PBPK for Food-Drug Interactions

1. What Is a Food-Drug Interaction?

A food-drug interaction occurs when food or a meal changes the pharmacokinetics of a drug relative to an appropriate fasted condition. For orally administered drugs, the effect may appear as a change in the rate of absorption, the extent of absorption, or both.

Food can alter the gastrointestinal environment in several ways. Meal ingestion can change gastric pH, gastric emptying, gastrointestinal transit, fluid volume, bile secretion, intestinal physiology, and splanchnic blood flow. These changes can interact with the physicochemical properties of the drug and the characteristics of its formulation.

The FDA's current 2026 guidance emphasizes the importance of evaluating food effects during development because food can increase or decrease drug absorption and therefore potentially alter safety and effectiveness. :contentReference[oaicite:1]{index=1}

2. Why Use PBPK for Food Effects?

A conventional food-effect study can establish that exposure differs between fed and fasted conditions. A mechanistic PBPK model can go one step further by representing the processes that generate that difference.

PBPK models integrate drug-specific information with physiological information. FDA describes PBPK as an approach that combines drug and system information within a mathematical modeling framework. :contentReference[oaicite:2]{index=2}

The advantage is therefore not simply computational complexity. The value of PBPK comes from connecting measurable drug properties and physiological mechanisms to clinical concentration-time behavior.

3. How Can Food Change Drug Exposure?

Food effects are often multifactorial. The dominant mechanism depends on the drug's physicochemical properties, formulation, dose, site of absorption, and susceptibility to gastrointestinal and hepatic processes.

Reviews of food effects emphasize that meal ingestion can modify gastric and intestinal pH, bile concentrations, GI transit, fluid characteristics, and other physiological factors. :contentReference[oaicite:3]{index=3}

4. Modeling Fed and Fasted Conditions

The central comparison in a food-effect PBPK analysis is usually between a fasted state and a fed state. The drug and formulation may be identical, while the physiological environment is changed.

Similarly, the food effect on peak exposure can be summarized as:

A ratio greater than one indicates higher exposure in the fed condition, while a ratio below one indicates lower exposure. The ratio itself, however, does not identify the mechanism.

A mechanistic PBPK model attempts to reproduce the observed ratio by changing physiologically meaningful inputs between fasted and fed states rather than simply inserting an empirical multiplier.

5. Food Effects Often Begin in the GI Tract

For an oral drug, systemic exposure depends on the amount of drug that becomes available for absorption and subsequently escapes intestinal and hepatic first-pass loss.

A useful conceptual decomposition is:

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

Food can influence one or several of these terms. For example, a meal can improve dissolution and therefore increase \(F_a\). A meal or specific food component can also alter intestinal enzyme or transporter activity, potentially affecting \(F_g\).

For a food-effect PBPK model, separating these mechanisms is valuable because an increase in systemic exposure caused by improved absorption has a different interpretation from an increase caused by reduced first-pass metabolism.

6. Food, Solubility, and Dissolution

One of the most important food-effect mechanisms for oral drugs is a change in the environment in which the drug dissolves.

For ionizable compounds, pH can substantially affect apparent solubility. A simplified conceptual relationship for a monoprotic weak base is:

For a monoprotic weak acid:

These equations are simplified representations rather than universal dissolution models. Real formulations may require more detailed treatment of salt form, precipitation, supersaturation, particle size, excipients, and the complete pH-solubility profile.

Food can change gastric pH and the timing of gastric re-acidification. These effects can alter the amount of drug that dissolves before the formulation reaches the small intestine.

Food-effect PBPK work has identified fed-state solubility, gastric re-acidification, dissolution, precipitation, and gastric emptying as important sources of uncertainty in some predictive models. :contentReference[oaicite:4]{index=4}

7. Bile, Lipids, and Poorly Soluble Drugs

Meals, particularly meals containing substantial fat, can change bile secretion and the intestinal environment. Bile components can enhance the apparent solubilization of some lipophilic compounds.

This can be particularly relevant for poorly water-soluble drugs. A fed-state environment may maintain more drug in a solubilized form, increasing the amount available for absorption.

The resulting mechanism can be represented conceptually as:

However, increased solubilization does not automatically mean increased systemic exposure. The drug must still cross the intestinal wall, and intestinal and hepatic first-pass processes can influence the final effect.

This is why PBPK modeling is useful: it allows multiple processes to operate simultaneously rather than assuming that one observed change represents the entire food effect.

8. Gastric Emptying and GI Transit

Food can delay gastric emptying and modify the movement of a dosage form through the gastrointestinal tract. This can change the timing of drug delivery to the small intestine.

For an immediate-release drug, delayed gastric emptying may shift the time at which drug reaches the primary absorption site. For a modified-release formulation, the interaction can be more complex because formulation release and physiological transit occur simultaneously.

A change in gastric emptying can therefore alter \(T_{\max}\) and \(C_{\max}\), while its effect on AUC depends on whether absorption is complete and on other simultaneous mechanisms.

9. Food Effects on Systemic Disposition

Not every food effect is confined to the gastrointestinal lumen. Food can also alter physiological conditions relevant to systemic disposition.

Meal ingestion can increase splanchnic blood flow, and food components can sometimes affect intestinal enzymes or transporters. Such mechanisms may influence first-pass availability and, depending on the drug, systemic exposure.

The appropriate mechanisms depend on the compound. A PBPK model should not include every theoretically possible food effect merely because it can be represented mathematically.

10. Anatomy of a Food-Effect PBPK Model

A food-effect PBPK model generally contains several interacting layers:

1. Drug properties: molecular weight, lipophilicity, ionization, solubility, permeability, and metabolic characteristics.
2. Formulation properties: dosage form, particle size, release behavior, dissolution, precipitation, and other product characteristics.
3. GI physiology: organ volumes, pH, fluid volumes, transit times, bile concentrations, and gastric emptying.
4. Absorption: dissolution, luminal drug concentration, permeability, and intestinal uptake.
5. First-pass processes: intestinal and hepatic metabolism and transport where relevant.
6. Systemic disposition: distribution, metabolism, excretion, and organ blood flows.

The fed and fasted simulations can then differ through physiologically meaningful inputs rather than through an arbitrary food-effect correction factor.

11. What Information Does the Model Need?

The quality of a food-effect PBPK prediction depends strongly on the quality and relevance of its inputs.

Not all parameters need to be estimated from the clinical food-effect study. Many should be informed independently from in vitro, literature, formulation, preclinical, or clinical data and then evaluated as part of model development and verification.

12. Translating the Model Into a Food Effect

After simulating both conditions, the model can generate exposure metrics such as AUC and \(C_{\max}\).

The predicted food effect can then be summarized as:

and:

These metrics are useful because they connect the mechanistic simulation back to the clinical quantities used in food-effect studies.

13. Worked Example: A Hypothetical Food Effect

Consider a hypothetical immediate-release oral drug. Suppose a fasted-state PBPK simulation predicts:

• AUC_fasted = 100 mg·h/L
• C_max,fasted = 10 mg/L
• T_max,fasted = 2.0 h

After changing the model to represent a fed-state physiological environment, suppose the model predicts:

• AUC_fed = 160 mg·h/L
• C_max,fed = 14 mg/L
• T_max,fed = 3.0 h

Step 1: Predicted AUC food effect

The model therefore predicts a 1.60-fold AUC in the fed state relative to fasting.

Step 2: Predicted C_max food effect

The predicted peak concentration is 1.40-fold higher in the fed condition.

Step 3: Change in T_max

The fed simulation predicts a one-hour delay in the time to peak concentration.

Step 4: Mechanistic interpretation

Suppose the model indicates that food increases intestinal solubilization while simultaneously delaying gastric emptying. The increased solubilization can increase the fraction available for absorption, increasing AUC, while delayed delivery can shift \(T_{\max}\).

14. Why Can Food Decrease Exposure?

Food effects are not necessarily positive. Food can sometimes reduce drug exposure.

Potential mechanisms include reduced dissolution under a particular fed-state environment, delayed delivery that interacts unfavorably with the drug or formulation, degradation or instability, altered precipitation, or other formulation-specific effects.

A mechanistic PBPK model can be particularly useful when the direction of the effect is unexpected because it allows competing mechanisms to be examined separately.

These patterns are descriptive. A mechanistic interpretation requires consideration of the drug, formulation, physiology, and model assumptions.

15. Why Does Meal Composition Matter?

Different meals can produce different gastrointestinal environments. Fat content is particularly important because meals with substantial fat can influence gastric emptying, bile secretion, and intestinal solubilization.

Current FDA guidance recommends a high-fat meal in food-effect studies for orally administered drugs under development because high-fat meals generally produce substantial effects on drug absorption; additional meal types may be evaluated when clinically relevant. :contentReference[oaicite:5]{index=5}

From a PBPK perspective, the important question is not simply whether a meal is labeled "high fat." The model needs to represent the physiological consequences that are relevant to the drug and formulation.

16. How Should a Food-Effect PBPK Model Be Evaluated?

A PBPK model should be evaluated against data that were not simply used to construct the model whenever possible. The goal is to determine whether the model is sufficiently predictive for its intended use.

1. Verify the drug model under baseline conditions. Establish that the model can describe relevant PK before adding the food-effect mechanism.
2. Characterize the absorption model. Evaluate dissolution, solubility, permeability, and formulation assumptions.
3. Establish fed and fasted physiology. Use defensible physiological inputs and document their sources.
4. Simulate the food-effect study. Predict concentration-time profiles and exposure metrics.
5. Compare predictions with observations. Examine AUC, C_max, T_max, and concentration-time profiles.
6. Investigate discrepancies. Determine whether uncertainty arises from dissolution, precipitation, gastric emptying, physiology, or other assumptions.
7. Assess sensitivity and uncertainty. Identify parameters that materially influence the predicted food effect.
8. Document the model. Clearly describe assumptions, data sources, parameter values, and model changes.

FDA's PBPK guidance recommends structured reporting of the executive summary, introduction, materials and methods, results, discussion, and appendices for PBPK analyses submitted to the agency. :contentReference[oaicite:6]{index=6}

17. Sensitivity Analysis: Which Mechanisms Matter?

Food-effect PBPK models can contain many parameters. Sensitivity analysis helps identify which assumptions have the greatest influence on predicted exposure.

For a parameter \(\theta\), a local sensitivity concept can be represented as:

where \(P\) is a model prediction such as AUC or C_max.

A large magnitude of sensitivity means that relatively small changes in the parameter can produce relatively large changes in the prediction.

Sensitivity analysis is especially useful when the observed food effect is difficult to reproduce. It can show whether the prediction is robust or whether it depends heavily on a small number of uncertain inputs.

18. A Practical PBPK Food-Effect Workflow

1. Define the question. Is the objective to explain an observed food effect, predict a food effect, evaluate a formulation, or support dosing instructions?
2. Characterize the drug. Assemble physicochemical, permeability, metabolism, transport, and formulation information.
3. Build the baseline model. Establish the drug's PK under conditions where clinical data are available.
4. Build the oral absorption model. Represent dissolution, solubility, precipitation, permeability, and GI transit as appropriate.
5. Define fasted physiology. Establish the baseline GI environment.
6. Define fed physiology. Introduce defensible changes in pH, gastric emptying, bile, fluid, transit, and other relevant variables.
7. Predict the food effect. Simulate concentration-time profiles under both conditions.
8. Compare AUC and C_max. Calculate predicted fed/fasted ratios.
9. Evaluate the mechanism. Determine which physiological and drug-specific processes drive the difference.
10. Perform sensitivity analysis. Identify important uncertainties.
11. Validate or qualify the model. Compare predictions against appropriate clinical and experimental observations.
12. Use the model cautiously. Distinguish predictions supported by evidence from extrapolations dependent on assumptions.

19. What PBPK Food-Effect Models Do Not Tell Us Automatically

A mechanistic model can provide valuable insight, but mechanistic detail does not guarantee predictive accuracy.

• A mechanistic model is still a model. Physiological processes are represented through assumptions and parameterizations.
• Input uncertainty matters. Poorly characterized solubility, precipitation, or fed-state physiology can limit predictions.
• Food effects can be formulation-specific. The same active ingredient can behave differently in different dosage forms.
• Different mechanisms can produce similar PK profiles. Identifiability can therefore be limited.
• Clinical variability remains important. Meal composition, timing, physiology, and individual characteristics can vary.
• Extrapolation requires evidence. A model validated for one meal, formulation, or population is not automatically validated for every other condition.
• Model acceptance is context-dependent. Regulatory use depends on the intended application and the quality, relevance, and reliability of the analysis. :contentReference[oaicite:7]{index=7}

20. PBPK and Regulatory Food-Effect Assessment

Food-effect studies remain an important component of clinical development for orally administered drugs. The FDA's current guidance, issued in May 2026, provides recommendations for sponsors conducting food-effect studies under INDs and using the information to support NDAs and supplements. :contentReference[oaicite:8]{index=8}

PBPK can complement these studies by providing a mechanistic framework for interpreting the observed food effect and exploring scenarios that are difficult to study empirically.

FDA's PBPK guidance describes applications of PBPK analyses across the drug-development lifecycle and notes that acceptance of PBPK results in place of clinical PK data is determined case by case based on the intended use and the quality, relevance, and reliability of the analysis. :contentReference[oaicite:9]{index=9}

For oral drug products, FDA has also published guidance concerning the use of PBPK analyses in biopharmaceutics applications, including oral drug product development and manufacturing changes. :contentReference[oaicite:10]{index=10}

Where to Go Next

A natural progression is to study mechanistic oral absorption models, followed by dissolution and precipitation modeling, GI transit models, gastric pH effects, intestinal metabolism and transport, and PBPK-based drug-drug interaction modeling.

The next tutorial can build directly on this framework by examining PBPK for CYP-Mediated Drug Interactions and showing how enzyme inhibition, induction, intestinal metabolism, hepatic clearance, and perpetrator exposure are incorporated into mechanistic simulations.

References

1. U.S. Food and Drug Administration. Assessing the Effects of Food on Drugs in INDs and NDAs – Clinical Pharmacology Considerations. Final Guidance for Industry, May 2026. FDA. :contentReference[oaicite:11]{index=11}
2. U.S. Food and Drug Administration. Physiologically Based Pharmacokinetic Analyses — Format and Content Guidance for Industry. September 2018. FDA. :contentReference[oaicite:12]{index=12}
3. U.S. Food and Drug Administration. The Use of Physiologically Based Pharmacokinetic Analyses — Biopharmaceutics Applications for Oral Drug Product Development, Manufacturing Changes, and Controls. September 2020. FDA. :contentReference[oaicite:13]{index=13}
4. Parrott N, et al. Predictive Performance of Physiologically Based Pharmacokinetic Models for the Effect of Food on Oral Drug Absorption: Current Status. CPT: Pharmacometrics & Systems Pharmacology. 2017. PubMed. :contentReference[oaicite:14]{index=14}
5. Heimbach T, et al. Food Effects on Oral Drug Absorption: Application of Physiologically-Based Pharmacokinetic Modeling as a Predictive Tool. European Journal of Pharmaceutics and Biopharmaceutics. 2020. PubMed. :contentReference[oaicite:15]{index=15}
6. Jamei M, et al. Use of Physiologically Based Pharmacokinetic Modeling for Predicting Drug-Food Interactions: Recommendations for Improving Predictive Performance of Low Confidence Food Effect Models. Clinical Pharmacology & Therapeutics. 2021. PubMed. :contentReference[oaicite:16]{index=16}
7. Abuhelwa AY, et al. Food, gastrointestinal pH, and models of oral drug absorption. European Journal of Pharmaceutics and Biopharmaceutics. 2017. PubMed. :contentReference[oaicite:17]{index=17}
8. Kambayashi A, et al. Food effects on gastrointestinal physiology and drug absorption. Drug Metabolism and Pharmacokinetics. 2023. PubMed. :contentReference[oaicite:18]{index=18}