Reviewing Data Integrated for PBPK Model Development to Predict Metabolic Drug-Drug Interactions: Shifting Perspectives and Emerging Trends

Abstract

Physiologically-based pharmacokinetics (PBPK) modeling is a robust tool that supports drug development and the pharmaceutical industry and regulatory authorities. Implementation of predictive systems in the clinics is more than ever a reality, resulting in a surge of interest for PBPK models by clinicians. We aimed to establish a repository of available PBPK models developed to date to predict drug-drug interactions (DDIs) in the different therapeutic areas by integrating intrinsic and extrinsic factors such as genetic polymorphisms of the cytochromes or environmental clues. This work includes peer-reviewed publications and models developed in the literature from October 2017 to January 2021. Information about the software, type of model, size, and population model was extracted for each article. In general, modeling was mainly done for DDI prediction via Simcyp^® software and Full PBPK. Overall, the necessary physiological and physio-pathological parameters, such as weight, BMI, liver or kidney function, relative to the drug absorption, distribution, metabolism, and elimination and to the population studied for model construction was publicly available. Of the 46 articles, 32 sensibly predicted DDI potentials, but only 23% integrated the genetic aspect to the developed models. Marked differences in concentration time profiles and maximum plasma concentration could be explained by the significant precision of the input parameters such as Tissue: plasma partition coefficients, protein abundance, or Ki values. In conclusion, the models show a good correlation between the predicted and observed plasma concentration values. These correlations are all the more pronounced as the model is rich in data representative of the population and the molecule in question. PBPK for DDI prediction is a promising approach in clinical, and harmonization of clearance prediction may be helped by a consensus on selecting the best data to use for PBPK model development.

Keywords: clinical setting; drug-drug interaction; metabolism; physiologically-based pharmacokinetics; precision dosing; transporters.

FIGURE 1

Overview of drug metabolism and transport in the liver. Drug metabolism is divided into phase 1 and phase 2 reactions. In phase 1 reactions, polar functional groups are unmasked or introduced to the molecules through oxidation, reduction and, hydrolysis. The so formed phase 1 metabolites can be readily excreted or can undergo subsequent conjugation reaction with hydrophilic moieties (phase 2 reactions). Transporters play a complemental role to the phase 1 and 2 assuring the phase 0 (uptake) and phase 3 (export) crucial to the drug elimination by metabolism.

FIGURE 2

PBPK model components. PBPK models are separated into three main components: the drug, the system and the trial. Drug data include physicochemical and experimental or predicted ADME data. System data include physiological data which relevant for the ADME properties of drugs. Trial data include information on trial design such as administration route, dose regimen or trial duration (adapted from (Jamei 2016))