Pascale Schmidt 1, Salomé Dumasdelage 2, André Dallmann 3,4, Verena Gotta 2
1 University of Basel (Basel, Switzerland), 2 University Children's Hospital Basel (Basel, Switzerland), 3 Bayer HealthCare SAS (Lille, Germany), 4 on behalf of: Model-Informed Drug Development, Research and Development, Pharmaceuticals, Bayer AG (Leverkusen, France)
Objectives: Physiologically based pharmacokinetic (PBPK) modelling is a tool increasingly used to support pediatric dosing decisions during drug development. In the clinical setting, PBPK models could also be useful to address pediatric dosing uncertainties, particularly in infants and neonates. For example, the Pseudomonas-active antibiotic cefepime – known for neurotoxicity risk at increased exposure – is licensed in Switzerland for infants >1 month, while dosing uncertainty is acknowledged for infants 1-2 months of age. We aimed to evaluate whether an existing adult-pediatric PBPK model for cefepime could be used to address dosing uncertainties in neonates and infants <2 months of age. Methods: An existing adult-pediatric PBPK model [1] was implemented in the software PK-Sim® (Open Systems Pharmacology Suite, version 11.3). First, the implementation was validated by comparison of model-predicted plasma clearance (CLpredicted, ml/min/kg) as main outcome of interest with previously published predicted [1] and clinically observed CL values (CLclinical) for a healthy adult population [2] and a pediatric population with broad age range (0.2 to 16.4 years) [3], respectively. Then, CLpredicted was generated for three pediatric sub-groups < 2 years of age (infants 6-24 months, 2-6 months, neonates < 1 month). CLpredicted was compared to CLclinical by computing the relative clearance difference as the ratio CLpredicted/CLclinical. To improve CLpredicted <2 years of age, the model structure was adjusted by incorporating a more mechanistic definition of renal drug clearance, with age-dependent maturation terms for passive glomerular filtration rate (pre-defined in the software; predicting passive renal clearance), and specific active tubular secretion (KTSspec, min-1, estimated; predicting active renal clearance). Results: Predicted pharmacokinetic values for healthy adults and pediatric population with broad age range, respectively, well aligned with previously published PBPK-predicted and observed clinical values: relative CL difference compared to predicted values was <3% in adults and <18% in children [1], and compared to observed clinical values <5% in adults [2] and <6% in children [3], respectively. However, increasing clearance overprediction with decreasing age was observed: CLpredicted / CLclinical was 1.0 for infants 6-24 months (< 1% difference), 1.37 for infants 2-6 months (37% overprediction) [3] and 3.4 in neonates < 1 month (240% overprediction) [4]. Incorporation of age-dependent, more mechanistic renal drug clearance allowed prediction of CLclinical with ≤ 5% error over age range from healthy adult to neonates (CLpredicted / CLclinical range: 0.99 – 1.05). The corresponding active renal clearance was predicted to be decreased by approximately 77% in infants 6-24 months, and by 93% to 97% in infants 2-6 months and neonates under 1 month, respectively, compared to the healthy adult population. Discussion: The evaluated published adult-pediatric PBPK model of cefepime was, in its initial form, not suitable for pharmacokinetic extrapolation to infants < 6 months and neonates. This highlights the need for careful pediatric PBPK model evaluation in this age group. A more mechanistic clearance definition to capture passive and active renal function maturation significantly enhanced its suitability for pediatric extrapolation. However, this required estimation of age-dependent tubular secretion, for which relevant transporters have not yet been described for cefepime. References: [1] M. Talha Zahid et al., Saudi Pharmaceutical Journal 2023 [2] R. H. Barbhaiya et al., Antimicrob Agents Chemother 1992 [3] M. D. Reed et al., Antimicrob Agents Chemother 1997 [4] E. Capparelli et al., Antimicrob Agents Chemother 2005
Reference: PAGE 34 (2026) Abstr 12127 [www.page-meeting.org/?abstract=12127]
Poster: Drug/Disease Modelling - Paediatrics