A unified population pharmacokinetic model of amikacin in paediatric patients from birth to 18 years for model-informed precision dosing
Najla Alabdulkarim, Charlotte Barker, Bairavi Indrakumar, James Hatcher, Orlagh McGarrity, Timothy Best, Iek Leng Cheng, Joseph F StandingAbstract
Background
Amikacin is widely used in paediatric patients, but developmental changes in drug disposition from birth through to adolescence make dose optimization challenging. Existing population pharmacokinetic (PopPK) models are typically developed for specific paediatric age groups, limiting their applicability across the full paediatric age range.
Objectives
To develop a unified paediatric PopPK model of amikacin capable of predicting drug concentrations from birth to 18 years of age and to evaluate its suitability for model-informed dosing.
Methods
Retrospective therapeutic drug monitoring data were obtained from 409 paediatric patients receiving intravenous amikacin at a tertiary children's hospital. A PopPK model was developed using data from 204 neonates, which incorporated mechanistic covariates describing growth and renal maturation, including allometric body-weight scaling, postmenstrual age (PMA), postnatal age (PNA) and age-adjusted serum creatinine. Predictive performance was evaluated for 205 older patients aged 1 month to 18 years, and compared with previously published paediatric amikacin PopPK models. The model developed using the neonatal data was subsequently refitted to the combined dataset to obtain the final parameter estimates.
Results
A two-compartment model adequately described amikacin pharmacokinetics. Clearance was influenced by body weight, PMA, PNA, age-adjusted serum creatinine and concomitant glycopeptide use. The model developed in neonates using mechanistic maturation and renal function covariates successfully predicted amikacin concentrations in older paediatric patients and showed lower prediction bias and improved precision than previously published paediatric amikacin models. The final parameter estimates were obtained by fitting the model developed on neonates to all patients.
Conclusions
A maturation-informed PopPK model successfully predicted amikacin concentrations from birth to 18 years and provides a unified framework for model-informed precision dosing across the paediatric age range.