A Novel Approach to Evaluating the Design of Pediatric Pharmacokinetic Studies Focused on Accurate Dose Selection
Zou Y, Nedelman J, Karlsson MO, Svensson EM
Clinical pharmacokinetics · 2025-06
Abstract
Background and objective In pediatric trial design, it is particularly essential to maximize data utilization and ensure robust designs while minimizing sample collection. A common criterion to justify pediatric pharmacokinetic study designs is based on parameter precision (PP) evaluation, recommended by the US Food and Drug Administration. Here, we propose an alternative approach to design evaluation based on accuracy for dose selection (ADS). Methods This work was conducted using a simulation-and-reestimation framework, based on a real-case scenario of designing the single-dose pharmacokinetic study of the anti-tuberculosis (TB) drug pretomanid, with the aim of selecting doses for the next multidose long-term study. The study powers were computed using the ADS approach under scenarios with (1) real-case conditions, (2) high variabilities, (3) available options of tablet doses for selections. The study power using a PP approach was computed to compare with the ADS approach. Results The ADS approach suggested that the design selected accurate doses with study power >80% in almost all dosing weight groups, whereas the PP approach found the design underpowered for clearance. The ADS-based power was decreased to ~65% in the smallest weight groups given high variability. Varying the options of dose levels affected the ADS-based power non-monotonically, although fewer levels generally yielded higher power. Conclusion The ADS approach practically evaluates the precision in dose selection, providing a directly relevant decision criterion for designing pediatric pharmacokinetic studies and could be an alternative for power evaluation when the study is focused on determining doses using discrete tablet sizes.
MeSH terms
- Humans
- Antitubercular Agents
- Dose-Response Relationship, Drug
- Models, Biological
- Research Design
- Computer Simulation
- Child