Population PBPK modeling using parametric and nonparametric methods of the Simcyp Simulator, and Bayesian samplers

Abstract

Physiologically-based pharmacokinetic (PBPK) models usually include a large number of parameters whose values are obtained using in vitro to in vivo extrapolation. However, such extrapolations can be uncertain and may benefit from inclusion of evidence from clinical observations via parametric inference. When clinical interindividual variability is high, or the data sparse, it is essential to use a population pharmacokinetics inferential framework to estimate unknown or uncertain parameters. Several approaches are available for that purpose, but their relative advantages for PBPK modeling are unclear. We compare the results obtained using a minimal PBPK model of a canonical theophylline dataset with quasi-random parametric expectation maximization (QRPEM), nonparametric adaptive grid estimation (NPAG), Bayesian Metropolis-Hastings (MH), and Hamiltonian Markov Chain Monte Carlo sampling. QRPEM and NPAG gave consistent population and individual parameter estimates, mostly agreeing with Bayesian estimates. MH simulations ran faster than the others methods, which together had similar performance. Study Highlights WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?No single software platform allows practitioners to compare parametric and nonparametric estimates of calibrated physiologically-based pharmacokinetic (PBPK) model parameters. WHAT QUESTION DID THIS STUDY ADDRESS?How do a maximum likelihood parametric (quasi-random parametric expectation maximization) and nonparametric (nonparametric adaptive grid estimation) algorithms, and two Bayesian numerical methods (Hamiltonian Markov Chain Monte Carlo and MH) compare in results and timing for calibration of a PBPK model?

Bayesian estimation in veterinary pharmacology: A conceptual and practical introduction

“…The Stan code developed for this model, and R code to call it, is provided publicly with the study code. It is important to note that Bayesian estimation has been conducted in various software environments relevant to pharmacometrics (Johnston et al, 2023;Wedagedera et al, 2022), but a detailed evaluation was beyond the scope of this work.…”

Advancing Precision Medicine: A Review of Innovative In Silico Approaches for Drug Development, Clinical Pharmacology and Personalized Healthcare

“…Optimizing clinical trial designs by identifying relevant covariates that influence drug PK streamlines trial enrollment, ensuring the inclusion of precisely the right patients [ 148 ], and enhances the accuracy and efficiency of clinical investigations, allowing advances in accurate parameter estimation for drug safety and efficacy [ 149 ]. Data-driven methods excel in predicting adverse events, guiding dose adjustments, and minimizing side effects.…”

Cross‐Species Extrapolation of Biological Data to Guide the Environmental Safety Assessment of Pharmaceuticals—The State of the Art and Future Priorities

“…In parallel with the experimental approach focused on a small number of species, modeling methods should, therefore, play a major role in the extrapolation of ADME profiles across species (Davies et al, 2020; Priority Research Question 4—Textbox 1). For example, Bayesian‐PBPK modeling (Krauss & Schuppert, 2016), multitask deep neural network models (Wenzel et al, 2019), and population‐based ADME simulators (Wedagedera et al, 2022) are just a few examples of advanced modeling methods currently used to predict human PK profiles and estimate their variability within human populations. We foresee that comparable modeling concepts could be applied to simulate the variability of uptake and ADME data across species for ecotoxicological purposes, either from a few fish species to many or from humans to non‐humans.…”