Steady-State Concentrations During Constant-Rate Infusion

Why the Infusion Rate Has a Threshold

With linear elimination, increasing concentration can increase the elimination rate without a fixed upper bound. Michaelis–Menten elimination is different: its rate approaches a maximum, $V_{\max}$. A constant infusion can therefore be balanced only if its input rate is below that maximum. This threshold determines whether a finite steady-state concentration exists at all.

The result is useful for interpreting constant-infusion experiments and for checking model predictions. In particular, a mathematically large steady-state concentration is not the same as no steady state: below the threshold a finite equilibrium exists, although it may be extremely high. At or above the threshold, there is no finite concentration at which input and elimination balance.

Model and Assumptions

Consider a well-mixed compartment with constant volume $V$. Drug enters at a constant rate $R_0$, and elimination follows the Michaelis–Menten rate law with maximum rate $V_{\max}$ and concentration constant $K_m$. Let $C(t)$ denote concentration at time $t$. Assume $V>0$, $K_m>0$, $V_{\max}>0$, and $R_0\geq 0$, with all rate and concentration units used consistently.

Definition: Under these assumptions, the concentration balance is $$ V\frac{dC}{dt}=R_0-\frac{V_{\max}C}{K_m+C}, $$ or, equivalently, $$ \frac{dC}{dt}=\frac{1}{V}\left(R_0-\frac{V_{\max}C}{K_m+C}\right). $$ A finite steady-state concentration $C_{\mathrm{ss}}$ is a finite, nonnegative concentration at which $dC/dt=0$, so the infusion rate exactly equals the elimination rate.

Here $R_0$ and $V_{\max}$ have units of amount per time, $V$ has units of volume, and $K_m$ and $C$ have units of amount per volume. Thus $dC/dt$ has units of concentration per time. The model describes a fixed set of parameters and a single concentration state; its conclusions apply when infusion and the stated elimination mechanism remain unchanged.

Deriving the Finite Steady State

At a steady state, set the rate of change to zero. The balance equation then requires $R_0=V_{\max}C_{\mathrm{ss}}/(K_m+C_{\mathrm{ss}})$. Solving this equation also reveals the condition for a finite solution.

R_0(K_m+C_{\mathrm{ss}})=V_{\max}C_{\mathrm{ss}} \quad\Longrightarrow\quad R_0K_m=(V_{\max}-R_0)C_{\mathrm{ss}}.

If $0\leq R_0<V_{\max}$, the solution is nonnegative and finite:

C_{\mathrm{ss}}=\frac{K_mR_0}{V_{\max}-R_0}.

If $R_0=0$, this gives $C_{\mathrm{ss}}=0$. If $R_0=V_{\max}$, the equation would require $V_{\max}K_m=0$, which contradicts the positive parameter assumptions. If $R_0>V_{\max}$, its algebraic solution would be negative, so it cannot be a nonnegative concentration. These cases are not merely limitations of the rearrangement: the rate equation shows that no finite nonnegative equilibrium exists when input reaches or exceeds $V_{\max}$.

Key result: A finite nonnegative steady-state concentration exists exactly when $0\leq R_0<V_{\max}$. For a positive infusion rate below the maximum elimination rate, $C_{\mathrm{ss}}=K_mR_0/(V_{\max}-R_0)$.

This equilibrium is stable under the model. Define the concentration rate function $f(C)=[R_0-V_{\max}C/(K_m+C)]/V$. Its derivative for $C\geq0$ is $f'(C)=-V_{\max}K_m/[V(K_m+C)^2]$, which is negative. Moreover, when a finite steady state exists, $f(C)>0$ below $C_{\mathrm{ss}}$ and $f(C)<0$ above it. Concentration therefore moves toward the equilibrium: it rises if it begins below it and falls if it begins above it. This is a local direction-of-change argument, not a claim that the concentration reaches equilibrium in a finite time.

Worked Examples

Worked Example: A Finite Steady State Below the Threshold

Suppose a hypothetical infusion delivers $R_0=3$ mg/h into a compartment with $V=18$ L. Let $V_{\max}=12$ mg/h and $K_m=5$ mg/L. The constant-parameter model applies, and $3<12$ mg/h, so a finite steady state is possible.

Substitute into the steady-state formula:

C_{\mathrm{ss}} =\frac{(5\ \mathrm{mg/L})(3\ \mathrm{mg/h})} {12\ \mathrm{mg/h}-3\ \mathrm{mg/h}} =\frac{15}{9}\ \mathrm{mg/L} =\frac{5}{3}\ \mathrm{mg/L} \approx 1.667\ \mathrm{mg/L}.

Check the balance using the unrounded concentration $5/3$ mg/L. The elimination rate is $12(5/3)/(5+5/3)=12(5/3)/(20/3)=3$ mg/h, exactly the infusion rate. The volume affects the speed of concentration change, but not the equilibrium obtained by balancing amount rates.

Worked Example: Near-Maximal Input Produces a Large Concentration

Consider a hypothetical drug with $V_{\max}=10$ mg/h and $K_m=4$ mg/L. Compare constant infusion rates of 8 mg/h and 9 mg/h. Both are below $V_{\max}$, so each has a finite steady state.

For $R_0=8$ mg/h, $C_{\mathrm{ss}}=(4)(8)/(10-8)=16$ mg/L. For $R_0=9$ mg/h, $C_{\mathrm{ss}}=(4)(9)/(10-9)=36$ mg/L. Increasing the input rate by 12.5% in this comparison increases the predicted steady-state concentration from 16 to 36 mg/L, a 125% increase. Both values satisfy their respective balances: the elimination rate at 16 mg/L is $10(16)/(4+16)=8$ mg/h; at 36 mg/L it is $10(36)/(4+36)=9$ mg/h.

The concentration-to-$K_m$ ratio makes the same behavior easy to see: $C_{\mathrm{ss}}/K_m=R_0/(V_{\max}-R_0)$. At 8 mg/h the ratio is $8/2=4$; at 9 mg/h it is $9/1=9$. As the input approaches the maximum rate from below, the denominator shrinks and the required concentration grows rapidly.

Behavior at and Above the Threshold

For finite $C\geq0$, the Michaelis–Menten elimination rate is less than $V_{\max}$, since $C/(K_m+C)<1$. At the threshold $R_0=V_{\max}$, therefore, input still exceeds elimination at every finite concentration. The concentration continues to rise, even though its rate of rise becomes smaller as concentration increases.

At the threshold, the differential equation simplifies to $dC/dt=V_{\max}K_m/[V(K_m+C)]$. Separating variables and integrating from $C_0$ at time $t_0$ gives

K_m(C-C_0)+\frac{C^2-C_0^2}{2} =\frac{V_{\max}K_m}{V}(t-t_0).

The left side increases without bound as $C$ increases, so concentration does not settle at any finite value. For large $t$, the quadratic term dominates, giving $C(t)$ growth proportional to $\sqrt{t}$, rather than convergence to a finite plateau. If $R_0>V_{\max}$, then \[ \lim_{C\to\infty}\frac{dC}{dt}=\frac{R_0-V_{\max}}{V}>0. \] Thus concentration ultimately grows approximately linearly, with limiting slope $(R_0-V_{\max})/V$.

Worked Example: Checking the Threshold and Above-Threshold Cases

Take hypothetical values $V=20$ L, $K_m=4$ mg/L, and $V_{\max}=10$ mg/h. First set $R_0=10$ mg/h and $C_0=0$. At this threshold, the integrated equation becomes $4C+C^2/2=(10)(4)t/20=2t$, with $C$ in mg/L and $t$ in hours. At $t=10$ h,

4C+\frac{C^2}{2}=20 \quad\Longrightarrow\quad C^2+8C-40=0 \quad\Longrightarrow\quad C=-4+\sqrt{56}\approx3.483\ \mathrm{mg/L}.

The other root is negative and is not a physical concentration. This value at 10 h is not a steady state; the derivative remains positive at every finite concentration. Now set $R_0=12$ mg/h and evaluate the rate at $C=20$ mg/L. Elimination is $10(20)/(4+20)=8.333$ mg/h, so

\frac{dC}{dt} =\frac{12-8.333\ldots}{20} =0.1833\ldots\ \mathrm{mg/L/h}.

This positive rate is consistent with input exceeding the maximum elimination rate. At very high concentrations, the rate approaches $(12-10)/20=0.1000$ mg/L/h. Neither case has a finite steady state.

Why Near-Threshold Predictions Need Care

The formula also quantifies sensitivity to the infusion rate. Differentiating with respect to $R_0$, while holding $K_m$ and $V_{\max}$ fixed, gives $dC_{\mathrm{ss}}/dR_0=K_mV_{\max}/(V_{\max}-R_0)^2$. This sensitivity increases sharply as $R_0$ approaches $V_{\max}$. Small errors in the infusion rate or in estimated $V_{\max}$ can then produce large changes in the predicted concentration.

A related issue is the time needed to get close to the equilibrium. Linearizing the rate equation near a sub-threshold steady state gives the local time scale $\tau=V(K_m+C_{\mathrm{ss}})^2/(V_{\max}K_m)$. This is a local approximation to the rate of approach, not a universal exact time to steady state. For the hypothetical values $V=20$ L, $K_m=4$ mg/L, and $V_{\max}=10$ mg/h used above, $R_0=8$ mg/h gives $C_{\mathrm{ss}}=16$ mg/L and $\tau=20(20)^2/(10\cdot4)=200$ h. At $R_0=9$ mg/h, $C_{\mathrm{ss}}=36$ mg/L and $\tau=20(40)^2/(10\cdot4)=800$ h. The sharp increase warns that an observed concentration may still be changing substantially even when a finite equilibrium exists.

For comparison, with linear elimination at rate $CL\cdot C$, where $CL$ is constant clearance, constant infusion has the finite equilibrium $C_{\mathrm{ss}}=R_0/CL$ for every finite $R_0$. Michaelis–Menten elimination differs because its rate has the ceiling $V_{\max}$. A numerical solution of the differential equation can describe the time course, but it cannot create a finite equilibrium when the rate balance has none. The steady-state formula is the direct check on existence; numerical trajectories should be interpreted in light of that check.

Diagnostics and Common Mistakes

Before interpreting an estimated steady state, verify that the model and inputs are compatible with the calculation. Confirm that infusion rate and $V_{\max}$ use the same amount-per-time units, that $K_m$ and concentration use the same concentration units, and that $V$ is expressed as a volume. Then compare $R_0$ with $V_{\max}$ before evaluating the formula. A predicted negative $C_{\mathrm{ss}}$, or a denominator $V_{\max}-R_0$ that is zero or negative, signals that the finite nonnegative steady-state formula does not apply.

Wrong reasoning: “The concentration is high, so elimination must exceed input and a steady state will eventually occur.” Correction: Michaelis–Menten elimination is bounded above by $V_{\max}$; if $R_0\geq V_{\max}$, no finite concentration can balance input.
Wrong reasoning: “A rate just below $V_{\max}$ should produce a concentration close to $K_m$.” Correction: $C_{\mathrm{ss}}/K_m=R_0/(V_{\max}-R_0)$; the ratio can be much greater than one near the threshold.
Wrong reasoning: “The formula contains $V$, so changing volume changes the steady-state concentration.” Correction: $V$ cancels from the equilibrium balance. It affects the transient rate of concentration change, not this steady-state value.
Wrong reasoning: “A concentration that still rises must mean the model has no steady state.” Correction: A sub-threshold model has a finite stable equilibrium, but the approach can be slow, especially near $V_{\max}$. Distinguish slow convergence from an infusion rate at or above the threshold.

These conclusions depend on the stated assumptions: constant infusion, fixed volume, and Michaelis–Menten elimination with unchanged parameters. If input or elimination changes over time, the balance equation and the interpretation of a steady state must be reconsidered. In a population model, the derivation is for a given set of individual parameter values; it does not imply that substituting population-average parameters gives the average individual steady state.

Check Your Understanding

Answer each question, then compare with the solutions below.

For $R_0=6$ mg/h, $V_{\max}=10$ mg/h, and $K_m=3$ mg/L, calculate the finite steady-state concentration and verify the elimination rate.
With $V_{\max}=10$ mg/h and $K_m=3$ mg/L, what happens to the finite steady-state concentration as $R_0$ approaches 10 mg/h from below?
For $R_0=11$ mg/h and $V_{\max}=10$ mg/h, can a finite steady state exist? What is the limiting concentration rate at high concentration if $V=25$ L?
Does doubling $V$, while leaving $R_0$, $V_{\max}$, and $K_m$ unchanged below the threshold, double $C_{\mathrm{ss}}$? Explain.

Solutions. (1) Since $6<10$, $C_{\mathrm{ss}}=(3)(6)/(10-6)=4.5$ mg/L. Elimination is $10(4.5)/(3+4.5)=45/7.5=6$ mg/h, matching input. (2) $C_{\mathrm{ss}}=3R_0/(10-R_0)$ mg/L, which grows without bound as $R_0$ approaches 10 mg/h from below; it remains finite for each strictly sub-threshold rate. (3) No. At high concentration the limiting rate is $(11-10)/25=0.0400$ mg/L/h, positive. (4) No. Volume cancels from the steady-state balance, so $C_{\mathrm{ss}}$ is unchanged; doubling volume reduces the instantaneous concentration rate for the same net amount input.