This is a multicenter, non-blinded, patient preference comparative trial for efficacy of deep brain stimulation (DBS) on drug-resistant epilepsy compared to continued vagal nerve stimulation (VNS) optimization in children with failed VNS. The two conditions being compared are therefore DBS (treatment) versus VNS (control). Fifty (50) patients will be recruited and enrolled in this pilot study (25 from The Hospital for Sick Children and 50 from CHU Sainte-Justine).

For children with epilepsy that have failed pharmacological and alternative treatments (drug-resistant epilepsy; [DRE]), surgical interventions may be considered. This includes vagus nerve stimulation (VNS) and deep brain stimulation (DBS). VNS for the treatment of DRE in children is an established and widely used treatment. Unfortunately, the positive response to VNS rate (>50% reduction in seizures) is not consistent, ranging from 26-55% in pediatric epilepsy patients. In children with failed VNS, defined as no improvement in seizure control after at least 1 year of treatment, they may undergo DBS as a recommended therapy. DBS is a safe and established treatment for various childhood neurological conditions and the indications for DBS in children continue to expand. It is currently unknown which children may benefit from DBS after failed VNS.

As a patient-preference randomized trial, patients and their parents will be introduced to the options of continuing with current VNS management (control arm) or trialing DBS (treatment arm). Patients and their parents will fall into three possible groups according to preference and willingness for randomization.

i) Patients with no strong preferences and consent to randomization ii) Patients with a preference, yet still consent to randomization iii) Patients who refuse randomization and opt for enrollment in a specific arm

Patients in the treatment arm will receive DBS of the centromedian nucleus. The centromedian nucleus is believed to reduce electrocortical activity in generalized epilepsy. The investigators hypothesize that stimulating this target will lead to a decrease in seizure severity and frequency in patients who have failed VNS; in comparison with patients who will have continued VNS treatment and optimization. Patients in the control arm will continue to be observed for 12 months with no change to their treatment.

Inclusion Criteria:

1. Female or Male patients 8 years of age and over but not including 18 year olds.
2. Diagnosis of drug-resistant epilepsy with failure after trial of two anti-epileptic medications (as defined by Kwan et al. 20093). All children screened for entry into the study will be re-diagnosed by a neurologist prior to entry.
3. Failure of vagal nerve stimulation, defined as the same or increased frequency and severity of seizures at 12 months or more after instigation and optimization of therapy. Failure is defined objectively and subjectively. Objective evidence includes caregiver logs, clinician assessment, or neuromonitoring if the clinician has documented a baseline status prior to instigation of vagal nerve stimulation. Subjective measures include family or patient opinion that seizure frequency or severity has not improved.
4. Parents or legal guardians, including caregivers, who are informed and able to give written consent.
5. Ability to comply with all testing, follow-ups and study appointments and protocols for 12 months following the end of the duration of the study.

Exclusion Criteria:

1. Substance dependence or abuse in the last 6 months, excluding caffeine and nicotine
2. Any contraindication to MRI scanning. A preoperative MRI scan is essential to planning DBS and therefore any contraindication to MRI is a contraindication to enrollment in the study.
3. Unwillingness or inability to return to SickKids for follow-up visits.

There is not a plan to make IPD available.