Tutorials › Physiologically-Based Pharmacokinetics (PBPK) › Muscle and Skin Kp from Tissue Composition

Home  /  Tutorials  /  Tissue-to-Plasma Partitioning: Prediction Methods

Tissue-to-Plasma Partitioning: Prediction Methods · Tutorial 130 of 799

Muscle and Skin Kp from Tissue Composition

Use phase-weighted calculations to compare muscle and skin Kp and identify how tissue composition and pH affect predictions for ionizable drugs.

Intermediate 12 min read

What You'll Learn

  • Define phase-specific contributions to a tissue-to-plasma partition coefficient.
  • Calculate and compare muscle and skin Kp from water, neutral-lipid, and phospholipid fractions.
  • Explain how ionization and tissue pH can alter water- and phospholipid-associated contributions.
  • Separate the effect of tissue composition from the effect of compound-specific phase partitioning.
  • Check composition inputs, assumptions, and common sources of misleading comparisons.

Why Muscle and Skin Can Give Different Kp Predictions

Muscle and skin are both perfused tissues, but their tissue-to-plasma partition coefficients need not be alike. The difference can reflect how much of each tissue is represented by water, neutral lipid, and phospholipid, as well as how the drug distributes into those phases. For an ionizable drug, tissue pH can also change the concentration associated with the water and phospholipid terms. A comparison is therefore not just a question of which tissue contains more lipid.

This tutorial uses a phase-weighted calculation to make those contributions explicit. The phase fractions and partition terms below are hypothetical, chosen to demonstrate the arithmetic rather than to represent reference values for human muscle or skin. In an application, use composition values and phase-partition terms that are compatible with the selected prediction method and species.

Definition: The tissue-to-plasma partition coefficient, \(K_{p,t}\), is the ratio of the total drug concentration in tissue \(t\) to the total drug concentration in plasma at distribution equilibrium. In a three-phase composition calculation, \(f_{w,t}\), \(f_{nl,t}\), and \(f_{ph,t}\) are the fractions of the modeled tissue assigned to water, neutral lipid, and phospholipid. The phase terms \(q_{w,t}\), \(q_{nl,t}\), and \(q_{ph,t}\) are the corresponding drug concentrations in those phases, each expressed relative to plasma total concentration. Thus \(K_{p,t}\) is dimensionless.

The Phase-Weighted Calculation

The calculation begins with a mass balance. If the total tissue concentration is the sum of concentrations contributed by the modeled phases, then each phase contributes its fraction of the tissue multiplied by its concentration relative to plasma. Dividing that tissue total by plasma total concentration gives:

$$ K_{p,t} = f_{w,t}q_{w,t} + f_{nl,t}q_{nl,t} + f_{ph,t}q_{ph,t}. $$

For this three-phase illustration, the fractions are normalized so that \(f_{w,t}+f_{nl,t}+f_{ph,t}=1\). This is a defined simplification, not a claim that all tissue mass consists only of these three components. Some implementations include additional modeled phases or use a different definition of their inputs; those terms must be retained when required by the chosen method. Do not silently discard a residual fraction or mix volume fractions with mass fractions.

The phase terms are drug-specific. Neutral compounds generally have a water term and lipid-associated terms determined by their phase partitioning. For ionizable drugs, the relevant terms must also reflect the compound’s ionization and the pH conditions used by the prediction method. Consequently, it is useful to write \(q_{j,t}\), rather than assume every phase term is identical in every tissue. If muscle and skin pH are treated as equal in a particular calculation, the terms may be shared; if tissue pH differs, the appropriate terms may differ.

Assumptions for the worked calculations: The modeled fractions use a common basis and sum to one; each phase is adequately represented by a single average concentration; the phase terms are equilibrium ratios to plasma total concentration; and the required pH and compound inputs are specified consistently. These calculations predict equilibrium partitioning, not the rate at which equilibrium is reached.

The equation follows directly from the tissue concentration balance. Let \(C_p\) be plasma total concentration and \(C_{j,t}\) the concentration in phase \(j\) of tissue \(t\). The phase contribution to the tissue average is \(f_{j,t}C_{j,t}\). Summing the three contributions and dividing by \(C_p\) gives:

$$ C_t=\sum_j f_{j,t}C_{j,t} \quad\Longrightarrow\quad \frac{C_t}{C_p} = \sum_j f_{j,t}\frac{C_{j,t}}{C_p} = \sum_j f_{j,t}q_{j,t}. $$

This derivation explains why a large phase fraction does not by itself guarantee a large contribution. Its effect also depends on \(q_{j,t}\). Conversely, a small phospholipid fraction can matter when the drug has a large phospholipid-associated term.

Worked Example: Neutral Drug in Muscle and Skin

Suppose a hypothetical neutral drug has \(q_w=1.0\), \(q_{nl}=20\), and \(q_{ph}=4.0\) in both tissues. Use the following modeled fractions: muscle \(f_w=0.80\), \(f_{nl}=0.05\), and \(f_{ph}=0.15\); skin \(f_w=0.65\), \(f_{nl}=0.20\), and \(f_{ph}=0.15\). Each set sums to 1.

For muscle:

$$ K_{p,\mathrm{muscle}} =(0.80)(1.0)+(0.05)(20)+(0.15)(4.0) =0.80+1.00+0.60=2.40. $$

For skin:

$$ K_{p,\mathrm{skin}} =(0.65)(1.0)+(0.20)(20)+(0.15)(4.0) =0.65+4.00+0.60=5.25. $$

Under these hypothetical inputs, predicted skin \(K_p\) is 5.25, compared with 2.40 for muscle. The difference is mainly the neutral-lipid contribution: it is 4.00 in skin and 1.00 in muscle. This is a calculation under the stated model and assumptions, not evidence that skin must have a higher measured \(K_p\) for every neutral drug.

Ionizable Drugs: Composition and pH Work Together

For an ionizable drug, composition still weights the phase contributions, but the phase terms may reflect ionization as well as neutral-species partitioning. The earlier material on Rodgers–Rowland predictions and tissue pH gradients establishes why ionized and neutral forms need not distribute alike and why tissue pH matters. Here the practical point is to carry those effects into the muscle-versus-skin comparison rather than treating an ionizable drug as if it had one universal lipid term.

If the tissues have different pH, their water-associated terms can differ even when the same compound is being considered. Phospholipid-associated terms may also depend on the model’s treatment of ionization and binding. A change in \(K_p\) between tissues can therefore arise from different fractions, different phase terms, or both. The calculation should make clear which source of difference is being compared.

Worked Example: A Basic Drug With Tissue-Specific Terms

Consider a hypothetical basic drug. Use the same muscle and skin fractions as above. Suppose a method, with its chosen pH assumptions already applied, supplies these terms: for muscle \(q_w=1.5\), \(q_{nl}=2.0\), and \(q_{ph}=12\); for skin \(q_w=1.2\), \(q_{nl}=2.0\), and \(q_{ph}=12\). The different water terms represent a hypothetical tissue-pH effect; the values are illustrative, not empirical.

For muscle:

$$ K_{p,\mathrm{muscle}} =(0.80)(1.5)+(0.05)(2.0)+(0.15)(12) =1.20+0.10+1.80=3.10. $$

For skin:

$$ K_{p,\mathrm{skin}} =(0.65)(1.2)+(0.20)(2.0)+(0.15)(12) =0.78+0.40+1.80=2.98. $$

Here muscle has the slightly higher predicted value, 3.10 versus 2.98, even though skin has the larger neutral-lipid fraction. The water contribution is 1.20 in muscle and 0.78 in skin, while the larger skin neutral-lipid fraction contributes 0.40 rather than 0.10. The resulting difference is small: the opposing contributions nearly offset. The direction cannot be inferred from the tissue names or from lipid content alone.

Worked Example: Separating Composition Effects From Phase-Term Effects

Use the basic-drug example’s muscle terms for both tissues to isolate composition alone: \(q_w=1.5\), \(q_{nl}=2.0\), and \(q_{ph}=12\). Muscle remains \(K_p=3.10\). For skin, holding the terms fixed gives:

$$ K_{p,\mathrm{skin,\ shared\ terms}} =(0.65)(1.5)+(0.20)(2.0)+(0.15)(12) =0.975+0.400+1.800=3.175. $$

With shared phase terms, the skin prediction is 3.175, or 0.075 higher than muscle. In the previous example, allowing the skin water term to differ changed its prediction to 2.98. Thus the composition difference alone favored skin under the shared-term calculation, while the hypothetical tissue-specific water term reversed the ranking. This decomposition is useful for interpretation, but it does not prove that pH alone caused the difference in a real system; the phase terms and fractions are model inputs, not independently measured causal effects.

Checks, Alternatives, and Limitations

Before interpreting a muscle–skin comparison, check the inputs and arithmetic. Fractions should be on the same basis, use compatible definitions, and sum to the intended total. The phase terms should correspond to the drug’s ionization properties and the pH assumptions of the calculation. Recomputing each phase contribution separately makes it easier to find a misplaced decimal point or an incorrectly reused term.

  • Check the fraction basis. A mass fraction and a volume fraction are not interchangeable without conversion. If a model requires additional phases, include them rather than renormalizing them away without justification.
  • Check pH assignment. Do not use a single tissue term for both tissues if the selected method specifies tissue-specific pH effects. Conversely, do not invent a pH difference that is not part of the stated inputs.
  • Check which term drives the result. Calculate \(f_jq_j\) for each phase. A large total can be dominated by one term, and a small change in fractions may have little effect if the corresponding \(q_j\) is small.
  • Check equilibrium and interpretation. \(K_p\) is an equilibrium concentration ratio. It does not specify tissue uptake speed, blood flow, or the concentration at an arbitrary sampling time.

A common mistake is to reason, “Skin has more neutral lipid, so its \(K_p\) must be larger.” That conclusion ignores the drug-specific phase terms, the water and phospholipid contributions, and any tissue-specific pH effects. The correction is to calculate each weighted contribution and compare the totals under explicit assumptions.

A composition-based prediction is one mechanistic approach, not a measurement. Different Kp prediction methods may use different phase representations and assumptions, so their numerical predictions need not agree. Measured tissue-to-plasma ratios can provide a comparison when suitable equilibrium data exist, but they also depend on study design and sampling. The calculation here is limited by its three-phase representation, the quality of composition inputs, and the adequacy of the phase terms. It should not be treated as a complete account of all tissue binding or distribution processes.

Key takeaway: Muscle and skin \(K_p\) predictions are phase-weighted sums. Tissue fractions determine the weights; drug-specific partitioning and, for ionizable drugs, tissue pH determine the phase terms. Compare both before explaining a difference.

Check Your Understanding

Use the phase-weighted definition and show the contribution of each phase where a calculation is requested.

  1. For a tissue with \(f_w=0.70\), \(f_{nl}=0.10\), and \(f_{ph}=0.20\), and terms \(q_w=1\), \(q_{nl}=5\), and \(q_{ph}=3\), calculate \(K_p\).
  2. If muscle and skin have identical fractions but different water terms for an ionizable drug, can their predicted \(K_p\) values differ? Explain.
  3. Why is “the tissue with more lipid must have the larger \(K_p\)” not a generally valid rule?
  4. In the basic-drug example, what is the skin prediction when muscle’s phase terms are used for both tissues, and what does that comparison isolate?

Solutions. (1) \(K_p=(0.70)(1)+(0.10)(5)+(0.20)(3)=0.70+0.50+0.60=1.80\). (2) Yes. The water contribution is \(f_wq_w\); when \(f_w\) is positive, a difference in \(q_w\) changes the weighted sum unless offset by another contribution. (3) Lipid fraction is only a weight. The lipid-associated term, water and phospholipid contributions, and any tissue-specific terms also affect the total. (4) The value is 3.175. Because the phase terms are held the same, this comparison isolates the effect of the different fractions within the illustrative model.

Previous

Next

Back to Tutorials