Steady-State DDI Designs and Accumulation

Why Steady State Matters in a Repeated-Dose DDI Study

When a drug is administered repeatedly, each dose may be given before the preceding dose has been completely eliminated. The resulting concentrations reflect both the latest dose and residual drug from earlier doses. A DDI study that measures a victim drug after repeated dosing therefore needs to specify more than the dose: it must define the dosing interval, the duration of repeated dosing, and exactly where samples fall relative to each dose.

For a steady-state comparison, the victim drug is ordinarily assessed under a consistent dosing regimen while the perpetrator condition is present and sufficiently established. Perpetrator drug concentrations alone do not always show that the interaction effect has stabilized: for example, an effect mediated by enzyme turnover may develop or resolve on a different timescale. The study schedule should therefore be based on the intended interaction condition as well as on drug accumulation.

Definition: Under a fixed repeated-dose regimen, steady state is the condition in which the concentration-time profile over each dosing interval repeats from one interval to the next. In a linear, time-invariant system, this profile is approached as contributions from earlier doses accumulate. It does not mean that concentration is constant within an interval.

Steady state is useful because it makes a full dosing-interval profile interpretable under the repeated regimen of interest. It does not, by itself, specify which exposure endpoint is appropriate; that depends on the study question and the defined exposure estimand. The earlier material on AUC and Cmax sampling is relevant when choosing how densely to sample the interval.

Accumulation Under a Simple Repeated-Dose Model

The basic accumulation calculation assumes linear, time-invariant pharmacokinetics, a constant dose and interval, and complete adherence. For a one-compartment model with instantaneous intravenous boluses, let \(D\) be the dose, \(V\) the volume of distribution, and \(k\) the first-order elimination rate constant. The concentration immediately after a single dose is \(D/V\), and the fraction remaining after time \(t\) is \(e^{-kt}\). For repeated doses every \(\tau\) hours, the steady-state concentration \(t\) hours after a dose is the sum of the residual concentrations from all doses, including the current dose.

C_{\mathrm{ss}}(t)=\frac{D}{V}\sum_{j=0}^{\infty} e^{-k(t+j\tau)} =\frac{D}{V}\frac{e^{-kt}}{1-e^{-k\tau}}, \qquad 0\leq t<\tau.

The sum is a geometric series with ratio \(e^{-k\tau}\), which is between zero and one when \(k>0\). Its sum is \(1/(1-e^{-k\tau})\). Thus the steady-state accumulation factor for the immediate postdose concentration is \(1/(1-e^{-k\tau})\), relative to the single-dose immediate postdose concentration \(D/V\). The trough just before the next dose is the concentration at \(t=\tau\) approached from within the interval:

C_{\mathrm{trough,ss}}=\frac{D}{V}\frac{e^{-k\tau}}{1-e^{-k\tau}}.

The elimination half-life is \(t_{1/2}=\ln(2)/k\). Accumulation is greater when the dosing interval is short relative to the half-life. Conversely, a long interval relative to the half-life allows more drug to be eliminated between doses and produces a larger peak-to-trough fluctuation. These relationships describe concentrations under the stated model; oral absorption and multicompartment distribution can produce different profile shapes.

Worked Example: How the Dosing Interval Changes Accumulation

Suppose a hypothetical drug follows the one-compartment intravenous bolus model, with \(t_{1/2}=8.00\) h and \(D/V=4.00\) mg/L. Compare dosing intervals of 8.00 h and 16.0 h. Assume linear elimination and that dosing continues long enough to reach steady state.

For an 8.00 h interval, \(\tau=t_{1/2}\), so \(e^{-k\tau}=0.5\). The postdose concentration at steady state is

C_{\mathrm{ss}}(0)=\frac{4.00}{1-0.5}=8.00\ \mathrm{mg/L}.

Immediately before the next dose, the fraction remaining from that postdose concentration is 0.5, giving \(C_{\mathrm{trough,ss}}=8.00(0.5)=4.00\) mg/L. The postdose accumulation factor is 2.00.

For a 16.0 h interval, two half-lives pass between doses, so \(e^{-k\tau}=0.25\). The postdose steady-state concentration is \(4.00/(1-0.25)=5.333\) mg/L, and the trough is \(5.333(0.25)=1.333\) mg/L. The accumulation factor is 1.333. Under these assumptions, the shorter interval produces more accumulation and less peak-to-trough fluctuation.

Oral Dosing and the Steady-State Trough

For oral dosing, absorption adds another process to elimination. A useful idealized model assumes one-compartment disposition, first-order absorption with rate constant \(k_a\), first-order elimination with rate constant \(k\), constant oral dose \(D\), constant bioavailability \(F\), and \(k_a\neq k\). It also assumes linear, time-invariant pharmacokinetics and no lag time. Under these assumptions, the concentration during a dosing interval at steady state is the sum of the oral concentration contributions from the current and all previous doses.

C_{\mathrm{ss}}(t)= \frac{F D}{V}\frac{k_a}{k_a-k} \left[ \frac{e^{-kt}}{1-e^{-k\tau}} - \frac{e^{-k_a t}}{1-e^{-k_a\tau}} \right], \qquad 0\leq t<\tau.

The two terms have distinct roles: the elimination term includes residual drug from prior doses, while the absorption term accounts for drug that has not yet entered the systemic circulation. A trough is sampled immediately before the next dose, at the end of the interval. The formula can be evaluated at \(t=\tau\) as the limiting concentration just before that next dose. It should not be read as the concentration after the new dose has been administered.

Worked Example: Calculating an Oral Steady-State Trough

Consider a hypothetical oral drug with \(F D/V=2.00\) mg/L, \(k=0.100\ \mathrm{h}^{-1}\), \(k_a=1.20\ \mathrm{h}^{-1}\), and \(\tau=12.0\) h. Assume the one-compartment, first-order absorption model above applies. The factor outside the brackets is

\frac{F D}{V}\frac{k_a}{k_a-k} =2.00\frac{1.20}{1.20-0.100} =2.181818\ \mathrm{mg/L}.

At the end of the interval, \(e^{-k\tau}=e^{-1.2}\), and \(e^{-k_a\tau}=e^{-14.4}\). Therefore,

C_{\mathrm{ss}}(12)=2.181818 \left[ \frac{e^{-1.2}}{1-e^{-1.2}} - \frac{e^{-14.4}}{1-e^{-14.4}} \right] \approx 2.181818(0.4310127603-0.0000005574) \approx 0.9403903\ \mathrm{mg/L}.

The predicted trough immediately before the next 12-hour dose is therefore 0.9404 mg/L to four decimal places. This result is conditional on the model and parameter values; it is not a general trough estimate for oral drugs.

Designing the Repeated-Dose DDI Phase

A practical design should connect the dosing schedule to the exposure profile being measured. If the target is victim exposure over a full dosing interval at steady state, give the victim drug on its intended repeated regimen and collect samples across a complete interval after the interaction condition is established. Record actual dose and sample times, rather than relying only on nominal clock times. A small timing deviation may matter particularly for a trough, which is defined by its position immediately before the next dose.

The perpetrator schedule needs the same precision. Specify its dose times and the duration before victim profiling begins. A simple first-order drug approaches its steady-state profile exponentially. After five half-lives, the remaining departure from its asymptotic value is \(2^{-5}=0.03125\), or 3.125% of the initial departure. This is a useful rule of thumb for ordinary drug accumulation, not a guarantee that the DDI effect is fully established. Mechanism-based inhibition, induction, active metabolites, nonlinear pharmacokinetics, or a slowly changing physiological process can require a different duration.

Worked Example: Aligning Victim and Perpetrator Timing

A hypothetical victim drug is prescribed every 12 h and has an elimination half-life of 6.93 h under the relevant condition. A perpetrator is given every 24 h and has a 24.0 h half-life. Suppose the planned interaction mechanism is known to stabilize when the perpetrator regimen and its effect have stabilized, and no slower turnover process is expected. Plan a full victim interval profile.

The victim elimination rate constant is approximately \(k=\ln(2)/6.93=0.1000\ \mathrm{h}^{-1}\). Five victim half-lives are \(5(6.93)=34.65\) h, so ordinary victim accumulation is expected to be close to its asymptote after about 35 h of uninterrupted dosing. Five perpetrator half-lives are \(5(24.0)=120\) h, so the perpetrator period should last at least about 120 h before the intended steady-state victim profile begins, subject to confirmation that the interaction effect itself has also stabilized.

For example, if both drugs are scheduled at 08:00, a 12-hour victim interval runs from 08:00 to the next 20:00 dose. A full profile would include samples across that interval, with the trough drawn immediately before the 20:00 dose. The protocol should also preserve and report the actual preceding and following dose times. If the interaction requires slower equilibration than simple perpetrator accumulation, the profile must be delayed accordingly; the five-half-life calculation alone would not justify starting it.

At steady state under linear pharmacokinetics, the AUC over one dosing interval has a useful mass-balance interpretation. Over a complete interval, the amount entering the systemic circulation is \(F D\). If the amount of drug in the body is the same at the beginning and end of that interval, the net change in body amount is zero. With clearance \(CL\), elimination over the interval is \(CL\times AUC_{\tau}\). Equating input and elimination gives:

AUC_{\tau,\mathrm{ss}}=\frac{F D}{CL}.

This result assumes linear clearance, a complete interval at steady state, and a consistent dose and bioavailability. It clarifies a key distinction: accumulation changes the concentration profile and the relation between peak and trough, but does not multiply the dose-normalized steady-state interval AUC by an accumulation factor. If the interaction changes clearance, bioavailability, or both, the interval AUC can change; the formula does not say that a DDI leaves exposure unchanged.

Diagnostics, Alternatives, and Limitations

Check the design and data for evidence that the intended repeated-dose condition was not achieved. Concentrations from successive intervals that continue to rise or fall suggest that steady state may not have been reached, doses were missed, or the system is not time-invariant. A trough that is not close to the scheduled predose time is not directly comparable with a correctly timed trough without accounting for the timing difference. Unexpectedly high or low troughs can also reflect incorrect dose records, sample handling problems, or model assumptions that do not represent the absorption and disposition processes.

The simple accumulation equations are not substitutes for a more suitable pharmacokinetic model when there is nonlinear elimination, time-varying clearance, multiple distribution phases, delayed absorption, or flip-flop kinetics. For those cases, repeated-dose concentration profiles may need to be simulated using the relevant model and verified against appropriately timed observations. A single trough is also not a full profile: it cannot establish the interval AUC or reliably characterize the peak. Conversely, dense sampling is not always necessary if the endpoint and sampling strategy have been justified for the study objective.

A related alternative is to measure exposure after a single dose rather than after repeated dosing. A single-dose design can be appropriate when the scientific question concerns single-dose exposure and the pharmacology permits it. It does not directly describe a repeated-dose steady-state profile when accumulation is material. The choice is therefore not “steady state is always better,” but whether the design matches the intended clinical regimen and estimand.

Common pitfall: It is incorrect to treat a predose sample taken at any convenient time as a steady-state trough. A trough is the concentration immediately before the next scheduled dose, and “steady state” requires a repeating interval profile under the relevant regimen and interaction condition. Confirm both the phase duration and the dose-to-sample timing.

Check Your Understanding

Try each question, then compare your reasoning with the concise solutions below.

For an intravenous bolus model with \(e^{-k\tau}=0.20\), what is the steady-state postdose accumulation factor relative to the single-dose postdose concentration?
A victim dose is scheduled every 8 h, and the trough sample is drawn 90 minutes before the next dose. Is it correctly timed as a trough? Explain.
Under linear steady-state pharmacokinetics, what is the interval AUC for a hypothetical drug with \(F D=150\) mg and \(CL=15\) L/h? Include units.
A perpetrator has a 12 h half-life. Under a simple first-order accumulation assumption, how long is five half-lives, and what fraction of the initial departure from the asymptote remains then?

Solutions. (1) \(1/(1-0.20)=1.25\). (2) No. A trough is taken immediately before the next dose; 90 minutes early is a different time point and should not be labelled a predose trough. (3) \(AUC_{\tau,\mathrm{ss}}=F D/CL=150\ \mathrm{mg}/(15\ \mathrm{L/h})=10.0\ \mathrm{mg\cdot h/L}\). (4) Five half-lives are \(5(12)=60\) h. The remaining fraction is \(2^{-5}=0.03125\), or 3.125%; this estimates ordinary accumulation and does not establish that an interaction mediated by a slower process has stabilized.