Clinical randomized controlled observer blinded add-on design. Additionally there will be a non-controlled follow-up phase of the study. Children (Age 4-18 years) with intractable epilepsy, and not eligible for resective surgery will be treated with VNS.

Aim of the study:

1. To evaluate tolerability and effectiveness of VNS in children with intractable epilepsy and cognitive and behavioural problems in a controlled study.
2. To evaluate the effect of VNS on the immune system which, in its turn, will lead to changes in the serotonin metabolic pathway
3. To link the therapeutic effect of VNS to changes in the serotonin (5HT) metabolic pathway.

In addition the investigators hope to detect some markers of immune and neurotransmitter function that enable us to predict 1) Neuronal cell loss in relation to cognitive decline 2) the response to therapeutic treatment of VNS.

Hypothesis:

The investigators aim to explore neuronal correlates for cognitive morbidity in children with intractable epilepsy and to relate this to morphologic changes, biochemical markers, and to epilepsy characteristics.

Correction of the "stressed" pro-inflammatory status of monocytes/macrophages via an electrical stimulation of the vagus nerve will prevent/ameliorate seizures as well as behavioural mood symptoms in children with refractory epilepsy, characterized by the "pro-inflammatory monocyte signature"

Background information:

Repetitive seizures lead to an increase of pro-inflammatory cytokines in the peripheral blood. Experimentally it has been shown that activation of inflammatory cytokines by the peripheral administration of a toxic agents causes sickness behaviour. Pro-inflammatory cytokines interfere with the catabolisation of a precursor of Serotonin (Tryptophan). Tryptophan is catabolised to an endogen NMDA receptor agonist. NMDA (an excitatory neurotransmitter) can lead to neuronal damage.

The Vagus Nerve (VN) plays an important role in the interaction between the neurotransmitter and immune system in which cytokines are crucial. Vagus nerve stimulation (VNS) has an effect on various amino-acid pools in the brain. Patients who respond to VNS show significantly increased serotonin metabolites in their cerebrospinal fluid. VNS is also associated with marked peripheral increases in pro-, and anti-inflammatory circulating cytokines. Given the complexity of the serotonergic system and its interaction with multiple neurotransmitter systems in the human brain it is not surprising to find that serotonin plays a role in the etiology and the course of affective disorders. It is to be expected that the serotonin-immune pathway also plays a role in the course of epilepsy.

Study population:

Children (Age 4-18 years) with intractable epilepsy, and not eligible for resective surgery

Study design:

Clinical randomized controlled observer blinded add-on design. Additionally, and as a non-controlled follow-up of the study, the active control group will receive therapeutic stimulation parameters. In a secondary analysis both VNS groups will be compared with the pre-surgical period (baseline)

Intervention: VNS The generator is implanted beneath the subcutaneous tissue in the upper chest region. The electrode is tunneled from the generator to the stimulation site in the neck. The system is programmed with a computer. The pulse width, output current, signal frequency and stimulation time are programmed telemetrically.

The study group is stimulated with the following parameters: Output current 0.25 milliampere (to be ramped up to max. 1.75 milliampere), Pulse width 0.5 milliseconds, Frequency 30 Hz, Duty cycle: 30 sec on 5 min off (duty cycle 10%).

The active control group is stimulated with: Output current 0.25 milliampere, Pulse width 0.1 milliseconds, Frequency 1 Hz, Duty cycle: 14 sec on 60 min off (duty cycle <0.5%)

Primary endpoint:

Seizure frequency reduction of 50% or more.

Main variables:

Demographics Epilepsy specific data (Seizure frequency (diary), Medication) Biochemical and neuro-immunological assessments in in peripheral blood and CSF(serotonin metabolites, Noradrenalin, GABA, Glutamate, pro-, and anti-inflammatory cytokines).

Neuropsychologic variables (mood, cognition, QoL )

Inclusion Criteria:

• Seizures not adequately controlled by anti-epileptic drugs of first or second choice with adequate and stable serum anticonvulsant concentrations
• Acceptable seizure regulation but intolerable side effects with adequate and stable serum anticonvulsant concentrations
• Not eligible for epilepsy surgery
• Age between 4 and 18 years
• Informed consent

Exclusion Criteria:

• Evidence of a progressive cerebral lesion, degenerative disorder, malignancy or a history with malignancy in the past 5 years
• Unstable medical disease (i.e. cardiovascular, hepatic, renal, musculoskeletal, gastrointestinal, metabolic, endocrine) in the last 2 years
• Documented history with generalized status epilepticus in the past three months
• High risks for complications (obstructive respiratory disease, gastric disorders, cardiac I disorders)
• A history of alcohol or drug abuse, of psychiatric disorder requiring electroconvulsive therapy, chronic use of major tranquillisers (neuroleptics, antidepressants, or MAO inhibitors) in the past 6 months
• Regularly treatment with antihistamines, metoclopramide or CNS-active compounds
• Treatment with an experimental drug during the past 30 days
• Subjects who are schizophrenic or have exhibited any psychotic symptomatology