Saturable Absorption

Why absorption can depend on dose

A first-order absorption model makes the input rate proportional to the amount remaining at the absorption site. That is a useful description when absorptive capacity is not limiting. If transport or dissolution has a finite capacity, however, a larger dose need not enter the systemic circulation at a proportionally larger rate. The input can approach a ceiling, while the excess drug remains at the absorption site for longer.

This distinction matters because concentration profiles reflect both the amount that ultimately reaches the circulation and the timing of its arrival. Under complete absorption and linear elimination, saturable absorption can substantially alter peak concentration and time to peak while leaving total exposure proportional to dose. A lower-than-expected exposure requires an additional explanation, such as incomplete absorption or a dose-dependent fraction reaching the circulation.

A capacity-limited absorption model

Let \(A_g(t)\) be the drug amount remaining at the absorption site, in mg, and let \(D\) be the administered oral dose in mg. A simple saturable absorption rate is a Michaelis–Menten function of the amount at that site. Here \(V_{\max,a}\) is the maximum absorption rate, in mg/h, and \(K_a\) is the depot amount at which the rate is half of its maximum, in mg.

Definition: In the saturable absorption model, the absorption-site amount and absorption rate satisfy

\frac{dA_g}{dt}=-R_{\mathrm{abs}}(t), \qquad R_{\mathrm{abs}}(t)=\frac{V_{\max,a}A_g(t)}{K_a+A_g(t)}, \qquad A_g(0)=D.

The systemic input rate is \(F_aR_{\mathrm{abs}}(t)\), where \(F_a\) is a dose-independent fraction reaching the systemic circulation. Thus the model allows a fixed fraction to be lost, but does not itself make that fraction depend on dose.

When \(A_g\ll K_a\), the rate is approximately \((V_{\max,a}/K_a)A_g\), so absorption is approximately first order, with \(k_a=V_{\max,a}/K_a\) in h\(^{-1}\). When \(A_g\gg K_a\), the rate is approximately \(V_{\max,a}\), nearly independent of the depot amount. The same dose can therefore move through a capacity-limited phase and then a more nearly first-order terminal phase as the depot empties.

For interpreting concentration, suppose the systemic amount is \(A_c(t)\), the apparent distribution volume is \(V\), and disposition is linear with clearance \(CL\). Then \(C(t)=A_c(t)/V\), and

\frac{dA_c}{dt}=F_aR_{\mathrm{abs}}(t)-CL\,C(t), \qquad \frac{dC}{dt}=\frac{F_aR_{\mathrm{abs}}(t)}{V}-kC(t), \qquad k=\frac{CL}{V}.

These equations assume a single well-mixed systemic compartment, first-order systemic elimination, fixed \(F_a\), and no competing loss from the depot other than the specified absorption process. They also assume that the administered dose is represented by the initial depot amount. More elaborate disposition structures can be used, but the absorption-rate interpretation remains the same.

How dose changes the absorption time

The depot equation gives a direct relationship between dose and the time required to reach a specified remaining amount. Separating variables gives

dt=-\frac{K_a+A_g}{V_{\max,a}A_g}\,dA_g =-\frac{1}{V_{\max,a}}\left(1+\frac{K_a}{A_g}\right)dA_g.

Integrating from \(A_g=D\) at time zero to \(A_g=qD\), where \(0<q<1\), gives the exact time \(t_q\) at which fraction \(q\) of the dose remains:

t_q=\frac{(1-q)D+K_a\ln(1/q)}{V_{\max,a}}.

The term \((1-q)D/V_{\max,a}\) grows in direct proportion to dose. Therefore, at doses large relative to \(K_a\), a greater amount must be absorbed while the input is near its capacity, extending the absorption period. This result concerns the depot trajectory; it does not alone determine the concentration peak, which also depends on systemic elimination.

Worked Example: Time Until 90% of the Dose Has Left the Depot

Consider a hypothetical dose \(D=200\) mg with \(V_{\max,a}=40\) mg/h and \(K_a=20\) mg. Assume the stated depot model applies and ask how long it takes for 90% of the dose to leave the absorption site. At that point \(q=0.10\).

Using the integrated depot equation,

t_{0.10} =\frac{(1-0.10)(200\ \mathrm{mg})+(20\ \mathrm{mg})\ln(1/0.10)} {40\ \mathrm{mg/h}} =\frac{180+46.05}{40}\ \mathrm{h} =5.651\ \mathrm{h}.

The units are hours because the numerator is an amount and the denominator is an amount per hour. The dose is ten times \(K_a\), so the initial absorption rate is already close to its capacity: \(40(200)/(20+200)=36.36\) mg/h. The result is consistent with a prolonged, nearly capacity-limited early input, followed by a slower terminal phase.

Peak concentration and total exposure

For any concentration maximum occurring at an interior time, the concentration derivative is zero. The peak condition is therefore that systemic input equals systemic elimination at that instant:

\left.\frac{dC}{dt}\right|_{t=t_{\max}}=0 \quad\Longrightarrow\quad F_aR_{\mathrm{abs}}(t_{\max})=CL\,C_{\max}.

A capacity-limited input can keep the early rate from increasing in proportion to dose. Additional drug may instead sustain input for longer. Consequently, peak concentration often increases less than proportionally with dose and the peak occurs later. “Less than proportional” is a common pattern, not a universal numerical rule independent of the disposition model and parameter values.

Total exposure has a different result under the stated assumptions. Integrating the systemic amount equation from time zero to infinity, with \(A_c(0)=A_c(\infty)=0\), gives

0=F_a\int_0^\infty R_{\mathrm{abs}}(t)\,dt -CL\int_0^\infty C(t)\,dt.

The depot equation implies \(\int_0^\infty R_{\mathrm{abs}}(t)\,dt=D\) when the depot empties completely. Thus

AUC_{0-\infty}=\int_0^\infty C(t)\,dt=\frac{F_aD}{CL}.

This is a mass-balance result, not a claim that concentration profiles are dose-scaled copies. Saturable absorption changes the time course, but with complete absorption, fixed \(F_a\), and linear disposition it does not change the dose-normalized total AUC. Incomplete follow-up can yield a smaller observed partial AUC because some input arrives later. Nonlinear disposition or dose-dependent \(F_a\) also removes the assumptions behind the result.

Worked Example: Low-Dose Approximation and the Peak

Suppose a hypothetical 5 mg dose has \(K_a=100\) mg and \(V_{\max,a}=100\) mg/h. Let \(F_a=0.80\), \(V=20\) L, and \(CL=4\) L/h, so \(k=CL/V=0.20\) h\(^{-1}\). Since \(D/K_a=0.05\), the depot is small relative to \(K_a\), and the first-order approximation is reasonable: \(k_a=V_{\max,a}/K_a=1.00\) h\(^{-1}\).

For this approximation, the concentration after a single oral dose is \(C(t)=(F_aD/V)\,[k_a/(k_a-k)](e^{-kt}-e^{-k_at})\). The peak time follows by setting the time derivative to zero, giving \(t_{\max}=\ln(k_a/k)/(k_a-k)\). Substitution gives

t_{\max}=\frac{\ln(1.00/0.20)}{1.00-0.20} =\frac{\ln(5)}{0.80}=2.012\ \mathrm{h}.

At that time, the dimensionless shape factor is \((1/0.80)(e^{-0.20(2.012)}-e^{-1.00(2.012)})=1.25(0.6687-0.1338)=0.6686\). Therefore,

C_{\max}\approx \frac{(0.80)(5\ \mathrm{mg})}{20\ \mathrm{L}}(0.6686) =0.1337\ \mathrm{mg/L}.

The linear-disposition AUC is \(F_aD/CL=(0.80)(5\ \mathrm{mg})/(4\ \mathrm{L/h})=1.000\) mg·h/L. These are approximate peak calculations because the exact saturable model is only approximated as first order in this low-dose regime. The AUC result, in contrast, follows exactly from complete absorption and linear disposition.

Worked Example: A High-Dose Capacity Limit

For a separate hypothetical case, take \(D=1000\) mg, \(K_a=10\) mg, \(V_{\max,a}=100\) mg/h, \(F_a=0.80\), \(V=20\) L, and \(CL=4\) L/h. Initially, \(R_{\mathrm{abs}}=100(1000)/(1010)=99.01\) mg/h, close to the 100 mg/h capacity. The systemic input is initially about 79.21 mg/h.

Since \(D/K_a=100\), a useful rough approximation is a systemic input near \(F_aV_{\max,a}=80\) mg/h while the depot is large, lasting on the order of \(D/V_{\max,a}=10\) h. Under an ideal constant-input approximation lasting \(T=10\) h, the concentration during input is \(C(t)=(80/4)(1-e^{-0.20t})\) mg/L. The peak at the end of this approximate input is

C_{\max,\mathrm{approx}} =\frac{80\ \mathrm{mg/h}}{4\ \mathrm{L/h}}(1-e^{-0.20(10)}) =20(1-e^{-2})=17.29\ \mathrm{mg/L}.

This is a limiting approximation, not the exact peak of the saturable differential-equation model: the actual rate declines as the depot empties, particularly near the end. It illustrates why a large dose can produce a prolonged input rather than a peak proportional to dose. The total AUC under the stated complete-absorption assumptions is nevertheless \(F_aD/CL=(0.80)(1000)/4=200\) mg·h/L.

Compared with the low-dose example, the dose and AUC are each 200 times as large. The illustrative peak is about \(17.29/0.1337=129.3\) times as large, rather than 200 times. The peak comparison is approximate and uses different dose-regime approximations; it demonstrates the mechanism, not a universal scaling law.

Diagnostics, alternatives, and limitations

A saturable absorption model is most useful when data support a dose-dependent input rate, not merely when a nonlinear curve can be fitted. Helpful evidence includes concentration profiles at multiple doses with sufficiently dense early sampling, a longer apparent input phase at higher doses, and peaks that rise less than proportionally while the dose-normalized AUC remains broadly consistent with complete absorption and linear disposition. These are patterns to investigate, not proof of a unique mechanism.

Compare the model with plausible alternatives. First-order absorption gives an input proportional to depot amount and may be adequate when observed doses remain well below \(K_a\). A zero-order input specifies a constant rate over a defined duration; it can approximate the high-depot phase, but does not by itself describe the smooth transition to a declining rate. Dose-dependent oral bioavailability changes the fraction reaching the circulation, whereas saturable absorption changes the rate at which that fraction arrives. The two can produce superficially similar dose-related concentration differences and may be hard to distinguish from sparse concentration data.

Common warning signs include systematic residuals around the peak or in the late absorption phase, estimated \(V_{\max,a}\) or \(K_a\) values that are poorly determined, and fitted predictions that imply an implausible input rate or absorption duration. If observed dose-normalized AUC falls as dose increases, the complete-absorption, fixed-\(F_a\), linear-disposition model cannot explain that trend by saturation of the input rate alone. Consider whether follow-up missed delayed absorption, whether the fraction absorbed changes, or whether disposition is nonlinear. A model should not be interpreted mechanistically just because it improves fit.

The parameters can be difficult to identify separately. At low doses, the data may mainly inform the ratio \(V_{\max,a}/K_a\), not both quantities individually. At high doses, observations may reveal the capacity \(V_{\max,a}\), but estimating the transition governed by \(K_a\) requires data while the depot passes through amounts near \(K_a\). Dose groups that do not span these regimes, or sampling that misses the absorption phase, provide limited information. Numerical integration of the differential equations is generally needed for fitting the full model; the approximations above are for interpretation and checking, not substitutes for evaluating the model over a concentration-time profile.

Common mistake

Common mistake: “Absorption is saturated, so the total AUC must fall at higher doses.” A capacity-limited rate can delay input and blunt the peak, but under complete absorption, fixed \(F_a\), and linear disposition, mass balance gives \(AUC_{0-\infty}=F_aD/CL\). A lower AUC per dose needs an additional explanation or a violation of those assumptions.

Key takeaway: Saturable absorption limits the instantaneous input rate, so increasing dose can extend absorption and raise peak concentration less than proportionally. With complete absorption and linear disposition, that timing change alone does not alter the dose-proportional total AUC.

Check Your Understanding

Answer each question, then compare with the solutions below.

For \(A_g\gg K_a\), what does the absorption rate approach, and what does that imply for the early input from a larger dose?
With fixed \(F_a\) and \(CL\), what is the predicted \(AUC_{0-\infty}\) for a fully absorbed dose \(D\)?
Suppose dose-normalized total AUC declines at higher doses. Is saturable absorption alone, under the model assumptions, a sufficient explanation?
Why can a low-dose data set provide weak information about \(V_{\max,a}\) and \(K_a\) separately?

Solutions. (1) The rate approaches \(V_{\max,a}\), so increasing the depot amount much further produces little increase in the early absorption rate; more drug is instead absorbed over a longer period. (2) \(AUC_{0-\infty}=F_aD/CL\), obtained by integrating the systemic amount balance when the depot empties and systemic amount returns to zero. (3) No. Under complete absorption, fixed \(F_a\), and linear disposition, saturation changes timing but not dose-normalized total AUC; consider incomplete absorption, dose-dependent \(F_a\), nonlinear disposition, or incomplete observation. (4) When \(A_g\ll K_a\), the rate is approximately \((V_{\max,a}/K_a)A_g\), so low-dose profiles primarily identify the ratio \(V_{\max,a}/K_a\), rather than the two parameters independently.

What You'll Learn