Covariates of amikacin disposition in a large pediatric oncology cohort

Objective: Amikacin pharmacokinetics (PK) in children display large variability due to maturational and disease-related covariates. The objective was to explore amikacin PK in a large pediatric oncology cohort, taking into account within-patient changes. Materials and methods: Clinical data and amikacin therapeutic drug monitoring (TDM) observations were collected retrospectively from children with an oncology diagnosis receiving amikacin during febrile neutropenia. Individual amikacin PK parameters were calculated using a 1-compartment model with instantaneous input and first-order output. This approach was selected based on a pragmatic study design using TDM from routine clinical care, with availability of 2 TDM samples per treatment episode. To explore covariates of clearance (Cl) and volume of distribution (Vd), linear mixed models were used, modelling a random effect for patient to account for clustering due to repeated measurements. Results: Based on 188 amikacin treatment episodes in 114 patients, median (interquartile range) amikacin Cl was 1.37 (1.05; 2.46) L/h and Vd 7.98 (5.66; 12.73) L. Height and creatinemia were significant covariates for Cl (marginal R-2 71.1%), while weight, height, and creatinemia determined Vd (marginal R-2 59.5%). Conclusion: We described extensive variability of amikacin PK in a large cohort of pediatric oncology patients, including within-patient changes across treatment episodes. Maturational covariates and creatinemia determined amikacin Cl and Vd, while primary non-maturational covariates were not significant. Our observations, based on combined clinical and PK data in children with oncology diagnoses, can be useful to feed dosing software programs to improve drug exposure in special populations.