Dose Proportionality and AUC Under Saturable Elimination

Why Dose Proportionality Changes

With linear elimination, doubling an IV dose doubles the area under the concentration-time curve (AUC), provided other pharmacokinetic conditions remain unchanged. Saturable elimination changes this relationship: as dose increases, the elimination pathway becomes less able to remove each additional unit of drug at the same rate. Consequently, AUC can rise more than proportionally with dose. This tutorial derives that result for a single IV bolus under a one-compartment Michaelis-Menten model.

The derivation concerns total AUC from the bolus until complete elimination. It does not require a detailed concentration-time profile. The key is to express the time spent across each concentration range in terms of the elimination rate, then integrate concentration over that time.

Definition: Let \(D\) be a single IV bolus dose, \(V\) a constant apparent distribution volume, and \(C(t)\) the drug concentration at time \(t\). Under Michaelis-Menten elimination, the elimination rate in amount per unit time is \(V_{\max}C/(K_m+C)\), where \(V_{\max}>0\) is the maximum elimination rate and \(K_m>0\) is the concentration at which the elimination rate is half of \(V_{\max}\). The single-dose AUC is \(\int_0^\infty C(t)\,dt\), assuming concentration ultimately approaches zero.

Model and Assumptions

For a bolus dose in a single well-mixed compartment, the initial concentration is \(C_0=D/V\). The amount in the compartment is \(VC(t)\), so its rate of change is the negative elimination rate. With \(V\) constant, the concentration equation is

V\frac{dC}{dt}=-\frac{V_{\max}C}{K_m+C}, \qquad C(0)=C_0=\frac{D}{V}.

This AUC result applies when the Michaelis-Menten pathway is the only elimination route; \(V_{\max}\), \(K_m\), and \(V\) do not change during the dose; and the concentration driving elimination is the same concentration represented by \(C(t)\). It also assumes no input after the bolus and no additional processes such as time-varying enzyme activity, changing binding, or a separate elimination pathway. These assumptions define a simplified system, not every PBPK application.

Deriving the Single-Dose AUC

For positive concentrations, rearrange the concentration equation to express elapsed time for a small concentration decrease. As concentration falls from \(C_0\) to zero, this gives the AUC directly.

dt=-\frac{V}{V_{\max}}\frac{K_m+C}{C}\,dC

Multiplying by \(C\) and integrating along the decline from \(C_0\) to zero yields

\begin{aligned} \mathrm{AUC} &=\int_0^\infty C(t)\,dt\\ &=-\frac{V}{V_{\max}}\int_{C_0}^{0}(K_m+C)\,dC\\ &=\frac{V}{V_{\max}}\left(K_mC_0+\frac{C_0^2}{2}\right). \end{aligned}

Substitute \(C_0=D/V\) to obtain the dose-AUC relationship:

\boxed{\mathrm{AUC} =\frac{K_mD}{V_{\max}}+\frac{D^2}{2VV_{\max}}}

The first term is proportional to dose; the second is proportional to dose squared. Both are positive for a positive dose, so increasing dose increases AUC, and the quadratic contribution makes AUC grow faster than dose. The equation has units of concentration multiplied by time: \(K_mD/V_{\max}\) has units \((\text{amount}/\text{volume})(\text{amount})/(\text{amount}/\text{time})\), and the second term has the same units.

Key result: In this model, dose-normalized AUC is \(\mathrm{AUC}/D=K_m/V_{\max}+D/(2VV_{\max})\). It increases linearly with dose, so a higher dose has a greater AUC per unit dose. For any two positive doses \(D_2>D_1\), AUC increases more than proportionally, provided the model parameters and assumptions remain unchanged.

Low- and High-Dose Interpretation

At low concentrations relative to \(K_m\), the Michaelis-Menten rate is approximately first order: \(V_{\max}C/(K_m+C)\approx (V_{\max}/K_m)C\). The corresponding linear clearance is \(CL=V_{\max}/K_m\), and the low-dose AUC is approximately \(D/CL=K_mD/V_{\max}\). This is the first term in the exact expression. At higher doses, the quadratic term contributes more, and dose-normalized AUC rises above the low-dose value.

The relative contribution of the quadratic term compared with the linear term is \(D/(2VK_m)\), or equivalently \(C_0/(2K_m)\). Thus, when the initial concentration is small relative to \(K_m\), the departure from dose proportionality is modest. As initial concentration increases, the departure becomes more pronounced. This comparison does not mean that the pathway is either wholly unsaturated or wholly saturated at a single threshold; it describes how the terms in the AUC expression compare.

A useful local summary is the elasticity of AUC with respect to dose: the percentage change in AUC associated with a small percentage change in dose. Differentiating the AUC expression gives

\frac{d\log(\mathrm{AUC})}{d\log D} =\frac{D}{\mathrm{AUC}}\frac{d(\mathrm{AUC})}{dD} =\frac{K_m+D/V}{K_m+D/(2V)}.

For positive dose this value is greater than 1 and less than 2. It approaches 1 in the low-dose limit and 2 when the quadratic term dominates. The value describes a local slope on a log-dose versus log-AUC scale; it is not a universal constant across doses.

Worked Example: Calculating AUC at One Dose

Consider a hypothetical one-compartment system with \(D=10\) mg, \(V=10\) L, \(K_m=2\) mg/L, and \(V_{\max}=20\) mg/h. Assume the model conditions above hold. The initial concentration is \(C_0=D/V=1\) mg/L. Use the exact expression:

\begin{aligned} \mathrm{AUC} &=\frac{K_mD}{V_{\max}}+\frac{D^2}{2VV_{\max}}\\ &=\frac{(2\ \mathrm{mg/L})(10\ \mathrm{mg})}{20\ \mathrm{mg/h}} +\frac{(10\ \mathrm{mg})^2}{2(10\ \mathrm{L})(20\ \mathrm{mg/h})}\\ &=1.00\ \mathrm{mg\cdot h/L}+0.250\ \mathrm{mg\cdot h/L}\\ &=1.25\ \mathrm{mg\cdot h/L}. \end{aligned}

The first-order low-dose approximation would give \(D/(V_{\max}/K_m)=10/(20/2)=1.00\) mg·h/L. The exact AUC is higher because the dose-dependent quadratic term contributes 0.250 mg·h/L.

Worked Example: Comparing AUC Across Doses

Use the same hypothetical system, but compare doses of 10 mg and 40 mg. The dose increases fourfold. At 10 mg, the AUC is 1.25 mg·h/L, as calculated above. At 40 mg:

\begin{aligned} \mathrm{AUC}_{40} &=\frac{(2)(40)}{20}+\frac{40^2}{2(10)(20)}\\ &=4.00+4.00\\ &=8.00\ \mathrm{mg\cdot h/L}. \end{aligned}

The AUC ratio is \(8.00/1.25=6.40\), whereas the dose ratio is \(40/10=4.00\). Thus, in this model, the AUC rises 6.40-fold when dose rises fourfold: the increase is more than proportional. Dose-normalized AUC rises from \(1.25/10=0.125\) to \(8.00/40=0.200\) mg·h/L per mg. Those values are consistent with the dose-normalized expression, which increases with dose.

Worked Example: Interpreting the Local Dose-AUC Slope

For a separate hypothetical system, let \(V=20\) L and \(K_m=5\) mg/L. Evaluate the local AUC elasticity at a dose of 40 mg. The \(V_{\max}\) value cancels from the elasticity formula, so it is not needed:

\begin{aligned} \frac{d\log(\mathrm{AUC})}{d\log D} &=\frac{K_m+D/V}{K_m+D/(2V)}\\ &=\frac{5+(40/20)}{5+(40/(2\cdot20))}\\ &=\frac{7}{6}=1.167. \end{aligned}

Near 40 mg, a 1% increase in dose corresponds approximately to a 1.167% increase in AUC under this model. This is a local approximation, not the exact AUC ratio for a large dose change. It indicates more-than-proportional scaling at this dose, but a smaller departure than the limiting elasticity of 2.

Comparisons, Diagnostics, and Limitations

A constant-clearance, first-order model is the main alternative for this comparison. It predicts \(\mathrm{AUC}=D/CL\), so dose-normalized AUC is constant and AUC ratios equal dose ratios. The Michaelis-Menten model instead predicts the linear-plus-quadratic expression derived above. A numerical integration of the concentration differential equation can also calculate AUC, but the closed-form expression is exact for the stated one-compartment model and makes the dose dependence explicit. A numerical result is useful when the model includes additional mechanisms that prevent this simple expression from applying.

Across dose levels, an increasing dose-normalized AUC is consistent with more-than-proportional exposure, but it is not by itself proof of Michaelis-Menten elimination. A practical check is to compare AUC per unit dose across otherwise comparable single-dose conditions. If it increases systematically, investigate saturation as one explanation, while also checking dose administration, sampling and AUC estimation, changes in distribution or binding, and time-dependent changes in elimination. The model should also be checked against whether the pathway plausibly dominates elimination and whether the same parameters can describe the dose range.

The closed-form result has important limits. It assumes the elimination-driving concentration is the measured compartment concentration, fixed volume, a single bolus, and one saturable elimination route. Multiple pathways, changing enzyme abundance, distribution between tissues, or dose-dependent bioavailability are not represented. In a PBPK model, the enzyme may act on a tissue concentration that differs from a systemic plasma concentration; in that setting, the single-compartment formula need not describe plasma AUC. The formula is therefore a mechanistic result for its stated model, not a general conversion from dose to AUC in every patient or PBPK structure.

Common Pitfall

A common error is to reason that a saturable pathway has a maximum rate and therefore AUC cannot continue increasing once the dose is high. The rate is capped near \(V_{\max}\), but the drug amount still has to be eliminated. A higher dose can therefore remain in the system longer, increasing the integrated concentration. In the model, this appears as the positive quadratic term \(D^2/(2VV_{\max})\). The cap on elimination rate does not cap AUC.

Key takeaway: For a single IV bolus with Michaelis-Menten elimination as the sole route, \(\mathrm{AUC}=K_mD/V_{\max}+D^2/(2VV_{\max})\). The dose-squared contribution explains why AUC per unit dose increases as dose rises, subject to the model assumptions.

Check Your Understanding

Use the stated single-compartment Michaelis-Menten model. Each problem includes its solution.

If \(D=12\) mg, \(V=6\) L, \(K_m=3\) mg/L, and \(V_{\max}=18\) mg/h, calculate the initial concentration and AUC. Solution: \(C_0=D/V=12/6=2\) mg/L. The AUC is \((3\cdot12)/18+12^2/[2(6)(18)]=2.00+0.667=2.67\) mg·h/L.
In the model, what is the low-dose approximation to AUC, and what clearance does it imply? Solution: When \(C\) is small relative to \(K_m\), \(\mathrm{AUC}\approx K_mD/V_{\max}=D/CL\), where the corresponding linear clearance is \(CL=V_{\max}/K_m\).
If dose is doubled, must AUC exactly double under this model? Solution: No. The expression contains a positive term proportional to \(D^2\), so for positive doses AUC increases more than proportionally. The exact ratio depends on the dose and model parameters.
Why does a maximum elimination rate not imply a maximum possible AUC? Solution: A higher dose can take longer to eliminate when the rate is capacity-limited. The concentration-time integral can therefore continue to increase; the AUC expression includes a positive dose-squared term.
Name one condition that would make the closed-form result inappropriate for a plasma AUC prediction. Solution: For example, if the enzyme acts on a tissue concentration different from plasma concentration, the one-compartment assumption that the elimination-driving and modeled concentrations are identical does not hold. A richer model may be required.