A comparative pharmacokinetic study of intravenous and intramuscular midazolam in patients with epilepsy.

The pharmacokinetics of midazolam, a water soluble 1,4-benzodiazepine, has been studied in 12 patients (11 male, 1 female; age range 19-57 years) with epilepsy. All patients were taking hepatic enzyme inducing antiepileptic drugs (AEDs) on a regular basis. Midazolam (5 mg) was administered intravenously and 1 week later midazolam was administered intramuscularly, the dose used being dependent on the sedative response to the intravenous dose (10 mg, n = 2; 7 mg, n = 8; 5 mg, n = 2). Serial blood samples were collected at timed intervals for 5-7 h. After intravenous administration initial distribution was rapid with a mean half-life (t 1/2 alpha) of 0.06 +/- 0.03 h followed by a terminal half-life (t 1/2 beta or gamma) of 1.5 +/- 0.3 h. Volume of distribution was 0.62 +/- 0.27 l/kg. After intramuscular administration midazolam was rapidly absorbed with peak serum concentrations achieved at 25 +/- 23 min. Two patients showed delayed absorption. Mean terminal half-life was 2.8 +/- 1.7 h. The absolute bioavailability of intramuscular midazolam was calculated in 11 patients as 87 +/- 18%. Sedation was rapid (less than 1-2 min) but transient (7-75 min) after intravenous and slower (2-30 min) and for a longer period (20-120 min) after intramuscular administration. Since intravenous administration of AEDs including diazepam is not always feasible in status epilepticus there are obvious advantages in having an effective intramuscular formulation. Our data suggest that midazolam may be such a drug.