Parallel Luminal Pathways Within a Segment

Why Represent Parallel Pathways?

A single gut segment need not behave as one uniform luminal population. Drug entering the segment may be distributed across subpopulations that differ in their physical state, dissolution rate, or exposure to transit and absorption. For example, one fraction might dissolve relatively quickly while another remains solid longer. A parallel-pathway model represents these populations separately while treating their absorption as a combined input from the segment.

The central bookkeeping rule is that splitting a dose creates separate states, not extra drug. If the dose assigned to the segment is \(D\), the initial amounts across all pathway states must add to \(D\). The segment’s absorption flux is then the sum of the absorption fluxes from those pathways.

States, Rates, and Dose Allocation

Consider two pathways, indexed by \(j=1,2\). Let \(S_j(t)\) be the solid drug amount and \(L_j(t)\) the dissolved luminal amount in pathway \(j\), both measured in mg. Let \(k_{d,j}\) be its first-order dissolution rate constant, \(k_{tS,j}\) the transit rate constant for solid drug, \(k_{a,j}\) the absorption rate constant for dissolved drug, and \(k_{tL,j}\) the transit rate constant for dissolved drug. Each rate constant has units of inverse time, such as h\(^{-1}\).

The model assumes that dissolution transfers mass from solid to dissolved drug within the same pathway. Solid drug can also leave by transit. Dissolved drug can either be absorbed or leave by transit. Under these assumptions, the pathway mass balances are:

The absorption flux from pathway \(j\) is \(J_j(t)=k_{a,j}L_j(t)\), in mg per unit time. Since the pathways contribute to the same segment, the segment-level flux is their sum:

At the time the dose enters the modeled segment, the initial states must satisfy:

A pathway may begin with solid drug, dissolved drug, or both. The allocation should reflect the model’s assumed entry state; a pathway with zero initial amount may still receive drug later only if an explicit input is added. In the model used here, there is no subsequent external input to the segment.

Combining Fluxes and Checking Conservation

The total luminal amount is \(M(t)=\sum_j[S_j(t)+L_j(t)]\). Adding the two state equations within a pathway cancels the dissolution transfer: dissolution removes mass from \(S_j\) and adds exactly the same amount to \(L_j\). Thus the only losses from the segment are absorption and transit.

In particular, \(k_{a,j}L_j\) is counted as absorption, while \(k_{tS,j}S_j\) and \(k_{tL,j}L_j\) are transit losses. This identity is a useful internal check: dissolution must not appear as a net loss from total luminal mass. It also makes clear why pathway fluxes are added rather than averaged. Each pathway has its own amount and rate, and the output is the sum of the resulting rates of absorption.

Eventual Absorption From a Pathway

For constant positive rate constants and no later input, define \(\lambda_{S,j}=k_{d,j}+k_{tS,j}\) and \(\lambda_{L,j}=k_{a,j}+k_{tL,j}\). A solid amount leaves its solid state through either dissolution or solid transit. The fraction of that amount that dissolves is \(k_{d,j}/\lambda_{S,j}\). Once dissolved, the fraction that is absorbed rather than transited is \(k_{a,j}/\lambda_{L,j}\). An initial dissolved amount faces only the second competition.

Consequently, the eventual absorbed amount from pathway \(j\) is:

This result can also be obtained by integrating the absorption flux. The solid state is \(S_j(t)=S_j(0)e^{-\lambda_{S,j}t}\). Integrating the solid-state balance over time shows that the amount dissolved is \(S_j(0)k_{d,j}/\lambda_{S,j}\); the remaining solid amount exits by transit. The dissolved amount available over its lifetime is its initial amount plus that dissolved amount. Of this, the fraction \(k_{a,j}/\lambda_{L,j}\) is absorbed, because absorption and dissolved transit are competing first-order exits. This gives the expression above, including the contribution from initial dissolved drug.

The result is an amount, not an instantaneous flux. To obtain total segment absorption, sum the eventual amounts from all pathways. It is not generally valid to combine pathway rates into a single arithmetic average: the amounts in the pathways change over time, so their contributions to the total flux change as well.

Initialization With Mixed States

A dose split may assign different physical states as well as different amounts to the pathways. The initial-dose check must include every state, not just the solid states or just one pathway. The next example shows why this check matters before any rate calculation.

Diagnostics, Alternatives, and Limitations

Several simple checks catch common implementation errors. First, verify that all initial pathway states add to the amount entering the segment. Second, confirm that each term in a state equation has units of amount per time; for example, \(k_{a,j}L_j\) has units h\(^{-1}\) times mg, or mg/h. Third, check that every rate is nonnegative and that the absorption flux is nonnegative. Finally, compare the numerical model’s cumulative absorbed amount with the competing-rate calculation when rates are constant and the model has no later input. A disagreement indicates a likely error in state assignment, rate application, or flux summation.

A common mistake is to average pathway rate constants first and then apply the average to the total amount. For example, an average absorption rate does not generally preserve the segment flux because the pathway amounts evolve differently. The appropriate quantity is \(k_{a,1}L_1(t)+k_{a,2}L_2(t)\), not an average rate multiplied by \(L_1(t)+L_2(t)\), unless special conditions make those expressions equal.

A single homogeneous pathway is a simpler alternative when there is no reason to distinguish subpopulations, or when the available information supports only one effective set of rates. It uses fewer states and parameters, but cannot represent distinct dissolution or transit behavior. Parallel pathways can represent such differences directly, at the cost of more parameters and assumptions. Neither representation is automatically more accurate; the choice should match the biological or formulation question and the information available.

The model is limited by its assumptions. Fixed first-order rates do not capture saturation, time-varying physiology, or interactions between pathways. Distinct pathways may also be difficult to justify from aggregate observations alone. A good fit does not by itself establish that the chosen subpopulations are uniquely identified or biologically distinct. The model’s strongest immediate contribution is transparent bookkeeping: each pathway’s state, flux, and dose contribution can be inspected separately, and the total output follows by addition.