Population Pharmacokinetic Modelling for Estimation of Remifentanil Metabolic-Ratio Using Non-steady-State Concentrations under Rapidly Adaptive Dosing

PurposeTo predict steady-state metabolite-to-drug concentration ratio (metabolic ratio) for analgesic drug remifentanil, using sparse non-steady-state data from patients with normal or impaired renal function during individualised, highly variable and rapidly adaptive intravenous infusion.MethodsA three-compartment joint parent-metabolite population pharmacokinetic model was developed using concentrations of remifentanil and its metabolite remifentanil acid from two trials. Renal function was included as an important mechanistic covariate. To address the large covariate effect and highly individualised and rapidly adaptive dosing, standardised visual predictive check was conducted on the observations and individualised visual predictive check was conducted on metabolic ratio estimates. The model was used to simulate metabolic ratio distribution in patients with various renal functions.ResultsThe model, including its covariate structure, adequately described the data. The predictive checks allowed informative model evaluation. The predictedmedian (10th - 90th percentile) of remifentanil metabolic ratio was 12.5 (2.4-58.2) for patients with normal or mildly impaired renal function, or 54.3 (12.8-218.4) for patients with moderately or severely impaired renal function.ConclusionsThe methodologies applied here allowed robust estimation of steady-state parameters using non-steady-state sparse data under highly variable adaptive dosing.