Abstracts

Rationale: Rectal diazepam has been an outpatient rescue therapy for seizure clusters for 2 decades but is limited by social considerations and PK variability. Diazepam nasal spray (Valtoco®) is now approved for acute treatment of seizure clusters in patients ≥6 years with epilepsy. Both product labels permit a second dose after 4 hr if needed. This schedule is based on safety and therapeutic estimates from the 1990s; more recent clinical diazepam nasal spray data included shorter dosing intervals, without safety signals. We developed a population PK model of diazepam nasal spray, identified sources of interindividual variability (IIV), and performed simulation studies to estimate diazepam exposures after a second dose within shorter dosing intervals.

Methods: Diazepam PK data were analyzed from 3 phase 1 studies in healthy volunteers or patients with epilepsy. All participants received ≥1 dose of diazepam nasal spray (5, 10, 15, or 20 mg) based on age (6-11 years and ≥12 years) and weight. Population PK models were run using nonlinear mixed-effects modeling, with first-order conditional maximum likelihood estimation with interaction in NONMEM. In the covariate model, all relevant covariates with observed bias were tested separately, and all significant covariates were collectively added for the full covariate model. Parameter-covariate relationships were tested with backward selection. An exposure-response analysis was planned for adverse events of interest after single and repeated doses in patients with epilepsy.

Results: The final dataset included PK measurements from 126 individuals (Table), some receiving >1 dosage. A two-compartment open PK model with first-order input and first-order elimination adequately fit the data. The model included clearance (CL), volume of distribution in central (V2) or peripheral compartments (V3), intercompartmental clearance (Q), and first-order absorption rate constant (k_a); weight was added as an allometric covariate. A final model that fit the observed PK was generated when population (volunteers or patients) was included as a covariate for k_a. Point estimates of IIV were 41.8%, 45.5%, 47.1%, 66.2%, and 36.2% for CL, V2, V3, Q, and k_a, respectively. Overall, >90% of the observed data fell within the range of the 5^th and 95^th percentiles of the predicted data. In a simulation using the final PK model, a second dose of diazepam nasal spray increased maximum concentration (~65%) and total exposure (~100%) (Figure). The predicted exposures from dosing intervals of 0.5–4 hr overlapped. There was no relationship between adverse events and number of doses, and so the exposure-response analysis was not performed.

Conclusions: The PK of diazepam nasal spray is adequately described by a two-compartment model with first-order absorption and first-order elimination. This pattern was similar in healthy volunteers and patients with epilepsy. Dosing intervals of ≤4 hr were not associated with more safety events and are predicted to lead to comparable diazepam exposures.

Funding: Please list any funding that was received in support of this abstract.: Neurelis, Inc.