Physiologically based pharmacokinetic (PBPK) modeling is a sophisticated computational approach that represents the body as a network of organ compartments connected by blood flow, each with physiological volumes and perfusion rates. Unlike classical compartmental models where structures are abstract and parameters are fitted to data, PBPK models incorporate real physiological and anatomical information, allowing prediction of drug disposition in tissues that cannot be sampled experimentally. This approach has become increasingly important in drug development and regulatory decision-making.
Model Structure
A whole-body PBPK model typically includes compartments for the liver, kidneys, lungs, heart, brain, adipose tissue, muscle, skin, bone, and other organs, each connected by arterial and venous blood flow. The lungs serve as the point of entry for inhaled drugs and are the site where venous blood becomes oxygenated, making them a central element in the circulatory loop. Each organ compartment is characterized by its volume, blood flow rate, and tissue-to-plasma partition coefficient for the drug.
Drug disposition within each organ is described by differential equations that account for drug delivery via arterial blood flow, partitioning into tissue, metabolism or excretion where applicable, and removal via venous blood flow. The equations are solved simultaneously to predict drug concentrations in plasma and all tissues over time. The model parameters include anatomical values, which are well established from the physiological literature, and drug-specific parameters, which are determined experimentally or predicted from physicochemical properties.
Organ Compartments and Blood Flow
The assignment of organ volumes and blood flows follows standard physiological values. Cardiac output, approximately 5 to 6 L per minute in a 70 kg adult, is distributed to organs according to their metabolic demand. The liver receives approximately 25% of cardiac output via the hepatic artery and portal vein. The kidneys receive approximately 20%, the brain approximately 12%, and muscle and skin together receive approximately 35%. Adipose tissue receives approximately 5% of cardiac output despite representing a large fraction of body mass, resulting in slow equilibration for lipophilic drugs.
Tissue partitioning is described by the tissue-to-plasma partition coefficient (Kp), which reflects the equilibrium concentration ratio between tissue and plasma. Kp values can be measured experimentally in animal studies, predicted from the drug’s physicochemical properties using mechanistic tissue composition models, or estimated from human clinical data. The accuracy of PBPK predictions depends heavily on the quality of these partition coefficient estimates.
Applications in Drug Development
PBPK modeling is used throughout the drug development lifecycle. During early discovery, PBPK models predict human pharmacokinetics from preclinical data, informing first-in-human dose selection and study design. The models incorporate in vitro data on metabolism, transport, and protein binding, along with in vivo animal pharmacokinetic data, to predict human clearance, volume of distribution, and concentration-time profiles.
In later development, PBPK models evaluate the impact of intrinsic and extrinsic factors on drug exposure. Intrinsic factors include organ impairment, age, genetics, and disease states. PBPK models can predict the effect of renal or hepatic impairment on drug exposure, guiding dose recommendations for patients with organ dysfunction without requiring dedicated clinical studies. Extrinsic factors include drug-drug interactions and food effects. PBPK models incorporating CYP enzyme inhibition or induction kinetics can predict the magnitude of interactions with coadministered drugs.
Regulatory Acceptance
Regulatory agencies including the US Food and Drug Administration and the European Medicines Agency have embraced PBPK modeling as a tool to support drug approval and labeling decisions. The FDA has published guidance on the use of PBPK models in regulatory submissions, and the number of submissions incorporating PBPK analyses has increased substantially. PBPK modeling has been accepted to waive clinical drug-drug interaction studies, to support pediatric dose selection, and to justify dosing recommendations in hepatic impairment.
Software Tools
Several commercial and open-source PBPK modeling platforms are available. Simcyp is the most widely used platform in the pharmaceutical industry, with extensive databases of drug parameters, physiological populations, and enzyme kinetics. GastroPlus offers integrated PBPK modeling with advanced absorption simulation capabilities. PK-Sim and the open-source platform Open Systems Pharmacology provide additional options. These tools allow users to simulate virtual clinical trials in populations with specified demographic, genetic, and physiological characteristics.
Example Use Cases
PBPK modeling has successfully guided dose selection for drugs that have complex pharmacokinetics or are difficult to study in certain populations. For example, PBPK models have been used to predict pediatric doses by incorporating age-dependent physiological changes in organ size, blood flow, and enzyme maturation. They have predicted the impact of pregnancy on drug clearance by incorporating gestational changes in renal blood flow, metabolic enzyme activity, and plasma protein levels. They have simulated drug exposure in patients with rare genetic polymorphisms to determine whether dose adjustment is needed.
The continued advancement of PBPK modeling, driven by improving computational power, expanding physiological databases, and increasing regulatory acceptance, is making this approach an integral part of modern pharmacokinetic science and rational drug development.
