Why Separate Flow From Partitioning?
A tissue concentration curve reflects both how rapidly drug is delivered by blood and how much drug the tissue holds relative to blood at equilibrium. These influences are related in the governing equation, but they leave different signatures in a concentration-time profile. With the arterial concentration held constant, changing blood flow changes the rate of approach to equilibrium. Changing the tissue-to-blood partition coefficient changes the equilibrium concentration and also changes the rate of approach.
The distinction matters when interpreting a curve or comparing model predictions. A tissue concentration that rises more slowly does not, by itself, prove that blood flow is lower: a larger partition coefficient can also slow equilibration. The plateau, the arterial input, and the other model quantities must be considered together.
Model and Assumptions
Consider one well-mixed tissue with volume \(V_t\), supplied by blood flow \(Q_t\). Let \(C_t(t)\) be tissue concentration, \(C_a(t)\) arterial blood concentration, and \(C_v(t)\) venous blood concentration leaving the tissue. Under the perfusion-limited assumption, drug equilibrates rapidly between tissue and blood at the tissue boundary. The tissue-to-blood partition coefficient \(K_{t:b}\) is the equilibrium ratio of tissue to blood concentration.
The model assumes a spatially uniform tissue concentration, instantaneous equilibration at the tissue-blood interface, and fixed \(Q_t\), \(V_t\), and \(K_{t:b}\) during the interval considered. It also assumes that the arterial concentration is known and that there is no tissue metabolism or other unmodeled source or sink. These assumptions make this a single-compartment perfusion-limited description; they should not be treated as guaranteed properties of a real tissue.
Substituting \(C_v=C_t/K_{t:b}\) into the mass balance gives
Here \(\lambda\) has units of inverse time. This form makes the two roles visible: \(K_{t:b}C_a\) is the equilibrium concentration under a constant arterial concentration, while \(\lambda\) governs the speed of approach to that value.
Deriving the Concentration-Time Curve
Suppose the arterial concentration is held at a constant value \(C_a\) from time zero onward. At equilibrium, the derivative is zero, so the model requires \(C_t=K_{t:b}C_a\). To describe the transient, define the deviation from equilibrium as \(Y(t)=C_t(t)-K_{t:b}C_a\). Because \(C_a\) is constant, the model gives \(dY/dt=-\lambda Y\).
If \(Y(0)=0\), then the tissue is already at equilibrium and \(Y(t)=0\) for all \(t\). If \(Y(0)\ne 0\), divide by \(Y\) and integrate from zero to \(t\):
Combining these cases gives the same solution, including when the initial tissue concentration equals equilibrium:
After one time constant, the remaining deviation from equilibrium is \(e^{-1}\), or about 36.8%, of its initial value. Thus the fraction of the total change completed by time \(t\) is \(1-e^{-t/\tau}\). This normalized fraction is useful because it separates the shape of the approach from the absolute equilibrium concentration.
Worked Example: Doubling Flow Changes the Approach Rate
Suppose a hypothetical tissue has \(V_t=1.20\ \text{L}\), \(K_{t:b}=2.00\), and a constant arterial concentration \(C_a=1.20\ \text{mg/L}\). Its initial tissue concentration is \(0\ \text{mg/L}\). Compare the predicted curve at \(t=4.00\ \text{min}\) when \(Q_t=0.300\ \text{L/min}\) and when flow is doubled to \(0.600\ \text{L/min}\). The partition coefficient and tissue volume are held fixed.
In both cases the plateau is \(K_{t:b}C_a=(2.00)(1.20)=2.40\ \text{mg/L}\). At the original flow,
At doubled flow,
Both curves approach \(2.40\ \text{mg/L}\), but the doubled-flow curve is farther along at four minutes. The time constants are \(8.00\ \text{min}\) and \(4.00\ \text{min}\), respectively. Under the stated assumptions, this is a rate change, not a change in the equilibrium concentration.
Worked Example: Changing Partitioning Changes the Plateau and Rate
Use the same hypothetical \(V_t=1.20\ \text{L}\), \(Q_t=0.600\ \text{L/min}\), \(C_a=1.20\ \text{mg/L}\), and initial concentration \(C_t(0)=0\). Compare \(K_{t:b}=2.00\) with \(K_{t:b}=4.00\).
For \(K_{t:b}=2.00\), the plateau is \(2.40\ \text{mg/L}\) and \(\lambda=0.2500\ \text{min}^{-1}\). At four minutes, \(C_t(4.00)=2.40(1-e^{-1})=1.517\ \text{mg/L}\).
For \(K_{t:b}=4.00\), the plateau is \((4.00)(1.20)=4.80\ \text{mg/L}\), while \[ \lambda=\frac{0.600}{(1.20)(4.00)} =0.1250\ \text{min}^{-1}. \] Thus the four-minute concentration is
The larger partition coefficient produces a higher plateau, but the approach is slower: its time constant is \(8.00\ \text{min}\), compared with \(4.00\ \text{min}\) for \(K_{t:b}=2.00\). At four minutes the higher-partition curve is only slightly above the other curve, despite having twice the plateau. This illustrates why a single early concentration is not enough to identify the mechanism.
Using a Curve to Separate the Effects
If arterial concentration is constant and the curve reaches a clear plateau, the plateau provides information about \(K_{t:b}\): \(K_{t:b}=C_{t,\mathrm{eq}}/C_a\). The normalized transient provides information about the rate parameter. Define the fraction of the total change completed by time \(t\) as
For \(0<f(t)<1\), rearranging gives \(\lambda=-\ln(1-f(t))/t\). If \(V_t\) is known, flow can then be recovered as \(Q_t=\lambda V_tK_{t:b}\). The fraction is undefined when the initial concentration already equals the plateau, because there is no observed change to normalize; such a curve contains no transient-rate information.
Worked Example: Estimating Partitioning and Flow From a Hypothetical Curve
A hypothetical tissue has known volume \(V_t=1.00\ \text{L}\). A constant arterial concentration of \(1.00\ \text{mg/L}\) is maintained. The tissue concentration starts at \(0\ \text{mg/L}\), approaches a measured plateau of \(3.00\ \text{mg/L}\), and reaches \(1.50\ \text{mg/L}\) at \(6.00\ \text{min}\). Assume the stated model and that the plateau and time-point measurements are exact for this calculation.
The plateau gives \(K_{t:b}=C_{t,\mathrm{eq}}/C_a=(3.00\ \text{mg/L})/(1.00\ \text{mg/L})=3.00\). At six minutes, the curve has completed half of its total change: \(f(6.00)=(1.50-0)/(3.00-0)=0.500\). Therefore,
Using \(Q_t=\lambda V_tK_{t:b}\),
Under the model assumptions, the plateau identifies the partition coefficient and the normalized rise identifies \(\lambda\); with the known volume, these yield the flow estimate. In real data, measurement error, uncertain arterial input, or failure to observe the plateau can make these estimates less certain.
Diagnostics, Alternatives, and Limitations
A useful diagnostic is to check whether the curve is compatible with a single exponential approach to a stable plateau. After estimating the plateau, plot or inspect the remaining deviation \(C_{t,\mathrm{eq}}-C_t(t)\) over time. Under this model, that deviation has an exponential form. Systematic departures—such as changing apparent rates across time, overshooting the plateau, or a plateau inconsistent with \(K_{t:b}C_a\)—suggest that one or more assumptions may not hold, or that the arterial input is not actually constant.
A complementary comparison is to plot the normalized fraction \(f(t)\) rather than raw tissue concentration. With the same \(V_t\) and \(K_{t:b}\), changing flow changes the rate at which \(f(t)\) rises but leaves the plateau unchanged. With a changed \(K_{t:b}\), the raw plateau changes and the rate parameter changes as well. If \(V_t\), \(K_{t:b}\), or the arterial input also differs between conditions, the comparison cannot be attributed to flow alone.
A common alternative is to compare concentrations at one selected time. That is often inadequate: different combinations of plateau and rate can give similar concentrations at an early time point. A curve observed through its plateau is more informative, and the normalized transient can then be assessed separately. Even with a full curve, estimates depend on the model assumptions and on adequate information about \(C_a\) and \(V_t\). In particular, a changing arterial concentration drives the tissue response over time, so a tissue curve alone cannot be interpreted as though it arose from a constant step input.
The central distinction is conditional, not absolute: with partitioning, tissue volume, and arterial concentration held fixed, flow changes the approach rate but not the plateau. Partitioning changes both the plateau and the approach rate. That distinction is a guide for interpreting this model, not proof that a real tissue follows it.
Check Your Understanding
Use the constant-arterial-concentration model. Solutions are included after the questions.
- If \(Q_t\) doubles while \(V_t\), \(K_{t:b}\), and \(C_a\) remain fixed, what happens to the time constant and the plateau?
- A tissue has \(C_a=0.800\ \text{mg/L}\) and \(K_{t:b}=2.50\). What is its equilibrium concentration under the model?
- For a tissue with \(V_t=1.00\ \text{L}\), \(K_{t:b}=2.00\), and \(Q_t=0.400\ \text{L/min}\), calculate \(\lambda\) and \(\tau\).
- Why can one early tissue concentration not usually distinguish a lower flow from a larger partition coefficient?
Solutions. (1) Since \(\tau=V_tK_{t:b}/Q_t\), doubling flow halves the time constant; since \(C_{t,\mathrm{eq}}=K_{t:b}C_a\), the plateau is unchanged. (2) \(C_{t,\mathrm{eq}}=(2.50)(0.800)=2.00\ \text{mg/L}\). (3) \(\lambda=0.400/[(1.00)(2.00)]=0.2000\ \text{min}^{-1}\), so \(\tau=1/\lambda=5.00\ \text{min}\). (4) Flow changes the approach rate, while a larger partition coefficient changes both the plateau and the rate; different combinations can therefore produce similar concentrations at a particular time. The arterial input and the curve’s plateau and transient must be considered together.