A Population Pharmacokinetic Model of Gentamicin in Pediatric Oncology Patients To Facilitate Personalized Dosing

Abstract

To ensure the safe and effective dosing of gentamicin in children, therapeutic drug monitoring (TDM) is recommended. TDM utilizing Bayesian forecasting software is recommended but is unavailable, as no population model that describes the pharmacokinetics of gentamicin in pediatric oncology patients exists. This study aimed to develop and externally evaluate a population pharmacokinetic model of gentamicin to support personalized dosing in pediatric oncology patients. A nonlinear mixed-effect population pharmacokinetic model was developed from retrospective data. Data were collected from 423 patients for model building and a further 52 patients for external evaluation. A two-compartment model with first-order elimination best described the gentamicin disposition. The final model included renal function (described by fat-free mass and postmenstrual age) and the serum creatinine concentration as covariates influencing gentamicin clearance (CL). Final parameter estimates were as follow CL, 5.77 liters/h/70 kg; central volume of distribution, 21.6 liters/70 kg; peripheral volume of distribution, 13.8 liters/70 kg; and intercompartmental clearance, 0.62 liter/h/70 kg. External evaluation suggested that current models developed in other pediatric cohorts may not be suitable for use in pediatric oncology patients, as they showed a tendency to overpredict the observations in this population. The final model developed in this study displayed good predictive performance during external evaluation (root mean square error, 46.0%; mean relative prediction error, Ϫ3.40%) and may therefore be useful for the personalization of gentamicin dosing in this cohort. Further investigations should focus on evaluating the clinical application of this model. KEYWORDS pharmacokinetics, gentamicin, NONMEM, pediatrics, oncology, pharmacometrics, population pharmacokinetics F ebrile neutropenia induced by chemotherapy is a common complication in pediatric oncology patients. Neutropenic patients are more susceptible to the development of infections (1). Sepsis is the primary cause of mortality and morbidity in pediatric oncology patients with febrile neutropenia (2). The rate of mortality due to sepsis is 1.6-fold higher for oncology pediatric patients than it is for other pediatric patients (3). Aminoglycoside antibiotics, such as gentamicin, in combination with other broad-spectrum antibacterial agents play an important role in managing infectious complications in these individuals and are used as second-line therapies when treating Gram-negative bacterial infections and when resistance to first-line agents develops (4).Gentamicin has a narrow therapeutic window and displays large pharmacokinetic variability. High levels of and prolonged exposure to gentamicin has been associated with nephro-and ototoxicity (5, 6). Pediatric oncology patients often receive long

Population Pharmacokinetic Modeling of Gentamicin in Pediatrics

“…Allometric weight scaling was included on all PK parameters. Several previously published models developed in pediatric populations with similar ages to those in this study were limited by failure to recognize the true compartmental structure or to include the impact of body size descriptor, physiological maturation of renal function, and the maturational changes of body composition as well as the deviation from normal renal function . In comparison with previous reports, the final model developed here displayed good predictive performance when it was applied to the model evaluation data set.…”

“…In the first approach to external evaluation, 5 population PK models of gentamicin developed for pediatric populations of similar age distribution as this data set were evaluated on the data set used for model evaluation (Table ) . These models, together with the final model from this study, were tested for their predictive performance with the external data set: pcVPC obtained from all 6 models generated during external evaluation are displayed in Figure .…”

“…Different body size descriptors were used in previous studies . Based on our model, TBW had better predictive performance than FFM and NFM ().…”

“…Simulated dose values suggest that for young children, particularly those less than 2 years old, an increase in the body weight–normalized dose is required to provide exposure comparable to that of an older child. For example, the model‐predicted dose for a 6‐month‐old child weighing 6.3 kg would be 3.1 mg/kg every 24 hours, which is a 6% to 20% dose decrease as compared to the model‐predicted dose of 3.3‐3.9 mg/kg once‐daily regimen using earlier models …”

“…Therefore, FFM was included in the final model. Llanos et al selected FFM as the body size descriptor because FFM fit the data best compared to TBW and NFM, even though TBW was able to explain larger between‐subject variabilities of gentamicin CL, V 1 , and V 2 . In their study there was no significant improvement (ΔOFV = –3.5) in model fitting between using TBW and FFM in the final covariate model …”