Diagnosing Inconsistent Binding And Blood Partition Data

Why apparently conflicting measurements need a basis check

PBPK inputs for plasma binding, red-cell partition, and tissue binding are often collected using different samples and assays. Two reported values can look contradictory even when they describe different concentration bases, or when one measurement has inadequate recovery. Conversely, a plausible-looking value can conceal a real inconsistency. The first task is therefore not to average the numbers or choose the one that best fits a model. It is to identify exactly what each measurement represents and test whether the measurements can be reconciled by mass balance.

This tutorial builds on the definitions of plasma fraction unbound, blood-to-plasma concentration ratio, and tissue fraction unbound. The focus here is using those quantities together as diagnostic constraints. The checks do not prove that an assay is correct; they identify discrepancies that require an explanation before the inputs are combined.

Definition: Let \(C_p\) denote total drug concentration in plasma and \(C_b\) denote total drug concentration in whole blood, each expressed per volume of its named sample. The blood-to-plasma ratio is \(R_{b:p}=C_b/C_p\). Let \(C_{RBC}\) be the total concentration in packed red cells, expressed per volume of packed red cells, and \(K_{RBC:p}=C_{RBC}/C_p\). Hematocrit, \(Hct\), is the red-cell volume fraction of whole blood. Plasma fraction unbound is \(f_{u,p}=C_{u,p}/C_p\), and tissue fraction unbound is \(f_{u,t}=C_{u,t}/C_t\), with concentrations defined on the corresponding tissue or homogenate basis.

The units in each ratio cancel, but the concentration bases do not become interchangeable. A value reported per millilitre of whole blood is not automatically comparable with one reported per millilitre of plasma or packed cells. Record the sample matrix, volume basis, total or unbound status, and assay conditions for every result before comparing values.

Use hematocrit to connect red-cell and whole-blood measurements

Whole blood contains plasma occupying fraction \(1-Hct\) of the sample volume and red cells occupying fraction \(Hct\). The total amount of drug per unit whole-blood volume is the sum of the amounts in those two portions. Dividing that amount by the whole-blood volume gives the concentration identity:

C_b=(1-Hct)C_p+Hct\,C_{RBC}

Dividing both sides by \(C_p\) gives the key cross-check:

R_{b:p}=(1-Hct)+Hct\,K_{RBC:p}

This relation is a mass-balance identity for compatible total concentrations measured in the same blood sample at the same time. It assumes that hematocrit is known for that sample, the reported concentrations use the specified volume bases, and the analyte measured in plasma, red cells, and whole blood is the same parent drug. It does not require a particular mechanism of red-cell uptake. It can fail as a comparison if the measurements came from different species, different conditions, or samples that were not at comparable distribution states.

Worked Example: Reconciling a Blood-to-Plasma Ratio

Suppose a hypothetical paired blood sample has \(Hct=0.45\), and the measured blood-to-plasma ratio is \(R_{b:p}=1.60\). What red-cell-to-plasma ratio would be required by the mass-balance identity? An independent red-cell assay reports \(K_{RBC:p}=2.20\). Are the results exactly consistent?

Rearrange the identity to solve for the red-cell ratio:

K_{RBC:p}=\frac{R_{b:p}-(1-Hct)}{Hct}

Substitute the reported values:

K_{RBC:p}=\frac{1.60-(1-0.45)}{0.45} =\frac{1.60-0.55}{0.45} =\frac{1.05}{0.45} =2.333

The whole-blood result therefore implies a red-cell concentration about 2.333 times the plasma concentration. In the other direction, the independent red-cell result predicts:

R_{b:p}=0.55+(0.45)(2.20)=0.55+0.99=1.54

The predicted ratio, 1.54, is 0.0600 below the measured value of 1.60. The measurements are not exactly consistent. Before concluding that one is erroneous, check whether the assays used paired samples, the same concentration units and analyte definition, and the same hematocrit. If the difference exceeds expected analytical and sampling variability, investigate the methods rather than silently forcing the values to agree.

A further derived quantity is sometimes used to express unbound drug on a whole-blood basis. Under equilibrium, with compatible definitions of unbound concentration and the same parent drug in both matrices, the usual relationship is \(f_{u,b}=f_{u,p}/R_{b:p}\). It is a consequence of converting the unbound concentration from a plasma basis to a whole-blood total-concentration basis; it is not an independent measurement. It should not be used to repair a blood ratio that already fails the hematocrit check.

Worked Example: Converting Plasma to Whole-Blood Fraction Unbound

For a hypothetical compound, suppose \(f_{u,p}=0.0800\) and a compatible blood assay gives \(R_{b:p}=1.60\). Under the equilibrium and concentration-basis assumptions above:

f_{u,b}=\frac{f_{u,p}}{R_{b:p}} =\frac{0.0800}{1.60} =0.0500

The implied whole-blood fraction unbound is 0.0500, or 5.00%. This is a calculated quantity, not a second observation of plasma binding. If another report gives a substantially different whole-blood fraction unbound, check whether it was calculated from the same blood-to-plasma ratio and whether both reports use compatible definitions. Differences in equilibrium, sample conditions, or concentration basis can invalidate the conversion.

Check plasma assay recovery before interpreting fraction unbound

In an equilibrium dialysis experiment, free drug crosses a membrane while plasma proteins are retained. When the relevant conditions permit equilibrium, and the buffer and plasma sides have compatible free-drug activities, the buffer concentration can estimate the unbound concentration in the plasma chamber. The fraction unbound is then obtained by dividing that estimate by the total plasma concentration. The interpretation depends on more than the ratio: loss to the membrane, apparatus, or precipitate can reduce the amount recovered and distort the result.

A useful recovery calculation compares the total amount measured after incubation across all relevant compartments with the initial amount added. Amounts, not concentrations alone, must be summed when compartment volumes differ. A simplified expression is:

\text{Recovery}= \frac{\sum_j C_jV_j}{A_0}

Here \(C_j\) and \(V_j\) are the measured concentration and volume for each compartment included in the assay, and \(A_0\) is the initial amount added. A complete protocol may require adjustments for sampling, replacement volumes, or other accounted sources and sinks. The equation is not a universal acceptance threshold: acceptable recovery depends on the validated method. Low recovery signals that an apparent binding result may be affected by loss, but does not by itself identify the mechanism.

Worked Example: Interpreting a Low-Recovery Dialysis Result

In a hypothetical dialysis experiment, 24 ng of drug is added to a plasma chamber of volume 0.20 mL. After incubation, the plasma-side total concentration is 60 ng/mL and the buffer-side concentration is 30 ng/mL. The buffer volume is also 0.20 mL. Calculate the apparent fraction unbound and recovery, then interpret them separately.

If the buffer concentration is a valid estimate of the plasma-side unbound concentration, the apparent fraction unbound is:

f_{u,p,\mathrm{app}}=\frac{30\ \mathrm{ng/mL}}{60\ \mathrm{ng/mL}}=0.500

The amount recovered on the plasma side is \((60\ \mathrm{ng/mL})(0.20\ \mathrm{mL})=12\ \mathrm{ng}\). The amount on the buffer side is \((30\ \mathrm{ng/mL})(0.20\ \mathrm{mL})=6\ \mathrm{ng}\). Thus:

\text{Recovery}=\frac{12\ \mathrm{ng}+6\ \mathrm{ng}}{24\ \mathrm{ng}} =\frac{18}{24}=0.750

The apparent fraction unbound is 0.500, while recovery is 75.0%. These are different diagnostics: the first is a concentration ratio, and the second is an amount balance. The fraction unbound should not be accepted uncritically just because it is between zero and one. Investigate missing mass, membrane or apparatus binding, degradation, precipitation, and whether equilibrium was reached. If the protocol's validated recovery criteria are not met, the binding estimate may be unreliable; recovery alone does not supply a corrected value for \(f_{u,p}\).

Keep tissue dilution correction separate from recovery

Tissue binding assays commonly measure fraction unbound in a diluted homogenate. Dilution reduces the concentration of binding components, so the measured homogenate fraction unbound is generally higher than the value expected in undiluted tissue. If the tissue is diluted by a factor \(D\), where \(D\) is the homogenate volume divided by the original tissue volume, the standard correction is:

f_{u,t}=\frac{1}{D\left(\frac{1}{f_{u,\mathrm{hom}}}-1\right)+1}

This correction assumes that dilution changes binding-site concentration as specified and that the measured homogenate fraction unbound is valid. It is not a correction for drug lost during processing. Recovery and dilution answer different questions: recovery asks whether measured amounts account for the amount added; dilution correction translates a valid binding measurement from diluted homogenate toward undiluted tissue.

Worked Example: Correcting a Tissue Homogenate Result

A hypothetical tissue homogenate is prepared at a fourfold dilution, so \(D=4\), and the measured homogenate fraction unbound is \(f_{u,\mathrm{hom}}=0.400\). Calculate the corrected tissue fraction unbound.

f_{u,t} =\frac{1}{4\left(\frac{1}{0.400}-1\right)+1} =\frac{1}{4(2.50-1)+1} =\frac{1}{6+1} =0.1429

The corrected value is 0.1429, or 14.29%. Reporting 0.400 as the undiluted tissue fraction unbound would overstate the value under the stated dilution model. If the assay also has poor recovery, the correction formula does not restore the missing drug; first establish whether the homogenate measurement is interpretable.

A practical sequence for diagnosing disagreement

When results conflict, use a sequence that preserves the distinction between a calculation error and an experimental disagreement. A single discrepancy may have several explanations, so document what each check establishes and what it cannot establish.

Write down each measurement basis.
Record matrix, concentration units, volume basis, total or unbound status, species, sample conditions, and whether the measurements were paired.

Check identities that link measurements.
For matched blood data, compare \(R_{b:p}\) with \((1-Hct)+Hct\,K_{RBC:p}\). Recalculate from the original concentrations if possible.

Inspect assay recovery and method conditions.
Reconstruct recovered amounts using concentrations and compartment volumes. Review equilibrium, stability, nonspecific binding, and sample handling.

Apply only the relevant transformations.
Use the homogenate dilution correction for tissue binding when its assumptions hold. Do not treat it as a recovery adjustment or use one matrix's fraction unbound as another's.

Classify what remains unresolved.
After basis and method checks, retain a discrepancy as unresolved measurement or biological variation rather than choosing an arbitrary value or forcing exact agreement.

The most useful comparison is often not between the final reported parameters, but between the underlying measured quantities and their transformations. For example, a blood-to-plasma ratio can be compared with a red-cell ratio only after accounting for hematocrit. A tissue homogenate result can be compared with an undiluted tissue estimate only after the dilution basis is clear. A plasma fraction unbound cannot be judged from a reported percentage without knowing whether it is an assay result, a model-based estimate, or a value altered by low recovery.

Common mistakes, alternatives, and limitations

Comparing concentrations on different volume bases. A packed-cell concentration is not a whole-blood concentration. Use the hematocrit-weighted identity rather than comparing the raw ratios as if the denominators were identical.
Using a different hematocrit from the one paired with the blood sample. The calculation may then show an apparent disagreement even if the measurements are otherwise compatible. Use matched sample data where possible.
Calling low recovery a binding correction. Low recovery warns of missing mass; it does not indicate how much drug was unbound. Investigate the source of loss or qualify the result.
Applying tissue dilution correction to a plasma assay. The correction is specific to diluted tissue homogenate under its assumptions. It does not reconcile unrelated assays.
Assuming a mass-balance identity proves assay validity. A set of numbers can satisfy the identity and still share a systematic error. Orthogonal measurements, such as a separately validated whole-blood assay, can help identify method-specific problems.

Two broad strategies are available when data disagree. One is a measurement-first audit: inspect raw concentrations, sample volumes, recovery, hematocrit, and assay conditions, then recompute each quantity. This is usually the clearest first approach because it tests whether the reported quantities are being compared correctly. Another is to fit or simulate a PBPK model using alternative candidate inputs and examine their consequences. That can show whether a difference matters for predictions, but a good model fit cannot establish which assay is correct; compensating errors may also obscure the source of conflict.

These checks have limits. The blood identity is exact only for compatible measurements and the stated compartment definitions. Real assays have uncertainty, and blood samples may not be at equilibrium. Plasma and red-cell measurements may differ because of temperature, pH, anticoagulant, incubation time, or drug instability. Likewise, tissue binding can vary by tissue preparation and assay conditions. The equations identify what must agree under stated conditions; they do not explain every disagreement or replace method validation.

Key takeaway: Reconcile concentration bases before comparing binding inputs. Use hematocrit-weighted mass balance for blood and red-cell data, evaluate assay recovery independently of fraction unbound, and apply tissue dilution correction only to valid diluted-homogenate measurements.

Check Your Understanding

Work through each question, then compare with the solutions below.

A paired sample has \(Hct=0.40\) and \(K_{RBC:p}=1.50\). What blood-to-plasma ratio is implied?
A measured \(R_{b:p}=1.30\) is reported at \(Hct=0.50\). What \(K_{RBC:p}\) is implied by mass balance?
An assay starts with 10 ng and measures 3 ng and 5 ng in its two compartments. What is the recovery? Does that number alone determine fraction unbound?
A tissue homogenate has \(f_{u,\mathrm{hom}}=0.25\) at a twofold dilution. Calculate the corrected \(f_{u,t}\).
Why should a low-recovery plasma binding result not be corrected using the tissue homogenate dilution equation?

Solutions. (1) \(R_{b:p}=(1-0.40)+(0.40)(1.50)=0.60+0.60=1.20\). (2) \(K_{RBC:p}=[1.30-(1-0.50)]/0.50=0.80/0.50=1.60\). (3) Recovery is \((3+5)/10=0.800\), or 80.0%. It does not determine fraction unbound, which requires appropriate concentration measurements and assay assumptions. (4) \(f_{u,t}=1/[2(1/0.25-1)+1]=1/[2(4-1)+1]=1/7=0.1429\). (5) The tissue equation corrects for dilution of tissue binding components; low recovery indicates unaccounted drug amount. They address different processes, so the cause of missing mass must be investigated separately.