Mechanistic Absorption Models in PBPK

1. What Is a Mechanistic Absorption Model?

A mechanistic absorption model describes how a drug moves from its administered dosage form through the gastrointestinal tract and ultimately enters the systemic circulation using explicit representations of relevant physical, chemical, and physiological processes.

In a conventional compartmental PK model, oral absorption may be summarized with a single first-order absorption rate constant \(k_a\). That approach can be useful, but it does not explicitly explain why absorption occurs at a particular rate.

A PBPK absorption model instead attempts to represent the processes that generate the observed input into the systemic circulation. Depending on the model, these can include dosage-form disintegration, dissolution, precipitation, luminal solubility, intestinal transit, permeability, surface area, enterocyte metabolism, and intestinal or hepatic first-pass extraction.

2. Why Move Beyond a First-Order \(k_a\)?

A first-order absorption model commonly represents systemic input as proportional to the amount of drug remaining at the absorption site:

This is compact and often useful for describing observed data. However, \(k_a\) is generally an effective parameter. It summarizes multiple processes without explicitly separating them.

For example, a slow observed absorption profile could arise from slow dissolution, low intestinal permeability, prolonged gastric emptying, precipitation after dissolution, or a combination of these processes. A single \(k_a\) does not necessarily distinguish among these mechanisms.

Mechanistic PBPK models attempt to make these processes explicit. Instead of asking only, "What value of \(k_a\) fits the data?", the model asks questions such as:

• How quickly does the dosage form release drug?
• How much drug dissolves in each intestinal segment?
• Is luminal concentration limited by aqueous solubility?
• Does drug precipitate after moving into a different intestinal environment?
• How readily does dissolved drug cross the intestinal membrane?
• How does intestinal transit determine the time available for dissolution and absorption?

3. The Gastrointestinal Tract as an Absorption System

After oral administration, a drug encounters a changing gastrointestinal environment rather than a single uniform absorption compartment.

A mechanistic model can divide the gastrointestinal tract into segments and allow drug to move between them according to physiological transit assumptions.

The result is a model in which the amount of drug available for absorption can change with both time and location.

4. Dissolution: From Solid Drug to Dissolved Drug

For an immediate-release oral solid dosage form, drug generally must become dissolved before it can cross the intestinal membrane. Dissolution therefore creates an important link between formulation properties and systemic exposure.

A simplified dissolution representation can be written as:

More mechanistic dissolution models may instead relate the dissolution rate to particle surface area, solubility, diffusional properties, and the concentration gradient between the dissolving solid and surrounding fluid.

A conceptual form of the Noyes–Whitney relationship is:

where \(A\) represents available particle surface area, \(C_s\) is the saturation solubility, and \(C\) is the dissolved concentration.

Why particle size matters

Smaller particles generally provide greater surface area for a given mass of material. Reducing particle size can therefore increase dissolution rate when dissolution is a relevant limiting process.

Why formulation matters

Salt form, crystal form, particle size, amorphous content, coatings, excipients, and release characteristics can all influence how rapidly drug becomes available in solution.

5. Solubility and Luminal Concentration

Once drug is dissolved, its concentration in the intestinal lumen cannot increase indefinitely. Solubility places an upper constraint on the dissolved concentration under the relevant conditions.

For a simplified system:

where \(S\) represents the relevant solubility under the modeled intestinal conditions.

For ionizable compounds, solubility can depend strongly on pH. A weak base, for example, may have substantially different apparent solubility in an acidic environment compared with a more neutral intestinal environment.

Consequently, a PBPK absorption model may need to represent:

• regional gastrointestinal pH;
• intrinsic solubility;
• ionization state;
• aqueous or luminal volume;
• bile-associated solubilization where relevant; and
• changes in solubility as the drug moves through the GI tract.

These properties can make absorption region-dependent. A drug that dissolves readily in one GI segment may encounter very different conditions after transit to another segment.

6. Precipitation and Supersaturation

Dissolution and precipitation are related but distinct processes. A drug can temporarily reach a dissolved concentration above its equilibrium solubility, creating a supersaturated state.

Supersaturation can occur, for example, when a formulation or drug form initially produces a high dissolved concentration and the intestinal environment subsequently changes.

A simplified conceptual precipitation model is:

where \((x)_+=\max(x,0)\), and \(C_{\mathrm{eq}}\) represents an equilibrium concentration.

The exact treatment of precipitation can differ substantially between mechanistic absorption models. Some models explicitly represent precipitation kinetics, while others use simpler approaches to constrain dissolved concentration.

7. Intestinal Permeability

After becoming dissolved, drug must cross the intestinal epithelial barrier to enter the portal circulation. The rate of this process depends on drug properties and physiological characteristics of the intestinal wall.

A simplified membrane-flux relationship is:

where \(J\) is flux and \(P_{\mathrm{eff}}\) is an effective permeability term.

In a mechanistic PBPK model, permeability can be connected to physicochemical properties such as lipophilicity, ionization, molecular size, and hydrogen-bonding characteristics, together with experimental permeability measurements where available.

Permeability should not be interpreted as the only determinant of absorption. A highly permeable compound can still have incomplete absorption if dissolution, solubility, transit, or other processes limit the amount available to permeate.

Passive versus transporter-mediated processes

Passive diffusion is only one possible mechanism. Some compounds interact with uptake or efflux transporters expressed in the intestinal epithelium. When these processes are scientifically important and adequately characterized, they can be incorporated into a more detailed model.

8. Gastric Emptying and Intestinal Transit

The drug cannot be absorbed from a GI segment before it reaches that segment. Transit therefore affects the time available for dissolution and permeation.

A simplified first-order transit process can be written as:

where \(A_i\) is the amount in intestinal segment \(i\) and \(k_{\mathrm{tr}}\) represents a transit rate.

More detailed models can represent different transit rates in different regions, including gastric emptying and segment-specific intestinal residence times.

Transit becomes particularly important when dissolution or absorption is slow. A drug that remains in the lumen for only a short period may not have enough time to dissolve or permeate completely.

9. Intestinal and Hepatic First-Pass Processes

Not all drug that crosses the intestinal lumen reaches systemic circulation unchanged. Drug may be metabolized within enterocytes and may subsequently undergo hepatic first-pass extraction.

A useful conceptual sequence is:

For an orally administered drug, systemic bioavailability can therefore be viewed conceptually as the product of several processes, including the fraction released and dissolved, the fraction absorbed across the intestinal wall, and the fraction escaping intestinal and hepatic first-pass loss.

A simplified expression is:

where \(F_a\) represents the fraction absorbed, \(F_g\) the fraction escaping intestinal loss, and \(F_h\) the fraction escaping hepatic first-pass extraction.

This decomposition is useful because two drugs can have the same systemic bioavailability for very different mechanistic reasons.

10. Solubility–Permeability Interactions

A useful way to understand oral absorption is to consider both the extent of dissolution and the ability of dissolved drug to permeate.

These categories are useful conceptual guides rather than complete mechanistic classifications. Real compounds can exhibit nonlinear behavior, pH-dependent solubility, transporter effects, precipitation, formulation effects, and regional differences.

11. From GI Processes to Differential Equations

A mechanistic absorption model can be constructed by defining drug amounts or concentrations in different physiological regions and writing equations for the transfer between them.

For a simplified GI segment:

Here:

• \(R_{\mathrm{in}}\) represents drug entering the segment;
• \(R_{\mathrm{diss}}\) represents dissolution of solid drug;
• \(R_{\mathrm{precip}}\) represents drug returning to the solid phase;
• \(R_{\mathrm{transit}}\) represents movement to another GI segment; and
• \(R_{\mathrm{abs}}\) represents transfer across the intestinal wall.

The systemic model then receives the absorbed drug according to the modeled intestinal and hepatic processes.

12. How Is a Mechanistic Absorption Model Built?

1. Define the formulation. Specify the administered dose, dosage form, release characteristics, and relevant solid-state properties.
2. Characterize drug physicochemistry. Gather information on molecular weight, ionization, lipophilicity, solubility, permeability, and other relevant properties.
3. Specify the GI physiology. Define gastrointestinal volumes, pH, transit, fluid conditions, and other physiological assumptions.
4. Represent dissolution and precipitation. Select an appropriate model for release into solution and changes in dissolved concentration.
5. Represent intestinal permeation. Link dissolved drug concentration to transfer across the intestinal wall.
6. Add intestinal metabolism or transport when necessary. Include these processes when evidence indicates they are important and parameters are sufficiently characterized.
7. Connect the gut to the systemic PBPK model. Absorbed drug enters portal circulation and then the hepatic and systemic compartments.
8. Evaluate the model. Compare predicted profiles and exposure with appropriate experimental observations.
9. Perform sensitivity analysis. Determine which parameters meaningfully control predicted exposure or absorption.

13. Worked Example: When Does Dissolution Limit Absorption?

Consider a hypothetical immediate-release oral drug with a dose of 100 mg. Suppose the intestinal absorption process is highly permeable, but the drug has relatively low aqueous solubility.

Assume that, under the modeled GI conditions:

• 100 mg is administered orally;
• the effective intestinal solubility is \(0.5\text{ mg/mL}\);
• the relevant intestinal fluid volume is \(250\text{ mL}\); and
• the drug is sufficiently permeable that dissolved drug can be absorbed efficiently.

Step 1: Estimate the equilibrium dissolved amount

A simple equilibrium upper bound is:

The nominal dose of 100 mg is below this simplified solubility capacity. Therefore, under these assumptions, equilibrium solubility alone would not necessarily prevent the entire dose from becoming dissolved.

Step 2: Consider a lower solubility

Now suppose the effective solubility is only \(0.2\text{ mg/mL}\).

Only 50 mg could be present as dissolved drug at equilibrium in this simplified static-volume calculation.

Step 3: Why this does not automatically mean \(F_a=0.5\)

The calculation is intentionally simplified. In an actual mechanistic PBPK absorption model, intestinal fluid volume changes, transit moves drug between segments, dissolved drug is continuously removed by absorption, pH can change, bile components can alter solubilization, and supersaturation or precipitation can occur.

Therefore, the result should not be interpreted as saying that exactly 50% of the dose will be absorbed.

14. Why Mechanistic Absorption Models Can Help Explain Food Effects

Food can change several gastrointestinal conditions simultaneously. Depending on the drug and formulation, relevant changes can include gastric emptying, GI fluid volume, pH, bile secretion, intestinal motility, and luminal solubilization.

A simple empirical model might represent a food effect by changing a parameter or applying an observed correction factor. A mechanistic model instead attempts to represent the physiological changes that generate the food effect.

The value of a mechanistic model is that multiple physiological changes can be represented simultaneously rather than treating "fed" and "fasted" as completely unrelated empirical states.

15. Formulation Development and Virtual Bioequivalence Questions

Mechanistic absorption models can also provide a framework for exploring how formulation changes may affect exposure.

Potential formulation variables include:

• particle size;
• salt or solid form;
• amorphous versus crystalline material;
• drug loading;
• release rate;
• coating or controlled-release properties; and
• excipients that alter dissolution or solubilization.

For example, reducing particle size may increase the surface area available for dissolution. A model can therefore propagate that formulation change through dissolution, luminal concentration, absorption, and ultimately systemic concentration.

This creates a conceptual chain:

Such predictions should be evaluated against appropriate experimental data. A mechanistic model is not a substitute for empirical validation, particularly when important formulation or physiological processes are uncertain.

16. Sensitivity Analysis: What Actually Controls Absorption?

One of the most useful applications of a mechanistic model is to determine which assumptions or parameters have the greatest influence on predicted exposure.

Candidate parameters might include:

• intestinal solubility;
• effective permeability;
• particle size;
• dissolution rate;
• precipitation rate;
• gastric emptying rate;
• intestinal transit time;
• intestinal metabolism;
• hepatic clearance; and
• luminal fluid or bile-related parameters.

A local sensitivity measure can be expressed conceptually as:

where \(Y\) is an output such as AUC or \(C_{\max}\), and \(p\) is a model parameter.

A large absolute sensitivity indicates that relatively small proportional changes in the parameter can produce relatively large proportional changes in the output.

17. Mechanistic Absorption Models vs. Simpler Approaches

The choice should depend on the scientific question. A mechanistic absorption model is most useful when the mechanisms being represented are relevant to the decision and the available information is sufficient to parameterize and evaluate the model.

18. Model Identifiability and Parameter Uncertainty

Mechanistic models contain more parameters than simple empirical absorption models. That additional detail creates both an advantage and a challenge.

Multiple parameter combinations can sometimes generate similar plasma concentration-time profiles. For example, a reduction in permeability and a reduction in the amount of dissolved drug could both reduce systemic exposure.

This means that a good fit to plasma concentrations does not necessarily prove that every mechanistic parameter has been uniquely identified.

External information—such as measured solubility, permeability, dissolution profiles, formulation characterization, or physiological measurements—can therefore be extremely valuable for constraining mechanistic models.

19. How Should a Mechanistic Absorption Model Be Evaluated?

Model evaluation should consider more than whether a simulated plasma profile visually resembles an observed curve.

1. Check input assumptions. Are physicochemical and physiological inputs supported by appropriate evidence?
2. Evaluate dissolution behavior. Does the model reproduce relevant dissolution observations where such data are available?
3. Evaluate absorption. Does the model reasonably describe observed \(C_{\max}\), \(T_{\max}\), AUC, and concentration-time profiles?
4. Evaluate multiple conditions. Can the model describe more than one dose, formulation, or physiological condition?
5. Examine parameter plausibility. Are estimated or assumed parameters physiologically and physically reasonable?
6. Perform sensitivity analysis. Identify predictions that depend strongly on uncertain assumptions.
7. Assess extrapolation carefully. Predictions outside the conditions used for development require additional support.

A model that predicts several independent datasets or conditions using a consistent parameterization provides more informative evidence than a model that only reproduces a single concentration-time profile.

20. Where Are Mechanistic Absorption Models Used?

Mechanistic absorption models can support a range of pharmacometric and drug-development questions.

21. What Mechanistic Absorption Models Do Not Tell Us Automatically

A mechanistic structure does not guarantee that every prediction is correct. Several limitations deserve particular attention.

• More detail does not automatically mean more accuracy. Incorrect assumptions can make a complex model misleading.
• Input uncertainty propagates into predictions. Uncertain solubility, permeability, transit, or formulation properties can materially affect results.
• Physiology varies between individuals. A population-average GI model may not represent every individual.
• Different mechanisms can compensate for one another. Multiple parameter combinations may produce similar plasma profiles.
• Experimental measurements are context-dependent. In vitro permeability or dissolution data may not translate directly to the in vivo environment.
• Extrapolation requires evidence. A model validated under one set of conditions may not automatically be valid under another.
• Model complexity has a cost. More parameters require more data, more assumptions, and more careful evaluation.

22. A Practical Workflow for Mechanistic Absorption Modeling

1. Start with the scientific question. Determine whether the question concerns dose, formulation, food, physiology, drug interactions, or another factor affecting absorption.
2. Characterize the drug. Gather physicochemical, dissolution, permeability, and solid-state information.
3. Characterize the formulation. Identify dosage-form properties that influence drug release and dissolution.
4. Define the GI physiology. Specify relevant pH, fluid, transit, and regional assumptions.
5. Build the absorption model. Connect dissolution, solubility, precipitation, transit, and permeability processes.
6. Add first-pass mechanisms. Represent intestinal and hepatic metabolism or transport when relevant.
7. Connect to systemic PBPK. Use the absorbed drug as input into the systemic distribution and elimination model.
8. Evaluate against data. Compare predictions with appropriate in vitro, preclinical, and clinical observations.
9. Perform sensitivity and uncertainty analysis. Identify influential parameters and important uncertainties.
10. Use the model for prediction. Simulate scenarios that answer the original scientific question, while clearly distinguishing model-based predictions from direct observations.

Where to Go Next

A natural progression is to study PBPK gastrointestinal physiology, followed by dissolution and precipitation modeling, intestinal permeability, transporter-mediated absorption, first-pass metabolism, and food-effect prediction.

From there, mechanistic absorption models can be connected to full PBPK models to examine how changes in absorption propagate through distribution, metabolism, clearance, and systemic exposure.

The next tutorial can build directly on this framework by examining physiological factors controlling oral drug absorption and showing how gastric emptying, intestinal transit, pH, fluid volume, bile components, and regional intestinal properties are represented quantitatively.

References