Why Plasma Binding Changes the Prediction
An in vitro assay measures metabolic turnover under specified incubation conditions. Scaling that turnover gives an estimate of intrinsic clearance, but the concentration used to define clearance matters. Enzymes act on unbound substrate; a clearance calculated using total concentration can therefore differ from one calculated using unbound concentration. Plasma binding introduces a second distinction when predicting hepatic clearance: only the unbound fraction in plasma is available to enter the liver in the simplified well-stirred model.
The key is to keep the concentration convention attached to each clearance value. If scaled intrinsic clearance is expressed relative to unbound concentration, plasma unbound fraction is applied when predicting hepatic clearance. If intrinsic clearance has already been expressed relative to total plasma concentration, that fraction is already incorporated. Applying it again counts plasma binding twice.
The distinction applies to the concentration basis, not to two different metabolic processes. For example, an in vitro clearance measured relative to total incubation concentration may require correction for incubation binding before it represents unbound intrinsic clearance. The earlier material on binding corrections in microsomal and hepatocyte incubations addresses that in vitro step. The plasma correction discussed here is separate: it concerns translating an unbound intrinsic clearance into a hepatic clearance prediction.
Plasma Binding in the Well-Stirred Liver Model
For a well-stirred liver, the model assumes that liver contents are mixed uniformly, hepatic blood flow is \(Q_H\), and metabolism is linear over the concentrations considered. In the plasma-based form below, the plasma unbound fraction is treated as constant and equilibrium between bound and unbound drug is assumed. Concentrations, blood flow, and clearances must also use compatible blood–plasma conventions; the simple expression assumes those have been made consistent.
The model’s structure can be justified with a steady-state mass balance. Let \(C_{in}\) and \(C_{out}\) denote compatible inlet and outlet total concentrations. The flow-mediated delivery to the liver is \(Q_H(C_{in}-C_{out})\). Under the model assumptions, metabolism is \(f_{u,p}CL_{int,u}C_{out}\). Equating delivery and metabolism gives
Rearranging yields \(Q_HC_{in}=(Q_H+f_{u,p}CL_{int,u})C_{out}\). The hepatic extraction ratio is \(E_H=(C_{in}-C_{out})/C_{in}\), and hepatic clearance is \(CL_H=Q_HE_H\). Substituting the mass-balance solution for \(C_{out}\) gives the stated equation. This derivation also shows why the unbound fraction enters with intrinsic clearance: it scales the metabolism term in the mass balance.
Worked Example: Scale an Unbound Clearance to Predict Hepatic Clearance
Suppose a hypothetical microsomal assay gives an apparent intrinsic clearance of \(12\ \mu\text{L}/\text{min}/\text{mg}\) when referenced to total incubation concentration. The incubation unbound fraction is \(0.40\). For this example, use \(40\ \text{mg}\) microsomal protein per gram of liver, a liver mass of \(1500\ \text{g}\), plasma unbound fraction \(f_{u,p}=0.050\), and hepatic blood flow \(Q_H=1.5\ \text{L}/\text{min}\). Assume the assay correction and scaling factors are valid and the well-stirred assumptions apply.
First correct the apparent assay clearance to an unbound-concentration basis:
Scale to the whole liver:
This is the scaled \(CL_{int,u}\), not yet a total-plasma-referenced clearance. Apply plasma binding once in the well-stirred equation:
The predicted hepatic clearance is approximately \(0.0849\ \text{L}/\text{min}\). As a check, the equivalent total-plasma-referenced intrinsic clearance is \(CL_{int,p}=0.050\times1.8=0.0900\ \text{L}/\text{min}\); using this value in the second form gives the same result.
Two Equivalent Conventions, Used Consistently
The unbound convention is often useful when the in vitro correction and scaling produce \(CL_{int,u}\). The total-plasma convention can be convenient when a value is already defined relative to total plasma concentration. They are interchangeable only when the binding fraction and concentration basis are clearly defined and the conversion is made once.
| Clearance value in hand | Meaning | Use in the well-stirred model |
|---|---|---|
| \(CL_{int,u}\) | Rate divided by unbound concentration | Use \(f_{u,p}CL_{int,u}\) in numerator and denominator |
| \(CL_{int,p}\) | Rate divided by total plasma concentration; already includes \(f_{u,p}\) | Use \(CL_{int,p}\) directly; do not multiply by \(f_{u,p}\) again |
Worked Example: Avoid Applying Plasma Binding Twice
A hypothetical compound has scaled \(CL_{int,u}=2.0\ \text{L}/\text{min}\), \(f_{u,p}=0.10\), and \(Q_H=1.2\ \text{L}/\text{min}\). Assume the model conditions above hold.
Using the unbound convention, the binding-adjusted term is \(0.10\times2.0=0.20\ \text{L}/\text{min}\), so
Alternatively, \(CL_{int,p}=0.20\ \text{L}/\text{min}\), and the total-plasma form gives the same \(0.1714\ \text{L}/\text{min}\). A common incorrect calculation treats \(0.20\) as though it were still unbound-concentration-referenced and multiplies by \(0.10\) again. That gives \(CL_H=(1.2\times0.020)/(1.2+0.020)=0.01967\ \text{L}/\text{min}\), an erroneous prediction because plasma binding was counted twice.
The opposite convention error is also consequential: putting \(CL_{int,u}=2.0\) directly into the model without \(f_{u,p}\) would give \(1.2\times2.0/(1.2+2.0)=0.750\ \text{L}/\text{min}\). The calculation is only meaningful when the clearance’s concentration basis is identified first.
Worked Example: Compare Compounds With Different Plasma Binding
Consider two hypothetical compounds with the same scaled unbound intrinsic clearance, \(CL_{int,u}=1.0\ \text{L}/\text{min}\), and hepatic blood flow \(Q_H=1.2\ \text{L}/\text{min}\). Compound A has \(f_{u,p}=0.050\); compound B has \(f_{u,p}=0.50\). Assume each fraction is constant at the relevant concentrations and the same well-stirred assumptions apply.
For compound A, \(f_{u,p}CL_{int,u}=0.050\ \text{L}/\text{min}\), so
For compound B, \(f_{u,p}CL_{int,u}=0.50\ \text{L}/\text{min}\), so
With the same unbound intrinsic clearance, the model predicts higher hepatic clearance for compound B because its larger unbound fraction increases the concentration available for metabolism. This is a model-based comparison, not evidence that binding alone determines observed clearance in every setting.
Checks, Limits, and Common Pitfalls
For positive \(Q_H\), \(f_{u,p}\), and \(CL_{int,u}\), the model predicts \(0\leq CL_H<Q_H\). This is a useful arithmetic check: a prediction above hepatic blood flow signals a calculation or unit problem under this model. The equation also has informative limiting behavior. If \(f_{u,p}CL_{int,u}\) is much smaller than \(Q_H\), the denominator is approximately \(Q_H\), so \(CL_H\) is approximately \(f_{u,p}CL_{int,u}\). If it is much larger than \(Q_H\), the denominator is approximately \(f_{u,p}CL_{int,u}\), so \(CL_H\) approaches \(Q_H\). These limits explain why clearance becomes increasingly flow-limited at high extraction.
A practical audit should record whether each input is based on total or unbound concentration, its units, and whether it is an assay value or a scaled whole-liver value. Check that \(f_{u,p}\) lies between zero and one, and that the plasma fraction refers to the relevant species and conditions. The equation’s numerator and denominator must use compatible clearance and flow units. If hepatic blood flow is in volume per time, the scaled intrinsic clearance must also be in volume per time.
This correction does not remove all uncertainty in a prediction. Plasma binding may vary with concentration or physiological conditions, violating the constant-\(f_{u,p}\) assumption. A plasma-based fraction may also be insufficient if the model inputs use blood concentrations and the blood–plasma relationship has not been handled consistently. Further, the well-stirred expression assumes linear metabolism and uniform mixing; it is not a substitute for a more appropriate model when those assumptions fail.
The main alternative within this calculation is not a different biological mechanism, but the equivalent total-plasma convention: convert \(CL_{int,u}\) to \(CL_{int,p}\) once and use the total-plasma form. A measured systemic clearance can also be compared with a prediction, but it is not interchangeable with intrinsic clearance: the former reflects the body-level outcome, whereas the latter is an input to a liver model. A mismatch should prompt checks of conventions and assumptions rather than an automatic extra plasma-binding correction.
- Wrong: “The scaled value is \(CL_{int,p}\), so I should multiply it by \(f_{u,p}\) in the liver model.” Correction: If it is already defined relative to total plasma concentration, it already contains the plasma fraction.
- Wrong: “The assay clearance is unbound just because metabolism involves unbound drug.” Correction: The clearance’s concentration basis is determined by how it was calculated; check the denominator and any incubation-binding correction.
- Wrong: “Any plasma unbound fraction can be used with any clearance value.” Correction: Use the fraction and concentration conventions that match the model, species, and conditions, and ensure blood–plasma inputs are consistent.
Check Your Understanding
Answer each question, then compare with the solutions below.
- A hypothetical compound has \(CL_{int,u}=0.80\ \text{L}/\text{min}\), \(f_{u,p}=0.25\), and \(Q_H=1.0\ \text{L}/\text{min}\). What hepatic clearance does the well-stirred model predict?
- A reported \(CL_{int,p}\) is \(0.12\ \text{L}/\text{min}\). Should it be multiplied by \(f_{u,p}\) again before using the total-plasma form? Explain.
- For positive intrinsic clearance and flow, can the well-stirred model predict \(CL_H>Q_H\)? Give the reason.
- State one assumption that may make a constant plasma-binding correction unreliable.
Solutions. (1) \(f_{u,p}CL_{int,u}=0.25\times0.80=0.20\ \text{L}/\text{min}\), so \(CL_H=(1.0\times0.20)/(1.0+0.20)=0.1667\ \text{L}/\text{min}\). (2) No. By definition, \(CL_{int,p}\) is already referenced to total plasma concentration and includes the plasma fraction. (3) No. Writing \(a=f_{u,p}CL_{int,u}>0\), \(CL_H=Q_Ha/(Q_H+a)\), and \(a/(Q_H+a)<1\), so \(CL_H<Q_H\). (4) For example, if \(f_{u,p}\) changes with concentration, the constant-fraction assumption does not hold.