A population pharmacodynamic Markov mixed‐effects model for determining remimazolam‐induced sedation when co‐administered with fentanyl in procedural sedation

The clinical effects of remimazolam (an investigational, ultra-short acting benzodiazepine being studied in procedural sedation) were measured using the Modified Observer's Assessment of Awareness/Sedation Scale (MOAA/S). The objective of this analysis was to develop a population pharmacokinetic/pharmacodynamic model to describe remimazolam-induced sedation with fentanyl over time in procedural sedation. MOAA/S from 10 clinical phase I-III trials were pooled for analysis, where data were collected after administration of placebo or remimazolam with or without concomitant fentanyl. A Markov model described transition states for 35,356 MOAA/S-time observations from 1071 subjects. Effect-compartment models of remimazolam and fentanyl linked plasma concentrations to the Markov model, and drug effects were described using a synergistic maximum effect (E_max ) model. Simulations were performed to identify the optimal remimazolam-fentanyl combination doses in procedural sedation. Fentanyl showed synergistic effects with remimazolam in sedation. Increasing age was related to longer recovery from sedation. Patients with body mass index greater than 25 kg/m² had ~30% higher rates of distribution from plasma to the effect site (keo), indicating a slightly faster onset of sedation. Simulations showed that remimazolam 5 mg was more appropriate than 4 or 6 mg when administered with fentanyl 50 μg. The model and simulations support that a combination of remimazolam 5 mg with fentanyl 50 μg is an appropriate dosing regimen and the dose of remimazolam does not need to be changed in elderly patients, but some elderly patients may have a longer duration of sedation.