Background: Seizures in childhood absence epilepsy (CAE) are typically easily controlled with anti-seizure medications (ASMs). Factors predictive of treatment resistance remain unclear. Our objectives were to assess prevalence of neuropsychiatric problems and factors influencing treatment resistance in a cohort of CAE at a single centre. Methods: We retrospectively reviewed patients with CAE diagnosed between January 1999 and December 2016 with at least 1-year follow-up. Treatment resistance was defined as failure to respond to two appropriate ASMs. Exclusion criteria included eyelid myoclonia with absence, myoclonic absence, and generalized tonic-clonic (GTC) seizures prior to absence seizures. Results: The study population comprised 164 patients (65 males) with mean age at seizures onset of 6.1years. 21% had treatment-resistant seizures. The first ASM was Ethosuximide (63.4%), Valproic acid (23.2%), and Lamotrigine (6.7%). 32.9% of children had learning disabilities, 28.7% ADHD, and 12.8 % anxiety.

A stepwise binary logistic regression analysis identified GTC seizures, learning disability (LD) and ongoing general spike and wave on EEG as predictors of treatment resistance. At last follow-up (average of 5.4 years), 43.3% of children were seizure-free off ASMs. Conclusions: 21% of children with CAE had treatment-resistant seizures. LD, GTC and failure of normalization of the EEG were associated with treatment resistance.