A July 1997 assessment of the American Academy of Neurology (AAN) Therapeutics and Technology Subcommittee reviewed vagus nerve stimulation (VNS) for epilepsy, and concluded that it was promising, but not yet established.1 Subsequent to that review, a second multicenter, randomized, controlled clinical trial has shown safety and efficacy of VNS for treatment of intractable partial seizures. Given the importance of this novel therapy for epilepsy, and the new information provided by a second major study, the Therapeutics and Technology Subcommittee requested an update on VNS. Use of VNS in epilepsy has recently been reviewed. [2][3][4] VNS uses intermittent stimulation of the left vagus nerve in the neck to reduce the frequency and intensity of seizures. Mechanism of action of VNS remains uncertain, but stimulation does not induce grossly visible alterations in the human EEG. 5 Recent studies suggest that metabolic activation of certain thalamic, brainstem, and limbic structures may be important in mediating the effect of VNS.6,7 Depletion of norepinephrine in the locus coeruleus attenuates the antiseizure affect of VNS. 8 After open-label studies, 9,10 sponsors of VNS engaged in a multicenter randomized study, named E03, 11 details of which were reviewed in the previous assessment.1 Results of the trial were positive, and the technology was considered promising, but the number of patients receiving VNS was insufficient to achieve Food and Drug Administration approval. Therefore, a second randomized, controlled, multicenter clinical trial, called E05, was undertaken.
12Update on controlled studies. The E05 study 12 of VNS evaluated 254 patients, ages 13 to 60 years, with intractable partial seizures. To be eligible, subjects had to have at least six complex partial, visible partial motor, or secondarily generalized seizures in the month before entry, and to be free from other complicating neurologic, psychiatric, or medical conditions. Individuals with prior VNS were excluded. After a 12-to 16-week baseline, the vagus nerve stimulator electrode was implanted around the left vagus nerve and connected subcutaneously to the subclavicular stimulator device. Two weeks after implantation, patients were randomly assigned to high-or lowstimulation groups. Current was increased over 2 weeks, and maintained for 3 months of treatment. The highstimulation group received stimulation with 500-µsec pulses at 30 pulses per second, on for 30 seconds and off for 5 minutes. Current was increased as tolerated over the next 16 weeks to a maximum of 3.5 mA. Patients receiving low stimulation as an active control received 130-µsec pulses at one per second, on for 30 seconds and off for 3 hours, with mA set to the point of patient perception. All patients were told that they could activate the stimulator with a hand-held magnet to produce a 30-second stimulation train at the start of a perceived seizure; however, only patients in the high-stimulation group actually received magnet-induced stimulation.The 94 patients receiving high ...