Tutorials › Bioavailability and Bioequivalence › Why AUC Is Less Peak-Sensitive Than Cmax

Home  /  Tutorials  /  Exposure Metrics: AUC, Cmax and Partial AUC

Exposure Metrics: AUC, Cmax and Partial AUC · Tutorial 40 of 399

Why AUC Is Less Peak-Sensitive Than Cmax

Learn how peak height and duration shape the different responses of AUC and observed Cmax, and when that distinction matters.

Intermediate 10 min read

What You'll Learn

  • Distinguish total AUC from observed Cmax and state what each summarizes.
  • Derive the area contribution of a triangular, localized peak.
  • Compare relative changes in AUC and Cmax when peak width changes.
  • Identify assumptions that make the comparison valid and recognize when they fail.
  • Interpret equal-AUC and equal-Cmax profiles without treating either metric as a complete description of profile shape.

Why a Tall Peak Need Not Add Much Exposure

A concentration profile records both how high concentrations rise and how long they remain elevated. Those features do not contribute equally to every exposure metric. A short, sharp increase can move the maximum concentration substantially while adding only a small amount of area under the curve. Conversely, a wider increase can add considerable area without raising the maximum any further.

This difference matters when interpreting profiles: a change in Cmax is primarily a change in peak height, whereas a change in AUC reflects the accumulated concentration over time. Neither metric describes every feature of the profile. The comparison below isolates the role of peak shape, assuming the profiles are measured and summarized on the same time interval.

Definition: For a concentration profile \(C(t)\) over \(0\leq t\leq T\), the total AUC over that interval is the integral \(\int_0^T C(t)\,dt\). It has units of concentration multiplied by time. The observed Cmax is the largest concentration recorded at the scheduled sampling times; it has units of concentration. It need not equal the true continuous-time maximum if the sampling schedule misses the peak.

AUC is an accumulated quantity: a concentration contributes according to both its magnitude and its duration. Cmax is an extreme-value summary: among the recorded concentrations, only the largest determines the metric. Thus, an increase concentrated in a narrow time region can have a large effect on Cmax but a relatively small effect on AUC. This is a comparison of the metrics, not a claim that AUC is unaffected by peaks.

Area and Height: A Simple Peak Model

Consider an existing profile that is constant at \(b\) concentration units over an interval of length \(T\). Add a triangular peak centered within that interval. Let the triangle rise \(h\) concentration units above the baseline and return to baseline over a total width \(w\). Assume \(0<w\leq T\), the peak is entirely within the interval, and its apex is included in the recorded samples.

The added area is the area of a triangle: its base is \(w\), and its height is \(h\). The baseline AUC is \(bT\), and the new maximum is \(b+h\). Therefore, the relative increases are:

$$ \Delta \mathrm{AUC}=\frac{hw}{2},\qquad \frac{\Delta \mathrm{AUC}}{\mathrm{AUC}_0}=\frac{hw}{2bT},\qquad \frac{\Delta C_{\max}}{C_{\max,0}}=\frac{h}{b}. $$

The relative AUC increase depends on both peak height and width. The relative Cmax increase depends on height, but not on width, provided the apex is still observed. Dividing the relative AUC change by the relative Cmax change gives:

$$ \frac{\Delta \mathrm{AUC}/\mathrm{AUC}_0} {\Delta C_{\max}/C_{\max,0}}=\frac{w}{2T}. $$

This ratio is the peak width relative to the full interval, with the factor of two arising from the triangular shape. When \(w\) is small compared with \(T\), the relative AUC change can be much smaller than the relative Cmax change. The result follows from the specified shape and baseline; it is not a universal conversion factor for arbitrary profiles.

Worked Example: A Narrow Peak on a Stable Baseline

Suppose a hypothetical concentration profile is at \(10\ \text{mg/L}\) over \(8\ \text{h}\). A triangular peak, fully captured by sampling, rises \(6\ \text{mg/L}\) above the baseline and has a total width of \(0.5\ \text{h}\). Calculate the AUC and Cmax changes.

The baseline AUC and Cmax are \(10\ \text{mg/L}\times 8\ \text{h}=80\ \text{mg}\cdot\text{h/L}\) and \(10\ \text{mg/L}\), respectively. The added triangular area is:

$$ \Delta \mathrm{AUC} =\frac{(6\ \text{mg/L})(0.5\ \text{h})}{2} =1.5\ \text{mg}\cdot\text{h/L}. $$

The new AUC is \(81.5\ \text{mg}\cdot\text{h/L}\), a relative increase of \(1.5/80=0.01875\), or \(1.875\%\). The new Cmax is \(10+6=16\ \text{mg/L}\), a relative increase of \(6/10=0.60\), or \(60.0\%\). In this hypothetical profile, a short peak produces a much larger percentage change in Cmax than in AUC.

When Profile Shape Changes the Comparison

The simple model helps explain two useful comparisons. First, peaks with equal area can have different heights if their widths differ. Second, peaks with equal height can have different areas if their widths differ. These are profile-shape differences: the peak metrics can separate even when one of them is held constant.

Worked Example: Equal Added AUC, Different Cmax

Two hypothetical profiles each have a constant baseline of \(5\ \text{mg/L}\) over \(10\ \text{h}\), giving a baseline AUC of \(50\ \text{mg}\cdot\text{h/L}\) and Cmax of \(5\ \text{mg/L}\). Profile A has a triangular peak \(4\ \text{mg/L}\) high and \(2\ \text{h}\) wide. Profile B has a triangular peak \(16\ \text{mg/L}\) high and \(0.5\ \text{h}\) wide. Assume each apex is sampled.

For Profile A, the added area is \((4\times2)/2=4\ \text{mg}\cdot\text{h/L}\), so total AUC is \(54\ \text{mg}\cdot\text{h/L}\). Its Cmax is \(5+4=9\ \text{mg/L}\). For Profile B, the added area is \((16\times0.5)/2=4\ \text{mg}\cdot\text{h/L}\), so its total AUC is also \(54\ \text{mg}\cdot\text{h/L}\). Its Cmax is \(5+16=21\ \text{mg/L}\).

The profiles have equal AUC but substantially different Cmax values. The narrow peak in Profile B reaches a much greater height while contributing the same area as Profile A's wider, lower peak. Equal total AUC therefore does not imply equal peak height or equal profile shape.

Worked Example: Equal Cmax, Different Added AUC

Now consider two hypothetical profiles with a constant baseline of \(4\ \text{mg/L}\) over \(6\ \text{h}\). Each has a triangular peak \(8\ \text{mg/L}\) above baseline, but one peak is \(0.5\ \text{h}\) wide and the other is \(2\ \text{h}\) wide. Both apexes are sampled.

The baseline AUC is \(4\times6=24\ \text{mg}\cdot\text{h/L}\), and each profile has Cmax \(4+8=12\ \text{mg/L}\). For the narrow peak, the added area is \((8\times0.5)/2=2\ \text{mg}\cdot\text{h/L}\), giving total AUC \(26\ \text{mg}\cdot\text{h/L}\). For the wider peak, the added area is \((8\times2)/2=8\ \text{mg}\cdot\text{h/L}\), giving total AUC \(32\ \text{mg}\cdot\text{h/L}\).

Cmax is identical, but total AUC differs by \(6\ \text{mg}\cdot\text{h/L}\). A peak-height summary alone does not show how long the concentration remained elevated.

Conditions, Diagnostics, and Limits

The area-versus-height argument is most direct when comparing profiles over the same time interval, with the same concentration units and a clearly specified baseline. In the triangular examples, the peak is an additive increase above that baseline, stays within the interval, and has a known width and height. To infer the stated Cmax change, the peak apex must be among the observed samples and must exceed every other sampled concentration. To infer the stated AUC change, the area of the change must be represented adequately by the profile or its calculation.

For a nonconstant baseline, the triangle still contributes \(hw/2\) of area if it is an additive triangular change. But the relative change in AUC depends on the original total AUC, while the relative change in Cmax depends on the original maximum and on where the peak occurs. If the peak is not at the original maximum, or a different time point remains the maximum, the increase in Cmax may be less than \(h\), or zero. If one profile rises in one region and falls in another, the net AUC change is the signed integral of those differences; a local peak increase can be offset elsewhere.

Several checks help identify when a simple peak comparison is not adequate:

  • Check whether the apex is observed. Observed Cmax is calculated from sampled concentrations. A narrow peak between scheduled samples may be missed, so the observed Cmax change can be smaller than the change in the true maximum. The earlier material on observed Cmax and sampling density addresses that distinction.
  • Check whether a small AUC change is credible. A narrow feature may be poorly represented by the available concentration-time information. The integral of the true profile and an AUC estimate from sampled data are not automatically identical; the earlier material on AUC sensitivity to sampling density is relevant to that issue.
  • Check the rest of the profile. A change in absorption or elimination may alter broad parts of the curve, not just a local peak. In that case, comparing peak height alone with a single triangular area can be misleading.
  • Check that the same interval is being compared. AUC values over different intervals do not answer the same question. This discussion concerns total AUC on a common, specified interval; it does not establish how a restricted time window would behave.

AUC and Cmax are complementary, not competing, summaries. AUC is useful for cumulative exposure over the interval, while Cmax highlights the observed highest concentration. A full profile can reveal timing and shape differences that either scalar summary hides. Neither metric, by itself, establishes why a profile changed or whether that change has a clinical consequence.

Common pitfall: “The AUC is nearly unchanged, so the peak must also be nearly unchanged.” This is incorrect: a narrow peak can add little area yet substantially raise the maximum. The reverse inference is also unsafe: equal Cmax values do not establish equal AUC. Compare the metric that matches the question, and inspect the profile when shape matters.

Check Your Understanding

Use the assumptions in each question; the solutions follow.

  1. A baseline of \(8\ \text{mg/L}\) lasts \(5\ \text{h}\). A fully observed triangular peak rises \(4\ \text{mg/L}\) above baseline and is \(0.5\ \text{h}\) wide. Find the relative increases in AUC and Cmax.
  2. Two triangular peaks have the same height above baseline and the same location, but one is twice as wide. What happens to their added areas and their Cmax values?
  3. Two profiles have equal total AUC. Must they have equal Cmax? Explain using the role of width.
  4. A narrow peak occurs between sampling times. Which conclusion about observed Cmax is unsafe, and why?

Solutions. 1. Baseline AUC is \(8\times5=40\ \text{mg}\cdot\text{h/L}\). Added area is \((4\times0.5)/2=1\ \text{mg}\cdot\text{h/L}\), so relative AUC increase is \(1/40=2.5\%\). Cmax rises from \(8\) to \(12\ \text{mg/L}\), a \(4/8=50\%\) increase. 2. The wider triangle has twice the area because its height is unchanged and its base doubles; both peaks have the same maximum if both apexes are observed. 3. No. Equal area can arise from a lower, wider peak or a taller, narrower peak; the worked equal-area comparison gives the same AUC with different Cmax values. 4. It is unsafe to conclude that observed Cmax equals the true maximum, or that observed Cmax captures the full peak change. The recorded maximum uses sampled values and may miss the between-sample apex.

Previous

Next

Back to Tutorials