Accumulation and Peak-Trough Profiles With Saturable Elimination

Why Repeated Doses Need a Nonlinear Profile

For linear elimination, a dose produces a concentration profile that is independent of other doses. The total profile after repeated doses can therefore be obtained by adding shifted single-dose profiles. With saturable elimination, the elimination rate depends nonlinearly on the concentration present at that time. A later dose changes that concentration and thus changes the rate at which both the new dose and the drug already in the body are eliminated. Adding isolated-dose profiles is no longer valid.

This tutorial considers repeated intravenous bolus doses in a single, well-mixed compartment. The volume \(V\) is constant; each dose \(D\) is given every \(\tau\) time units; and elimination follows a Michaelis–Menten rate law. There is no absorption delay, distribution between compartments, or additional elimination pathway. These assumptions isolate the effect of saturation on peak, trough, and accumulation. A real PBPK model may have more processes, but the same nonlinear principle applies.

The model assumes the same elimination parameters and dosing interval throughout, with no time-varying enzyme capacity. Since the elimination rate approaches \(V_{\max}\) rather than increasing without bound, elimination becomes relatively slow at high concentrations. At low concentrations, the rate is approximately proportional to concentration: \(V_{\max}C/(K_m+C)\approx (V_{\max}/K_m)C\) when \(C\) is much smaller than \(K_m\).

Concentration Decline Between Boluses

Separating variables gives an exact relation between the concentration at the start of a dose interval, \(C_{\mathrm{start}}\), the concentration \(C(t)\) later in that interval, and elapsed time. The relation is useful for calculating troughs without treating elimination as a constant-rate process.

Here, \(C_{\mathrm{start}}\) is the post-dose concentration for the interval being considered. To justify the expression, rearrange the differential equation as \(dt=-(V/V_{\max})(1+K_m/C)\,dC\). Integrating from \(C_{\mathrm{start}}\) to \(C(t)\) yields the displayed equation. The concentration at a specified time is generally obtained by solving this equation numerically; the logarithmic term means the decline is not a single exponential.

At a periodic steady dosing profile, let \(C_{\min}\) be the trough and \(C_{\max}\) the peak. The bolus raises concentration by \(D/V\), so \(C_{\max}=C_{\min}+D/V\). Applying the decline equation over one full interval gives a direct steady-state relationship.

Solving for the trough gives

A finite positive solution requires \(\tau V_{\max}/V>D/V\), equivalently \(D< V_{\max}\tau\). This condition says the dose per interval must be less than the maximum amount that could be eliminated during that interval. At equality, the limiting elimination capacity matches the dose, but the rate is strictly below \(V_{\max}\) at every finite concentration. The concentration therefore does not settle to a finite periodic profile.

Steady-State Profiles and the Feasibility Condition

At steady state, each interval begins and ends at the same trough, with the bolus producing the same peak on every cycle. The steady-state equation makes the effects of dose and interval explicit: the dose determines the jump between trough and peak, while the interval determines how far concentration can decline before the next dose. As \(\tau\) shortens at a fixed dose, both the steady trough and peak generally rise. As \(\tau\) lengthens, the trough falls and the peak approaches the concentration increment from a single dose.

Why Linear Superposition Fails

For linear first-order elimination, \(dC/dt=-kC\), the concentration after a bolus declines exponentially. Repeated bolus profiles can then be added because the differential equation is linear: the response to the sum of doses equals the sum of responses to each dose. For a fixed interval, the steady linear peak and trough are \(C_{\max}= (D/V)/(1-\exp(-k\tau))\) and \(C_{\min}=C_{\max}\exp(-k\tau)\). These formulas rely on a concentration-independent elimination constant \(k\).

Under Michaelis–Menten elimination, the rate \(V_{\max}C/(K_m+C)\) is not a constant times \(C\). In general, the decline from a combined concentration is not the sum of the declines that would occur from each component separately. A high post-dose concentration can also be eliminated at a smaller fraction of its amount per unit time than a low concentration, extending the tail and increasing the trough from later doses. This is why the single-dose profile cannot simply be replicated and summed.

Practical Calculation, Diagnostics, and Limitations

A practical calculation can proceed by identifying the dose-induced concentration jump, applying the integrated decline relation across each interval, and checking whether successive troughs stabilize. At steady state, the explicit trough formula is available when \(D<V_{\max}\tau\). Otherwise, a finite steady periodic peak–trough profile is not possible under this model. For more complicated inputs or additional processes, numerical integration of the model equations can track dose jumps and concentration changes directly; it still must use the nonlinear rate law rather than superpose isolated-dose profiles.

Diagnostics include checking that concentration decreases between boluses, jumps by exactly \(D/V\) at each dose, and has a lower trough than peak within a cycle. A computed steady trough that is negative, undefined, or based on an exponent at or below zero signals failure of the finite-steady-state condition, not a concentration prediction. If a numerical solution gives increasing troughs despite \(D<V_{\max}\tau\), check units, dose timing, and the implementation of the rate law. Model-data disagreement may also reflect assumptions that do not hold, such as a changing enzyme capacity or distribution outside the single compartment.

Constant infusion is a different input pattern: it continuously supplies drug rather than creating bolus peaks and troughs. It can inform how saturable elimination balances a sustained input, but it does not make repeated bolus profiles additive. Likewise, a linear model may be a useful approximation when concentrations remain far below \(K_m\), but it cannot represent the capacity limit at high concentrations. The present one-compartment result also does not specify how saturable metabolism is distributed across tissues; that requires a model that represents the relevant tissue processes.

Common Pitfall

A frequent error is to calculate the concentration after one dose, shift that curve by each dosing time, and add the curves to predict repeated-dose accumulation. That reasoning assumes each dose experiences the same linear elimination process regardless of the concentrations already present. With saturable elimination, the total concentration determines the rate, so the elimination of a new dose depends on residual drug. Instead, propagate the total concentration through each interval using the nonlinear equation, or solve the steady-state peak–trough relation when its assumptions and feasibility condition hold.