Comparisons of PK-Sim and R program for physiologically based pharmacokinetic model development for broiler chickens and laying hens: meloxicam as a case study

Abstract Physiologically based pharmacokinetic (PBPK) models play a critical role in evaluating drug residue concentrations and estimating withdrawal intervals (WDIs) for food-producing animals. These models are facilitated by various programming software (e.g. R program) and predefined PBPK platforms, such as Open Systems Pharmacology (OSP) suite integrated by PK-Sim and Mobi, which offers a user-friendly graphical interface. Both R and OSP are open-source software. However, there is a lack of comparative analyses of both platforms and their potential impact on PBPK models. This study aims to evaluate the influence of different platforms on PBPK workflow, parameters selection, and output results, which is exemplified via a case study for meloxicam in chickens in both platforms. Our findings indicate that while the choice of PBPK platforms affected the workflow and input parameters, the predictive performance of established models remained consistent across both platforms. Both platforms predicted meloxicam pharmacokinetics in plasma and tissues accurately across different exposure scenarios. The PBPK-estimated WDIs under various dosing regimens from both platforms were quite similar. Notable differences between OSP suite and R were primarily observed during sensitivity analysis and parameter identification processes, especially the time consumption. This study offers insight into software variances and their implications for translating PBPK modeling knowledge between users of 2 platforms. Also, it provides a PBPK model structure template implemented in both software platforms for food safety and risk assessment in poultry and a detailed tutorial on expanding the model structure in PK-Sim and Mobi.