When Exposure Rises More Than Dose
With linear elimination and unchanged bioavailability, a dose increase produces a proportional increase in AUC: doubling the dose doubles the area under the concentration–time curve. A different pattern—AUC increasing more than proportionally, so that dose-normalized AUC rises with dose—can be consistent with saturable elimination. The interpretation matters: at higher concentrations, elimination may be less efficient per unit concentration because a capacity-limited pathway is approaching its maximum rate.
This pattern is evidence to investigate, not proof of a specific mechanism. Dose-related changes in bioavailability, differences between dose groups, incomplete AUC measurement, or misspecified distribution can also alter the apparent dose–exposure relationship. A useful assessment therefore combines the exposure pattern with concentration–time behavior, study design, and comparisons between plausible models.
Linear and Capacity-Limited Elimination
For linear elimination from a one-compartment model, the elimination rate is proportional to concentration. If systemic clearance \(CL\) is constant, the rate is \(CL\,C\), where \(C\) is concentration. For an IV dose \(D\), complete AUC measurement and linear pharmacokinetics imply \(AUC=D/CL\). Thus dose-normalized AUC is constant, and apparent clearance does not depend on dose.
A standard capacity-limited alternative is Michaelis–Menten elimination. Let \(V\) be the apparent central volume, \(V_{\max}\) the maximum elimination rate, and \(K_m\) the concentration at which the elimination rate is half of \(V_{\max}\). The model assumes a well-mixed one-compartment system and elimination rate \(V_{\max}C/(K_m+C)\). Under an IV bolus, the concentration equation is:
Here \(V_{\max}\) has units of amount per time, while \(K_m\) has units of concentration. When \(C\) is much smaller than \(K_m\), the elimination rate is approximately \((V_{\max}/K_m)C\), which behaves like linear elimination with clearance \(V_{\max}/K_m\). When \(C\) is much greater than \(K_m\), the elimination rate approaches \(V_{\max}\): increasing concentration then produces relatively little increase in the amount eliminated per unit time.
The instantaneous clearance corresponding to the Michaelis–Menten elimination rate is the rate divided by concentration. It is therefore not constant:
As concentration increases, this quantity decreases. This is a concentration-dependent clearance within the model, distinct from a fixed individual clearance parameter in a linear model. Its units are volume per time.
Why Dose-Normalized AUC Increases
The dose–AUC relationship follows by integrating the model rather than assuming a constant clearance. After an IV bolus, \(C(0)=C_0=D/V\), and concentration declines toward zero. Rearranging the differential equation gives a time increment in terms of a concentration increment:
AUC is \(\int_0^\infty C\,dt\). Substitute the expression for \(dt\), and change the concentration limits from \(C_0\) down to zero into an integral from zero up to \(C_0\):
The last line uses \(C_0=D/V\). The first term is proportional to dose, but the second is proportional to dose squared. Consequently, dose-normalized AUC increases with dose under these assumptions:
The corresponding dose-over-AUC apparent clearance is \(D/AUC\), so it decreases as dose increases. This derivation is specific to an IV bolus in the stated one-compartment model; it should not be transferred unchanged to other dosing routes or dosing regimens.
Worked Example: Calculating AUC at Two IV Bolus Doses
Consider a hypothetical one-compartment drug with \(V=20\) L, \(V_{\max}=100\) mg/h, and \(K_m=5\) mg/L. Assume the model and complete-profile conditions above hold. Calculate AUC and dose-normalized AUC at IV bolus doses of 100 mg and 200 mg.
For 100 mg, \(C_0=D/V=100/20=5\) mg/L. Substitution into the derived formula gives:
Thus, \(AUC/D=7.50/100=0.0750\ \mathrm{h/L}\), and \(D/AUC=100/7.50=13.3\) L/h.
For 200 mg, \(C_0=200/20=10\) mg/L:
Here, \(AUC/D=20.0/200=0.100\ \mathrm{h/L}\), and \(D/AUC=200/20.0=10.0\) L/h. Doubling the dose from 100 mg to 200 mg increases AUC by a factor of \(20.0/7.50=2.67\), not by a factor of 2. In this hypothetical model, the increase in dose-normalized AUC and decrease in apparent clearance are consistent with saturable elimination.
Reading Concentration-Dependent Clearance
A dose–AUC pattern summarizes the whole profile. Concentration-dependent clearance provides a second, related model prediction. In the hypothetical example, \(CL(C)=100/(5+C)\) L/h when \(C\) is entered in mg/L. At \(C=1\) mg/L, this is \(100/6=16.7\) L/h; at \(C=5\) mg/L, it is \(100/10=10.0\) L/h; and at \(C=20\) mg/L, it is \(100/25=4.00\) L/h. The model therefore predicts lower instantaneous clearance at higher concentration. As concentration falls late in the profile, clearance approaches \(V_{\max}/K_m=20\) L/h, and elimination approaches its low-concentration, approximately linear behavior.
This concentration pattern should be evaluated against the profile and the sampling design, not inferred from a single concentration or a pooled plot alone. For example, one dose group could have a different distribution of body size or other patient characteristics than another. A population model should distinguish systematic dose or concentration behavior from between-individual differences and should check that the available data actually span informative concentrations. If most observations are far below \(K_m\), the data may show little curvature and provide weak information to distinguish the Michaelis–Menten model from linear elimination.
Worked Example: Comparing a Linear Prediction With the Capacity-Limited Model
Use the same hypothetical parameter values as above. Suppose a linear model is calibrated to the 50 mg IV dose. The Michaelis–Menten model gives \(C_0=50/20=2.5\) mg/L and:
A linear model passing through this dose–AUC point has \(AUC/D=3.125/50=0.0625\) h/L, so it predicts \(AUC=0.0625D\). At 200 mg it predicts \(0.0625(200)=12.5\) mg·h/L. The Michaelis–Menten prediction at 200 mg is 20.0 mg·h/L, as calculated above. Relative to the capacity-limited prediction, the linear prediction is lower by \(20.0-12.5=7.50\) mg·h/L, or \(7.50/20.0=37.5\%\).
This comparison illustrates a diagnostic question: does a constant-clearance model systematically underpredict exposure at higher doses when calibrated to lower-dose behavior? It does not establish that Michaelis–Menten elimination is the correct explanation. The comparison would need to use observed data, account for uncertainty and the study design, and assess whether other model assumptions are plausible.
Diagnostics, Alternatives, and Limits
A practical assessment begins with comparable dose groups or, where possible, within-individual dose information. Plot AUC against dose and AUC divided by dose against dose; also inspect concentration–time profiles across the dose range. A rising dose-normalized AUC is consistent with nonlinear pharmacokinetics, while a decline in \(D/AUC\) is the equivalent inverse pattern. In profiles with informative concentration coverage, check whether a constant-clearance model shows systematic dose- or concentration-related discrepancies that a capacity-limited model reduces. The pattern should be reproducible and larger than the uncertainty in the exposure estimates.
The main alternatives have different implications:
| Approach | Key implication | Useful caution |
|---|---|---|
| Linear elimination | For IV dosing, AUC is proportional to dose and \(D/AUC\) is constant. | A narrow dose range may make small departures from proportionality hard to detect. |
| Michaelis–Menten elimination | Clearance falls as concentration rises; under the stated IV bolus model, dose-normalized AUC rises with dose. | Estimates of \(V_{\max}\) and \(K_m\) may be weakly determined if observations do not cover the curved concentration range. |
| Empirical dose-specific exposure or clearance summaries | Describe observed departures from proportionality without asserting a specific elimination mechanism. | Summaries do not by themselves explain the pattern or reliably predict outside the observed dose range. |
Important failure modes include comparing doses administered to different patient groups without accounting for their differences, relying on AUC estimates that do not capture the full profile, and treating a decrease in \(D/AUC\) as uniquely diagnostic of saturable elimination. For non-IV dosing, a dose-dependent change in \(F\) can also change dose-normalized AUC, even if systemic elimination is linear. Differences in dose timing or incomplete concentration coverage can further distort comparisons. The investigation should therefore check that the route, dosing conditions, and exposure calculation are comparable and that a competing explanation is not more plausible.
The Michaelis–Menten form is a mechanistic simplification, not a universal law. It assumes a particular capacity-limited pathway and fixed parameters; more than one elimination pathway or time-varying processes may require a different model. Conversely, apparent curvature in a small or poorly distributed data set may not justify extra parameters. Model choice should consider fit across the observed profiles, parameter precision, and whether the model gives credible predictions over the dose range of interest.
Check Your Understanding
Answer each question, then compare with the solutions below.
- Under linear IV pharmacokinetics, what happens to dose-normalized AUC when the dose doubles?
- For \(V_{\max}=60\) mg/h, \(K_m=3\) mg/L, and \(V=10\) L, calculate the Michaelis–Menten AUC for a 60 mg IV bolus using the derived formula.
- With \(V_{\max}=60\) mg/h and \(K_m=3\) mg/L, calculate instantaneous clearance at concentrations of 3 mg/L and 9 mg/L.
- Why does an increase in dose-normalized AUC after oral dosing not, on its own, establish saturable elimination?
Solutions. (1) It remains constant, because \(AUC=D/CL\) and \(AUC/D=1/CL\) when \(CL\) is constant. (2) \(AUC=(3)(60)/60+60^2/[2(10)(60)]=3+3=6.00\) mg·h/L. (3) At 3 mg/L, \(CL(C)=60/(3+3)=10.0\) L/h; at 9 mg/L, \(CL(C)=60/(3+9)=5.00\) L/h. (4) Dose-dependent bioavailability can also change AUC relative to administered dose. The route and exposure pattern must be assessed before attributing the change to elimination.