Optimizing phenobarbital dosing in critically ill patients with refractory and superrefractory status epilepticus using a population pharmacokinetic model

Abstract

Objective: Current weight-based dosing fails to account for pharmacokinetic variability in refractory and superrefractory status epilepticus (RSE, SRSE). However, understanding pharmacokinetics in critically ill patients with varying degrees of organ dysfunction can improve both safety and efficacy. Hence, this study aims to quantify key pharmacokinetic variabilities to enable individualized dosing in RSE and SRSE.

Methods: Patients with RSE and SRSE admitted to a neurointensive care unit of a tertiary academic center were retrospectively screened for therapeutic drug monitoring (TDM) samples of phenobarbital. Demographics, laboratory data, comedication, and dosing history were collected from electronic health records. Using a nonlinear mixed effects modeling approach via MONOLIX, a population pharmacokinetic model was developed. Optimal dosing regimens were simulated based on estimated parameters, with target attainment calculated for trough plasma phenobarbital concentrations within 18-40 mg/L.

Results: Thirty-seven patients contributed 301 TDM samples. Oral bioavailability (96%), volume of distribution (V; 34.3 L), and total body clearance (CL; .38 L/h) were consistent with nonintensive care literature data. Ideal body weight (IBW) implemented as allometric scaling was the only significant covariate improving model fit, demonstrating a positive correlation with the required phenobarbital dose. Simulations identified optimal 12-h dosing strategies. Oral and intravenous dosing showed minor differences in loading doses but identical maintenance doses, with no significant impact on target attainment for both administration methods. As shown by the coefficients of variation (CVs), intensive care patients exhibited high interindividual (81.36% CV on V, 41.36% CV on CL) and interoccasion variability (36.85% CV on CL), resulting in low target attainment in simulated patients (~40%).

Significance: The pharmacokinetic model characterized phenobarbital pharmacokinetics in patients with RSE and SRSE, showing high oral bioavailability and IBW's impact on V and CL. High pharmacokinetic variability led to low target attainment. Model-informed precision dosing might improve target attainment in the future.

Keywords: individualized dosing; intensive care unit; model‐informed precision dosing; neurointensive care; nonlinear mixed effects modeling; therapeutic drug monitoring.