Abstracts

RATIONALE: Routine overnight EEG monitoring in the evaluation of children with autism remains controversial. Landau-Kleffner syndrome(LKS), acquired epileptic aphasia, is characterized by language deterioration associated with seizures or paroxysmal EEG abnormalities. The clinical distinctions between LKS and autism with language regression are blurred. The EEG abnormality in LKS is central to its diagnosis, and is characterized by bilateral spike-wave discharges, usually maximally posterior temporal or peri-Sylvian, with significant activation during slow wave sleep. The clinical utility of overnight EEG monitoring in this patient population, which represents a technically challenging group to study and in which sedation is often employed to complete these studies, is uncertain.
METHODS: The video-EEG telemetry experience at Children[ssquote]s National Medical Center from 9/96-3/00 was reviewed. Patients undergoing evaluation for LKS were identified and their telemetry findings analyzed for presence of interictal or ictal abnormalities, location of interictal discharges, and activation during slow wave sleep.
RESULTS: 894 patients were monitored over 3.5 years; 334(37%) were recorded to investigate for LKS. Of these, 83(25%) had an abnormal EEG, with epileptiform potentials in 63(19%). The distribution of epileptiform activity was most commonly temporal 39%, but also occipital 21%, frontal 18%, parietal 13%, and generalized 6%. Lateralization revealed bilaterally symmetric involvement in 37%, exclusively left hemispheric 19%, left hemisphere predominant 16%, exclusively right hemispheric 14%, right hemisphere predominant 14%. Sleep yielded epileptiform activity in 56 studies (89%). During sleep, 23 patients (36%) had discharges only during stage 2, and 33 (52%) had discharges during stages 2 and delta sleep. No study revealed electrical status epilepticus of slow wave sleep (ESES). Of the 63 studies with epileptiform potentials, 41(65%) had normal background.
CONCLUSIONS: There is a 19% incidence of epileptiform activity in patients being monitored overnight for the evaluation of LKS. Most discharges are temporal but there is no predominant localization or lateralization. We identified no cases of significant activation of interictal epileptiform discharges, when present, during slow wave sleep. The incidence of LKS in the setting of autistic spectrum patients with or without language regression appears extremely low and overnight EEG monitoring appears warranted only in highly selected cases.