1. What Is Hepatic Clearance in PBPK?
Hepatic clearance describes the removal of drug by the liver. In a physiologically based pharmacokinetic model, hepatic clearance is not simply entered as a single empirical number. Instead, the model can connect physiological properties of the liver with drug-specific properties such as protein binding, intrinsic metabolic clearance, and transporter activity.
This distinction is central to PBPK. A conventional PK analysis may estimate an apparent systemic clearance from observed concentration-time data. A PBPK model attempts to explain that clearance mechanistically by representing the processes that determine drug extraction by the liver.
PBPK models connect drug-specific properties and physiological characteristics to hepatic extraction and ultimately to systemic concentration and exposure.
2. Why Is the Liver Important for Drug Clearance?
The liver receives a substantial fraction of cardiac output through the hepatic artery and portal vein and contains a large complement of metabolic enzymes and transport proteins. Consequently, hepatic processes can make major contributions to the systemic clearance of many drugs.
Drug removal by the liver can involve several distinct processes:
- Metabolism by enzymes such as CYP and UGT families.
- Hepatic uptake through transporter-mediated or passive processes.
- Biliary excretion of unchanged drug or metabolites.
- Intracellular distribution that influences access to metabolic or excretory pathways.
- Efflux from hepatocytes back toward blood or into bile.
A simple PBPK liver model may represent only intrinsic metabolic clearance. More mechanistic models can explicitly represent uptake, intracellular concentrations, metabolism, biliary excretion, and efflux.
3. Hepatic Clearance and the Extraction Ratio
Hepatic clearance can be related to hepatic blood flow and the fraction of drug extracted by the liver.
where:
- \(CL_H\) is hepatic clearance,
- \(Q_H\) is hepatic blood flow, and
- \(E_H\) is the hepatic extraction ratio.
The extraction ratio is the fraction of drug entering the liver that is removed during a single pass through the organ.
where \(C_{in}\) and \(C_{out}\) represent the appropriate incoming and outgoing blood concentrations under the assumptions of the model.
This relationship immediately establishes an important physiological constraint: hepatic clearance cannot exceed hepatic blood flow in the standard organ-clearance framework.
4. The Well-Stirred Liver Model
The well-stirred model is one of the most widely used approaches for representing hepatic clearance. It treats the liver as a kinetically well-mixed organ and relates hepatic clearance to hepatic blood flow, unbound fraction, and intrinsic clearance.
In its common form:
where:
- \(Q_H\) = hepatic blood flow,
- \(f_{u,B}\) = fraction of drug unbound in blood, and
- \(CL_{int,H}\) = hepatic intrinsic clearance.
The quantity \(f_{u,B}CL_{int,H}\) represents the unbound intrinsic capacity of the liver to remove drug before the limitation imposed by hepatic blood flow is applied.
The well-stirred model represents hepatic elimination as the interaction between delivery to the liver and the liver's intrinsic ability to remove drug.
5. Low- and High-Extraction Drugs
The well-stirred equation becomes particularly intuitive when considering limiting cases.
Low extraction
If \(f_{u,B}CL_{int,H}\) is much smaller than hepatic blood flow:
then:
In this regime, hepatic clearance is strongly influenced by the drug's intrinsic elimination capacity and unbound fraction.
High extraction
If intrinsic unbound clearance is much larger than hepatic blood flow:
then:
The liver is highly efficient at removing drug from the blood delivered to it, so clearance becomes primarily flow-limited.
| Characteristic | Low extraction | High extraction |
|---|---|---|
| Approximate clearance | \(CL_H\approx f_{u,B}CL_{int,H}\) | \(CL_H\approx Q_H\) |
| Major determinants | Intrinsic clearance and unbound fraction | Hepatic blood flow |
| Sensitivity to enzyme activity | Often substantial | Often attenuated |
| Sensitivity to blood flow | Often smaller | Often substantial |
| Protein-binding influence | Can be important | May be less direct for total hepatic clearance |
These classifications are useful conceptual tools, but actual drugs exist along a continuum rather than in two perfectly separated categories.
6. What Is Intrinsic Clearance?
Intrinsic clearance, \(CL_{int,H}\), represents the liver's inherent capacity to eliminate drug before the limitation imposed by hepatic blood flow is incorporated into the organ-clearance model.
Intrinsic clearance can arise from multiple processes, including:
- Metabolic enzyme activity.
- Transporter-mediated uptake.
- Biliary secretion.
- Other intracellular elimination pathways.
In a simplified metabolic model, intrinsic clearance may be treated as the combined capacity of the relevant enzymes to metabolize unbound drug.
For example, if several independent metabolic pathways contribute to elimination:
This additive representation is useful when the pathways behave approximately independently under the assumptions of the model.
7. Why Does the Unbound Fraction Matter?
Many hepatic elimination processes depend on the fraction of drug that is available in an unbound form. The fraction unbound in blood is denoted \(f_{u,B}\).
Under the well-stirred model, the combination \(f_{u,B}CL_{int,H}\) determines the intrinsic component of hepatic extraction:
For a low-extraction drug, increasing the unbound fraction can therefore increase hepatic clearance if intrinsic clearance remains unchanged.
This can seem counterintuitive because increased protein binding can also reduce the concentration of unbound drug. The key is that total concentration, unbound concentration, intrinsic clearance, and hepatic extraction are related through the model. One should therefore avoid interpreting protein binding independently of the clearance model.
8. Why Blood-to-Plasma Partition Matters
PBPK hepatic clearance calculations are often expressed in terms of blood concentrations and hepatic blood flow. However, experimental drug-binding measurements are frequently reported in plasma.
The relationship between blood and plasma concentrations depends on the hematocrit and the extent to which drug partitions into red blood cells. A blood-to-plasma ratio can therefore be important when translating plasma measurements into the quantities used by a blood-based hepatic clearance model.
If a model is formulated using blood concentrations, the drug-specific binding and concentration inputs should be internally consistent with that formulation.
| Quantity | Role in hepatic PBPK |
|---|---|
| Plasma concentration | Common experimental measurement |
| Blood concentration | Often the concentration used with hepatic blood flow |
| Blood-to-plasma ratio | Connects blood and plasma concentration scales |
| \(f_{u,B}\) | Fraction unbound in blood |
9. From In Vitro Intrinsic Clearance to Human Hepatic Clearance
One of the major uses of mechanistic PBPK is in vitro-to-in vivo extrapolation (IVIVE). In vitro experiments can provide information about intrinsic clearance, which can then be scaled to the human liver.
Typical experimental systems include:
- Human liver microsomes.
- Suspended or plated human hepatocytes.
- Recombinant enzymes.
- Transporter-expressing systems.
- Other mechanistic cellular assays.
A simplified IVIVE workflow is:
A simplified IVIVE workflow converts in vitro measurements into a scaled intrinsic clearance and then embeds that quantity within a physiological liver model.
The scaling step can involve liver weight, hepatocellularity, microsomal protein abundance, enzyme abundance, transporter abundance, nonspecific binding, and other factors depending on the experimental system and PBPK framework.
10. Representing Hepatic Metabolism
A PBPK model may represent metabolism at different levels of mechanistic detail.
At the simplest level, metabolism can be summarized by a single intrinsic clearance:
A more detailed model can separate individual pathways:
Such decomposition can be useful when a drug is affected by enzyme inhibition, induction, genetic variation, disease-related changes, or changes in enzyme abundance.
| Level of representation | Example | Potential use |
|---|---|---|
| Aggregate | Total \(CL_{int,H}\) | General PK prediction |
| Pathway-specific | CYP3A4 + CYP2D6 | DDI and enzyme-specific questions |
| Mechanistic enzyme model | Enzyme abundance and kinetics | Changes in physiology or enzyme activity |
| Full liver model | Uptake + metabolism + biliary excretion + efflux | Complex transporter-enzyme systems |
11. When Transporters Matter
For some drugs, hepatic clearance cannot be adequately represented by metabolism alone. Transporters can determine how efficiently drug enters hepatocytes, leaves them, or is secreted into bile.
A more mechanistic liver representation can therefore distinguish processes such as:
- Passive diffusion into hepatocytes.
- Active uptake through transporters.
- Metabolic clearance within hepatocytes.
- Biliary secretion.
- Basolateral efflux back to blood.
Conceptually:
The exact mathematical relationship depends on the structural model. Extended clearance models can explicitly represent these processes rather than combining them into a single intrinsic clearance term.
12. Beyond the Well-Stirred Model
The well-stirred model is useful and widely applied, but it is not the only possible representation of hepatic elimination.
| Model | Basic idea | Potential advantage |
|---|---|---|
| Well-stirred model | Liver behaves as a well-mixed compartment | Simple and interpretable representation of flow and intrinsic clearance |
| Parallel-tube model | Drug concentration changes progressively along sinusoidal flow | Represents concentration gradients through the liver |
| Dispersion model | Represents nonideal mixing and dispersion | More flexible representation of hepatic flow |
| Extended clearance model | Separates uptake, metabolism, biliary excretion, and efflux | Useful for transporter-enzyme systems |
The choice should be driven by the scientific question and the available information. Greater mechanistic detail is not automatically better if the additional parameters cannot be identified or supported by data.
13. Hepatic First-Pass Extraction and Oral Bioavailability
Hepatic clearance becomes particularly important after oral administration because drug absorbed from the gastrointestinal tract can reach the liver through the portal circulation before entering systemic circulation.
The hepatic first-pass extraction fraction can be represented conceptually as:
where \(F_H\) is the fraction escaping hepatic first-pass extraction under the corresponding model assumptions.
Systemic oral bioavailability can be represented as:
where:
- \(F_A\) is the fraction absorbed,
- \(F_G\) is the fraction escaping intestinal loss, and
- \(F_H\) is the fraction escaping hepatic first-pass extraction.
Thus, a change in hepatic intrinsic clearance can influence oral exposure even if absorption itself has not changed.
14. How Hepatic Clearance Influences Exposure
For a linear IV system, systemic exposure is inversely related to total clearance:
If hepatic clearance represents a substantial portion of total clearance:
then changes in \(CL_H\) can propagate to systemic exposure.
For oral dosing, exposure also depends on bioavailability:
Because hepatic clearance can affect both \(F\) through first-pass extraction and \(CL\) through systemic elimination, the net effect of changing hepatic processes can be more complicated than simply increasing or decreasing a single clearance parameter.
15. Worked Example: Calculating Hepatic Clearance
Consider a hypothetical drug with the following properties:
- Hepatic blood flow: \(Q_H=90\) L/h
- Fraction unbound in blood: \(f_{u,B}=0.20\)
- Hepatic intrinsic clearance: \(CL_{int,H}=180\) L/h
Step 1: Calculate unbound intrinsic clearance
Step 2: Apply the well-stirred model
Step 3: Calculate the extraction ratio
So the model predicts an extraction ratio of approximately 0.29, meaning roughly 29% extraction under the assumptions of the model.
Step 4: Calculate the hepatic availability fraction
The corresponding fraction escaping hepatic first-pass extraction is therefore approximately 71%.
16. What Happens If Hepatic Parameters Change?
PBPK models make it possible to explore how changes in physiological or drug-specific parameters affect hepatic clearance.
| Change | Expected effect in a low-extraction setting | Expected effect in a high-extraction setting |
|---|---|---|
| Increase \(CL_{int,H}\) | Often increases \(CL_H\) | May have a smaller effect once flow limitation dominates |
| Increase \(f_{u,B}\) | Can increase \(CL_H\) | May have a smaller effect on total \(CL_H\) |
| Increase \(Q_H\) | Often smaller effect | Can increase \(CL_H\) |
| Decrease enzyme activity | Can substantially reduce \(CL_H\) | Effect can be attenuated if extraction remains flow-limited |
| Decrease hepatic blood flow | Often modest direct effect | Can substantially reduce \(CL_H\) |
These statements describe the behavior of the idealized model. Real drugs may have multiple clearance pathways, nonlinear kinetics, active transport, time-dependent inhibition, or disease-related changes that alter the response.
17. Hepatic Clearance and Drug-Drug Interactions
One major application of mechanistic hepatic clearance models is the prediction of drug-drug interactions (DDIs).
An inhibitor can reduce the intrinsic clearance of a victim drug through inhibition of a metabolic enzyme or transporter. An inducer can increase enzyme or transporter activity or abundance.
For a metabolic pathway:
an inhibitor may reduce one pathway while leaving others unchanged.
The resulting change in systemic clearance depends on how important that pathway is to the overall elimination of the drug and whether the drug is low or high extraction.
18. Hepatic Impairment
Hepatic disease can alter several physiological determinants of drug disposition simultaneously. Depending on the disease and drug, relevant changes can include hepatic blood flow, liver size, hepatocyte abundance, enzyme activity, transporter activity, plasma protein concentrations, and other physiological characteristics.
A PBPK model can represent these changes as modifications to physiological and drug-specific parameters rather than applying a single empirical clearance multiplier.
| Potential change | Possible PBPK consequence |
|---|---|
| Reduced hepatic blood flow | May reduce flow-limited hepatic clearance |
| Reduced enzyme abundance/activity | May reduce intrinsic metabolic clearance |
| Altered transporter activity | May change hepatic uptake or efflux |
| Altered plasma protein concentrations | May change unbound fraction |
| Reduced functional liver mass | May reduce total hepatic intrinsic capacity |
The net effect is drug-specific because the relative importance of these mechanisms depends on extraction, binding, metabolic pathways, transport processes, and other disposition characteristics.
Regulatory PBPK analyses for hepatic impairment therefore require explicit characterization of the assumptions and physiological changes represented by the model. The FDA issued a draft guidance on pharmacokinetics in patients with impaired hepatic function in September 2026; it is currently a draft and is not for implementation.
19. A Practical Workflow for Modeling Hepatic Clearance
- Define the scientific question. Determine whether the model is intended to predict systemic PK, oral exposure, DDIs, hepatic impairment, or another outcome.
- Characterize drug disposition. Identify metabolic enzymes, transporters, binding, blood-to-plasma partitioning, and renal or other elimination pathways.
- Measure or estimate intrinsic clearance. Use appropriate in vitro systems and characterize experimental variability.
- Perform IVIVE. Scale the in vitro intrinsic clearance to an in vivo hepatic intrinsic clearance using a justified methodology.
- Select the hepatic model. Determine whether a well-stirred model is sufficient or whether a more mechanistic liver model is needed.
- Parameterize physiology. Specify hepatic blood flow, liver size, hepatocellularity, and other relevant physiological inputs.
- Integrate the liver into the whole-body PBPK model. Combine hepatic elimination with distribution, renal clearance, absorption, and other relevant processes.
- Evaluate the model. Compare predictions with observed clinical PK data and investigate systematic discrepancies.
- Perform sensitivity analysis. Determine which assumptions and parameters have the greatest influence on the predicted exposure.
- Apply the model. Use the evaluated model for simulation, scenario analysis, or prediction within an appropriate domain of applicability.
20. Why Model Identifiability Matters
Mechanistic detail is valuable only when the available data and prior knowledge can support it. If several unknown parameters affect the concentration-time profile in nearly indistinguishable ways, the model may be poorly identifiable.
For example, a change in observed hepatic clearance could potentially arise from changes in:
- Hepatic blood flow.
- Unbound fraction.
- Intrinsic metabolic clearance.
- Transporter-mediated uptake.
- Biliary excretion.
- Other parallel elimination pathways.
Clinical concentration data alone may not uniquely determine all of these quantities.
21. Evaluating a Hepatic Clearance Model
A PBPK model should be evaluated against relevant observations rather than judged solely by whether its simulated curve looks plausible.
Useful evaluation questions include:
- Does the model reproduce observed IV clearance?
- Does it reproduce oral exposure and first-pass behavior?
- Does it reproduce the concentration-time profile across relevant doses?
- Are observed and predicted metabolite profiles consistent when metabolites are modeled?
- Does the model reproduce known effects of enzyme or transporter inhibition?
- Does it behave reasonably when physiological parameters are changed?
- Are important discrepancies systematic rather than random?
Regulatory PBPK analyses should also clearly document the modeling strategy, data sources, assumptions, model development, verification, and application. The FDA's PBPK guidance recommends a structured report covering these elements.
22. What Hepatic Clearance Models Do Not Tell Us Automatically
Even a detailed hepatic PBPK model remains a mathematical representation of biological processes. Several limitations are important.
- Intrinsic clearance is model-dependent. Its interpretation depends on the hepatic model and scaling framework used.
- In vitro systems are not complete human livers. Experimental conditions may differ substantially from the in vivo environment.
- Protein binding can be difficult to characterize. Different experimental methods and physiological assumptions can produce different estimates.
- Transporter systems can be complex. Uptake and efflux pathways can interact with metabolism and intracellular drug concentrations.
- Physiology varies among individuals. Population averages do not necessarily represent every individual.
- Diseased physiology can be difficult to parameterize. Multiple physiological changes may occur simultaneously.
- Model complexity increases uncertainty. Additional pathways introduce additional parameters and assumptions.
- Good predictive performance does not prove that every mechanistic assumption is correct.
23. Empirical Clearance vs. Mechanistic Hepatic Clearance
| Feature | Empirical PK clearance | Mechanistic PBPK clearance |
|---|---|---|
| Primary representation | Observed whole-body PK parameter | Physiological and drug-specific processes |
| Typical source | Clinical concentration-time data | In vitro, in vivo, literature, physiology, and clinical data |
| Intrinsic clearance | Usually not explicitly represented | Can be explicitly represented |
| Hepatic blood flow | Usually implicit | Explicit physiological input |
| Enzyme activity | Usually implicit | Can be explicitly represented |
| Transporters | Usually implicit | Can be mechanistically represented |
| Scenario simulation | Limited by empirical parameterization | Can simulate physiological and mechanistic changes |
| Interpretability | Strong at the whole-body level | Potentially stronger mechanistic interpretation, conditional on assumptions |
The two approaches are complementary. Empirical PK provides important clinical observations, while PBPK provides a framework for connecting those observations to physiological and drug-specific mechanisms.
24. Key Takeaways
- Hepatic clearance describes drug removal by the liver and is a major component of systemic disposition for many drugs.
- In PBPK, hepatic clearance can be connected to hepatic blood flow, protein binding, intrinsic clearance, and transporter processes.
- The well-stirred liver model is commonly expressed as \(CL_H=Q_Hf_{u,B}CL_{int,H}/(Q_H+f_{u,B}CL_{int,H})\).
- Hepatic extraction is related to clearance by \(CL_H=Q_HE_H\).
- Low-extraction drugs are often more sensitive to intrinsic clearance and unbound fraction, whereas high-extraction drugs can become flow-limited.
- Intrinsic clearance represents the liver's inherent capacity to eliminate drug before the limitation imposed by hepatic blood flow.
- IVIVE connects in vitro intrinsic clearance measurements to an estimated in vivo hepatic intrinsic clearance.
- Protein binding and blood-to-plasma partitioning can materially affect the translation of experimental measurements into PBPK parameters.
- Transporters can be important when hepatic uptake, efflux, or biliary excretion contributes substantially to disposition.
- Hepatic clearance can affect oral exposure through both systemic elimination and hepatic first-pass extraction.
- Changes in enzyme activity, transporter activity, hepatic blood flow, liver size, or protein binding can propagate through a PBPK model to alter predicted exposure.
- More mechanistic detail is not automatically better: the model should contain enough detail to answer the scientific question while remaining supported by data.
- Independent experimental information is particularly valuable for identifying mechanistic hepatic parameters.
- PBPK predictions remain conditional on the physiological assumptions, drug-specific parameters, scaling methods, and domain of applicability of the model.
25. References
- U.S. Food and Drug Administration. Physiologically Based Pharmacokinetic Analyses — Format and Content Guidance for Industry. 2018. Final guidance.
- U.S. Food and Drug Administration. The Use of Physiologically Based Pharmacokinetic Analyses — Biopharmaceutics Applications for Oral Drug Product Development, Manufacturing Changes, and Controls. 2020. Draft guidance.
- Wilkinson GR, Shand DG. A physiological approach to hepatic drug clearance. Clinical Pharmacology & Therapeutics. 1975;18:377–390.
- Pang KS, Rowland M. Hepatic clearance of drugs. I. Theoretical considerations of a "well-stirred" model and a "parallel tube" model. Journal of Pharmacokinetics and Biopharmaceutics. 1977.
- Pang KS, Rowland M. Hepatic clearance of drugs. II. Experimental evidence for acceptance of the "well-stirred" model over the "parallel tube" model. Journal of Pharmacokinetics and Biopharmaceutics. 1977.
- Jones HM, Rowland-Yeo K. Basic concepts in physiologically based pharmacokinetic modeling in drug discovery and development. CPT: Pharmacometrics & Systems Pharmacology. 2013;2:e63.
- Edginton AN, Willmann S. Physiologically based pharmacokinetic modeling. In: Drug Discovery and Development applications of PBPK and related approaches.
- Korzekwa K, et al. Process and system clearances in pharmacokinetic models: our basic clearance concepts are correct. Drug Metabolism and Disposition. 2023.
- U.S. Food and Drug Administration. Pharmacokinetics in Patients with Impaired Hepatic Function: Study Design, Data Analysis, and Impact on Dosing and Labeling. Draft Level 1 Guidance. 2026.
Where to Go Next
A natural progression from hepatic clearance is to study mechanistic absorption models in PBPK, followed by intestinal and hepatic first-pass extraction, IVIVE, enzyme and transporter-mediated drug-drug interactions, and hepatic impairment models.
These topics build directly on the central PBPK idea: drug-specific experimental information can be combined with physiological system information to predict concentration and exposure across clinical scenarios.