Choosing Inhibitor Concentrations for Static Screening

Why the Concentration Choice Matters

A static drug–drug interaction (DDI) screen combines an inhibitor concentration with an inhibition constant to estimate whether the inhibitor could reduce victim-drug metabolism. The arithmetic is simple; choosing the concentration is not. Total plasma concentration, unbound plasma concentration, hepatic-inlet concentration, and intestinal concentration represent different exposure approximations, and they do not answer precisely the same question.

For reversible inhibition, the concentration should represent inhibitor available at the enzyme site and use the same concentration basis as the inhibition constant. A plasma total concentration is not automatically an appropriate substitute for an unbound concentration, and a high nominal intestinal concentration is not automatically the free concentration inside an enterocyte. Static screening is useful for prioritization, but the selected input and its assumptions must remain visible in the interpretation.

Definition: The inhibitor concentration input, \(I\), is the concentration used in a static inhibition calculation. It should approximate inhibitor exposure at the relevant enzyme site and be on a concentration basis compatible with the inhibition constant, \(K_i\). The unbound fraction in plasma, \(f_{u,p}\), is the fraction of measured total plasma inhibitor concentration that is unbound. Under a simple equilibrium assumption, \(I_u=f_{u,p}I_{\mathrm{total}}\).

Concentration Choices and Their Meaning

A total systemic concentration, often a maximum plasma concentration, is readily observed or predicted. It includes both protein-bound and unbound inhibitor. If inhibition potency is expressed as an unbound \(K_i\), using total plasma concentration directly overstates the concentration available to interact with the enzyme whenever \(f_{u,p}<1\). Total concentration can be used only when the potency and concentration are expressed on a compatible basis and that basis is appropriate to the intended model.

The unbound systemic concentration is commonly estimated as \(f_{u,p}C_{\max,\mathrm{total}}\). It is a more direct approximation to freely available inhibitor in systemic plasma, but it is not necessarily the concentration at every enzyme site. It may be an adequate screening input when systemic unbound exposure is a reasonable proxy for exposure at the enzyme, and when binding is not so uncertain that the conversion becomes unreliable.

For orally administered inhibitors, portal blood can receive newly absorbed drug before the liver has extracted it. A simplified maximum hepatic-inlet estimate therefore adds a portal contribution to a systemic maximum. One common screening approximation is

I_{\mathrm{in,max,total}} = C_{\max,\mathrm{total}} + \frac{F_aF_gk_a\,\mathrm{Dose}}{Q_h}.

Here \(F_a\) is the fraction absorbed, \(F_g\) is the fraction escaping intestinal metabolism, \(k_a\) is a first-order absorption-rate constant, and \(Q_h\) is hepatic blood flow. With dose in amount units, \(k_a\) in inverse time, and \(Q_h\) in volume per time, the added term has units of amount per volume. If the calculation assumes the same plasma binding in portal and systemic blood, the estimated unbound inlet concentration is \(f_{u,p}I_{\mathrm{in,max,total}}\). This is a screening approximation, not a resolved concentration-time profile; in particular, it can be conservative because the systemic maximum and the portal absorption contribution need not peak simultaneously.

An intestinal input is intended to represent exposure near intestinal enzymes. A simple conservative estimate for an oral dose is the dose divided by an assumed luminal volume, often 250 mL in screening conventions. This gives a nominal concentration, not a measured free enterocyte concentration. Dissolution, precipitation, intestinal transit, uptake, binding, and metabolism can all make local enzyme exposure differ from that estimate.

Candidate input	What it represents	Key caution
Total systemic concentration	Bound plus unbound inhibitor in systemic plasma	Do not compare it directly with an unbound \(K_i\) as if all inhibitor were available.
Unbound systemic concentration	Unbound inhibitor in systemic plasma	May not represent a portal or intestinal enzyme site.
Hepatic-inlet concentration	Systemic concentration plus an estimated portal contribution	Depends on absorption and flow assumptions and may be a conservative maximum.
Intestinal concentration	A local luminal or enterocyte exposure approximation	A nominal dose-based value is not automatically an unbound enzyme-site concentration.

Static Reversible-Inhibition Prediction

For a simple hepatic clearance screen, let \(f_m\) be the fraction of the victim drug's baseline clearance mediated by the inhibited pathway. Assume that inhibition is reversible and competitive, that the victim has linear pharmacokinetics, that the inhibitor concentration is represented by one fixed screening value, and that other clearance pathways are unchanged. Use an unbound inhibitor concentration \(I_u\) with an unbound inhibition constant \(K_{i,u}\), or use another consistently defined pair of concentration and potency measures.

Under these assumptions, inhibition reduces the affected pathway's clearance to \(1/(1+I_u/K_{i,u})\) of its baseline value. The remaining fraction \(1-f_m\) is unchanged. The predicted clearance ratio is therefore the sum of the unchanged and inhibited contributions. For the same victim dose and linear pharmacokinetics, AUC is inversely proportional to clearance, so the predicted AUCR is the reciprocal:

\mathrm{AUCR} = \frac{\mathrm{AUC}_{\mathrm{with\ inhibitor}}} {\mathrm{AUC}_{\mathrm{without\ inhibitor}}} = \frac{1} {(1-f_m)+\dfrac{f_m}{1+I_u/K_{i,u}}}.

This equation makes the influence of concentration explicit. Increasing \(I_u/K_{i,u}\) increases the predicted AUCR, while the effect is limited when \(f_m\) is small because the unaffected clearance fraction remains. It is a screening calculation for the specified clearance pathway; it does not combine hepatic and intestinal effects or provide a time-dependent prediction.

Derivation: Normalize baseline victim clearance to 1. The inhibited pathway contributes \(f_m\), and the unaffected pathways contribute \(1-f_m\). With reversible inhibition, the affected contribution becomes \(f_m/(1+I_u/K_{i,u})\). Thus the clearance ratio is \((1-f_m)+f_m/(1+I_u/K_{i,u})\). At fixed dose under linear pharmacokinetics, \(\mathrm{AUC}\propto 1/\mathrm{CL}\), so the AUC ratio is the reciprocal of this clearance ratio.

Worked Example: Converting Total Systemic to Unbound Concentration

Suppose a hypothetical inhibitor has a total systemic \(C_{\max}\) of \(1.20\ \mu\mathrm{M}\) and \(f_{u,p}=0.10\). Its relevant unbound inhibition constant is \(K_{i,u}=0.50\ \mu\mathrm{M}\). The victim's \(f_m\) for the affected pathway is 0.70. Calculate the unbound concentration and the static AUCR.

First convert the concentration:

I_u=f_{u,p}C_{\max,\mathrm{total}} =(0.10)(1.20\ \mu\mathrm{M}) =0.120\ \mu\mathrm{M}.

Then \(I_u/K_{i,u}=0.120/0.50=0.240\). Substitution gives

\mathrm{AUCR} = \frac{1}{(1-0.70)+\dfrac{0.70}{1+0.240}} = \frac{1}{0.300+\dfrac{0.70}{1.240}} = \frac{1}{0.864516} = 1.157.

Under the stated assumptions, this screening calculation predicts an AUC about 1.157 times the baseline AUC, or a 15.7% increase. It is not a probability statement and does not establish that the clinical interaction will have exactly this magnitude.

How an Inlet Input Changes the Screening Result

Worked Example: Comparing Systemic and Hepatic-Inlet Inputs

Use the hypothetical inhibitor and victim from the previous calculation. Suppose \(C_{\max,\mathrm{total}}=1.20\ \mu\mathrm{M}\), \(f_{u,p}=0.10\), \(F_a=0.90\), \(F_g=0.80\), \(k_a=1.00\ \mathrm{h}^{-1}\), dose \(=100\ \mu\mathrm{mol}\), and \(Q_h=90\ \mathrm{L/h}\). Estimate a maximum total hepatic-inlet concentration and compare its unbound-input AUCR with the systemic unbound-input result.

The portal contribution is

\frac{F_aF_gk_a\,\mathrm{Dose}}{Q_h} = \frac{(0.90)(0.80)(1.00\ \mathrm{h}^{-1})(100\ \mu\mathrm{mol})} {90\ \mathrm{L/h}} = 0.800\ \mu\mathrm{mol/L} = 0.800\ \mu\mathrm{M}.

The approximate maximum total inlet concentration is \(1.20+0.800=2.00\ \mu\mathrm{M}\). Applying the assumed \(f_{u,p}=0.10\) gives \(I_{\mathrm{in,max},u}=0.200\ \mu\mathrm{M}\). With \(K_{i,u}=0.50\ \mu\mathrm{M}\), the ratio is \(0.200/0.50=0.400\), and

\mathrm{AUCR}_{\mathrm{inlet}} = \frac{1}{0.300+\dfrac{0.700}{1+0.400}} = \frac{1}{0.300+0.500} = 1.250.

The systemic unbound input gave 1.157, whereas this approximate hepatic-inlet input gives 1.250. The difference follows from the larger assumed unbound concentration at the liver inlet, not from a change in \(f_m\) or \(K_{i,u}\). The estimate should be interpreted in light of the simplified absorption and binding assumptions, including the assumption that the peak terms can be added.

Intestinal Inputs Are Local Screening Signals

Worked Example: A Dose-Based Intestinal Estimate

Consider a hypothetical oral inhibitor dose of \(1.00\ \mathrm{mg}\) with molecular weight \(500\ \mathrm{g/mol}\). Estimate its nominal concentration in 250 mL. Suppose a separate hypothetical assessment estimates the unbound concentration near the intestinal enzyme as \(1.60\ \mu\mathrm{M}\), and a matched unbound \(K_{i,u}\) is \(0.500\ \mu\mathrm{M}\). Calculate a local inhibition ratio.

The dose is \(0.00100\ \mathrm{g}\), so the amount is \(0.00100\ \mathrm{g}/(500\ \mathrm{g/mol})=2.00\times10^{-6}\ \mathrm{mol}=2.00\ \mu\mathrm{mol}\). Dividing by \(0.250\ \mathrm{L}\) gives a nominal concentration of

C_{\mathrm{gut,nominal}} = \frac{2.00\ \mu\mathrm{mol}}{0.250\ \mathrm{L}} = 8.00\ \mu\mathrm{M}.

That \(8.00\ \mu\mathrm{M}\) is not the unbound enzyme-site concentration. For the local calculation, use the separately supplied, matched estimate \(I_{\mathrm{gut},u}=1.60\ \mu\mathrm{M}\). The simple reversible-inhibition ratio is

R_{\mathrm{gut}} = 1+\frac{I_{\mathrm{gut},u}}{K_{i,u}} = 1+\frac{1.60}{0.500} = 4.20.

This ratio indicates potential local inhibition under the assumed conditions. It is not an AUCR of 4.20: translating intestinal enzyme inhibition into a systemic victim AUC change requires information about intestinal metabolism and the victim's disposition, not just this local ratio.

Diagnostics, Alternatives, and Limitations

Before interpreting a screen, check that the concentration and potency have compatible units and binding conventions. A concentration in \(\mu\mathrm{M}\) cannot be compared directly with a \(K_i\) in \(\mathrm{mg/L}\) without conversion. Likewise, a total plasma concentration divided by an unbound \(K_i\) is not a consistent free-concentration comparison unless a justified conversion has been made. If \(f_{u,p}\) is uncertain, report how that uncertainty changes the unbound input rather than presenting a falsely precise result.

The systemic unbound input is often the clearest simple choice for a systemic enzyme screen. The hepatic-inlet estimate is an alternative when presystemic portal exposure could matter and the necessary absorption and flow inputs are available. The intestinal dose-over-volume estimate is a deliberately simplified local screen; an estimate of unbound enterocyte exposure is more site-relevant when it is supported. These alternatives are not a menu of interchangeable values. Choose according to enzyme location and available evidence, and keep hepatic and intestinal conclusions distinct unless a model explicitly links them.

A static calculation also compresses changing inhibitor exposure into one value. It cannot describe the timing of inhibition, the full concentration profile, or uncertainty caused by absorption and distribution. Large predicted ratios should prompt scrutiny of concentration assumptions and model scope, not automatic acceptance as an expected clinical effect. A more detailed organ-specific or time-varying analysis may be needed when those assumptions determine the decision.

Pitfall: using total plasma concentration with an unbound \(K_i\). The incorrect reasoning is that all measured inhibitor is available to bind the enzyme. The correction is to convert to unbound concentration when appropriate, or use a potency measure defined on the same justified total-concentration basis.
Pitfall: calling the dose-over-volume value an intestinal free concentration. It is a nominal screening estimate. The correction is to label it as such and avoid treating it as enterocyte exposure without supporting assumptions or data.
Pitfall: interpreting a local gut ratio as a whole-body AUCR. The ratio describes a local inhibition signal only. The correction is to use an appropriate model of victim disposition before drawing a systemic exposure conclusion.

Check Your Understanding

Answer each question, then compare your reasoning with the brief solutions.

A hypothetical inhibitor has \(C_{\max,\mathrm{total}}=0.750\ \mu\mathrm{M}\) and \(f_{u,p}=0.20\). What is its estimated unbound systemic concentration?
For a victim with \(f_m=0.60\), \(I_u=0.300\ \mu\mathrm{M}\), and \(K_{i,u}=0.300\ \mu\mathrm{M}\), calculate the static AUCR.
A hypothetical \(2.00\ \mathrm{mg}\) dose has molecular weight \(400\ \mathrm{g/mol}\). What nominal concentration results from dividing the dose by 250 mL?
Why is a high intestinal inhibition ratio not, by itself, a prediction of the same-sized increase in systemic victim AUC?

Solutions. (1) \(I_u=(0.20)(0.750)=0.150\ \mu\mathrm{M}\). (2) \(I_u/K_{i,u}=1\), so \(\mathrm{AUCR}=1/[0.40+0.60/2]=1/0.70=1.429\). (3) \(0.00200\ \mathrm{g}/(400\ \mathrm{g/mol})=5.00\ \mu\mathrm{mol}\); \(5.00\ \mu\mathrm{mol}/0.250\ \mathrm{L}=20.0\ \mu\mathrm{M}\). (4) It is a local inhibition signal; systemic AUC also depends on the victim's intestinal contribution and whole-body disposition.