Dose Individualization in Major Burns and Capillary Leak

Why Burns and Capillary Leak Complicate Dose Selection

In major burns and capillary leak, a dose selected from a patient's usual weight or a single renal-function estimate may not produce the expected concentration. Capillary permeability and fluid resuscitation can expand extracellular fluid and change the apparent distribution volume of some drugs. At the same time, renal clearance may be increased in some patients, particularly during periods of hyperdynamic circulation. These effects are distinct: a larger distribution volume can lower concentrations after a dose, while higher clearance can lower concentrations over time.

The changes are not uniform across drugs or stable over the course of illness. Hydrophilic drugs that distribute mainly in extracellular fluid may be particularly affected by fluid expansion, but the size and direction of the effect depend on the drug and the patient's condition. A patient's fluid balance, renal function, and treatment may also change rapidly. Thus, a calculation is a model-based prediction under stated assumptions, not a universal dosing rule.

Definition: The apparent volume of distribution \(V\) relates the amount of drug in the body to its measured plasma concentration under a specified pharmacokinetic model. Clearance \(CL\) is the volume of plasma from which drug is removed per unit time. In a one-compartment model with first-order elimination, the elimination rate constant is \(k=CL/V\), and the half-life is \(t_{1/2}=\ln(2)/k\). These are model parameters, not direct measurements of fluid compartments or organ blood flow.

Separate Distribution Changes From Clearance Changes

For an intravenous loading dose in a one-compartment model with immediate mixing, the dose needed to reach a specified initial concentration is \(D_L=C_{\text{target}}V\). A larger \(V\) therefore requires a larger loading dose to reach the same modeled concentration. This relationship does not mean that every patient with edema needs a proportionally larger dose: the relevant \(V\) is drug-specific, and the model may not capture delayed distribution between tissues and plasma.

Maintenance dosing has a different dependence. At steady state, the average concentration over a dosing interval is determined by dose rate divided by clearance for continuous input, or by dose per interval divided by clearance for repeated dosing. If clearance rises while the regimen stays fixed, average exposure falls. Increasing a loading dose does not correct that continuing loss; maintenance input must be considered separately.

Augmented renal clearance (ARC) refers to renal drug elimination that is greater than expected for the clinical context. A measured creatinine clearance above about 130 mL/min/1.73 m² is often used as a screening threshold, but definitions vary and this value is not a universal dosing boundary. Serum creatinine can be difficult to interpret in critical illness, including when fluid dilution or reduced creatinine generation affects the result. A timed urine measurement may help assess renal clearance, but it also has collection and timing limitations. Neither an estimated nor measured creatinine clearance is identical to the clearance of every renally eliminated drug.

For a drug with substantial renal elimination, higher renal clearance can shorten half-life and reduce concentrations between doses. Fluid expansion and ARC can occur together, but they do not have to: \(V\) and \(CL\) should be assessed as separate, potentially changing quantities. Since \(k=CL/V\), an increase in both does not imply a predictable net change in half-life without their relative magnitudes.

Worked Example: Loading Dose With an Expanded Distribution Volume

Suppose a hypothetical intravenous drug is described by a one-compartment model. A target initial concentration is 12 mg/L. The model-estimated \(V\) is 24 L before fluid expansion and 36 L after expansion. Assume immediate mixing, linear pharmacokinetics, and that these volumes are appropriate for this drug and patient.

Before expansion, the model-based loading dose is:

D_L=C_{\text{target}}V=(12\ \text{mg/L})(24\ \text{L})=288\ \text{mg}.

With the expanded volume, the corresponding calculation is:

D_L=(12\ \text{mg/L})(36\ \text{L})=432\ \text{mg}.

The model predicts that 432 mg, rather than 288 mg, would be needed to produce the same initial concentration under these assumptions. This is an illustration of the effect of \(V\), not a clinical recommendation. It does not determine a maintenance regimen, which depends on clearance and the dosing objective.

Predicting Concentrations During Repeated Infusion

A one-compartment model with constant clearance, constant volume, linear elimination, and intravenous infusion gives a transparent way to see how \(CL\) and \(V\) affect concentrations. Suppose a dose \(D\) is infused over \(T_{\text{inf}}\) hours every \(\tau\) hours. Let the infusion rate be \(R_0=D/T_{\text{inf}}\), and assume \(T_{\text{inf}}\leq\tau\). At steady state, the end-of-infusion concentration and the concentration immediately before the next dose are:

C_{\max,ss}=\frac{R_0}{CL}\frac{1-e^{-kT_{\text{inf}}}}{1-e^{-k\tau}}, \qquad C_{\min,ss}=C_{\max,ss}e^{-k(\tau-T_{\text{inf}})}, \qquad k=\frac{CL}{V}.

These expressions assume that the model parameters remain constant across intervals and that doses are administered as specified. The average steady-state concentration over an interval is \(D/(CL\tau)\); the area under the concentration-time curve over that interval is \(D/CL\). A concentration measured at another time cannot be compared directly with a predicted peak or trough unless the sampling time is accounted for.

Derivation: During an infusion at rate \(R_0\), the concentration contribution after \(t\) hours is \((R_0/CL)(1-e^{-kt})\). At the end of an infusion, the residual concentration from the preceding interval has decayed by \(e^{-k\tau}\). If \(C_{\max,ss}\) is the end-of-infusion concentration, steady state therefore satisfies \(C_{\max,ss}=(R_0/CL)(1-e^{-kT_{\text{inf}}})+C_{\max,ss}e^{-k\tau}\). Rearranging gives the stated peak equation. During the time from the end of infusion to the next dose, concentration declines exponentially for \(\tau-T_{\text{inf}}\) hours, giving \(C_{\min,ss}=C_{\max,ss}e^{-k(\tau-T_{\text{inf}})}\).

Worked Example: Effect of Higher Clearance During Continuous Infusion

A hypothetical drug is given by continuous intravenous infusion at 40 mg/h. Assume a one-compartment model, linear elimination, and a constant \(CL\) of 4 L/h. The predicted steady-state concentration is:

C_{ss}=\frac{\text{infusion rate}}{CL} =\frac{40\ \text{mg/h}}{4\ \text{L/h}}=10\ \text{mg/L}.

Now suppose the estimated clearance rises to 6.5 L/h, with the infusion rate unchanged. The predicted concentration becomes:

C_{ss}=\frac{40\ \text{mg/h}}{6.5\ \text{L/h}} =6.154\ \text{mg/L}.

If 10 mg/L were a specified hypothetical target and the model remained valid, the required infusion rate would be \(10\ \text{mg/L}\times6.5\ \text{L/h}=65\ \text{mg/h}\). This calculation isolates the effect of clearance; it does not establish that 10 mg/L is an appropriate target for a real drug or patient.

Solving an Illustrative Dose Adjustment

A practical model-informed calculation begins by stating the pharmacokinetic structure, the concentration objective, and the times at which concentrations are defined. The example below uses a target trough solely to demonstrate the arithmetic. A real target must be justified for the drug and clinical indication, and any adjustment must account for safety, measured concentrations, and the patient's evolving condition.

Worked Example: Adjusting an Intermittent Infusion After Fluid and Renal Changes

Consider a hypothetical drug infused intravenously over 1 hour every 8 hours. In a patient with changing fluid status, a one-compartment model estimates \(V=30\) L and \(CL=6.5\) L/h. Assume these values are constant over the modeled interval, elimination is first order, the drug is infused as scheduled, and steady state has been reached. The current dose is 600 mg, and the illustrative objective is a steady-state trough of 8 mg/L.

First calculate \(k=CL/V=6.5/30=0.2167\ \text{h}^{-1}\). The current infusion rate is \(600\ \text{mg}/1\ \text{h}=600\ \text{mg/h}\). Substituting into the steady-state peak equation:

C_{\max,ss} =\frac{600}{6.5} \frac{1-e^{-(0.2167)(1)}}{1-e^{-(0.2167)(8)}} \approx 21.84\ \text{mg/L}.

The trough, 7 hours after the end of the infusion, is:

C_{\min,ss}=21.84e^{-(0.2167)(7)} \approx 4.79\ \text{mg/L}.

With the same interval, infusion duration, and patient parameters, the trough is proportional to dose because the model is linear. The dose that would produce an 8 mg/L trough is therefore:

D_{\text{new}}=600\ \text{mg}\times\frac{8\ \text{mg/L}}{4.79\ \text{mg/L}} \approx 1{,}002\ \text{mg}.

Rounding for illustration, a 1,000 mg dose predicts a trough of about 7.99 mg/L and an end-of-infusion concentration of about 36.40 mg/L. The result depends on the assumed \(CL\), \(V\), sampling time, steady state, and target. It should not be interpreted as a recommended dose. In particular, meeting a trough target does not by itself establish that the associated peak or total exposure is suitable.

Monitoring, Alternatives, and Failure Modes

This model-based approach is useful because it makes the roles of \(V\) and \(CL\) explicit. It also makes predictions checkable against measured concentrations. When monitoring is available, record the exact dose and infusion times and interpret the sample using its actual time relative to dosing. If the observed concentration differs substantially from prediction, reassess the input data and model rather than reflexively scaling the dose.

Changing physiology: A steady-state calculation can mislead if fluid balance, renal function, or treatment changes across the dosing intervals. Re-estimate or update parameters when the clinical state changes.
Uncertain clearance: A renal-function estimate is a proxy, not a direct measurement of drug clearance. Check the measurement method, units, timing, and plausibility; consider whether ARC is present rather than assuming ordinary renal function.
Distribution not captured by one compartment: Early concentrations may reflect distribution phases that the one-compartment model cannot represent. The predicted peak can be especially sensitive to when the sample was drawn.
Wrong concentration measure: Total and unbound concentrations are not interchangeable when protein binding changes. If the clinical objective concerns unbound exposure, a total-concentration target may not answer the relevant question.
Wrong target or objective: A trough, peak, average concentration, and exposure over time are different quantities. Matching one does not guarantee that the others are appropriate.

The main alternatives are not mutually exclusive. A conventional dose based on weight or renal function is straightforward but can miss rapidly changing \(V\) or drug clearance. A population pharmacokinetic model can provide an initial estimate while accounting for patient characteristics, but predictions remain uncertain when the patient differs from the model's data or the clinical state changes quickly. Therapeutic drug monitoring can anchor an individual estimate to observed concentrations, provided timing and assay interpretation are sound. Model-informed adjustment combines the model and measurements; it does not remove uncertainty or replace clinical judgment.

A common mistake is to respond to a low concentration by increasing the loading dose and assuming the problem is solved. A loading dose primarily addresses the concentration associated with distribution volume; it does not compensate for persistently elevated clearance during maintenance. The correction is to identify whether the objective concerns initial concentration, ongoing exposure, or both, then evaluate \(V\) and \(CL\) separately. In major burns and capillary leak, that assessment may need to be repeated as fluid shifts and renal function evolve.

Key takeaway: Fluid expansion may increase the apparent distribution volume and change the dose needed to reach an initial concentration. Augmented clearance may lower ongoing exposure and increase the maintenance input needed for a specified model-based target. Separate these effects, state the assumptions, and check predictions against appropriately timed observations.

Check Your Understanding

Answer each question, then compare with the solutions below.

In a one-compartment model, \(V\) increases from 20 L to 30 L. What happens to the loading dose required for the same target initial concentration, assuming all other assumptions hold?
A hypothetical continuous infusion is 80 mg/h. Predict the steady-state concentration at \(CL=5\) L/h and at \(CL=8\) L/h.
Why is increasing a loading dose not, by itself, a solution to augmented renal clearance?
Name two reasons an observed concentration might disagree with a one-compartment steady-state prediction.

Solutions. (1) Since \(D_L=C_{\text{target}}V\), the required loading dose increases by a factor of \(30/20=1.5\), or 50%, for the same target. (2) The predictions are \(80/5=16\) mg/L and \(80/8=10\) mg/L, respectively. (3) A loading dose addresses the initial concentration associated with \(V\); augmented clearance increases ongoing drug removal, so maintenance input and exposure must be considered. (4) For example, the patient's clearance may have changed since estimation, the sample may have been taken at a different time than assumed, the patient may not be at steady state, or distribution may not be adequately represented by one compartment.