and in 6 others a diagnosis of ESES was made based on the overnight study. Therefore, at least 15% of studies yielded new information not evident on routine EEG, with potential to change clinical management. A history of regression in language or cognitive skills, older age at regression, and the presence of clinical seizures tended to increase the likelihood of epileptiform abnormalities on overnight EEG.
OBJECTIVE: Readers will be able to discuss the utility of overnight EEG in the diagnosis of LKS or ESES.