Two-Compartment IV Bolus Amplitudes From Initial Conditions

Why Initial Conditions Determine the Amplitudes

A two-compartment IV bolus concentration profile has a fast distribution phase and a slower terminal phase. Its two exponential rate constants determine how quickly those phases decline, but they do not by themselves specify the concentration contributed by each phase. The amplitudes are determined by the concentration immediately after the bolus and the initial rate of change of that concentration.

This provides a direct way to connect the biexponential curve to the central compartment at time zero. It also gives a useful consistency check: amplitudes obtained from a standard two-compartment model with positive parameters should be positive and should add to the initial concentration.

Here \(\alpha\) is the fast rate constant and \(\beta\) the slow rate constant. The terms “fast” and “slow” refer to how rapidly each exponential declines, not to the size of its amplitude. For a positive dose and positive compartment volumes and rate constants, the standard two-compartment model gives positive amplitudes.

Deriving the Amplitude Formulas

At time zero, both exponential terms equal one. Differentiating the profile gives its slope at any time, and evaluating that derivative at zero supplies a second equation. Together, the two initial conditions are sufficient to solve for \(A\) and \(B\).

Solving the first equation for one amplitude, \(B=C_0-A\), and substituting into the slope equation yields

Rearranging gives \(A=(s_0+\beta C_0)/(\beta-\alpha)\). Equivalently, define the positive initial fractional decline rate \(r_0=-s_0/C_0\). For a positive initial concentration, this converts the slope equation to \(\alpha A+\beta B=r_0 C_0\). Solving the two linear equations gives the more interpretable formulas:

These formulas are derived from the biexponential form and its initial conditions; they do not require a numerical curve-fitting algorithm. The amplitudes sum to \(C_0\), and their rate-weighted sum equals the magnitude of the initial slope:

For both amplitudes to be positive, the initial fractional decline rate must lie between the two exponential rates. This follows directly from the formulas: because \(C_0>0\) and \(\alpha-\beta>0\), \(A>0\) requires \(r_0>\beta\), while \(B>0\) requires \(r_0<\alpha\).

Connection to the Central Compartment

The initial slope has a specific compartmental interpretation. Let \(A_c(t)\) and \(A_p(t)\) be the central and peripheral amounts, and let \(V_c\) be the central volume. For a central IV bolus, the peripheral compartment initially contains no drug. With first-order elimination from the central compartment and first-order exchange between compartments, the central amount balance is

Since \(C(t)=A_c(t)/V_c\), the initial concentration is \(C_0=A_c(0)/V_c\), and the corresponding initial slope is \(s_0=-(k_{10}+k_{12})C_0\). Therefore \(r_0=k_{10}+k_{12}\). The initial decline reflects both elimination from the central compartment and distribution out of it; it is not simply the elimination rate constant \(k_{10}\).

Worked Calculations

Using and Checking Initial-Condition Estimates

In an ideal mathematical model, \(C_0\) and \(s_0\) are exact values. In data analysis, however, the slope is usually not observed directly. It may be estimated from very early measurements or inferred from an estimated compartmental model. A difference between two measured concentrations is not automatically the slope at time zero: it is a finite-interval average change, and the biexponential curve is generally changing slope over that interval.

The initial-condition calculation is most useful when the quantities refer to the same central concentration profile and use consistent time and concentration units. For example, if the rate constants are in \(\mathrm{h^{-1}}\), the slope must be in concentration per hour. An immediate post-bolus concentration is also a model-based value at \(t=0\); an actual sample taken later may already reflect distribution.

There are alternatives to calculating amplitudes from initial conditions. A biexponential curve can be fitted directly, estimating amplitudes and rates together; this uses the full set of observations rather than just the initial concentration and slope. Alternatively, amplitudes can be derived from a compartmental parameterization. The initial-condition approach is an algebraic relationship and a consistency check, not a substitute for fitting data when the initial slope is uncertain.

Each approach has limitations. Direct fitting can be sensitive to sparse or noisy data, while an initial-condition calculation can be especially sensitive to slope error because it uses a local rate at time zero. The algebra also becomes numerically unstable when \(\alpha\) and \(\beta\) are very close, since the denominator \(\alpha-\beta\) is then small. A negative calculated amplitude can indicate measurement noise, mismatched units, imprecise early-time information, or a mismatch between the assumed two-compartment model and the data. It should prompt investigation rather than be interpreted as a positive compartmental contribution.

Common Mistakes

• Using a positive slope sign. A declining concentration has \(s_0<0\). Define \(r_0=-s_0/C_0\) to use a positive decline rate, and retain the negative sign in \(s_0=-r_0C_0\).
• Confusing \(r_0\) with \(k_{10}\). Initially, distribution out of the central compartment also contributes to the decline, so \(r_0=k_{10}+k_{12}\), not just \(k_{10}\).
• Switching the fast and slow rates. Keep \(\alpha>\beta\); the amplitude paired with \(\alpha\) is \(A\), and the one paired with \(\beta\) is \(B\).
• Rounding inconsistently. Substituting rounded rates may not reproduce amplitudes calculated from higher-precision rates. Use the same displayed inputs throughout and report an appropriate number of significant figures.
• Treating any early finite difference as the initial slope. A measurement interval is not generally infinitesimal. State how the slope was obtained and recognize that its uncertainty affects the amplitudes.