Scaling pharmacokinetics to estimate the “first dose in children”: comparing allometric scaling and PBPK
A. Strougo (1,2), T. Eissing (3), S. Willmann (3), A. Yassen (2), M. Danhof (1), J. Freijer (2)
(1) Division of Pharmacology, Leiden/Amsterdam Center for Drug Research, Leiden, The Netherlands, (2) Global Clinical Pharmacology & Exploratory Development, Astellas Pharma Europe, Leiderdorp, The Netherlands, (3) Bayer Technology Services GmbH, Systems Biology and Computational Solutions, Leverkusen, Germany
Introduction: For the estimation of “first dose in children” scaling of the pharmacokinetics (PK) from adults to children is required. Two of the most frequent applied methodologies are physiologically-based pharmacokinetic modeling (PBPK) and allometric scaling in combination with maturation of clearance (CL) for early life. Both methodologies are accepted by regulatory authorities, but it remains unclear if they are interchangeable. This investigation aimed to compare CL predictions when using both methodologies for morphine, paracetamol and various hypothetical drugs with different pharmacokinetic characteristics.
Methods: PK-Sim® was used to create models for more than 100 hypothetical drugs with different PK properties. These pharmacokinetic characteristics were generated by combining different lipophilicities, low, medium and high clearance liver blood flow, liver diffusion and free fractions for compounds being metabolised by either UGT2B7 or UGT1A6/sulfation. Hypothetical drugs with unrealistic PK properties were excluded from the comparisons. The population wrapper in MoBi® was used to scale the PK to both adults and children. The simulated CL in adults was subsequently used to fit a population model using NONMEM 7. Finally, the estimated CL and its respective inter-individual variability were scaled to children using allometry in combination with published maturation functions. In addition, PBPK model predictions for morphine and paracetamol in children were evaluated by comparison with models developed to describe the paediatric PK data.
Results: In children younger than 1-2 years of age, allometry in combination with published maturation functions yields CL predictions that differ up to a factor 5 from PBPK-based predicted clearance. These differences are related not only to the clearance pathways but also to the extraction ratio of the hypothetical drugs. Evaluation of PBPK predictions for morphine and paracetamol showed that the predicted CL ratios were within a factor two-range and that the respective maturation functions were not always in agreement with the liver enzyme(s) activity.
Conclusion: This investigation provides insights into the physiological meaning of the maturation functions indicating that its use in the scaling of the pharmacokinetics of other drugs requires improvement. Until the current methodologies for scaling PK from adults to children have been extensively validated with paediatric clinical data, methodological uncertainty of the predictions should be considered in the risk-benefit assessment of the “first dose in children”.