Abstracts

RATIONALE: The conventional treatment strategy in newly diagnosed childhood epilepsy does not take into account prognostic factors of the child to be treated. By assessing the prognosis, identifying relevant variables at intake and during the early course and comparing the results of short-term and long-term follow-up we may enable physicians to design different treatment strategies for children with a poor or with a good prognosis.
METHODS: In 466 children out of the cohort of the Dutch Study of Epilepsy in Childhood, we studied the prognosis 2 years after intake [ref]. The outcome after 5 years of these 466 children was defined in terms of terminal remission (TR). The relation of intake and early (6-month) follow-up variables with the outcome after 5 years was determined with univariate and multivariate logistic regression analysis. The outcomes after 2 and 5 years were correlated, and possible reasons for discrepancies were looked for.
RESULTS: 5 Children were lost and 8 died, leaving 453 followed for 5 years. Of these, 345 (74%) attained a TR of at least 1, 290 (62%) of at least 2 years. Of the 108 children with a TR [lt] 1 year, 24 had a seizure frequency of at least 1 per month during the 5th year. 66 Children (14%) were not treated with AEDs (94% achieved a TR [gt] 1 year), 211 (45%) were treated with only 1 AED (TR [gt] 1 year 84%) and 189 (41%) with more than 1 drug either sequentially or as polytherapy (TR [gt] 1 year 56%). In the multivariable analysis, remote symptomatic or cryptogenic etiology, post-ictal signs, history of febrile convulsions and not achieving a 3-month remission during the first 6 months after intake were associated with a TR [lt] 1 year after 5 years. 28 Children (6%) had a worse outcome after 5 than after 2 years. Older age at onset and remote symptomatic etiology were associated with this course. A better outcome after 5 than after 2 years (112 children [24%]) was associated with simple partial seizures. On the other hand, 232 children (50%) had an unchanging benign course with a TR [gt] 1 year after 2 and 5 years, and 81 children (17%) did neither achieve a 1-year remission after 2 nor after 5 years. The variables predictive for such a course were largely identical to the variables that were significant in the entire cohort.
CONCLUSIONS: 65% Of children have a TR of at least 2 years, and 3/4 of at least 1 year after 5 years of follow-up. Etiology and early course were the variables best predicting the outcome. A change of the course of the epilepsy for better or for worse is associated with other variables than those predicting outcome in the entire cohort.
Ref. Arts WFM, Geerts AT, Brouwer OF et al. The early prognosis of epilepsy in childhood: the prediction of a poor outcome. The Dutch Study of Epilepsy in Childhood. Epilepsia 1999;40:726-34.
Support: Dutch National Epilepsy Foundation (grant no[ssquote]s A72, A86 and 96-12)