Cholestasis and Biliary Drug Disposition

Why Cholestasis Requires More Than a Clearance Multiplier

Cholestasis is impaired formation or flow of bile. It can change the movement of drugs and metabolites from hepatocytes into bile, the drainage of bile, or both. These are related but not interchangeable changes: canalicular transport moves a compound into bile, whereas bile flow carries bile and its contents onward. A PBPK model that represents only one of these processes may predict the wrong biliary output, liver retention, or systemic exposure.

The relevant change depends on the compound. A drug may be secreted into bile by a canalicular transporter, metabolized in the liver and then secreted as a metabolite, or undergo little biliary elimination. Cholestasis does not imply that every drug’s biliary transport or hepatic enzyme activity falls by the same proportion. The model should connect a disease-related change to the specific process supported by the available evidence.

The earlier material on translating renal and hepatic disease phenotypes into PBPK changes provides the general principle of linking a disease phenotype to a physiological or biochemical input. Here, that principle is applied specifically to bile formation, drainage, and biliary transport. The central modeling task is to decide which of these processes is impaired and to represent it at a level supported by the data.

Representing Transport and Bile Drainage

A compact model can track the amount of compound in hepatocytes, \(A_H\), and the amount in a bile compartment, \(A_B\). Let \(C_{H,u}\) denote the unbound concentration in hepatocytes that drives canalicular transport. Let \(J_{\mathrm{can}}\) be the rate of transport across the canalicular membrane, and let \(Q_{\mathrm{bile}}\) be bile flow. In a well-mixed bile compartment of volume \(V_B\), bile drainage is represented by \(Q_{\mathrm{bile}}C_B\), where \(C_B=A_B/V_B\).

This mass balance separates entry into bile from drainage out of the modeled bile compartment. A simple saturable representation of canalicular transport is:

Here, \(V_{\max,\mathrm{can}}\) is the maximum transport rate and \(K_{m,\mathrm{can}}\) is the unbound hepatocyte concentration at which transport reaches half of that maximum. If concentrations are well below \(K_{m,\mathrm{can}}\), the expression is approximately linear: \(J_{\mathrm{can}}\approx (V_{\max,\mathrm{can}}/K_{m,\mathrm{can}})C_{H,u}\). At higher concentrations, transport approaches capacity and cannot increase proportionally with concentration.

Cholestasis can be represented as a change in \(Q_{\mathrm{bile}}\), in one or more transport parameters, or in both, when evidence supports those changes. Reduced \(Q_{\mathrm{bile}}\) slows drainage in this compartment. Reduced \(V_{\max,\mathrm{can}}\) represents lower effective canalicular transport capacity; a change in \(K_{m,\mathrm{can}}\) instead changes the concentration dependence of transport. These parameters should not be treated as interchangeable.

A simpler model may represent biliary elimination directly as \(J_{\mathrm{bile}}=CL_{\mathrm{bile}}C_{H,u}\), where \(CL_{\mathrm{bile}}\) is an apparent biliary clearance. This can be appropriate when the data support an approximately linear relationship and the bile compartment’s transient dynamics are not needed. It combines transport and downstream handling into an effective term, so it cannot by itself distinguish a transport-capacity change from a drainage change.

Linking Biliary Changes to Systemic Exposure

For a parent drug, reduced biliary loss can increase the amount remaining available for other pathways or in the systemic circulation. In a linear model with constant systemic clearance components, write total clearance as the sum of biliary and nonbiliary clearances. The relationship below applies when clearance is linear, the dose is given intravenously, and the drug is followed until the administered amount has been eliminated.

The AUC relation follows from a mass balance. If \(A\) is the amount of drug in the modeled body and elimination is \(CL_{\mathrm{total}}C\), then \(dA/dt=-CL_{\mathrm{total}}C\) after an instantaneous intravenous dose. Integrating from the initial amount \(A(0)=\mathrm{Dose}\) until the amount has returned to zero gives \(\mathrm{Dose}=CL_{\mathrm{total}}\int_0^\infty C(t)\,dt\). The integral is AUC, which yields the stated equation. If biliary elimination is only one part of total clearance, a proportional reduction in \(CL_{\mathrm{bile}}\) therefore does not produce the same proportional increase in AUC.

For oral dosing, the simple linear relation becomes \(\mathrm{AUC}=F\cdot\mathrm{Dose}/CL_{\mathrm{total}}\), where \(F\) is systemic bioavailability. The intravenous example holds \(F=1\). If cholestasis also changes absorption or presystemic availability, that change must be represented separately rather than attributed automatically to altered biliary clearance.

Worked Examples: Flow, Transport, and Metabolites

Choosing a Model and Checking Its Behavior

The appropriate representation depends on the question and available data. An apparent \(CL_{\mathrm{bile}}\) is economical for approximately linear, aggregate disposition. An explicit transporter model is more informative when concentration dependence or transporter capacity matters. A bile compartment is useful when transient retention and drainage delay matter. These approaches can be combined, but additional parameters should be supported by data or justified physiological information.

Check the model’s behavior rather than relying only on a final AUC. Confirm that the drug or metabolite mass balance closes: administered or formed amount should be accounted for by remaining amounts, metabolism, excretion, and any modeled return pathways. Inspect predicted hepatocyte concentrations, biliary transport rates, bile-compartment amounts, and systemic concentrations over time. A large increase in liver amount after a modest parameter change may indicate a transport bottleneck or an inconsistent scaling of the relevant compartments.

Interpret biliary entry, biliary drainage, and net body elimination separately. Transport into bile is not necessarily irreversible loss: if intestinal reabsorption is included, some biliary material can return to the systemic circulation. Conversely, if the model treats all biliary output as permanent elimination, that assumption should be justified for the compound and model purpose. Data on bile flow, biliary recovery, parent and metabolite concentrations, and systemic concentration-time profiles can constrain different parts of the pathway; one type of observation does not automatically identify all of them.

A common failure is to reduce bile flow and biliary clearance by the same factor without checking the equations. If bile flow only controls the emptying term \(Q_{\mathrm{bile}}C_B\), it changes the residence time and amount in the bile compartment; it does not directly reduce \(J_{\mathrm{can}}\). A second failure is to apply a generic cholestasis multiplier to every biliary substrate. Transport depends on compound-specific processes, so a disease effect on one substrate does not establish the same effect for another.