Foetal and neonatal exposure prediction and dosing evaluation for ampicillin using a physiologically‐based pharmacokinetic modelling approach

Abstract

Aims: Ampicillin is frequently used in neonates for the treatment of sepsis and as an intrapartum prophylaxis option for Group B Streptococcus. Pharmacokinetic data to guide ampicillin dosing in neonates and during the intrapartum period are limited.The objective of this study was to build a physiologically-based pharmacokinetic (PBPK) model to characterize the disposition of ampicillin in neonates and foetuses and to inform corresponding optimal dosing regimens.Methods: An adult ampicillin PBPK model was first developed using the Simcyp ® simulator. The adult model was then scaled to neonates by accounting for maturational changes in physiological parameters and age-dependent drug disposition or extended to a pregnancy model for mothers and foetuses. Models were verified using collected mean or individual-level concentration data from the literature. Results:The developed adult PBPK model included elimination via glomerular filtration, OAT3-mediated tubular secretion and biliary excretion as well as hepatic metabolism, and 89.8% of the observed mean concentrations in adults were within a 2-fold range of model mean predictions. Most of the observed individual-level observations in neonates (78.4%) and foetuses (about 65% in two studies) were within the 90% prediction intervals. The recommended 50 mg/kg every 8 h (q8h) ampicillin regimen achieved the 75% fraction time of total drug concentration above minimum inhibitory concentration (T > MIC) target for an MIC ≤8 mg/L in >90% virtual neonates, and 1 g ampicillin for pregnant women provided adequate foetal exposure (>0.25 mg/L) for 4 h prior to delivery.Conclusions: A PBPK model was developed to characterize ampicillin's disposition in neonates, pregnant women, and foetuses, and the model supported optimal dosing evaluation in these vulnerable populations.

Physiologically Based Pharmacokinetic Modeling in Neonates: Current Status and Future Perspectives

“…Ampicillin is a beta-lactam antibiotic frequently used to treat early and late-onset neonatal sepsis [ 248 , 249 ]. Li and Xie developed a PBPK model of ampicillin in neonates and pregnant women during the intrapartum period and indicated that 50 mg/kg ampicillin q8h was effective and safe for newborns, and that 1 g of ampicillin was adequate for intrapartum prophylaxis [ 188 ].…”

Physiologically Based Pharmacokinetic Modeling of Small Molecules: How Much Progress Have We Made?

“…described the application of PBPK modeling to pregnant populations to either predict drug disposition in the pregnant mother or to evaluate the fetal exposure and potential toxicokinetics of the therapeutic agents in the fetus (Table 2). The majority (n=5) of the pregnancy PBPK models were constructed using Simcyp simulator Bukkems et al, 2022;Peng et al, 2022;Li and Xie, 2023), but other software packages such as R (n=2, (Chang et al, 2022;Kapraun et al, 2022)), PK-Sim (n=2, (Alsmadi, 2023;Liu et al, 2023))…”

“…Four studies stated that model performance was evaluated based on visual comparison of predicted and observed data Chang et al, 2022;Alsmadi, 2023;Liu et al, 2023), while one study assessed the correlation between predicted and observed data (Kapraun et al, 2022). Two studies used the 90% predictive interval as the range that observed data needed to be within (Alsmadi, 2023;Li and Xie, 2023) and one specified a visual comparison of the observed mean plasma concentration-time curve and the predicted 5-95 th percentile (Coppola et al, 2022). Two of the papers Peng et al, 2022) specifically noted that the observed plasma concentration-time curve had to be within the 5 th and 95 th percentile of the simulated.…”

“…Similarly two papers (Bukkems et al, 2022;Peng et al, 2022) established a stringent criteria for predicted pharmacokinetics parameters with 0.8-1.25 fold (bioequivalence) or 0.7-1.3-fold (narrow therapeutic index), respectively. The 2-fold range of pharmacokinetics parameters was also used in two studies (Li and Xie, 2023;Liu et al, 2023).…”

“…Taken together these models show the increasing trust in PBPK modeling in evaluating drug disposition during pregnancy. A notable advancement was the incorporation of the mechanistic kidney model into pregnancy PBPK modeling for renally cleared drugs Coppola et al, 2022;Li and Xie, 2023) although it was noted that limited information is available for physiological changes in the kidney during pregnancy and for how transporter expression changes in the kidney during pregnancy. Several of the papers (Bukkems et al, 2022;Peng et al, 2022) also used either perfused placenta studies or mechanistic modeling of the placenta to better predict maternal-fetal distribution.…”