Abstracts

RATIONALE: Over a dozen drugs and vagus nerve stimulation are now available for the treatment of epilepsy. To randomly attempt each therapeutic option as mono-or adjunctive- therapy would take decades. Clinical goals are to accurately predict a patient[ssquote]s response to therapy and to combine therapeutics [dsquote]rationally[dsquote]. We review our experience with the combined use of levetiracetam (LEV) and vagus nerve stimulator (VNS).
METHODS: We identified all patients treated concomitantly with VNS followed by LEV. Thirty-two patients were identified. Two patients were excluded: one developed behavioral problems on LEV and one was non-compliant. LEV therapy was added in cases where VNS had no benefit within 6 months, if reduction of seizure frequency was transient, or if seizure freedom was not achieved.
RESULTS: With VNS plus LEV, six (20%) of patients became seizure-free for 5 months to five years. One had seizure recurrence at 10 months. Another seven (23%) had a greater than 90% decrease in seizures for five to 11 months. Eight (27%) had a 25-90% decrease in seizures and nine (30%) had no significant benefit. Response to VNS was highly predictive of response to LEV(Correlation Coefficient 0.7). All patients becoming seizure-free on VNS/LEV had a greater than 50% decrease in seizure frequency to VNS though this was not sustained in one patient, who also had a non-sustained response to LEV. Of the six patients with a greater than 90% reduction in seizure frequency on LEV, three were nearly seizure-free for a time with VNS, two had a sustained 50% decrease in seizure frequency to VNS, and one had minimal change in seizure frequency but marked decrease in intensity and duration. In contrast, eight of nine patients having no response to VNS also had no response to LEV. An intermediate response to VNS typically produced an intermediate response to LEV. Four patients with a 25% response to LEV had either a 25% response to VNS or a non-sustained 50% decrease in seizure frequency. Three patients with a non-sustained response to LEV had a nonsustained response to VNS. One patient with a sustained 50% seizure reduction to VNS had only a 75% declining to 50% seizure reduction to LEV.
CONCLUSIONS: The reduction in seizure frequency to VNS appears predictive of response to LEV. The mechanism of such an interaction is unknown. Presently, it is unclear if the actions of these therapies combined is additive, a simple synergy, or if VNS foretells or [dsquote]enables[dsquote] LEV efficacy. In the latter scenario reversing the order of the therapies, with LEV first would not be predictive of VNS efficacy. However, in the case of an additive interaction, a trial of LEV could provide information on the likelihood of successful outcome with VNS