Worked Problems in Saturable Elimination

Putting Saturable Elimination to Work

The Michaelis–Menten elimination model is useful only when its consequences are carried through consistently: the rate depends on concentration, while the amount eliminated also depends on the volume in which that concentration is defined. This tutorial works through four applications—bolus decline, constant infusion, dose-dependent AUC, and repeated dosing—using one-compartment calculations that can also serve as checks on a PBPK tissue or plasma compartment.

The calculations assume a well-mixed compartment with constant volume \(V\), concentration \(C\), and amount \(A=VC\). Elimination is the only process changing the amount, and its parameters remain fixed over the calculation. In particular, there is no changing enzyme abundance, interacting drug, or parallel elimination route. These are model assumptions, not guarantees about a real system.

The fraction of the maximum elimination rate used at concentration \(C\) is \(C/(K_m+C)\). Thus \(C/K_m\) helps describe the regime, but \(K_m\) is not a universal boundary between “unsaturated” and “saturated”: the fractional rate rises continuously with concentration. At \(C=K_m\), for example, the elimination rate is one-half of \(V_{\max}\).

Bolus Decline and the Integrated Profile

For a bolus dose \(D\), the initial concentration is \(C_0=D/V\). Separation of variables gives an exact relationship between elapsed time and the concentration reached. This is a convenient way to calculate a time without treating the changing elimination rate as constant.

This relationship applies to a decline from \(C_0\) to a positive \(C\), with \(V\), \(V_{\max}\), and \(K_m\) constant. It is not a single exponential: the logarithmic term and the linear concentration-difference term both contribute. At low concentration relative to \(K_m\), the logarithmic contribution dominates; at high concentration relative to \(K_m\), the linear contribution becomes more important.

Constant Infusion: A Capacity Check Comes First

During infusion at a constant rate \(R\), the concentration changes according to input minus elimination. At a finite steady state, these rates must balance. This yields a useful capacity check before calculating any concentration.

A finite positive \(C_{\mathrm{ss}}\) requires \(0<R<V_{\max}\). If \(R=V_{\max}\), the elimination rate approaches the input rate only as concentration grows without bound; there is no finite steady state. If \(R>V_{\max}\), elimination cannot balance input at any concentration. The formula must not be used at or above capacity.

Dose-Dependent AUC After a Bolus

AUC is the area under the concentration-time curve. For a bolus followed by complete elimination, it can be derived by integrating concentration over time rather than assuming that clearance is constant. From the no-input differential equation, \(dt=-(V/V_{\max})(K_m+C)dC/C\). Multiplying by \(C\) and integrating from \(C_0\) down to zero gives

This derivation assumes that the bolus is the only input and that the concentration eventually returns to zero. Using \(C_0=D/V\), the AUC contains a term proportional to dose and a term proportional to dose squared. Consequently, AUC is not generally dose-proportional under saturable elimination. In the low-concentration limit, the squared term is relatively small and the model approaches first-order elimination; as dose and \(C_0\) rise, the squared term matters more.

Repeated Boluses: Solving for Peak and Trough

For repeated bolus doses separated by interval \(\tau\), the concentration jumps by \(D/V\) at each dose and then declines between doses. At steady state, let \(C_{\mathrm{trough}}\) be the concentration just before a dose and \(C_{\mathrm{peak}}\) the concentration just after it. The jump condition and the integrated decline jointly determine the profile.

Unlike a first-order model, the steady-state trough cannot generally be found by multiplying the dose jump by a constant accumulation factor. The interval equation is nonlinear in the trough and is usually solved numerically. A finite steady-state cycle also requires average input \(D/\tau\) to be less than \(V_{\max}\); otherwise elimination cannot remove the full dose over each interval at any finite profile.

How to Check a Calculation and Choose an Approach

Several checks make these calculations more reliable. First, check dimensions: \(V/V_{\max}\) has units of volume·time/amount, so multiplying it by a concentration gives time; the AUC expression has units of concentration·time. Second, check direction: a higher infusion rate should require a higher steady-state concentration, and the predicted bolus concentration must decline rather than rise. Third, compare the input with capacity. An infusion at or above \(V_{\max}\), or a repeated-dose average input at or above \(V_{\max}\), cannot have a finite steady state under this model.

When solving repeated-dose equations numerically, the trough must be positive and the peak must exceed it by \(D/V\). Substituting the resulting peak and trough back into the integrated equation should reproduce \(\tau\). A numerical solver that returns a negative trough, violates the dose jump, or fails the time check has not produced a valid profile. In complex PBPK models, numerical integration of differential equations is often the practical alternative; it can accommodate changing volumes, multiple tissues, and additional processes, but it does not remove the need to check parameter units, mass balance, and model assumptions.

The first-order approximation is another useful comparison, not a universal replacement. When \(C\ll K_m\), the rate is approximately \((V_{\max}/K_m)C\), which has a constant proportionality to concentration. When \(C\gg K_m\), the rate approaches \(V_{\max}\), so concentration can decline approximately linearly with time over that range. Between these limits, use the full Michaelis–Menten expression rather than forcing either approximation. The introductory material on concentration-dependent clearance and dose proportionality provides the broader context; these worked calculations show how the same concentration dependence enters specific dosing problems.

A common mistake is to treat \(K_m\) as a sharp concentration threshold and label all concentrations above it “fully saturated.” At \(C=K_m\), the rate is only half of \(V_{\max}\); at \(C=2K_m\), it is two-thirds, not 100 percent. The correction is to calculate the fractional capacity \(C/(K_m+C)\), state the concentration relative to \(K_m\), and describe the regime as an approximation when justified. A further limitation is that this model uses fixed kinetic parameters. If observed data show changing capacity or multiple routes, a mismatch should prompt model checking rather than an automatic reinterpretation of \(K_m\).