Time to relapse after epilepsy surgery in children: AED withdrawal policies are a contributing factor.

Aim. It was recently suggested that early postoperative seizure relapse implicates a failure to define and resect the epileptogenic zone, that late recurrences reflect the persistence or re-emergence of epileptogenic pathology, and that early recurrences are associated with poor treatment response. Timing of antiepileptic drugs withdrawal policies, however, have never been taken into account when investigating time to relapse following epilepsy surgery. Methods. Of the European paediatric epilepsy surgery cohort from the "TimeToStop" study, all 95 children with postoperative seizure recurrence following antiepileptic drug (AED) withdrawal were selected. We investigated how time intervals from surgery to AED withdrawal, as well as other previously suggested determinants of (timing of) seizure recurrence, related to time to relapse and to relapse treatability. Uni- and multivariable linear and logistic regression models were used. Results. Based on multivariable analysis, a shorter interval to AED reduction was the only independent predictor of a shorter time to relapse. Based on univariable analysis, incomplete resection of the epileptogenic zone related to a shorter time to recurrence. Timing of recurrence was not related to the chance of regaining seizure freedom after reinstallation of medical treatment. Conclusion. For children in whom AED reduction is initiated following epilepsy surgery, the time to relapse is largely influenced by the timing of AED withdrawal, rather than by disease or surgery-specific factors. We could not confirm a relationship between time to recurrence and treatment response. Timing of AED withdrawal should be taken into account when studying time to relapse following epilepsy surgery, as early withdrawal reveals more rapidly whether surgery had the intended curative effect, independently of the other factors involved.