Author(s): Goldsmith IL, Zupanc ML, Buchhalter JR
Abstract Share this page
Abstract PURPOSE: To determine if using more stringent criteria for cryptogenic Lennox-Gastaut syndrome (LGS) would result in an improved prognosis for that group. Cryptogenic, symptomatic, and non-cryptogenic LGS patients without etiology (indeterminate) were compared with respect to seizure and cognitive outcome. METHODS: Retrospective chart review was performed on 245 patients seen at the Mayo Clinic Rochester from 1976 to 1997, with a diagnosis of either LGS or slow spike wave on EEG. LGS was confirmed in 107 (64 male, 43 female) patients. This group was divided into cryptogenic, symptomatic, and indeterminate groups containing 23, 47, and 37 patients, respectively. In this study, cryptogenic patients all had normal development before onset of LGS, absence of dysmorphic features, normal neurologic examination, and normal magnetic resonance (MRI) brain imaging. Of the 107 patients, 74 had >/=3 years of follow-up. RESULTS: LGS onset in the 107 patients occurred at a median age of 4.0 years (range, 0.6-28.9 years). When last seen, 63\% of those with symptomatic LGS had more than three seizures a day compared with 50\% of cryptogenic and 34\% of indeterminate patients. The most common seizure types were tonic (77\%), atypical absence (61\%), and generalized tonic-clonic (56\%). Only three patients, all part of the indeterminate group, were seizure free at last follow-up. CONCLUSIONS: Using stringent criteria in defining the cryptogenic subgroup resulted in no significant difference in seizure outcome. Individuals with a normal cognitive outcome did not segregate into one etiologic subgroup, but did have LGS onset at an older age.
This article was published in Epilepsia
and referenced in Journal of Neurology & Neurophysiology