Extrapolation of Midazolam Disposition in Neonates Using Physiological‐Based Pharmacokinetic/Pharmacodynamic Modeling

Abstract There is a shortage of data in clinical studies of neonatal populations, which often utilize extrapolation strategies and model simulation techniques to support drug development and clinical applications. This study established an adult physiological‐based pharmacokinetic/pharmacodynamic (PBPK/PD) model in the modeling software and extended it to neonates using pediatric extrapolation strategies based on maturation formulas for plasma albumin and CYP3A4/5. The neonatal model was then utilized to simulate midazolam dosage regimens for sedation in newborns. Individualized validation for adults indicated that 95.1% of predicted concentration values and all area under curve (AUC) values fell within a 2‐fold range. The extrapolated neonatal model showed 84.4% of predicted concentrations within 2‐fold, an absolute average fold error (AAFE) valuen <2, and an average fold error (AFE) value between 0.5 and 1.5, confirming the model's adequacy. The validated neonatal PBPK/PD model indicated that virtual term neonates maintained the target plasma concentration for 25 hours with the recommended dosage (0.06 mg/kg/h, intravenous infusion 12 hours). Premature infants may require a slightly higher dose than the label's recommendation (0.03 mg/kg/h, intravenous infusion 12 hours). Our findings recommend this research strategy based on extrapolation and model simulation for drug dose prediction and optimization in the neonatal population.