Abstracts

RATIONALE: Two specific epileptic syndromes have a sleep activated epileptiform EEG consisting of spike and wave discharges in greater than 85% of NREM sleep: Landau-Kleffner Syndrome (LKS) and continuous spike and wave during slow wave sleep (CSWS). LKS is characterized by language regression and CSWS by neuropsychologic (NP) regression. The sleep-activated EEG pattern may occur independently of a specific epileptic syndrome. Because effective therapy early on in the course may minimize neurocognitive dysfunction, we have analyzed our cases to determine the best therapy (TH) regimen.
METHODS: We identified 12 children with a sleep-activated EEG, all with greater than 85% spike waves during sleep, with follow-up ranging from 2 to 6 years. All had cranial MRI and serial sleep EEGs and most had cognitive testing. Initial TH with standard AEDs was followed by prednisone (PRED), if needed, at 2 mg/kg/day, with tapering doses monthly over 6 months. Other therapies included ACTH, the ketogenic diet (KGD), intravenous gammaglobulin (IVIG), lesionectomy, vagal nerve stimulator (VNS), or multiple subpial transection (MST).
RESULTS: There were 6 males and 6 females. 7 were classified as symptomatic and 5 idiopathic. The age of onset was between 14 months and 8 years, presenting with either seizures, or language or behavioral regression. Seizure types were focal in 8, generalized in 10, and mixed in 6. Five children had LKS and 7 had CSWS. Neuropsychologic abnormalities seen included global deterioration in 5, behavioral problems in 10, hyperkinesia in 8, and language impairment in 9. 9 had focal EEG abnormalities. Word deafness occurred in all with LKS and in 6/7 with CSWS. One child responded to AEDS only (valproate), after early recognition of CSWS. After TH with Pred (N=6), either alone or with IVIG (N=2), or high-dose valium (N=2), CSWS disappeared in 5 cases. CSWS recurred in 4 cases: in 3 cases between the 3rd and 5th months, and one case recurred after 15 months, who responded to diamox but not a repeat course of PRED. 4/8 with bilateral or multifocal spike wave discharges in the waking state required multiple TH modalities. ACTH was used in 2 and KGD in 4, without response. 4 underwent a surical procedure (2 VNS, 1 lesionectomy, and 1 MST).
CONCLUSIONS: PRED has had the best TH response for CSWS, but there is a high risk of recurrence which may require another course of PRED. Epilepsy surgery may be beneficial in selected cases. CSWS may be difficult to control when the waking EEG shows multifocal or bilateral epileptiform activity.