One-year versus three-year treatment of children with uncomplicated epilepsy : A prospective study

Abstract: One-year versus Three-year Treatment of Children with Uncomplicated Epilepsy. A prospective study. Gunnar Braathen Department of Pediatncs, Huddinge University Hospital The main purpose of this prospective study was to investigate whether one-year treatment was equally efficient, regarding the remission rate, as three-year treatment of children with uncomplicated epilepsy. Our aim was also to find any prognostic factors related to the remission rate which would enable us to predict the outcome for the individual child. The duration of treatment was randomized to one year (Group 1) or three years (Group 11) before initiation of drug treatment in 234 children with newly diagnosed epilepsy, aged 2-16 years, with no neurological disabilities. Seizures and epileptic syndromes were classified according to the well-established ILAE system. Partial epilepsies were found in 137 children (59%), 68 children (29%) had generalized epilepsies and in 29 children (12%) it was uncertain whether the epilepsy was partial or generalized. Six main diagnostic groups were identified: simple parbal seizures (SPS 8%), complex partial seizures (CPS, 23%), benign partial epilepsy with centrotemporal (rolandic) spikes (BECT, 9%), generalized tonic-clonic seizures as the only ictal manifestation in children with rolandic spikes (2°GTCS-rol, 13%), absence epilepsy (AE, 12%) and pnmarily generalized tonic-clonic seizures (1°GTCS, 19%). The drug therapy could be completed in 161 children (78%), according to the planned schedule and the criterion of freedom from seizures during the last six months of treatment. The children were followed up for 2-10 years after the end of treatment. Relapses occurred in 60 children (37%). The relapse rate was 47% in Group I and 29% in Group ll (p <0.05). The relapse rate in children with CPS was significantly higher in Group I than in Group 11 (p <0.05). If children with CPS were not considered, the difference in outcome between the groups was not statistically significant. Children with BECT and SPS had the highest remission rates, children with CPS had the lowest and children with other types of epilepsy had remission rates in between. The EEG was recorded before treatment and at regular intervals during treatment. The existence of generalized irregular spike-and-wave (SW) activity after one year of treatment was associated with a poor outcome, as compared with children without these discharges (p <0.01). Persistent, regular 3-Hz SW after six months of treatment in children with absence epilepsy was also associated with a high relapse risk. Other epileptiform activities or background abnormalities did not influence the outcome. Prognostic factors of importance to the outcome were analysed by univariate and multiple regression analyses in 161 children from whom treatment was withdrawn. Factors found to have a positive influence on the outcome were (I) epilepsy type of BECT or SPS, (2) favourable age at seizure onset (younger than 10 years in children without rolandic spikes and older than 10 years in children with rolandic spikes) and (3) EEG recordings after one year of treatment free from irregular SW. These variables constituted a prognostic model with a simple scoring system, by which each child could receive a total score of 0-7. Children with a total score of 5-7 after one year of treatment had a relapse rate of 27% in Group I and 3% in Group 11. Discontinuation of the treatment could be recommended for these children at that time, while children with a total score of 0-2 had a high relapse risk (90% and 57% in Groups I and 11, respectively) and should continue treatment for at least two more years. In order to look for subtle, neurological impairments due to the drug therapy 19 children treated with carbamazepine (CBZ) were studied with a standardized test of fine and gross motor proficiency (Bruininks- Oseretsky test). The test was performed on two occasions on the same child, first during treatment and then after six months, when treatment had been completely withdrawn for three months. Significant improvements were found in response speed (p <.0.05), in all fine motor subtests together (p <0.01) and in the total test battery (p <0.05). A comparison group was tested on two occasions during treatment with CBZ. No improvements were found in this group. The results indicate side effects from CBZ treatment on motor proficiency and on fine motor functions in particular. The annual incidence of childhood epilepsy in the catchment area of Huddinge University Hospital was studied. The mean incidence during the period 1990-92 was 53 per 100,000 children, which was in accordance with the experience from similar studies. The results of this study showed that 65% of a population of children with epilepsy fulfilled the inclusion cnteria of the principal study. ISBN 91-628-2359-0 Huddinge 1997