Abstracts

Rationale: A seizure free interval or "honeymoon period" after initiation of antiepileptic drugs (AEDs) has been described.1 Likewise, an increased seizure free interval has been described following reinstitution of AEDs following discontinuation or reduction during admission for video encephalography (EEG). We sought to examine whether a similar "honeymoon period" was present after abrupt withdrawal of AEDs in patients reporting daily seizures prior to admission for video EEG. Methods: We reviewed all patients admitted to the Cook Children's Epilepsy Monitoring Unit for an epilepsy surgery evaluation since January 2010 that reported at least one seizure type that occurred daily (target seizure) prior to admission. Patients not on AEDs or with nonepileptic seizures were excluded. Demographic data, epilepsy characteristics, and treatment history were abstracted from the medical record. We compared time to target seizure occurrence between patients with complete discontinuation of AEDs, those with partial wean of AEDs, and those remaining on baseline AEDs throughout their video EEG. We also analyzed for effect of patient characteristics on time to target seizure. Results: One hundred and fourteen patients with daily seizures were identified. Thirty-one were excluded for lack of consent and 12 met exclusion criteria; thus, 71 patients were included in the sample (mean age 8.3years). Eleven (15%) had generalized epilepsy, 43 (61%) localization-related epilepsy, and 17 (24%) mixed-type epilepsy. Thirteen (18%) were on 1 AED, 30 (42%) were on 2, and 28 (40%) were on > 3 AEDs at the time of admission. Thirty seven (52%) patients had AEDs completely discontinued, 14 (20%) had a partial AED changes, and 20 (28%) had no change. There were no significant differences in patient characteristics between groups. Between group comparisons indicated a significant difference between the percent of patients with discontinued AEDs that had a seizure on day 1 (62.2%) and those with no change to AEDs at all (95.0%). While the median days to seizure in all groups was 0, the mean time to seizure occurrence was 0.51 for those with complete AED withdrawal, 0.86 for those with partial changes, and 0.05 for those with no change (F(2,68)=3.714, p=.029). Conclusions: This study demonstrates abrupt withdrawal of AEDs in patients with daily seizures results in a period of seizure freedom. The time to seizure occurrence was delayed by approximately one day in some patients when AEDs were discontinued upon admission. In patients with daily seizures undergoing video EEG, alteration of AED regimens is likely unnecessary and may result in prolongation of time until seizures occur, unless less frequently experienced seizure types are targeted for evaluation. The presence of a "honeymoon period" following initiation and withdrawal of AEDs presents the possibility that intermittent dosing of AEDs may be a method to improve efficacy for some patients with frequent seizures. Additional studies examining this phenomenon are warranted.