Background: Arbekacin (ABK) is an effective aminoglycoside antibiotic agent against methicillin-resistant Staphylococcus aureus (MRSA) used with the recommendation of therapeutic drug monitoring (TDM) in Japanese clinical practice guidelines for TDM1). Since differences in age and body weight (BW) can affect ABK pharmacokinetics (PK), empirical Bayesian methods using population PK (PPK) models with age and/or BW as covariates are useful for TDM dose adjustment. In this study, among three ABK PPK models reported for pediatric patients (Models 1-3 2-4)), the most appropriate model was selected for TDM of Japanese pediatric patients and the recommended dosing regimen based on the TDM guidelines were evaluated from efficacy and safety points of view by the simulation using the selected model.
Aim: The aim of this study was to select the most appropriate among previously reported PPK models for TDM of Japanese pediatric patients and for the evaluation of dosing regimen based on TDM guidelines.
Methods: All data were collected retrospectively from the pediatric patients admitted to Toho University Omori Medical Center who were treated with ABK due to MRSA infection or suspicion. By using the data, prediction-corrected visual predictive check (pcVPC) was carried out for the external validation of the previously reported PPK models with the covariates of the age, BW, height, postconceptional age (PCA), serum creatinine (Scr). Root mean square prediction error (RMSE) (%) of observed values and individually predicted concentrations (IPRED) for Models 1-3 were compared. For the evaluation of dosing regimen based on the TDM guidelines, the percentage of subject numbers within the target peak (15-20 mg/L) and less than the trough concentrations (<2 mg/L) were estimated for the simulated ABK concentrations in the subjects with various age and Scr. This study was approved by ethical committees of Toho University Omori Medical Center.
Results and discussion: Ten observed ABK plasma concentrations were retrospectively obtained from patient medical records from age ranged from 2 weeks to 3 years. Almost all observed concentrations adjusted by covariates were within the 90th percentile of concentrations simulated by Models 1-3. However, pcVPC plots were essentially distributed around the 50th percentile for Model 3 which achieved the lowest RMSE (%). From the estimated percentage of subject numbers for the target peak and trough concentrations, it was suggested that the dosing regimen based on the TDM guidelines might be partially insufficient.
Conclusions: Model 3 performed best for estimating CL and Vd values of ABK concentrations for pediatric patients in our hospitals. The evaluation procedure identified the most appropriate among three previously reported models. In pediatric, at least 6 mg/kg/day should be given once daily to reach the target peak.
References:
- Kenji O et al., J Infect Chemother., 20:1 (2014)
- Kimura T et al., Antimicrobial Agents Chemotherapy., 48:1159 (2004)
- Kimura T et al., Japanese Journal of Chemotherapy., 51:18 (2003)
- Sato N et al., Jpn J Ther Drug Monit., 24:86-97 (2007)
This work was supported by KAKENHI 18K14993.