Tutorials › Pharmacometrics › Intracellular Drug Concentration Models
Pharmacokinetics · Cellular Distribution

Intracellular Drug Concentration Models

Learn how pharmacokinetic models describe drug movement from the systemic circulation into cells, how intracellular concentrations can differ from plasma concentrations, and how transport, binding, and cellular processes shape intracellular exposure.

Intermediate PK Modeling Intracellular PK Pharmacometrics
01 · The big picture

1. Why Model Intracellular Drug Concentrations?

Traditional pharmacokinetic models often describe drug concentrations in plasma or blood. However, the biological effect of a drug frequently occurs somewhere else: inside a cell, at a membrane, within an organelle, or at a specific intracellular target.

This creates an important distinction between systemic exposure and target-site exposure. A drug can have a measurable plasma concentration without producing the same concentration inside the cells that determine its pharmacologic effect.

Plasma systemic concentration Extracellular fluid / tissue concentration Cell intracellular concentration target The site of pharmacologic action may be different from the site of measurement

An intracellular PK model extends the usual plasma-to-tissue framework by explicitly representing drug movement into cells and, when needed, intracellular binding or loss.

Core idea: the concentration measured in plasma is not automatically the concentration experienced by an intracellular target. An intracellular model provides a mathematical framework for describing the relationship between systemic concentration and intracellular exposure.
02 · Model structure

2. From Plasma to the Intracellular Space

The simplest intracellular model adds one or more compartments to a conventional PK model. A central compartment can represent plasma, while additional compartments represent extracellular tissue and intracellular drug.

The exact structure depends on the scientific question. In some applications, a direct plasma-to-intracellular model may be sufficient. In others, extracellular tissue, cellular uptake, intracellular binding, active efflux, or intracellular metabolism may need to be represented explicitly.

Compartment Typical interpretation Potential information
Plasma Systemic circulation Observed plasma concentration and systemic exposure
Extracellular Interstitial or extracellular fluid Drug concentration surrounding cells
Intracellular Cellular drug pool Concentration available to intracellular targets
Bound intracellular Drug associated with intracellular binding sites Free-versus-total intracellular exposure
Organelle Specific intracellular location Subcellular exposure when the mechanism requires it

Not every model needs all of these compartments. Adding a compartment should correspond to a biological or pharmacologic hypothesis that can be supported by the available data.

03 · A simple intracellular model

3. A One-Compartment Intracellular Model

Consider a simplified model with a plasma concentration \(C_p(t)\) and an intracellular concentration \(C_i(t)\). Assume that drug enters the intracellular compartment from plasma with a first-order uptake process and leaves the intracellular compartment with a first-order loss process.

A simple differential equation is:

\[ \frac{dC_i(t)}{dt}=k_{\mathrm{in}}C_p(t)-k_{\mathrm{out}}C_i(t) \]

Here, \(k_{\mathrm{in}}\) represents the apparent rate of intracellular drug entry and \(k_{\mathrm{out}}\) represents the apparent rate of intracellular loss.

Parameter Interpretation
\(C_p(t)\) Plasma or systemic drug concentration
\(C_i(t)\) Intracellular drug concentration
\(k_{\mathrm{in}}\) Apparent rate constant for intracellular entry
\(k_{\mathrm{out}}\) Apparent rate constant for intracellular loss

This equation is deliberately simple. It does not claim that cellular transport is always first-order. Instead, it provides a useful starting point for understanding how an intracellular concentration can lag behind, persist longer than, or otherwise differ from the plasma concentration.

04 · Equilibrium

4. What Happens at Steady State?

Suppose plasma concentration is held constant at \(C_p\). At steady state, intracellular concentration no longer changes, so:

\[ \frac{dC_i}{dt}=0 \]

Applying this condition to the simple model gives:

\[ 0=k_{\mathrm{in}}C_p-k_{\mathrm{out}}C_i \]

and therefore:

\[ C_i=\frac{k_{\mathrm{in}}}{k_{\mathrm{out}}}C_p \]

The ratio \[ K_i=\frac{C_i}{C_p} \] can therefore be interpreted, within this simplified model and under the stated conditions, as an intracellular-to-plasma concentration ratio.

Important: the intracellular-to-plasma ratio is not necessarily a universal constant. It can depend on transport, binding, ionization, metabolism, experimental conditions, and whether the system has actually reached equilibrium.
05 · Time course

5. Why Intracellular Concentration Can Lag Behind Plasma

Even when plasma concentration changes rapidly, intracellular concentration may respond more slowly. The reason is that drug must cross one or more barriers before reaching the intracellular space.

For a constant plasma concentration \(C_p\), the simple model has the solution:

\[ C_i(t)=C_{i,0}e^{-k_{\mathrm{out}}t} +\frac{k_{\mathrm{in}}}{k_{\mathrm{out}}}C_p \left(1-e^{-k_{\mathrm{out}}t}\right) \]

If the intracellular concentration initially equals zero, this becomes:

\[ C_i(t)=\frac{k_{\mathrm{in}}}{k_{\mathrm{out}}}C_p \left(1-e^{-k_{\mathrm{out}}t}\right) \]

The characteristic time scale is related to \(1/k_{\mathrm{out}}\). A smaller \(k_{\mathrm{out}}\) produces slower approach to the equilibrium concentration in this simple model.

Plasma Intracellular Time Concentration

A simplified illustration of intracellular equilibration. The intracellular profile may be delayed relative to the plasma profile when cellular uptake is not instantaneous.

06 · Cellular transport

6. Passive Diffusion and Active Transport

Drug movement across the cell membrane can involve several mechanisms. A useful model should reflect the mechanisms that are important for the drug and the scientific question.

Passive diffusion

For a sufficiently simple system, passive diffusion can be represented as movement proportional to the concentration difference across the membrane:

\[ \text{Rate}_{\mathrm{diffusion}} \propto C_{\mathrm{out}}-C_{\mathrm{in}} \]

This produces a tendency toward concentration equilibration. The actual relationship may depend on membrane permeability, surface area, lipophilicity, ionization, and other physicochemical properties.

Active uptake

Transporters can produce cellular uptake that is not adequately described by a simple linear diffusion process. A saturable uptake process can, for example, be represented using a Michaelis-Menten-type expression:

\[ \mathrm{Rate}_{\mathrm{in}} = \frac{V_{\max,\mathrm{in}}C_{\mathrm{out}}} {K_{m,\mathrm{in}}+C_{\mathrm{out}}} \]

At low concentrations, the process can behave approximately linearly. As concentration increases, transporter capacity becomes limiting and the uptake rate approaches \(V_{\max,\mathrm{in}}\).

Active efflux

Efflux transporters can move drug out of cells. A corresponding saturable efflux term might be written as:

\[ \mathrm{Rate}_{\mathrm{out}} = \frac{V_{\max,\mathrm{out}}C_i} {K_{m,\mathrm{out}}+C_i} \]

Including both uptake and efflux mechanisms can help explain intracellular concentrations that are substantially different from extracellular concentrations.

07 · Free and total concentration

7. Intracellular Binding: Free Versus Total Drug

The total intracellular drug concentration is not necessarily equivalent to the concentration that interacts with a molecular target. Drug may bind to proteins, membranes, organelles, or other intracellular components.

A simple binding model can be written as:

\[ D+P \rightleftharpoons DP \]

where \(D\) represents drug, \(P\) represents an intracellular binding site, and \(DP\) represents the drug-binding-site complex.

For a simple reversible binding process, the dissociation constant is:

\[ K_d=\frac{[D][P]}{[DP]} \]

The distinction between free and total intracellular concentration becomes particularly important when the pharmacologic effect depends on the free drug concentration near the target rather than the total amount of drug measured in the cell.

Modeling question: before interpreting an intracellular concentration, ask what was actually measured. Total cellular drug, unbound intracellular drug, and target-site concentration are not necessarily the same quantity.
08 · Physicochemical effects

8. Ionization and Intracellular Accumulation

A drug's physicochemical properties can strongly influence intracellular distribution. Weak acids and weak bases can behave differently across cellular membranes because different compartments may have different pH environments.

For an ionizable drug, the neutral form may cross a lipid membrane more readily than the ionized form. Once inside a compartment with a different pH, the relative fractions of ionized and unionized drug can change.

This can produce intracellular accumulation even when the drug does not have a conventional active uptake transporter.

Factor Potential effect on intracellular exposure
Membrane permeability Controls how readily drug crosses the cell membrane
Drug ionization Influences membrane partitioning and distribution across pH gradients
Intracellular pH Can alter the fraction of ionized drug
Binding Can reduce free intracellular drug relative to total drug
Transporters Can enhance uptake or promote efflux
Intracellular metabolism Can remove parent drug or generate intracellular metabolites
09 · Mechanistic structure

9. A Mechanistic Intracellular PK Model

A more explicit model can represent drug movement through extracellular and intracellular spaces separately.

Suppose \(A_e\) is the amount of drug in an extracellular compartment and \(A_i\) is the intracellular amount. A simple linear model might be:

\[ \frac{dA_e}{dt} = \mathrm{Input} -k_{ei}A_e +k_{ie}A_i -k_eA_e \]
\[ \frac{dA_i}{dt} = k_{ei}A_e -k_{ie}A_i -k_{mi}A_i \]

Here:

  • \(k_{ei}\) describes movement from extracellular to intracellular space.
  • \(k_{ie}\) describes movement from intracellular to extracellular space.
  • \(k_e\) represents elimination from the extracellular/systemic compartment.
  • \(k_{mi}\) represents intracellular loss such as metabolism or another first-order process.

If the extracellular and intracellular compartment volumes are \(V_e\) and \(V_i\), concentrations can be obtained from:

\[ C_e=\frac{A_e}{V_e}, \qquad C_i=\frac{A_i}{V_i} \]

This formulation highlights an important distinction: a concentration model and an amount model are mathematically related, but mass balance is often most naturally expressed using amounts.

10 · Mass balance

10. Why Mass Balance Matters

Intracellular PK models should obey conservation of mass unless drug is explicitly added to or removed from the system through an input or elimination process.

For example, if drug moves reversibly between extracellular and intracellular compartments:

\[ \text{Extracellular} \overset{\text{uptake}}{\longrightarrow} \text{Intracellular} \]

the corresponding intracellular gain should be represented as an extracellular loss, and vice versa.

This is one reason differential-equation models are useful: each process can be represented as a flux connecting compartments.

Practical principle: every intracellular gain should have a corresponding source, and every intracellular loss should have a corresponding destination or elimination mechanism.
11 · Target-site exposure

11. From Intracellular Concentration to Target-Site Exposure

An intracellular compartment can sometimes serve as a useful approximation to the location of pharmacologic action. However, the intracellular concentration may still not be identical to the concentration immediately surrounding the molecular target.

For example, a drug may enter the cytosol but subsequently partition into an organelle or bind to an intracellular protein before interacting with its target.

A hierarchical model might therefore be written as:

\[ C_p(t) \rightarrow C_e(t) \rightarrow C_i(t) \rightarrow C_{\mathrm{target}}(t) \rightarrow E(t) \]

Each additional layer adds biological interpretation, but also introduces additional parameters and data requirements.

Key modeling trade-off: mechanistic detail is useful only when the available data can support the additional processes and parameters.
12 · Parameter estimation

12. How Are Intracellular Model Parameters Estimated?

Parameter estimation follows the same general logic as other pharmacometric models: observed concentration data are compared with model predictions, and parameters are estimated using an appropriate statistical or likelihood-based approach.

  1. Define the biological question. Decide whether the objective is to characterize uptake, intracellular persistence, target exposure, or another process.
  2. Define the model structure. Determine which compartments and transport processes are required.
  3. Identify measured quantities. Distinguish plasma, extracellular, total intracellular, free intracellular, and target-site measurements.
  4. Specify the observation model. Measurement error can differ substantially between matrices and assay types.
  5. Estimate parameters. Fit the model to the available data using an appropriate estimation framework.
  6. Evaluate identifiability. Determine whether the data contain enough information to estimate the parameters separately.
  7. Evaluate diagnostics. Examine residuals, prediction performance, parameter plausibility, and other model diagnostics.

A major challenge is that intracellular concentration data are often much sparser than plasma concentration data. As a result, several mechanistic parameters may produce similar predictions over the observed time range.

13 · Identifiability

13. Why Identifiability Can Be Difficult

Suppose an intracellular model contains separate uptake and efflux parameters: \(k_{\mathrm{in}}\) and \(k_{\mathrm{out}}\). If only a single steady-state intracellular-to-plasma ratio is observed, the data may strongly inform their ratio without separately identifying both parameters.

At steady state:

\[ \frac{C_i}{C_p} = \frac{k_{\mathrm{in}}}{k_{\mathrm{out}}} \]

The observation therefore constrains a combination of parameters rather than necessarily providing independent information about each one.

Time-course measurements can provide additional information because the approach toward equilibrium contains information about the underlying rates.

Data available Potential information
Single intracellular concentration Limited information about intracellular exposure
Intracellular-to-plasma ratio Information about relative distribution under the observed conditions
Multiple intracellular time points Information about the dynamics of intracellular entry and loss
Plasma + extracellular + intracellular data Greater ability to separate sequential distribution processes
Free and total intracellular measurements Additional information about intracellular binding
14 · Worked example

14. Worked Example: Estimating Intracellular Exposure

Consider a hypothetical drug for which plasma concentration is approximately constant at 10 mg/L over a period of observation. Suppose a simplified intracellular model has:

  • \(k_{\mathrm{in}}=0.30\ \mathrm{h}^{-1}\)
  • \(k_{\mathrm{out}}=0.10\ \mathrm{h}^{-1}\)
  • \(C_i(0)=0\)

Step 1: Calculate the steady-state intracellular concentration

\[ C_{i,ss} = \frac{k_{\mathrm{in}}}{k_{\mathrm{out}}}C_p = \frac{0.30}{0.10}(10) = 30\text{ mg/L} \]

Under this simplified model, the predicted steady-state intracellular concentration is therefore 30 mg/L.

Step 2: Calculate the concentration after 5 hours

Because \(C_i(0)=0\), use:

\[ C_i(t) = \frac{k_{\mathrm{in}}}{k_{\mathrm{out}}}C_p \left(1-e^{-k_{\mathrm{out}}t}\right) \]

At \(t=5\) hours:

\[ C_i(5) = 30\left(1-e^{-0.10(5)}\right) \approx 30(1-0.6065) \approx 11.80\text{ mg/L} \]

Step 3: Compare intracellular and plasma concentrations

\[ \frac{C_i(5)}{C_p} = \frac{11.80}{10} \approx1.18 \]

The model therefore predicts that intracellular concentration has exceeded plasma concentration after five hours, but it has not yet reached its predicted steady-state value of 30 mg/L.

Interpretation: this example illustrates how a mechanistic intracellular model can distinguish the magnitude of intracellular exposure from the rate at which intracellular exposure develops.
15 · Nonlinear processes

15. When Intracellular PK Becomes Nonlinear

Intracellular drug concentrations may become nonlinear when transport, binding, metabolism, or other processes become saturated.

For example, consider saturable intracellular uptake:

\[ \mathrm{Rate}_{\mathrm{in}} = \frac{V_{\max}C_e}{K_m+C_e} \]

At low extracellular concentration:

\[ C_e\ll K_m \quad\Rightarrow\quad \mathrm{Rate}_{\mathrm{in}} \approx \frac{V_{\max}}{K_m}C_e \]

Thus the process behaves approximately linearly at sufficiently low concentrations. At high concentration:

\[ C_e\gg K_m \quad\Rightarrow\quad \mathrm{Rate}_{\mathrm{in}} \approx V_{\max} \]

The uptake pathway is then approaching its maximum capacity. Consequently, intracellular exposure may increase less than proportionally as extracellular exposure increases.

16 · Population models

16. Intracellular Models in Population Pharmacokinetics

Intracellular models can be embedded within population PK or pharmacometric models. Instead of estimating one parameter value for every individual, a population model can describe typical intracellular kinetics together with between-subject variability.

For example, an uptake parameter might be modeled as:

\[ k_{\mathrm{in},j} = k_{\mathrm{in,pop}}e^{\eta_{\mathrm{in},j}} \]

where \(k_{\mathrm{in,pop}}\) is the population-typical value and \(\eta_{\mathrm{in},j}\) represents an individual-specific deviation.

Covariates can also be incorporated when there is a scientific rationale for expecting them to influence intracellular distribution.

Population component Potential role
Typical uptake Describes the population-average cellular entry process
Between-subject variability Represents differences among individuals
Covariates Can explain systematic differences in intracellular kinetics
Residual variability Represents unexplained differences between observations and predictions
17 · PK → PD

17. Linking Intracellular Concentration to Pharmacodynamics

One of the strongest reasons to model intracellular concentration is that the pharmacologic effect may be more directly related to intracellular exposure than to plasma exposure.

A mechanistic sequence can therefore be represented as:

\[ \text{Dose} \rightarrow C_p(t) \rightarrow C_i(t) \rightarrow E(t) \]

For example, an \(E_{\max}\)-type relationship could use intracellular concentration:

\[ E(t) = E_0+ \frac{E_{\max}C_i(t)} {EC_{50}+C_i(t)} \]

In this framework, the intracellular PK model determines the time-varying exposure available to the pharmacodynamic model.

This can be particularly useful when plasma concentration is a poor surrogate for the concentration that actually drives the pharmacologic response.

18 · Applications

18. Where Are Intracellular Drug Models Useful?

Intracellular concentration models can be useful whenever the site of pharmacologic action is separated from the conventional PK sampling compartment.

  • Intracellular anti-infective drugs: when pathogens reside within cells and intracellular exposure is relevant to efficacy.
  • Oncology: when drug action occurs inside tumor cells and cellular exposure may differ from plasma exposure.
  • Immunology: when intracellular signaling pathways are affected by drug exposure.
  • Transporter-mediated disposition: when uptake and efflux determine cellular exposure.
  • Target-mediated pharmacology: when intracellular target engagement depends on local drug concentration.
  • Organelle-targeted drugs: when distribution beyond the cytosol is important.
  • Pharmacodynamic modeling: when an effect compartment based only on plasma concentration does not adequately represent the biological delay.
19 · Interpretation

19. What Intracellular Models Do Not Tell Us Automatically

An intracellular model can provide a useful mechanistic description, but its parameters should not automatically be interpreted as direct measurements of individual biological processes.

  • A compartment is a model construct. An intracellular compartment may represent an aggregate cellular space rather than a specific physical location.
  • Apparent transport parameters may combine mechanisms. A fitted uptake rate may reflect several biological processes simultaneously.
  • Total intracellular concentration may not equal free drug. Binding and sequestration can produce important differences.
  • Observed concentration does not automatically establish target exposure. Additional intracellular or subcellular processes may intervene.
  • Parameter identifiability can be limited. Sparse intracellular sampling may support only combinations of parameters.
  • Model predictions depend on assumptions. Predictions outside the conditions represented by the data can be sensitive to model structure.
Modeling principle: intracellular PK models are most informative when the model structure, measured concentrations, and biological interpretation are kept explicitly connected.
20 · Practical workflow

20. A Practical Workflow for Intracellular PK Modeling

  1. Define the scientific question. Decide whether the goal is to describe intracellular exposure, explain a PK/PD delay, characterize transport, or predict target-site concentrations.
  2. Identify the relevant matrices. Determine whether plasma, extracellular, intracellular, free intracellular, or target-site concentrations are available.
  3. Explore the data. Compare plasma and intracellular concentration-time profiles and look for delays, accumulation, or nonlinear behavior.
  4. Start with the simplest plausible model. Add extracellular, intracellular, binding, or transport compartments only when scientifically justified.
  5. Represent important mechanisms. Consider passive diffusion, active uptake, efflux, binding, metabolism, and other processes when supported by the evidence.
  6. Evaluate identifiability. Determine which parameters can actually be estimated from the available sampling design.
  7. Estimate parameters and assess diagnostics. Examine goodness of fit, residuals, prediction performance, and biological plausibility.
  8. Link intracellular exposure to PD when appropriate. Use the modeled intracellular concentration as the driver when there is a mechanistic rationale.
  9. Simulate relevant scenarios. Explore how changes in dose, transport, clearance, or other parameters affect intracellular exposure.
  10. Clearly separate measurement from inference. Identify which concentrations were directly observed and which were model-predicted.
21 · Study design

21. Designing Studies to Inform Intracellular Models

The quality of an intracellular model depends strongly on the information contained in the study design. If intracellular samples are collected at only one time point, it may be difficult to characterize intracellular kinetics.

When feasible, sampling across the relevant time course can help distinguish rapid equilibration from delayed intracellular accumulation.

Design feature Why it can matter
Multiple intracellular time points Provides information about the rate of intracellular accumulation and loss
Matched plasma measurements Allows intracellular exposure to be interpreted relative to systemic exposure
Pre-dose sampling Helps characterize baseline intracellular concentration
Post-dose follow-up Provides information about intracellular persistence and washout
Different dose levels Can help reveal nonlinear transport or binding
Free and total measurements Can inform intracellular binding and target-relevant exposure

The optimal sampling design depends on the expected kinetics and the parameters that need to be estimated. Sampling should therefore be planned around the scientific question rather than added only after the data have already been collected.

22. Key Takeaways

  • Intracellular drug concentration is not necessarily the same as plasma concentration.
  • An intracellular PK model can describe the movement of drug from systemic circulation into cells and its subsequent loss or retention.
  • A simple model can represent intracellular uptake and loss using \(k_{\mathrm{in}}\) and \(k_{\mathrm{out}}\).
  • At steady state in a simple linear model, the intracellular-to-plasma concentration ratio is determined by the relative uptake and loss rates.
  • Intracellular concentration can lag behind plasma concentration because cellular distribution takes time.
  • Transporters can produce uptake or efflux that may be linear or saturable depending on concentration and mechanism.
  • Total intracellular drug is not necessarily equivalent to free intracellular drug or concentration at the molecular target.
  • Ionization, intracellular pH, binding, transport, and intracellular metabolism can all influence cellular exposure.
  • More mechanistic intracellular models require more data and can create additional identifiability challenges.
  • Intracellular PK models can provide an important bridge between systemic pharmacokinetics and pharmacodynamic response.
  • The most useful model is not necessarily the most detailed model; it is the model that is adequate for the scientific question and the available data.
Next step

Where to Go Next

A natural progression is to study nonlinear intracellular transport, followed by transporter-mediated uptake and efflux, intracellular binding, target-mediated drug disposition, mechanistic tissue distribution, and PK/PD models driven by intracellular concentrations.

These models provide a foundation for more advanced pharmacometric approaches in which plasma exposure is connected to tissue, cellular, and molecular target exposure rather than treating plasma concentration as the final pharmacologic driver.

Related Tutorials

Clinical Biostats

← Back to Pharmacokinetics Tutorials