Schmitt Method for Tissue-to-Plasma Partitioning

Why the Schmitt Method Uses Composition and Drug Properties

A tissue-to-plasma partition coefficient, \(K_p\), describes the equilibrium ratio of drug concentration in a tissue to that in plasma, on a specified concentration basis. For a total-concentration \(K_p\), both numerator and denominator include drug that is bound as well as unbound. A mechanistic prediction therefore needs more than a drug’s lipophilicity: it must represent how much drug can reside in the different phases of tissue and plasma.

The Schmitt method combines physicochemical properties with the amounts of relevant phases in tissue and plasma. Depending on the implementation, those phases may include water, neutral lipids, phospholipids, and protein-associated binding. Drug properties determine the phase-specific capacities; tissue composition determines how much each capacity contributes. The result is a composition-based prediction, not a claim that every molecular interaction in a tissue has been measured.

This equation is a mass-balance statement: the amount accommodated by each phase is added to obtain the total distribution capacity. The phase capacities are where drug properties enter. In an application, they are estimated from properties such as lipophilicity and pKa, along with the model’s representation of phase interactions and binding. The tissue and plasma composition fractions must use compatible definitions and concentration bases.

Assumptions and Calculation

The composition-based calculation assumes that the relevant tissue and plasma phases are adequately represented by the model, and that their composition inputs apply to the species and physiological condition being modeled. It also assumes that the drug has reached an equilibrium distribution for which the predicted phase capacities are meaningful. A calculated equilibrium \(K_p\) is not automatically the concentration ratio at an arbitrary sampling time.

For an uncharged compound, the relative contributions of water, lipid, phospholipid, and protein-associated phases are determined by their phase capacities. For an ionizable compound, the capacity calculation must account for the fractions of the drug in each ionic state at the relevant pH. A weak base and a weak acid can therefore have different phase capacities in tissue and plasma when those compartments have different pH values. The relevant ionization fractions follow from pKa and pH; they are not interchangeable between compartments.

Binding affects total \(K_p\) because bound drug contributes to the total concentration. In a mechanistic phase calculation, binding is represented through the capacity assigned to the relevant phase or binding pool. If an implementation instead uses an unbound partition coefficient and explicit unbound fractions, those quantities must be combined consistently. Applying an additional binding correction to a capacity that already includes binding would count the same effect twice.

Ionization and pH-Dependent Contributions

For an ionizable drug, a phase capacity can depend on pH because pH determines the relative amounts of neutral and ionized drug. For example, for a monoprotic weak base \(BH^+ \rightleftharpoons B+H^+\), the ionized-to-neutral ratio in an aqueous compartment is \(10^{pKa-pH}\). Relative to the neutral aqueous concentration, total aqueous concentration is therefore \(1+10^{pKa-pH}\) times as large. This relationship is a species-balance result; it does not imply that ionized drug partitions into every nonaqueous phase as readily as neutral drug.

In a Schmitt calculation, the ionization state is used when assigning the effective capacities for the relevant phases. An aqueous contribution can reflect both species, while lipid and phospholipid contributions depend on how the method represents their interactions with neutral and charged forms. Binding contributions also depend on the model representation. It is important not to replace all phase capacities with a single octanol-water partition coefficient: doing so discards the distinctions the composition-based calculation is intended to capture.

For a weak acid, the Henderson–Hasselbalch relationship has the opposite pH direction: the ionized-to-neutral ratio is \(10^{pH-pKa}\). For compounds with more than one ionizable group, the species fractions require the appropriate multistate ionization calculation. The Zwiterion case likewise requires attention to the distinct charged species; treating the molecule as if it were simply a neutral compound or a single monoprotic ion can misrepresent its capacities.

Binding and the Total-versus-Unbound Distinction

A total \(K_p\) can be high because tissue has substantial binding capacity, even when the unbound concentration in tissue is not especially high. The standard relationship between total and unbound partitioning makes this distinction explicit. Let \(f_{u,p}\) be the unbound fraction in plasma and \(f_{u,t}\) the unbound fraction in tissue. If \(K_{p,u}\) is the ratio of unbound tissue to unbound plasma concentration, then

This identity follows by writing unbound concentration as unbound fraction times total concentration in each compartment and taking the tissue-to-plasma ratio. It is useful as a consistency check, but it is not permission to apply a second binding adjustment when binding is already represented in the predicted phase capacities.

Derivation, Diagnostics, and Alternatives

The capacity-ratio equation follows directly from the model’s mass balance. Let the common reference concentration be \(c_0\). In compartment \(x\), the amount of drug associated with phase \(j\), per unit compartment volume, is represented as \(v_{j,x}a_{j,x}c_0\). Summing across phases gives the predicted total concentration on the model’s common basis. The ratio of tissue to plasma predictions cancels \(c_0\):

A practical diagnostic is to inspect the contributions, not just the final ratio. Check whether tissue and plasma fractions are valid for the chosen implementation, whether fractions use the same basis, and whether their sum is consistent with that basis. Confirm that logP or distribution data, pKa, and pH refer to compatible conditions. For ionizable drugs, verify the pH-specific species fractions. Finally, identify whether binding is already included in the phase capacities before applying any unbound-fraction adjustment.

Large or implausible predictions can arise from a dominant lipid or binding term, a unit or fraction error, inconsistent plasma and tissue definitions, or an inappropriate equilibrium assumption. A tissue sample may also include blood; comparing that measured total concentration with a model tissue value requires attention to the reference phases and blood contribution. If transporters, active uptake, metabolism, or specific high-affinity binding materially affect distribution, a simple composition-based equilibrium estimate may not capture them.

The Schmitt method belongs to a family of composition-based approaches. The Poulin–Theil method provides a composition-based route particularly associated with neutral-compound partitioning, while Rodgers–Rowland methods make explicit ionization and ionic interactions with tissue components in their own parameterization. These approaches differ in their phase definitions, assumptions, and treatment of ionized drug; their outputs are not interchangeable simply because each is reported as \(K_p\). The Schmitt method’s advantage is a structured link between physicochemical inputs and tissue composition. Its limitation is that the prediction remains conditional on those inputs and on the adequacy of its phase representation. It does not, by itself, establish that the drug has reached equilibrium or resolve every tissue-specific binding or transport mechanism.

Common Pitfall

• Wrong reasoning: “The tissue capacity is much larger than the plasma capacity, so the unbound tissue concentration must be proportionally higher.”
• Correction: The capacity ratio predicts total-concentration \(K_p\) under the model assumptions. Tissue binding can raise total tissue concentration without raising the unbound concentration by the same factor. Use \(K_{p,u}=K_p f_{u,t}/f_{u,p}\) when the unbound ratio is needed, and avoid double-counting binding already represented in the model.
• Another warning: Do not use tissue pH for the plasma species fractions, or plasma pH for tissue fractions. For ionizable drugs, the compartments can have different ionization balances.