Segment-Specific Permeability and Absorption Windows

Why Location Changes Absorption Opportunity

A drug’s absorption opportunity depends not only on whether it contacts the gut wall, but also on where it is available. Effective permeability can differ between gut segments, as can the surface area and luminal volume that determine how permeability translates into a segmental absorption rate. Transit also matters: even a segment with appreciable permeability may provide limited opportunity if drug moves onward quickly.

The earlier material on transit–absorption competition established how first-order absorption competes with first-order transit. Here that competition is applied to a regional permeability profile. The aim is to identify where absorption is locally favored and to compare the opportunity available to drug entering at different locations. This is not the same as attributing a total absorbed amount to each region: that requires tracking the amount reaching and absorbed in each segment.

Definition: In a segment-specific absorption model, \(k_{a,i}\) is the first-order absorption rate constant for segment \(i\), and \(k_{t,i}\) is its first-order transit rate constant to the next segment. Both have units of inverse time. An absorption window is a gut region in which the model assigns appreciable opportunity for absorption relative to transit. Its boundaries depend on the regional rate profile and on the operational criterion used; they are not necessarily fixed anatomical boundaries.

From Effective Permeability to Local Opportunity

A common segmental representation links the effective permeability to the absorption rate using the segment’s effective absorptive area \(A_i\) and luminal volume \(V_i\):

k_{a,i} = \frac{P_{\mathrm{eff},i} A_i}{V_i}

Here \(P_{\mathrm{eff},i}\) has units of length per time, \(A_i\) has units of area, and \(V_i\) has units of volume, so \(k_{a,i}\) has units of inverse time. This is an effective, model-level rate: it summarizes the assumed permeability and geometry rather than describing every microscopic transport process. The mapping is appropriate when the dissolved luminal drug is treated as well mixed within a segment and absorption is represented as a linear loss.

In the simplest competition model, drug in segment \(i\) has two possible first-order exits: absorption at rate \(k_{a,i}\), or transit at rate \(k_{t,i}\). The local probability of absorption before transit is therefore:

p_i = \frac{k_{a,i}}{k_{a,i}+k_{t,i}}

This probability is conditional on drug entering the segment and on the model assumptions. It is not the same as \(k_{a,i}\): a rate is expressed per unit time, whereas \(p_i\) is dimensionless. A useful local comparison is the ratio \(k_{a,i}/k_{t,i}\). When the two rates are equal, the two competing exits are equally likely, so \(p_i=0.5\). A modeler could use that as a simple operational criterion for an absorption window, but the threshold is a modeling choice, not a universal physiological definition.

Assumptions: The local probability formula assumes well-mixed dissolved drug, constant segment-specific rates during the modeled interval, and independent first-order absorption and transit hazards. Absorption and transit are the only competing exits. The formula does not itself represent dissolution-limited input, degradation, return from tissue to lumen, saturation, or within-segment spatial gradients.

Why the Competition Formula Holds

The formula follows directly from the competing exponential waiting times. If absorption and transit have constant rates \(k_{a,i}\) and \(k_{t,i}\), the probability that neither has occurred by time \(t\) is the product of their survival probabilities. Thus the probability density that absorption occurs first at time \(t\) is \(k_{a,i}\) times the probability that both events have not yet occurred:

f_{\mathrm{abs},i}(t) = k_{a,i}\exp(-k_{a,i}t)\exp(-k_{t,i}t) = k_{a,i}\exp\!\left[-(k_{a,i}+k_{t,i})t\right]

Integrating this density over all possible times gives the probability of absorption before transit:

p_i = \int_0^\infty f_{\mathrm{abs},i}(t)\,dt = k_{a,i}\int_0^\infty \exp\!\left[-(k_{a,i}+k_{t,i})t\right]dt = \frac{k_{a,i}}{k_{a,i}+k_{t,i}}

The result makes the location dependence explicit: changing permeability, area, volume, or transit changes the local absorption opportunity. It also shows why comparing permeability alone can be inadequate. The relevant competition uses the resulting absorption rate and the transit rate, not permeability in isolation.

Worked Example: Identifying a Local Absorption Window

Consider two hypothetical gut segments. In segment A, \(P_{\mathrm{eff},A}=8.0\times10^{-5}\) cm/s, \(A_A=1500\) cm², \(V_A=120\) cm³, and \(k_{t,A}=1.2\) h\(^{-1}\). In segment B, \(P_{\mathrm{eff},B}=1.5\times10^{-5}\) cm/s, \(A_B=900\) cm², \(V_B=180\) cm³, and \(k_{t,B}=0.45\) h\(^{-1}\). Calculate each local absorption probability and use \(p_i\geq0.5\) as an explicitly chosen window criterion.

Segment A. First calculate the absorption rate in seconds: \(k_{a,A}=(8.0\times10^{-5}\ \mathrm{cm/s})(1500\ \mathrm{cm}^2)/(120\ \mathrm{cm}^3)=0.0010\ \mathrm{s}^{-1}\). Converting to hours gives \(0.0010\times3600=3.60\ \mathrm{h}^{-1}\). Therefore, \(p_A=3.60/(3.60+1.20)=3.60/4.80=0.750\).

Segment B. The rate is \(k_{a,B}=(1.5\times10^{-5}\ \mathrm{cm/s})(900\ \mathrm{cm}^2)/(180\ \mathrm{cm}^3)=7.5\times10^{-5}\ \mathrm{s}^{-1}\). In hours this is \(7.5\times10^{-5}\times3600=0.270\ \mathrm{h}^{-1}\). Thus \(p_B=0.270/(0.270+0.450)=0.270/0.720=0.375\).

Under the selected \(0.5\) criterion, segment A is inside the operational absorption window and segment B is outside it. This conclusion concerns the local chance of absorption before transit, conditional on entering each segment. It does not say how much drug will actually be absorbed in either region.

Comparing Different Entry Locations

A compound released or made available in a proximal segment can encounter a different sequence of local opportunities from one that becomes available farther downstream. In the chain model, a drug entering segment \(i\) is absorbed there with probability \(p_i\); with probability \(1-p_i\), it transits onward. The probability of reaching the next segment is therefore \(1-p_i\), under the same competition assumptions.

For comparing entry locations, the opportunity over a specified set of segments can be calculated by following the conditional probabilities along the chain. This is an application of sequential transit–absorption competition, not a separate estimate of each segment’s contribution to an observed total.

Worked Example: Comparing Proximal and Distal Entry

Suppose segment A has \(k_{a,A}=1.5\) h\(^{-1}\) and \(k_{t,A}=0.5\) h\(^{-1}\). The next segment, B, has \(k_{a,B}=0.20\) h\(^{-1}\) and \(k_{t,B}=0.80\) h\(^{-1}\). Compare the probability of absorption before leaving segment B for drug entering at A with the probability for drug entering directly at B. Treat B as the end of the window.

The local probabilities are \(p_A=1.5/(1.5+0.5)=0.75\) and \(p_B=0.20/(0.20+0.80)=0.20\). Drug entering at A is absorbed in A with probability \(0.75\). The fraction of the entering drug that transits from A to B is \(1-p_A=0.25\); conditional on reaching B, its chance of absorption there is \(0.20\). Thus the probability of absorption before leaving B, given entry at A, is \(0.75+(0.25)(0.20)=0.80\).

Drug entering directly at B has only B’s local opportunity in this window, so its probability is \(p_B=0.20\). Under these assumed rates, the modeled opportunity before leaving B is 0.80 for entry at A and 0.20 for entry at B. The contrast arises because proximal entry exposes the drug to segment A before it reaches B. It does not establish that proximal entry always yields more absorption; the result depends on the rates and on the window being compared.

Worked Example: The Same Absorption Rate, Different Transit

Two hypothetical segments have the same \(k_a=0.40\) h\(^{-1}\), but different transit rates. Segment C has \(k_{t,C}=0.40\) h\(^{-1}\), whereas segment D has \(k_{t,D}=1.20\) h\(^{-1}\). Calculate and compare their local absorption probabilities.

For C, \(p_C=0.40/(0.40+0.40)=0.40/0.80=0.500\). For D, \(p_D=0.40/(0.40+1.20)=0.40/1.60=0.250\). The same absorption rate therefore corresponds to different local opportunity when transit differs. If \(k_a\) arose from the same permeability–geometry calculation in both segments, it would still be incorrect to infer equal absorption probabilities without checking the transit rates.

This comparison also cautions against describing a window using permeability alone. A region may have a substantial effective permeability but a lower modeled local probability if transit is faster. Conversely, a modest absorption rate can yield a meaningful probability when transit is sufficiently slow.

Diagnostics, Alternatives, and Limits

A regional profile is useful only if its rates and assumptions are internally consistent. Before interpreting a window, check that \(P_{\mathrm{eff},i}A_i/V_i\) has units of inverse time, that the time units match those used for \(k_{t,i}\), and that the stated local probability lies between zero and one. Check the probability against limiting behavior: if \(k_{a,i}=0\), then \(p_i=0\); if \(k_{t,i}=0\) while \(k_{a,i}>0\), then \(p_i=1\); and if both rates are positive and equal, then \(p_i=0.5\). A result outside these bounds signals an arithmetic or unit error.

The regional profile should also be examined for implausible discontinuities. A sharp change in \(k_a\) may reflect a genuine modeling choice, but it can also arise from inconsistent geometry inputs, a unit conversion error, or assigning a permeability measured under different conditions to one segment. A smooth-looking profile is not proof that the underlying inputs are valid, and abrupt changes should be justified rather than automatically smoothed away.

A uniform-\(k_a\) model is simpler and can be appropriate when regional differences are not supported by the intended application or available inputs. Its cost is that it suppresses location-specific absorption opportunity. A segment-specific model makes those differences explicit but requires regional parameters and can create a false impression of precision if the inputs are weak. More spatially resolved models can represent within-segment gradients, but they require additional structure and information beyond a well-mixed compartmental description.

The local competition probability is a model-based comparison, not a direct measurement of a biological probability for every molecule in a real gut. The calculation assumes constant rates and a well-mixed dissolved state. If dissolution, formulation release, nonlinear transport, changing physiology, or spatial gradients substantially affect availability, the simple local calculation may not describe the relevant process. In particular, an absorption window for dissolved drug does not by itself establish where a solid formulation will dissolve or release its contents.

Common Mistakes

Treating permeability as the absorption rate. Permeability has units of length per time; the segmental rate also depends on area and volume. Use \(k_{a,i}=P_{\mathrm{eff},i}A_i/V_i\) when that mapping is assumed.
Comparing \(k_a\) values without checking transit. The local probability depends on both rates, \(k_a/(k_a+k_t)\), not on \(k_a\) alone.
Mixing seconds and hours. Convert rates to a common time unit before calculating the probability.
Calling a threshold a physiological boundary. A criterion such as \(p_i\geq0.5\) is an operational definition for a model comparison, not a universal anatomical fact.
Confusing local opportunity with absorbed amount. The local probability is conditional on entering a segment. The amount available there depends on what happened upstream.

Key takeaway: Segment-specific effective permeability becomes a local absorption rate only after accounting for geometry. Local absorption opportunity then depends on that rate relative to transit. Because those rates vary by location, the gut segment where drug becomes available can materially change the opportunity predicted by the model.

Check Your Understanding

For each question, calculate or explain the result using the stated model assumptions.

A segment has \(P_{\mathrm{eff}}=2.0\times10^{-5}\) cm/s, \(A=1000\) cm², \(V=100\) cm³, and \(k_t=0.50\) h\(^{-1}\). Find \(k_a\) in h\(^{-1}\) and the local absorption probability.
Two segments have \(k_a=0.60\) h\(^{-1}\). Their transit rates are \(0.30\) h\(^{-1}\) and \(1.20\) h\(^{-1}\). Which has greater local absorption opportunity, and what are the probabilities?
Drug enters segment A with \(p_A=0.40\), then segment B with \(p_B=0.50\). What is the probability of absorption before leaving B, given entry at A?
Why is it not sufficient to label the region with the largest \(P_{\mathrm{eff}}\) as the region of greatest absorption opportunity?

Solutions. (1) \(k_a=(2.0\times10^{-5})(1000)/100=2.0\times10^{-4}\) s\(^{-1}\). Converting gives \(2.0\times10^{-4}\times3600=0.720\) h\(^{-1}\). The local probability is \(0.720/(0.720+0.50)=0.590\), to three significant figures. (2) For \(k_t=0.30\), \(p=0.60/(0.60+0.30)=0.667\). For \(k_t=1.20\), \(p=0.60/(0.60+1.20)=0.333\). The segment with lower transit has greater local opportunity. (3) Absorption occurs in A with probability 0.40. Drug transits to B with probability \(1-0.40=0.60\), and then is absorbed there with conditional probability 0.50. The total probability is \(0.40+(0.60)(0.50)=0.70\). (4) \(P_{\mathrm{eff}}\) alone does not determine \(k_a\), which also depends on area and volume, and local opportunity additionally depends on transit. The relevant comparison is \(k_a/(k_a+k_t)\), subject to the model assumptions.