Representing Pharmacogenetic Transporter Variation

Why Transporter Variation Needs Its Own Representation

A pharmacogenetic change in an uptake transporter is not the same model input as a change in enzyme activity. For a drug taken up from blood into hepatocytes, reduced transporter function can lower the rate of hepatic entry. That can reduce hepatocyte concentrations at a given plasma concentration, while also reducing hepatic removal and raising systemic concentrations under a fixed dose. These effects describe different parts of the same process, so a PBPK representation should keep them distinct.

SLCO1B1 encodes OATP1B1, a transporter expressed on the sinusoidal membrane of hepatocytes. For a substrate whose hepatic uptake depends substantially on OATP1B1, alleles with reduced function can be represented as changes in uptake capacity. The translation from allele to capacity is not universal: it depends on the functional evidence, assay context, substrate, and model calibration. The activity values below are hypothetical and illustrate the translation, not clinical values for a particular allele or drug.

From Allele Function to Uptake Capacity

A practical representation begins with functional evidence for each allele and then states how the two alleles combine. For example, an assay or calibrated model may assign a reference-function allele a factor of 1 and a reduced-function allele a factor below 1. Under an explicitly assumed additive dosage model, the genotype factor is the mean of its two allele factors. If the reference uptake capacity is \(CL_{\mathrm{upt,ref}}\), genotype-specific capacity is that reference value multiplied by the genotype factor.

This mean-of-two-alleles rule is a modeling assumption, not a biological law. It is reasonable as a transparent starting point when each allele contributes comparably and there is no evidence of dominance or a nonadditive effect. If functional measurements instead provide genotype-level activity, those data should inform the genotype factor directly. When evidence is uncertain, alternative plausible factors can be examined rather than presenting one mapping as certain.

A population should generally be represented as a set of genotype-specific capacities with genotype weights, not only as a single average. The resulting distribution can be discrete across genotype groups, with additional within-group variability if supported by data. A population mean can summarize that distribution, but it does not preserve the different capacities an individual subject may have.

A Minimal Uptake–Efflux–Metabolism Model

To show how uptake capacity affects exposure, consider a linear model with unbound plasma concentration \(C_{p,u}\) and unbound hepatocyte concentration \(C_{h,u}\). Transporter-mediated uptake is \(CL_{\mathrm{upt}} C_{p,u}\). Hepatocyte drug leaves through a lumped efflux process with clearance \(CL_{\mathrm{back}}\) and is metabolized with intrinsic clearance \(CL_{\mathrm{met}}\). All three clearances are expressed in the same volume-per-time units. The model assumes constant parameters, well-mixed compartments, linear processes, and no saturation or time-dependent regulation.

Here \(V_h\) is the effective hepatocyte distribution volume. At a steady state, setting the derivative to zero gives the ratio of hepatocyte to plasma unbound concentration. Metabolic elimination is then the metabolic clearance multiplied by hepatocyte concentration. Together, these equations give an effective intrinsic clearance from unbound plasma concentration to hepatic metabolism.

The second expression follows by substituting the steady-state concentration ratio into the metabolic rate: \(CL_{\mathrm{met}}C_{h,u} =CL_{\mathrm{int,eff}}C_{p,u}\). It shows why lowering uptake can reduce the effective hepatic metabolic clearance even when the enzyme clearance itself is unchanged. It also shows why the hepatocyte-to-plasma ratio and systemic clearance should not be treated as interchangeable outputs: the ratio describes partitioning at a given plasma concentration, whereas clearance describes removal from the plasma-linked system.

Representing a Genotype-Weighted Capacity Distribution

Suppose a hypothetical cohort has genotype weights of 0.49 for \(*1A/*1A\), 0.42 for \(*1A/*5\), and 0.09 for \(*5/*5\). These are assumed weights for this example, not estimates for a real population. Using the capacities above produces three genotype-specific uptake values. The weighted mean is a useful summary, but it is not a substitute for retaining the genotype-specific values when predicting exposure.

Worked Substrate Consequences for Hepatic and Systemic Exposure

The next example uses the same hypothetical allele factors and a substrate with \(CL_{\mathrm{back}}=30\ \mathrm{L/h}\), \(CL_{\mathrm{met}}=120\ \mathrm{L/h}\), \(f_u=0.10\), and \(CL_{\mathrm{other}}=2.0\ \mathrm{L/h}\). Assume a single 100 mg intravenous dose, linear kinetics, and complete AUC over the relevant elimination period. Under these assumptions, total-plasma AUC equals dose divided by total systemic clearance. Hepatocyte unbound AUC is approximated by multiplying unbound plasma AUC by the steady-state concentration ratio from the model.

Interpretation, Diagnostics, and Alternatives

A useful diagnostic is to examine both the concentration ratio \(C_{h,u}/C_{p,u}\) and the systemic exposure for each genotype. In this model, reducing uptake capacity must reduce the ratio when efflux and metabolism are unchanged. Under a fixed dose, systemic exposure rises if the reduced effective hepatic clearance lowers total clearance. If nonhepatic clearance dominates, the systemic change may be small even when the hepatocyte-to-plasma ratio changes substantially. If simulated hepatic concentration rises as uptake capacity falls at a fixed plasma concentration, check the parameter mapping, units, and equations.

A second check is to compare transporter and enzyme effects conceptually. A change in enzyme intrinsic clearance changes \(CL_{\mathrm{met}}\); a transporter variant changes \(CL_{\mathrm{upt}}\) in this model. Either can alter effective hepatic metabolism, but they need not have the same effect on hepatocyte concentration. Replacing an uptake effect with an enzyme activity multiplier can therefore yield misleading tissue exposure predictions even if systemic clearance appears similar.

The allele-factor approach is compact and makes genotype assumptions visible. An alternative is to estimate genotype-specific uptake parameters from functional uptake assays or clinical data, using the assay-to-model relationship explicitly. Direct estimation can be more substrate-relevant, but may have limited precision or transportability across assay systems. Neither approach should treat an in vitro activity ratio as automatically equal to an in vivo uptake-clearance ratio; scaling and model validation are required.

Important failure modes include assigning a single unqualified function value to an allele with uncertain evidence, assuming additive dosage without checking it, and applying a transporter factor to every substrate regardless of transporter dependence. The model also omits saturation, transporter expression changes, competing substrates, hepatic blood-flow effects, and altered binding or metabolism. Where these processes matter, the simple linear representation may not predict either tissue or systemic exposure adequately. The introductory material on genetic variation as a PBPK input provides the broader parameter-mapping context; here the key additional requirement is to represent uptake as distinct from metabolism.