Reading Plasma and Biliary Data for Transporter Mechanisms

Reading the Pattern Before Naming a Mechanism

A plasma concentration profile describes drug exposure in the systemic circulation; it does not directly show how much drug entered hepatocytes or what happened after entry. Biliary recovery adds evidence about one intracellular exit route, while urinary recovery records parent drug reaching urine. Interpreted together, these measurements can help distinguish a reduction in hepatic entry from a redistribution of intracellular drug between routes. They do not, by themselves, prove which transporter changed.

The central comparison is between amounts recovered by each route and the plasma exposure that could have driven those amounts. A decrease in biliary recovery alone is ambiguous: less drug may have entered the liver, or a smaller share of intracellular drug may have been directed to bile. Metabolite formation data can help distinguish those explanations. Urinary recovery provides a useful check on systemic disposition, but it is not a direct measurement of hepatic uptake.

Use the same dose, observation window, and sampling conventions when comparing conditions. If plasma protein binding changes, unbound plasma exposure may be the more appropriate driver, so the indices should instead use \(AUC_{p,u}\), the unbound plasma AUC. In the examples below, binding is assumed unchanged and total plasma AUC is used. Complete collection, reliable metabolite quantification, and approximately linear, time-invariant kinetics are also assumed unless stated otherwise.

A Minimal Model for Interpreting Route Patterns

Consider a simplified linear model in which hepatic uptake flux is proportional to plasma concentration: \(J_{in}(t)=CL_{in}C_p(t)\). Once inside the hepatocyte, drug can be directed to biliary excretion, metabolism, or another fate. Let \(q_b\) and \(q_m\) be the fractions of hepatic entry ultimately routed to measured bile and measured metabolism, respectively. If these fractions remain constant and routing is instantaneous over the observation period, so that \(J_b(t)=q_bJ_{in}(t)\) and \(J_m(t)=q_mJ_{in}(t)\), integrating the route fluxes gives:

Dividing by plasma AUC gives \(CL_b=q_bCL_{in}\) and \(CL_m=q_mCL_{in}\). Thus, if both route indices decline by a similar proportion while their ratio remains stable, the pattern is consistent with lower hepatic entry with broadly unchanged relative routing. In contrast, a changed \(A_b/A_m\) ratio indicates a changed balance between the measured routes, provided that recovery and measurement are comparable.

This ratio cancels the uptake term in the simplified model. That is a useful diagnostic, not a guarantee: a changed ratio can also arise from incomplete recovery, altered metabolite formation or stability, saturation, or mismatched sampling. Other intracellular fates may mean that \(q_b+q_m<1\); the ratio compares the two measured routes and does not describe all drug disposition.

Worked Examples

Practical Checks and Failure Modes

Before comparing route patterns, verify that the plasma profile and recovery measurements refer to compatible conditions and time windows. AUC is obtained from the concentration–time profile, typically by an appropriate numerical integration and, when justified, a terminal extrapolation. AUC truncation can distort every exposure-normalized index. Biliary collections that stop early, incomplete urine collection, or metabolites not included in the assay can create apparent route shifts that are measurement artifacts.

Check recovery arithmetic explicitly. If the administered dose is 100 mg, measured routes cannot sum to 112 mg without an explanation such as double-counting, inconsistent molar-to-mass conversion, or assay error. Conversely, a large unrecovered fraction makes route comparisons less decisive. Metabolite amounts should be put on a consistent parent-drug equivalent basis before adding them to parent recoveries. Also check that a reported biliary amount represents drug excreted into bile rather than drug remaining in liver or gallbladder at the sampling endpoint.

• Changing unbound fraction: A change in binding can alter the relation between total plasma AUC and transporter-accessible concentration. Compare unbound exposure or state the limitation.
• Nonlinear or time-varying processes: Saturation, induction, inhibition that changes over time, or concentration-dependent routing can violate the constant-fraction model.
• Changing renal handling: Urinary recovery can change because renal clearance changes, even if hepatic entry does not. The urinary index is a check on systemic route disposition, not a hepatic uptake assay.
• Incomplete metabolite accounting: A fall in a measured metabolite may reflect altered downstream metabolite clearance or missing analytes rather than reduced formation.
• Different experimental settings: Biliary data from an animal, perfused liver, or hepatocyte system may not be directly comparable with plasma data from another setting without justified scaling and matched conditions.

Compared with reading plasma AUC alone, pairing exposure with bile and metabolite outcomes can distinguish a change in overall route indices from a shift in relative measured routing. Compared with relying on total dose recovery alone, exposure-normalized indices account for changes in systemic exposure. Neither comparison is as mechanistically specific as a well-supported dynamic PBPK model fitted to informative plasma and route data; such a model also depends on identifiable parameters, credible measurements, and appropriate assumptions. The earlier material on uptake and metabolism parameter identifiability is relevant: multiple parameter combinations can generate similar observed patterns.

A Common Mistake

A sound report separates observation from interpretation: first state the measured AUC and route amounts, then calculate the indices and ratios, and finally describe which mechanisms are consistent with the pattern and which alternatives remain. This distinction matters because a transporter claim can otherwise be stronger than the data support.