Effective Permeability and Intestinal Absorption

From Dissolved Drug to Enterocyte Uptake

Dissolution provides drug in the intestinal lumen, but dissolved drug is not necessarily absorbed at the same rate that it becomes available. The rate of transfer across the luminal boundary depends on both the concentration driving force and how readily the intestinal interface permits transport. In a PBPK model, effective intestinal permeability represents that transfer in a form that can be combined with the segment’s luminal concentration and available surface area.

The term “effective” matters. It is a model-scale parameter for uptake from the lumen under a specified representation of the gut, not a universal molecular constant. It should not be confused with the apparent permeability reported by a cell or membrane assay. The latter describes transport in that assay’s particular setup; it is not automatically the parameter required by an intestinal PBPK compartment.

For the uptake boundary, let \(C_{\mathrm{L,free}}\) be the free dissolved drug concentration in the lumen. Let \(C_{\mathrm{E}}^*\) be the enterocyte-side concentration expressed on the same concentration basis as the luminal driving force. This equivalent concentration allows the two sides of the boundary to be compared without assuming that their raw compartment concentrations have identical reference scales. With \(A_{\mathrm{int}}\) denoting the effective intestinal surface area, the net flux density and total transfer rate are

Here \(J_{\mathrm{L}\rightarrow\mathrm{E}}\) has units of amount per area per time, and \(R_{\mathrm{L}\rightarrow\mathrm{E}}\) has units of amount per time. A positive concentration difference gives net uptake; a negative difference gives net transfer in the opposite direction. Under a sink approximation, enterocyte-side concentration is negligible relative to luminal concentration, so \(C_{\mathrm{E}}^*\approx 0\) and \(R_{\mathrm{L}\rightarrow\mathrm{E}}\approx P_{\mathrm{eff}}A_{\mathrm{int}}C_{\mathrm{L,free}}\). The sink approximation is a modeling assumption, not part of the definition of permeability.

How Permeability Enters a PBPK Mass Balance

Suppose a luminal segment is well mixed, has fixed volume \(V_{\mathrm{L}}\), and contains free dissolved amount \(A_{\mathrm{L,free}}\), with \(C_{\mathrm{L,free}}=A_{\mathrm{L,free}}/V_{\mathrm{L}}\). If the enterocyte-side concentration is negligible and no other process changes the free luminal amount during the interval considered, the luminal mass balance is

This equation shows why permeability is often said to control the absorption rate: with area and volume fixed, it sets the first-order rate constant for removal from the free luminal pool. It does not imply that total luminal drug always follows a single exponential. Dissolution, precipitation, transit, binding, and other represented processes can replenish or remove drug and alter the time course. In a coupled PBPK model, those processes must appear as their own mass-balance terms.

Deriving the Sink-Condition Absorption Time Course

A useful check on the compartment equation is to solve it under its stated assumptions. Let \(k_{\mathrm{abs}}=P_{\mathrm{eff}}A_{\mathrm{int}}/V_{\mathrm{L}}\). This quantity has units of inverse time, since permeability has units of length per time and area divided by volume has units of inverse length. The mass balance becomes \(dA_{\mathrm{L,free}}/dt=-k_{\mathrm{abs}}A_{\mathrm{L,free}}\).

Effective Permeability and Apparent Assay Permeability

Apparent permeability, commonly denoted \(P_{\mathrm{app}}\), is calculated from transport measured in a specified in vitro assay. For a typical donor-to-receiver assay, its operational calculation is

where \(A_{\mathrm{assay}}\) is the assay membrane area, \(C_{\mathrm{donor}}\) is the donor concentration used for normalization, and \(dQ_{\mathrm{receiver}}/dt\) is the measured rate of appearance in the receiver. It also has units of length per time. Matching units does not make \(P_{\mathrm{app}}\) interchangeable with \(P_{\mathrm{eff}}\): the assay and PBPK parameter refer to different experimental or model systems, with potentially different barriers, surface-area definitions, concentration conditions, and sources of loss or retention.

Diagnostics, Alternatives, and Limitations

A permeability implementation should pass simple physical and numerical checks before it is used to interpret absorption. First, confirm units: \(P_{\mathrm{eff}}A_{\mathrm{int}}C\) must have units of amount per time. Second, verify signs and mass conservation: a positive luminal-to-enterocyte transfer rate must reduce luminal amount by exactly the amount added to the receiving compartment at that interface. Third, inspect limiting behavior. With zero free luminal concentration and a sink assumption, uptake must be zero; with zero permeability or zero area, transfer must also be zero. If the model predicts uptake when all three conditions eliminate the driving force, its flux expression or bookkeeping is inconsistent.

A common warning sign is an apparent absorption rate that changes when concentration units are changed, even though the physical state is unchanged. This often indicates that permeability, concentration, or area units have been mixed. Another is an implausibly large uptake rate caused by applying an effective area twice—for example, using an area-scaled transfer coefficient and then multiplying by area again. Finally, if luminal amount falls faster than enterocyte amount rises, or vice versa, check whether another explicitly modeled process accounts for the difference; otherwise the interface violates mass conservation.

The main alternative to a permeability-driven flux is a direct first-order absorption term, such as \(R=k_{\mathrm{abs}}A_{\mathrm{L,free}}\). Under the well-mixed sink assumptions, these forms are mathematically equivalent when \(k_{\mathrm{abs}}=P_{\mathrm{eff}}A_{\mathrm{int}}/V_{\mathrm{L}}\). The permeability form makes the roles of area and concentration explicit; a direct rate constant can be simpler when only the lumped removal rate is needed. Neither form is automatically more mechanistic: the choice depends on what the model needs to represent and which parameters can be supported.

An effective coefficient is a lumped description. It cannot, on its own, identify which physical barrier controls uptake, describe distinct transport routes, or determine the drug’s later enterocyte disposition. A fixed value may also be inadequate if the represented permeability, area, or concentration driving force changes substantially with location or conditions. Those limitations should be addressed through a richer model only when the data and intended predictions justify the added structure; an effective parameter should not be interpreted as a direct measurement of every underlying process.