Abstracts

Rationale: To understand effects of withdrawing Valproic acid (VPA) on the epileptogenic network in patients with intractable epilepsy. In presurgical evaluation of patients with intractable epilepsy, EEG seizure onsets and initial clinical signs are of great significance in the localization of the epileptogenic zone. In patients for possible epilepsy surgery, it is common practice to taper off some of their maintenance antiepileptic drugs (AEDs) during scalp video-EEG (VEEG) and intracranial VEEG to obtain optimal number of clinical seizures for presurgical evaluations and resective surgery. Valproic acid (VPA) is one of the effective AEDs and is used for both localization-related and generalized epilepsies. However, VPA can disturb hemostasis through its effects on platelet counts and functions to significantly increase the risk of bleeding. Therefore at The Hospital for Sick Children, we tapered off VPA at least 10 to 14 days before a planned surgery with invasive VEEG monitoring. We analyzed the difference of seizures on VEEG between on and off VPA to understand the mechanisms of VPA over the epileptogenic network.Methods: We included patients with intractable epilepsy, who completed presurgical evaluations and underwent resective surgeries. The patients were taking one or more anticonvulsants including VPA before the surgery. VPA was tapered over one to two week period and was completely off at least four days before the repeat scalp VEEG study and at least 10 days before invasive VEEG. We performed two scalp VEEG on and off VPA and invasive VEEG off VPA, using subdural grid with or without depth electrodes. We analyzed ictal EEG onset, latency between EEG onset and clinical onset, propagation of EEG for secondary generalization in seizures with similar clinical semiology. Results: We identified two patients with the inclusion criteria. The pathology consisted of tuberous sclerosis and cortical dysplasia one each. Both patients received one or two other anticonvulsants in addition to VPA. The number of seizures obtained during off and on VPA was not significantly different. There was a decreased latency from ictal EEG onset to clinical onset, more than 50% shorter on scalp VEEG off VPA compared to those on VPA. When seizures became secondarily generalized, the duration between initial clinical seizures and secondary generalization became, on the average, 50% shorter off VPA compared those on VPA. In the tuberous sclerosis case with two types of partial seizures with or without secondary generalization on VPA, they merged into a single type of partial seizures with secondary generalization off VPA. Conclusions: The withdrawal of VPA accelerated the propagation from ictal EEG onset to ictal symptomatogenic period. Furthermore, the clinical seizures quickly became secondarily generalized. Vice versa, VPA inhibited and controlled spreading of ictogenic network or inactivated part of the epileptogenic network.