PBPK for Pediatric Populations

1. Why Do Pediatric PBPK Models Matter?

Pediatric physiologically based pharmacokinetic (PBPK) modeling uses a mechanistic representation of the body to predict how drug concentrations may change across pediatric ages and developmental stages.

Children are not simply smaller adults. Development can alter body size, organ size, tissue composition, blood flow, plasma proteins, glomerular filtration, tubular function, metabolic enzyme activity, transporter expression, and gastrointestinal physiology. These changes can affect both the exposure and disposition of a drug.

A pediatric PBPK model attempts to represent those changes explicitly rather than relying only on an adult dose scaled by body weight.

2. Why Can Drug Disposition Differ Across Pediatric Ages?

Several biological processes change during development. Importantly, they do not all mature at the same rate.

These developmental processes create an important distinction between growth and maturation. Growth refers primarily to changes in physical size, whereas maturation refers to age-dependent changes in physiological function.

3. Growth and Maturation Are Not the Same Thing

A major strength of pediatric PBPK is that it can represent growth and maturation separately.

Growth

Growth changes quantities such as body weight, organ weight, blood volume, and tissue volumes. These changes can often be represented using age-dependent physiological scaling relationships.

For example, a simplified organ-weight relationship can be represented conceptually as:

As the child grows, organ mass and tissue volumes generally increase. A PBPK model can translate these changes into changes in the size of individual physiological compartments.

Maturation

Maturation describes changes in functional capacity that are not explained by size alone. Examples include development of renal filtration and the ontogeny of metabolic enzymes.

A simplified maturation function can be written as:

In this representation, \(M(\text{age})\) approaches the adult reference value as the relevant physiological process matures.

4. Enzyme and Transporter Ontogeny

Ontogeny describes the development of biological systems over the lifespan. In pediatric PBPK, ontogeny is especially important when a drug depends on metabolic enzymes or transporters whose activity changes with age.

Consider a metabolic pathway that contributes substantially to clearance. A conceptual expression for age-dependent intrinsic clearance might be:

Here, \(M_{\text{enzyme}}(\text{age})\) represents the developmental maturation of the relevant enzyme pathway.

For a drug primarily metabolized by one pathway, an incorrect ontogeny assumption can propagate directly into the predicted pediatric exposure.

Not every enzyme or transporter follows the same developmental trajectory. Therefore, pediatric PBPK should use pathway-specific evidence when available rather than assuming that all metabolic capacity changes according to a single universal maturation curve.

5. Renal Function and Pediatric Clearance

Renal elimination can be particularly important in neonates and young infants because renal function changes substantially during early development.

For a drug eliminated primarily by filtration, a simplified relationship is:

where \(f_u\) is the unbound fraction and \(GFR\) is the glomerular filtration rate.

In a pediatric PBPK model, the relevant renal physiological quantities can be represented as functions of age, body size, and developmental maturation.

For drugs that undergo active renal secretion or reabsorption, the model may additionally need transporter or tubular-process components. The importance of these mechanisms depends on the drug and the evidence supporting each pathway.

6. Hepatic Clearance in Pediatric PBPK

Many drugs depend on hepatic metabolism. PBPK models can represent hepatic clearance mechanistically by combining hepatic blood flow, protein binding, intrinsic metabolic capacity, and liver physiology.

A commonly used conceptual framework is the well-stirred hepatic model:

where:

• \(Q_H\) is hepatic blood flow,
• \(f_u\) is the unbound fraction in blood or plasma as appropriate to the model, and
• \(CL_{\text{int}}\) represents intrinsic hepatic clearance.

In pediatric PBPK, each of these components can potentially differ from adult values because of developmental physiology.

For example, intrinsic clearance may depend on enzyme ontogeny, while hepatic blood flow and liver size depend on developmental physiology and growth.

7. Pediatric Body Composition and Distribution

Drug distribution can also change during development because the relative proportions of total body water, extracellular water, fat, lean tissue, and individual organs change with age.

For a simplified compartmental relationship:

changes in the effective distribution volume can therefore influence concentration even when the administered dose is identical.

PBPK models go further by representing individual tissues or organs rather than using one global volume of distribution.

These effects are particularly relevant when an adult model is extrapolated to neonates or infants, where body composition can differ substantially from adults.

8. Pediatric Absorption

Development can also affect the absorption of orally administered drugs. Pediatric PBPK models may account for changes in gastrointestinal physiology, depending on the drug and the intended application.

Potentially relevant variables include:

• gastric pH,
• gastric emptying and intestinal transit,
• intestinal surface area,
• bile acid concentrations,
• intestinal fluid composition,
• intestinal metabolic enzymes,
• intestinal transporters, and
• developmental changes in first-pass metabolism.

For a simple first-order absorption model:

where \(A_g\) is the amount in the gastrointestinal absorption compartment and \(k_a\) is the absorption rate constant.

A mechanistic PBPK model can replace this simple representation with more detailed descriptions of dissolution, permeability, intestinal metabolism, and transit when those mechanisms are scientifically relevant.

9. Pediatric Age Groups

Pediatric PBPK applications often need to distinguish developmental stages because physiological changes are not uniform across childhood.

Age should therefore be treated as more than a simple label. Depending on the application, relevant descriptors may include chronological age, body weight, height, gestational age, postmenstrual age, developmental stage, and pubertal status.

10. What Does a Pediatric PBPK Model Contain?

A typical PBPK model separates the system into physiological compartments representing blood and tissues. Drug-specific properties determine how the compound moves through those compartments.

The pediatric model is therefore a combination of two major information streams:

1. Drug-specific information: physicochemical properties, protein binding, permeability, metabolism, transport, and other compound characteristics.
2. Population-specific physiology: organ size, blood flow, body composition, enzyme and transporter activity, renal function, and other age-dependent characteristics.

11. How Is an Adult PBPK Model Scaled to Children?

A common pediatric PBPK workflow begins with a model that has been developed and evaluated using adult data. The model is then adapted using pediatric physiology and developmental information.

Conceptually:

For example, a hepatic metabolic pathway might be represented as:

where \(F_{\text{size}}\) represents scaling related to physiological capacity and \(F_{\text{ontogeny}}\) represents developmental maturation of the pathway.

The exact mathematical implementation depends on the PBPK platform, drug, metabolic pathway, available data, and physiological model.

12. Pediatric PBPK Versus Simple Allometric Scaling

Simple allometric approaches often express a parameter as a function of body weight:

where \(b\) is an allometric exponent.

This can provide a useful empirical scaling relationship, but it does not explicitly represent the mechanisms responsible for developmental changes.

PBPK does not automatically outperform simpler approaches for every application. Its main advantage is that it can represent specific physiological mechanisms when those mechanisms are important to the question being addressed.

13. Worked Example: Scaling Hepatic Clearance Across Age

Consider a hypothetical drug for which hepatic metabolism is an important component of systemic clearance. Suppose an adult model estimates an intrinsic hepatic clearance of 100 L/h.

Assume, purely for illustration, that a particular pediatric physiological model implies:

• a size-related hepatic capacity factor of 0.30, and
• an enzyme maturation factor of 0.60.

The illustrative pediatric intrinsic clearance would then be:

Step 1: Separate size and maturation

The first factor represents the smaller physiological capacity associated with the child's size. The second represents the fact that the metabolic pathway has not reached the assumed adult level of functional activity.

Step 2: Apply the scaled intrinsic clearance

Step 3: Combine with hepatic physiology

If hepatic blood flow and protein binding are also represented in the model, the scaled intrinsic clearance can be inserted into a hepatic clearance relationship such as the well-stirred model:

The resulting clearance is therefore not obtained simply by multiplying an adult clearance by the child's body weight. It emerges from the interaction of physiological blood flow, binding, organ capacity, and metabolic activity.

14. Why Neonates Require Special Attention

Neonatal PBPK can be particularly challenging because several physiological systems are simultaneously undergoing rapid development.

For neonates, the model may need to distinguish between:

• gestational age,
• postnatal age,
• postmenstrual age,
• body weight,
• organ growth,
• renal maturation,
• hepatic enzyme ontogeny, and
• changes in protein binding and body composition.

For a preterm infant, chronological age alone may not adequately describe developmental physiology. A PBPK model may therefore use developmental age metrics that separate maturation associated with gestation from maturation occurring after birth.

15. Why Pediatric Patients Can Show Substantial Variability

Age is only one source of variability in pediatric pharmacokinetics.

Two children of the same chronological age may differ in body size, organ function, developmental stage, disease status, concomitant medications, and other characteristics that affect drug disposition.

A deterministic pediatric PBPK simulation describes a specified physiological scenario. When patient variability is important, additional approaches such as population simulation or sensitivity analysis can be used to characterize a range of possible exposures.

16. Incorporating Pediatric Covariates

Pediatric PBPK models can incorporate covariates that influence physiology or drug disposition.

A simplified clearance representation might be written as:

Here, the subscript \(i\) denotes an individual child.

The advantage of a mechanistic framework is that covariates can sometimes be connected to specific physiological processes rather than included solely as empirical statistical predictors.

For example, body weight may influence organ size, while age may influence enzyme ontogeny. A disease covariate may affect renal or hepatic function depending on the disease and available evidence.

17. Pediatric PBPK and Drug-Drug Interactions

PBPK can be particularly useful when pediatric dosing is affected by enzyme or transporter inhibition or induction.

A simplified conceptual chain is:

When an inhibitor is present, the model can incorporate its concentration-dependent effect on a metabolic pathway. When an inducer is present, the model may represent increased enzyme abundance over time.

The pediatric dimension matters because the baseline activity of the affected pathway may differ across ages. Consequently, the same perpetrator drug can produce different interaction magnitudes in different developmental populations.

18. A Practical Pediatric PBPK Modeling Workflow

1. Define the scientific question. Determine whether the objective is dose selection, exposure prediction, formulation assessment, drug interaction prediction, or another application.
2. Characterize the drug. Assemble physicochemical, binding, permeability, metabolism, transport, and elimination information.
3. Develop or select the adult PBPK model. Where possible, establish that the drug model adequately describes relevant adult observations before pediatric extrapolation.
4. Define the pediatric population. Specify age range, body size, developmental stage, and other relevant characteristics.
5. Specify growth parameters. Represent age-dependent body and organ sizes using an appropriate physiological framework.
6. Specify maturation functions. Incorporate evidence-supported developmental changes in renal function, enzymes, transporters, or other relevant systems.
7. Adjust body composition and tissue physiology. Ensure that distribution-related physiological parameters are appropriate for the pediatric population.
8. Simulate pediatric exposure. Generate concentration-time profiles and exposure metrics across the intended age range.
9. Compare with available pediatric observations. Where clinical data exist, evaluate whether the model reproduces observed concentration and exposure behavior.
10. Perform sensitivity and uncertainty analyses. Identify physiological assumptions that materially influence the prediction.
11. Use the qualified model for simulation. Apply the model to dosing scenarios or populations that are scientifically justified by the model's qualification and intended scope.

19. How Should a Pediatric PBPK Model Be Evaluated?

A pediatric PBPK model should be evaluated against observations whenever suitable data are available. Model qualification is more than determining whether a single prediction is close to one observed value.

Useful evaluation questions include:

• Does the model reproduce the observed concentration-time profile?
• Does it reproduce exposure metrics such as AUC and C_max?
• Does it behave reasonably across multiple pediatric age groups?
• Are predicted clearance and distribution parameters physiologically plausible?
• Are discrepancies systematic in particular developmental stages?
• Which assumptions contribute most to prediction uncertainty?
• Does the model remain appropriate when extrapolated to the intended population?

20. Sensitivity Analysis in Pediatric PBPK

Sensitivity analysis helps determine which model inputs have the greatest influence on predicted exposure.

Suppose predicted AUC is represented by:

A sensitivity analysis can vary individual physiological or drug-specific parameters and quantify the resulting change in predicted exposure.

This is particularly valuable in pediatric PBPK because some developmental parameters are difficult to measure directly. Sensitivity analysis helps distinguish assumptions that are consequential from those that have little impact on the prediction.

21. Using Pediatric PBPK for Dose Selection

One of the major applications of pediatric PBPK is to simulate candidate doses and determine whether they are expected to produce exposures within a clinically relevant target range.

Conceptually:

For example, simulations can compare several candidate doses across multiple pediatric weight or age groups.

PBPK predictions should support—not replace—the broader clinical, pharmacological, and regulatory evidence used in pediatric dose selection.

22. Weight-Based Versus Age-Dependent Dosing

Clinical dosing often needs to balance physiological precision with practical implementation. A mechanistic model may predict a continuous dose requirement, while the clinical regimen may need discrete dose bands.

For example, a simulation might estimate an idealized dose:

The final clinical regimen may then use practical dose bands or available dosage strengths.

PBPK can therefore be used to investigate whether a simplified dosing rule adequately approximates the exposure generated by a more individualized physiological model.

23. Limitations of Pediatric PBPK

Pediatric PBPK models can integrate large amounts of mechanistic information, but the predictions remain conditional on the quality of that information.

• Physiological uncertainty: developmental parameters may not be known precisely for every age and population.
• Ontogeny uncertainty: enzyme and transporter maturation may be pathway-specific and incompletely characterized.
• Drug-specific uncertainty: some parameters such as tissue partitioning or transporter contribution may be poorly characterized.
• Population heterogeneity: disease, nutrition, concomitant medications, and other factors can produce variability not represented in a reference physiological population.
• Model structural uncertainty: alternative mechanistic assumptions may produce different predictions.
• Data limitations: sparse pediatric observations can make model evaluation difficult.
• Extrapolation: predictions for age groups or clinical conditions outside the model's qualification range require additional justification.

24. Pediatric PBPK Versus Population PK

PBPK and population PK are complementary approaches rather than competing definitions of pharmacokinetic modeling.

In practice, the approaches can be used together. For example, PBPK can provide mechanistic priors or physiological hypotheses, while population PK can quantify variability using observed clinical concentration data.

25. When Is Pediatric PBPK Particularly Useful?

Pediatric PBPK can be especially informative when developmental physiology is expected to materially influence drug disposition or when pediatric clinical data are limited.

Potential applications include:

• first-in-pediatric dose selection,
• dose selection across multiple pediatric age groups,
• neonatal and infant exposure prediction,
• formulation or route comparisons,
• drug-drug interaction prediction,
• enzyme induction or inhibition scenarios,
• drugs with substantial renal or hepatic maturation effects,
• extrapolation from adult pharmacokinetic information, and
• simulation of alternative pediatric dosing regimens.

The value of PBPK increases when the model contains mechanisms that are relevant to the question and when those mechanisms can be supported by credible data.

26. An End-to-End Pediatric PBPK Workflow

1. Define the drug and clinical question.
2. Characterize drug-specific properties.
3. Develop and qualify the adult PBPK model where appropriate.
4. Select the pediatric physiological framework.
5. Represent age-dependent organ sizes and blood flows.
6. Incorporate body-composition changes.
7. Incorporate renal maturation.
8. Incorporate enzyme and transporter ontogeny where relevant.
9. Represent pediatric gastrointestinal physiology when absorption is important.
10. Simulate the intended pediatric population.
11. Compare predictions with available pediatric observations.
12. Perform sensitivity and uncertainty analyses.
13. Evaluate alternative dosing regimens.
14. Document assumptions, qualification, and limitations.

This workflow emphasizes an important feature of PBPK: the model should be developed around the scientific question, not around the desire to include every possible physiological detail.

Where to Go Next

A natural progression is to study enzyme and transporter ontogeny in PBPK, followed by pediatric renal maturation, pediatric hepatic clearance, age-dependent body composition, pediatric drug-drug interaction modeling, and PBPK-based dose selection.

The next tutorial can build directly on this framework by examining how developmental changes in CYP enzymes and other metabolic pathways are translated into age-dependent intrinsic clearance within a PBPK model.

References

These references provide starting points for understanding pediatric PBPK, developmental physiology, pediatric pharmacokinetics, and regulatory use of mechanistic modeling. Specific model applications should use the most current applicable regulatory guidance and drug-specific evidence.