Worked Example of Cross-Species Microsomal IVIVE

From an Animal Microsomal Assay to a Human Estimate

Microsomal intrinsic clearance is measured per amount of microsomal protein, whereas a liver-scale estimate depends on the amount of microsomal protein in the liver. Moving between these scales requires more than multiplying by a liver mass: the measured rate may need correction for substrate binding in the incubation, and animal and human microsomes may not have the same activity per milligram of protein. Each factor should be visible in the calculation so that the resulting human estimate can be interpreted appropriately.

This tutorial focuses on an estimate of hepatic intrinsic clearance, not systemic clearance. The calculation does not by itself account for plasma binding, hepatic blood flow, or other processes needed to predict whole-body pharmacokinetics. It also does not make an animal assay a direct measurement of human enzyme activity. Instead, it combines an animal measurement with stated cross-species and scaling assumptions.

The cross-species ratio is essential to interpret. If paired human and animal microsomal measurements are available, it can be estimated as \(R_{H/A}=CL_{\mathrm{int,u},H,\mathrm{per\,mg}}/CL_{\mathrm{int,u},A,\mathrm{per\,mg}}\). If no such comparative measurement is available, setting \(R_{H/A}=1\) is an assumption that unbound activity per milligram is equal across species; it is not an empirical finding. The incubation binding correction for the animal assay does not establish human incubation binding or human microsomal activity.

Why the Scaling Formula Has These Factors

The calculation proceeds in distinct steps. First, correct the measured animal clearance for the fraction of substrate available unbound in the incubation. Next, apply the cross-species ratio to represent the assumed or measured change in unbound activity per milligram of protein. Finally, scale protein-normalized clearance by the amount of microsomal protein in a human liver.

The units show why the last multiplication is appropriate: \[ \frac{\mathrm{mL}}{\mathrm{min}\cdot\mathrm{mg}} \times \frac{\mathrm{mg}}{\mathrm{g}} \times \mathrm{g} = \frac{\mathrm{mL}}{\mathrm{min}}. \] Here \(MPPGL_H\) is in mg microsomal protein/g liver and \(LW_H\) is in g liver. The ratio \(R_{H/A}\) and unbound fraction are dimensionless. This derivation also shows why using liver mass in kilograms without converting units, or applying the protein content twice, produces an incorrect result.

Worked Cross-Species Estimate

Alternative Scaling-Factor Choices

Microsomal protein content and liver mass are inputs, not universal constants. Published or laboratory-specific values can differ because of the source population, tissue preparation, measurement method, or convention used to define the scaling factor. The following comparison holds the binding correction and cross-species ratio from the worked example fixed, so that only human \(MPPGL_H\) and liver mass change.

A second comparison isolates the effect of the incubation binding correction. With all other values from the first example fixed, suppose the measured animal \(f_{\mathrm{u,inc},A}\) were 0.30 rather than 0.60. The binding-corrected animal value would be \(0.0120/0.30=0.0400\) mL/min/mg. Applying \(R_{H/A}=0.75\), \(MPPGL_H=45.0\) mg/g, and \(LW_H=1500\) g gives: \[ 0.0400\times0.75\times45.0\times1500 =2025\ \mathrm{mL/min} =2.025\ \mathrm{L/min}. \] Thus, halving the measured unbound fraction doubles the corrected estimate, if the other inputs remain fixed. This arithmetic is a sensitivity comparison, not a reason to substitute an unmeasured binding value.

Assumptions, Diagnostics, and Limits

The scaling calculation is interpretable only if its inputs correspond to the assay and species being modeled. Before using it, verify that apparent clearance is normalized to microsomal protein, that the incubation unbound fraction was measured or justified under the relevant assay conditions, and that the cross-species ratio compares compatible unbound-clearance quantities. The \(MPPGL_H\) and \(LW_H\) values must use the units in the formula and refer to a consistent human liver scaling convention.

• Check the assay scale. A value already expressed per gram of liver is not a per-milligram microsomal clearance and should not be multiplied by \(MPPGL_H\) again.
• Check the binding correction. Because \(0<f_{\mathrm{u,inc},A}\leq1\), the corrected value cannot be smaller than the apparent value under this correction. A very small measured fraction can produce a large estimate, so inspect the measurement and its uncertainty rather than treating the correction as exact.
• Check the species bridge. Confirm whether \(R_{H/A}\) is measured from comparable incubations and whether both values were corrected consistently. If it is set to one without comparative data, label the equality-of-activity assumption explicitly.
• Check the units and scale. If using a liver mass in kilograms, convert it to grams before multiplying by \(MPPGL_H\) in mg/g. Keep total-liver estimates distinct from estimates normalized per kilogram of body weight.
• Check plausibility without over-interpreting it. A high intrinsic-clearance estimate can flag a need to review the inputs and assumptions, but it is not by itself a prediction of high systemic clearance. The calculation has not incorporated other determinants of in vivo disposition.

The introductory material on incubation binding and microsomal scaling establishes why the assay value and the scaled value are not interchangeable. Here, the important additional issue is cross-species transfer. The main alternative to applying an animal-based estimate is to use human microsomal data directly, with human-specific binding and scaling inputs where available. That avoids assuming animal activity represents human activity, although it does not eliminate assay or factor uncertainty. A second option is to retain the animal assay but use a measured comparative ratio; this is more informative than assuming \(R_{H/A}=1\), provided the paired measurements are comparable. When alternatives are unavailable, report scenario estimates for plausible factors rather than presenting one assumption-dependent number as definitive.

Common Mistake

A frequent error is to apply \(MPPGL_H\) and human liver mass directly to the uncorrected animal assay value, then describe the result as a human estimate. This omits incubation binding and silently assumes that animal and human unbound activity per milligram are equal. The corrected approach is to divide by the animal incubation unbound fraction, apply an explicit \(R_{H/A}\) (including a clearly labeled value of one if assumed), and then multiply by human protein content and liver mass. The result remains an intrinsic-clearance estimate subject to those assumptions, not a direct human observation.