by Anders Lindahl, Ph.D., and Jan Neelissen, Ph.D.
Oral physiologically based pharmacokinetics (PBPK) in silico models focus on the link between a molecule’s physiochemical properties and formulation parameters to plasma exposure within animal species or a patient. The extent of modeling depends on the amount of data available, but it can be a useful tool during early drug development (figure 1) and improves as more data becomes available.
Figure 1. Schematic overview of the different stages of drug development with focus on formulation and human pharmacokinetic prediction
Early-Stage Candidate Selection
At the candidate selection phase, when little data is available, integrating API characteristics (pKa, logP, solubility) and permeability (Caco-2 or MDCK) combined with gastrointestinal physiology (gastric emptying, transit time, etc.) allows for the prediction of the fraction absorbed (Fa) after oral administration of a solution or suspension.
PBPK modeling at this stage can identify molecule(s) with the best absorption properties or choose the molecule(s) whose absorption challenges are addressable by formulation technologies. Using parameter sensitivity analysis, formulation, and API properties—such as particle size, solubility, or permeability—can be modelled to support candidate selection, potentially reducing the need for iterative formulation and in vivo pharmacokinetic (PK) development studies.
Incorporating in vivo animal/human metabolic clearance data into the PBPK model allows estimation of the first pass loss (FPL) in the liver. A high FPL in the liver will limit oral bioavailability and cannot be solved with formulation design so having this data early is vital.
The first animal species PBPK model guides selection of appropriate formulation technology for initial in vivo animal PK studies. These provide accurate data to update the PBPK model, allowing estimates of volumes of distribution (Vd), clearances (Cl) and calculation of absolute bioavailability.
With Vd and Cl information added into a PBPK model, parameter sensitivity analysis can be used again to adjust the oral exposure prediction to the best fit possible, allowing the model to estimate Fa, the fraction of drug escaping liver metabolism (Fh), FPL and bioavailability (F).
Having obtained a good first animal species PBPK model, it can be validated against PK data from a second species. If the predicted oral profile fits the measured data in the second animal species, confidence in predictability for human absorption is increased.
Confirming Formulation Technology Selection
A human PBPK model can also predict how a solid dosage form would perform. If Fa is predicted to be too low, the effect of particle size reduction can be modelled, and if necessary, a solubility enabling technology adopted, such as lipid-based formulations or amorphous dispersions. These formulations can also have in vitro dissolution profiles generated with different prototypes, which can be integrated in the PBPK model. This is valuable when designing the PK study and choosing the right formulation.1
More advanced estimations of Vd and CI may then be performed using different methods,2-5 and inputted to the human PBPK model. Combining validated animal PBPK models with estimates of human Vd and Cl results in a more reliable human PBPK model for the next phases.
Optimizing Formulation Development
Once a suitable formulation technology is selected, new formulations will be made to optimize the current prototype, including in vitro dissolution profiles, which can be loaded into the PBPK model to predict the outcome, followed by in vivo PK studies in an animal species, with the goal to pick the optimum for first in human studies. Finally, the human PBPK model will be updated with human PK data, allowing other assessments like the design of modified release products, where a release profile can be derived that matches the desired in vivo PK profile.
Ideally, PBPK models should be built as early as possible in drug development as they can guide selecting the optimal formulation technology to increase systemic exposure.
Conclusions
Early PBPK modeling helps minimize development risk by identifying molecule liabilities, facilitates product development decision-making, saving development time and resources. It is a viable tool that supports patient-centric product development.
References
- Jamei et al. 2020. Current status and future opportunities for incorporation of dissolution data in PBPK modeling for pharmaceutical development and regulatory applications: OrBiTo consortium commentary. European Journal of Pharmaceutics and Biopharmaceutics. 2020 (155), 55-68.
- Lombardo et al. 2004. Prediction of Human Volume of Distribution Values for Neutral and Basic Drugs. 2. Extended Data Set and Leave-Class-Out Statistics. J Med Chem 47:1242-1250.
- Rodgers and Rowland; 2007. Mechanistic approaches to volume of distribution predictions: understanding the processes. Pharm Res. 24(5):918-933.
- Sohlenius-Sternbeck et al. 2010. Evaluation of the human prediction of clearance from hepatocyte and microsome intrinsic clearance for 52 drug compounds. Xenobiotica, 40:9, 637-649.
- Mahmood 2007. Application of allometric principles for the prediction of pharmacokinetics in human and veterinary drug development. Advanced Drug Delivery Reviews. 59, 1177-1192.