Simplified Dosing Regimens for Gentamicin in Neonatal Sepsis

Abstract

Background: The effectiveness of antibiotics for the treatment of severe bacterial infections in newborns in resource-limited settings has been determined by empirical evidence. However, such an approach does not warrant optimal exposure to antibiotic agents, which are known to show different disposition characteristics in this population. Here we evaluate the rationale for a simplified regimen of gentamicin taking into account the effect of body size and organ maturation on pharmacokinetics. The analysis is supported by efficacy data from a series of clinical trials in this population. Methods: A previously published pharmacokinetic model was used to simulate gentamicin concentration vs. time profiles in a virtual cohort of neonates. Model predictive performance was assessed by supplementary external validation procedures using therapeutic drug monitoring data collected in neonates and young infants with or without sepsis. Subsequently, clinical trial simulations were performed to characterize the exposure to intra-muscular gentamicin after a q.d. regimen. The selection of a simplified regimen was based on peak and trough drug levels during the course of treatment. Results: In contrast to current World Health Organization guidelines, which recommend gentamicin doses between 5 and 7.5 mg/kg, our analysis shows that gentamicin can be used as a fixed dose regimen according to three weight-bands: 10 mg for patients with body weight <2.5 kg, 16 mg for patients with body weight between 2.5 and 4 kg, and 30 mg for those with body weight >4 kg. Conclusion: The choice of the dose of an antibiotic must be supported by a strong scientific rationale, taking into account the differences in drug disposition in the target patient population. Our analysis reveals that a simplified regimen is feasible and could be used in resource-limited settings for the treatment of sepsis in neonates and young infants with sepsis aged 0-59 days.

Keywords: bacterial infection; dosing optimization; gentamicin; modeling and simulation; neonatal sepsis; pharmacokinetics; resource-limited and remote setting.

FIGURE 1

Population pharmacokinetics of gentamicin in a large cohort of preterm and term neonates, as described by Fuchs et al. (2014) Absorption was assumed to be instantaneous and suitable for the description of gentamicin profiles following intramuscular administration.

FIGURE 2

(A) Workflow for the implementation of clinical trial simulations aimed at the evaluation of simplified dosing regimens for gentamicin in neonatal sepsis. (B) Design characteristics for the study protocol used in the clinical trial simulations. (C) Simulation scenarios evaluated for the identification of a simplified dosing regimen based on weight bands and fixed doses.

FIGURE 3

Predicted gentamicin AUC in sepsis patients aged between 0–59 days. Estimates are summarized according to the weight bands used in the original AFRINEST/SATT trials. Hinges represent 25^th and 75^th percentiles (respectively, Q1 and Q3), whiskers represent Q1 − 1.5IQR and Q3 + 1.5IQR, respectively, where IQR is the interquartile range. All the subjects outside this range are represented by the dots (N = 9,994).

FIGURE 4

Predicted gentamicin peak (left) and trough (right) concentrations in sepsis patients aged between 0 and 59 days. Estimates are summarized according to the weight-bands used in the original AFRINEST/SATT trials. Hinges represent 25^th and 75^th percentiles (respectively, Q1 and Q3), whiskers represent Q1 − 1.5IQR and Q3 + 1.5IQR, respectively, where IQR is the interquartile range. All the subjects outside this range are represented by the dots (N = 9,994). Dashed lines represent the threshold values for peak and trough concentrations of 10 and 2 mg/L, respectively.

FIGURE 5

(Left) Predicted gentamicin concentration vs. time profile in sepsis patients aged between 0 and 59 days. Each panel compares the pharmacokinetic profiles in the target population after the proposed regimen (red) with those obtained after higher (upper panel) and lower doses (lower panel) of the WHO recommended regimen (blue). Solid line depicts the median profile; shaded area represents the 90% prediction intervals. Time 0 is used as reference for the first dose (Right) Systemic exposure, expressed as area under the concentration vs. time curve. Hinges represent 25^th and 75^th percentiles (respectively, Q1 and Q3), whiskers represent Q1 − 1.5IQR and Q3 + 1.5IQR, respectively, where IQR is the interquartile range. All the subjects outside this range are represented by the dots (N = 9,994). Overall, the weight-banded regimens result in similar exposure ranges, with a slight trend to lower values in the highest weight band.

FIGURE 6

Predicted gentamicin peak (left) and trough (right) concentrations in sepsis patients aged between 0 and 59 days stratified according to the weight bands for the proposed simplified regimen. Panels show how the simplified regimen compares to the 2015 WHO recommendations. Hinges represent 25^th and 75^th percentiles (respectively, Q1 and Q3), whiskers represent Q1 − 1.5IQR and Q3 + 1.5IQR, respectively, where IQR is the interquartile range. All the subjects outside this range are represented by the dots (N = 9,994). Dashed lines represent the threshold values for peak and trough concentrations of 10 and 2 mg/L, respectively.