Modeling Perpetrators with Both Inhibitory and Inductive Effects

Why One Perpetrator Can Have Two Opposing Effects

A perpetrator drug may inhibit a CYP enzyme quickly while also increasing the amount of that enzyme more slowly. The victim substrate can therefore experience inhibition soon after perpetrator exposure rises, followed by a changing balance as enzyme abundance responds. After repeated dosing, perpetrator concentration and enzyme abundance need not peak or decline together. A single “net interaction” value can conceal this time dependence.

The useful modeling distinction is between enzyme activity per unit of enzyme and how much enzyme is present. Reversible inhibition changes the former directly; induction changes the latter over time. This tutorial combines those processes for one CYP pathway. It assumes that the perpetrator concentration at the site of CYP effect is available as an input profile; it does not specify how that site concentration is obtained.

Definition: Let $C(t)$ be the unbound perpetrator concentration at the CYP effect site, $E(t)$ the enzyme abundance relative to its baseline value, $K_i$ the inhibition constant, $E_{\max}$ the maximum induction increase above baseline, $EC_{50}$ the concentration producing half of that induction increase, and $k_{\mathrm{deg}}$ the first-order enzyme degradation rate constant. Under a reversible competitive-inhibition model and an indirect-response induction model, the CYP activity relative to baseline is $$ R_{\mathrm{CYP}}(t)=\frac{E(t)}{1+C(t)/K_i}, $$ and enzyme abundance changes according to $$ \frac{dE(t)}{dt}=k_{\mathrm{deg}}\left[1+\frac{E_{\max}C(t)}{EC_{50}+C(t)}-E(t)\right]. $$

Here $E(0)=1$ means baseline enzyme abundance, and $R_{\mathrm{CYP}}=1$ means baseline pathway activity. In the inhibition term, $K_i$ and $C(t)$ must use the same concentration units. The induction target is $1+E_{\max}C/(EC_{50}+C)$: $E_{\max}$ is the maximum fractional increase above baseline, not the maximum total enzyme abundance. The degradation rate and half-life are related by $k_{\mathrm{deg}}=\ln(2)/t_{1/2,\mathrm{enzyme}}$.

Combining the Processes Without Losing Their Time Courses

The combined activity expression uses the same enzyme abundance for both processes: the induced enzyme is still subject to reversible inhibition. It is not generally correct to calculate a net effect by subtracting an “inhibition percentage” from an “induction percentage.” Under the stated competitive model, inhibition scales catalytic activity by $1/(1+C/K_i)$; induction scales enzyme abundance by $E(t)$. Their effects therefore multiply.

The induction equation is a first-order response toward a concentration-dependent target. If $C(t)=C$ is constant over an interval, define the target abundance $E_{\mathrm{ss}}=1+E_{\max}C/(EC_{50}+C)$. Solving the differential equation gives:

E(t)=E_{\mathrm{ss}}+\left[E(0)-E_{\mathrm{ss}}\right]e^{-k_{\mathrm{deg}}t}.

The solution starts at the abundance present at the beginning of the interval and approaches the target exponentially. After one enzyme half-life, it has completed half the distance from its starting abundance to that target. If concentration changes between dosing phases, the target changes too, but $E(t)$ remains continuous: enzyme abundance does not jump instantly when concentration changes.

Worked Example: Immediate Inhibition and Delayed Induction

Suppose a hypothetical perpetrator is held at $C=2$ concentration units. Let $K_i=1$ concentration unit, $EC_{50}=1$ concentration unit, $E_{\max}=4$, and the enzyme half-life be 24 hours. Initially, $E(0)=1$. Calculate CYP activity immediately, after 24 hours, and at the eventual constant-concentration steady state.

Immediate effect. Enzyme abundance has not yet changed, so $E(0)=1$, while the inhibition multiplier is $1/(1+2/1)=1/3$. Thus $R_{\mathrm{CYP}}(0)=1/3\approx0.3333$: activity is one-third of baseline.

Induction target. At $C=2$, the target abundance is $$ E_{\mathrm{ss}}=1+\frac{4(2)}{1+2}=1+\frac{8}{3}=\frac{11}{3}. $$ Because 24 hours is one enzyme half-life, the abundance after 24 hours is halfway from 1 to $11/3$: $$ E(24)=\frac{11}{3}+\left(1-\frac{11}{3}\right)\frac{1}{2} =\frac{7}{3}\approx2.333. $$ The activity at that time is $$ R_{\mathrm{CYP}}(24)=\frac{7/3}{1+2/1}=\frac{7}{9}\approx0.7778. $$ The fraction $7/9$ is retained through the calculation; rounding enzyme abundance first and then dividing can produce an avoidable rounding discrepancy.

Long-term steady state. If the same concentration is maintained long enough, abundance approaches $11/3$, giving $$ R_{\mathrm{CYP,ss}}=\frac{11/3}{3}=\frac{11}{9}\approx1.222. $$ Thus the same perpetrator produces net inhibition initially and net induction eventually in this hypothetical example. The result depends on the specified parameters and sustained concentration; it is not a general prediction about perpetrators.

Repeated Dosing: Enzyme Abundance Carries Over Between Doses

For repeated dosing, the perpetrator concentration may rise and fall within each dosing interval. At every time point, the reversible inhibition term responds to the current concentration in this model. Enzyme abundance instead reflects the preceding concentration history. A convenient way to calculate $E(t)$ is to solve the constant-concentration equation separately over each interval during which $C(t)$ is approximated as constant, using the preceding interval’s ending abundance as the next interval’s starting abundance.

The calculation can also be written as a recurrence. For an interval of duration $\Delta t$ with constant target $E_{\mathrm{ss}}$, let $a=e^{-k_{\mathrm{deg}}\Delta t}$. Then:

E_{\mathrm{end}}=E_{\mathrm{ss}}+(E_{\mathrm{start}}-E_{\mathrm{ss}})a.

This recurrence makes the carryover explicit. It is also the basis for finding a periodic enzyme-abundance profile: apply the interval updates across one whole dosing interval, then choose the starting abundance for which the end-of-interval abundance equals the start.

Worked Example: A Two-Phase Repeated-Dose Profile

Use the hypothetical parameters from the previous example. Suppose a 24-hour dosing interval is approximated by $C=2$ concentration units for the first 8 hours and $C=0.5$ concentration units for the next 16 hours. The corresponding induction targets are $E_{\mathrm{ss,H}}=11/3$ and $E_{\mathrm{ss,L}}=7/3$. With a 24-hour enzyme half-life, the interval multipliers are $a=e^{-(\ln 2)(8/24)}=0.7937$ and $b=e^{-(\ln 2)(16/24)}=0.6300$.

To find the repeating profile, call the abundance at the start of the high-concentration phase $E_0$. After the high phase and then the low phase, the abundance is $$ E_{\mathrm{next}}=\frac{7}{3}+\left[\frac{11}{3}+\left(E_0-\frac{11}{3}\right)a-\frac{7}{3}\right]b. $$ At periodic steady state $E_{\mathrm{next}}=E_0$. Solving this linear equation gives $$ E_0= \frac{\frac{7}{3}(1-b)+b\frac{11}{3}(1-a)}{1-ab} \approx2.680. $$ Here $ab=e^{-k_{\mathrm{deg}}24}=0.5$, as expected for a full 24-hour interval. Applying the high-phase update gives $$ E_{\mathrm{high,end}}=\frac{11}{3}+\left(2.680-\frac{11}{3}\right)(0.7937)\approx2.884. $$

At the start of the high phase, the activity is approximately $2.680/(1+2)=0.8933$ of baseline. At the end of that phase it is approximately $2.884/3=0.9613$. When concentration falls to 0.5, inhibition is weaker, and activity rises immediately to approximately $2.884/(1+0.5)=1.923$. By the end of the low phase, abundance has returned to approximately 2.680, so activity is approximately $2.680/1.5=1.787$. The perpetrator’s concentration changes immediately at the phase boundary in this approximation; enzyme abundance does not. The shift in inhibition plus the slower abundance response explains why activity varies across the interval.

From CYP Activity to Victim-Substrate Clearance

If the affected CYP pathway accounts for fraction $f_m$ of the victim’s baseline total clearance, and all other clearance pathways remain unchanged, the total clearance ratio at a particular time is $$ \frac{CL(t)}{CL_0}=(1-f_m)+f_mR_{\mathrm{CYP}}(t). $$ Here $CL_0$ is baseline total clearance. This expression weights the affected pathway’s activity by its baseline contribution; it does not assume that the entire victim clearance is mediated by that CYP.

Worked Example: Clearance Implications at Acute and Induced States

Suppose a hypothetical victim has $f_m=0.80$ through the CYP pathway in the first worked example. Immediately after the perpetrator concentration reaches 2 units, $R_{\mathrm{CYP}}=1/3$. The total clearance ratio is $$ \frac{CL}{CL_0}=(1-0.80)+0.80\left(\frac{1}{3}\right) =0.20+0.2667=0.4667. $$ At the constant-concentration induction steady state, $R_{\mathrm{CYP}}=11/9$, so $$ \frac{CL}{CL_0}=0.20+0.80\left(\frac{11}{9}\right) =1.1778. $$

If victim pharmacokinetics are linear and bioavailability is unchanged, AUC over a dosing interval at steady state is inversely proportional to total clearance. Under those additional conditions, the long-term AUC ratio for the constant-concentration state would be $1/1.1778\approx0.8490$, a decrease of about 15.10% from baseline. The acute clearance ratio alone is not a dynamic AUC prediction for the transition period: abundance is changing, and a valid AUC comparison must account for the time-varying disposition across the relevant interval.

Assumptions, Diagnostics, and Model Choices

The combined model is useful when the same perpetrator concentration drives both reversible inhibition and induction of the same CYP pathway, but it rests on explicit assumptions. The inhibition expression represents reversible competitive inhibition with a concentration-independent $K_i$; the induction target follows a saturable $E_{\max}$ relationship; enzyme abundance turns over with a single first-order rate constant; and the victim’s CYP-mediated clearance is proportional to that abundance and activity. The simple pathway-clearance calculation additionally assumes that other clearance pathways are unaffected. These are model choices to test against evidence, not universal properties of all drug interactions.

Useful diagnostics focus on whether the two time courses are being represented plausibly. Check that $E(t)$ remains continuous at concentration changes, moves toward the current induction target, and does not exceed the target during a constant-concentration interval when starting below it. Check that $R_{\mathrm{CYP}}(t)$ responds to concentration changes through the inhibition term even when $E(t)$ changes little. Simulated activity that changes instantaneously with concentration through both inhibition and abundance, or that remains fixed despite a changing perpetrator profile, signals a likely implementation error.

A static calculation can be a reasonable alternative when concentrations and enzyme abundance are effectively constant over the question of interest. It is simpler, but cannot describe delayed induction or changing inhibition within a dosing interval. A time-varying model with the separate mechanisms is more informative when onset, offset, or repeated dosing matters, but it requires a concentration-time input and kinetic parameters. A single empirical net multiplier is simpler still, but can obscure why the interaction changes direction over time and may fail when the dosing schedule changes.

The model also has limits. If inhibition and induction occur in different tissues, involve different CYP enzymes, or depend on different local concentrations, a single shared $C(t)$ and pathway factor may be inappropriate. A single enzyme-turnover rate may not capture multiple response times, and the assumed induction function may not describe the available concentration range. Uncertainty in concentration profiles and parameters can materially change both the magnitude and timing of predicted effects. These limitations should be evaluated before interpreting a predicted net DDI as a reliable clinical outcome.

Common pitfall: Do not add an inhibition percentage to an induction percentage, or apply an instantaneous induction multiplier to represent delayed enzyme turnover. In this model, reversible inhibition scales the activity of the enzyme present, while induction changes $E(t)$ according to its turnover equation. Calculate $E(t)$ first, then combine it with inhibition using $R_{\mathrm{CYP}}(t)=E(t)/(1+C(t)/K_i)$.

Check Your Understanding

Answer each question using the stated model assumptions; brief solutions follow.

If $E(t)=2$, $C(t)=K_i$, and both values are in compatible concentration units where required, what is $R_{\mathrm{CYP}}(t)$?
Under a constant concentration, what happens to $E(t)$ as time becomes long?
Why does enzyme abundance not jump when perpetrator concentration changes between dosing phases?
If $f_m=0.60$ and the CYP activity ratio is 0.50, what is the total clearance ratio, assuming other pathways are unchanged?

Solutions. (1) $R_{\mathrm{CYP}}=2/(1+1)=1$, equal to baseline activity. (2) It approaches the concentration-dependent target $1+E_{\max}C/(EC_{50}+C)$. (3) The differential equation describes finite first-order turnover, so $E(t)$ is continuous even when its target changes. (4) The ratio is $0.40+0.60(0.50)=0.70$, or 70% of baseline total clearance.