Effect of intracerebroventricular continuous infusion of valproic acid versus single i.p. and i.c.v. injections in the amygdala kindling epilepsy model.

Two protocols were tested to assess anticonvulsant efficacy and drug concentrations after intracerebroventricular (i.c.v.) continuous valproic acid (VPA) infusion, as compared with acute injections in the kindling epilepsy model. Protocol 1: amygdala-kindled rats were injected via intraperitoneal (i.p.) and i.c.v. routes with varying doses of VPA and tested for seizure intensity, afterdischarge and seizure duration, ataxia and sedation. Concentrations of VPA were determined by immunofluorescence in the brain, plasma, cerebrospinal fluid (CSF) and liver in matching rats. Protocol 2: amygdala-kindled rats were implanted with osmotic minipumps containing a VPA solution in saline and connected to intraventricular catheters for 7 days. Seizure threshold, latency and duration, afterdischarge duration, ataxia and sedation were recorded daily before, during, and until 5 days after VPA infusion. In matching animals, CSF, brain, plasma and liver VPA concentration was determined. Acute i.c.v. VPA injection suppressed seizures with a remarkable ataxia and sedation. However, continuous i.c.v. infusion controlled generalised and even focal seizures without producing important side effects, high plasma levels or hepatic drug concentrations. In conclusion, continuous i.c.v. VPA infusion may protect against kindled seizures by minimising ataxia and sedation, and achieving suitable intracerebral, yet low plasma or hepatic drug concentrations, thus avoiding potential systemic toxicity.