Epilepsy in infancy study: a population based study on epilepsies with onset in the first two years of life

Children with the onset of epilepsy under the age of two years are a heterogeneous group with particularly poor outcome reported by studies mostly originating from specialist hospital / clinic settings. \nA population based study was designed to determine the incidence of epilepsy onset under the age of two years and to ascertain a cohort for long-term follow up. This cohort observation aimed to determine usefulness of the current international classification of epilepsies in this age group, associated structural brain abnormalities, predictors of developmental function close to diagnosis; and factors that are associated with a longitudinal change of developmental scores after 12 months. \nCases were prospectively identified from residents in 15 boroughs of North London involving hospital and community based paediatricians. Information was obtained either by questionnaire anonymously or from clinical assessments of children enrolled in the observational cohort. 57 cases were enrolled giving a crude annual incidence of 54 (95% CI 41-69) /100.000 children under the age of 2 years. A two source capture recapture model determined an adjusted annual incidence of 56.3 – 88.5 (95% CI) / 100.000 <= 2 years. \nClinical assessment of children in the observational cohort at baseline and 12 months follow up included neurodevelopmental evaluation using standardised tools, central review of EEG’s and neuroimages. Data of all subjects enrolled were independently evaluated by two paediatric neurologists. \nAlthough most cases were classified under epilepsy syndrome groups, a specific epilepsy syndrome diagnosis could not be allocated in a third of cases and there was only moderate inter-rater agreement (kappa scores: 0.48, 0.5). \nReview of neuroimages of 52 children (91%) demonstrated a high yield of aetiologically relevant abnormalities in 26 (50%) with common occurrence of developmental malformations identified in 11 (43%). \nMultivariate regression analysis showed that 'abnormal neurological examination' and 'presence of interictal discharges on EEG' significantly and independently predicted lower developmental function close to diagnosis. There was no significant difference between initial and developmental function on follow up after adjusting for initial infantile spasms, normal or abnormal initial EEG, seizure status at follow up, structural brain abnormalities, and antiepileptic medication taken, suggesting that the initial status determined the function after 12 months follow up.