Accounting for Route and Bioavailability in Starting-Dose Calculations

Why Route Matters in a Starting-Dose Calculation

A dose is the amount administered, not necessarily the amount that reaches the systemic circulation. An intravenous dose enters systemic circulation directly; an oral dose must dissolve, be absorbed, and pass through the gut wall and liver before systemic exposure occurs. These processes can make the same administered dose produce very different concentrations and exposure.

When the toxicology route and planned clinical route differ, a starting-dose calculation must account for this mismatch. The adjustment depends on the exposure measure that limits the dose. For total exposure measured by AUC, bioavailability can often be used in a direct calculation if pharmacokinetics are linear. For peak concentration, absorption rate and the concentration-time profile matter as well as the fraction absorbed.

The administered dose and the systemically available dose are therefore distinct: the latter is \(F\) times the former. First-pass metabolism is one reason oral \(F\) may be low, but incomplete dissolution or absorption, degradation in the gut, and transport back into the gut can also reduce it. A route-specific formulation can change these processes, so bioavailability estimates should correspond to the formulation and conditions intended for clinical use.

AUC-Based Route Adjustment

Under linear pharmacokinetics, with complete AUC measurement and clearance \(CL\) that does not depend on dose, the relationship between administered dose and AUC is:

This relationship says that AUC is proportional to systemic dose and inversely proportional to clearance. It supports a route adjustment when the dose-limiting exposure is AUC. It does not imply that two routes have the same peak concentration, time to peak, or local tolerability.

A useful derivation follows directly from the model. For the clinical route, \(AUC_{\text{target}}=F_{\text{human}}Dose_{\text{human}}/CL_{\text{human}}\). Multiplying both sides by \(CL_{\text{human}}\), then dividing by \(F_{\text{human}}\), gives:

If the target is to match an animal AUC, that measured AUC can be used as \(AUC_{\text{target}}\). If a margin is intended, the target must be set accordingly before calculating the dose. The animal and human AUCs must use compatible units, and the human clearance and bioavailability must describe the planned clinical route and population.

An animal dose expressed in mg/kg or converted to a human equivalent dose by body-surface-area scaling is not itself an AUC. Body-surface-area conversion adjusts a dose across species; it does not establish that systemic exposure is matched, nor does it correct a difference between the animal and human routes. Whenever adequate PK data are available, an exposure-based comparison is more directly informative for a route mismatch.

Why Bioavailability Alone Does Not Match Cmax

AUC summarizes total systemic exposure over time. \(C_{\max}\), by contrast, is the highest concentration reached. Two routes can give the same AUC but different \(C_{\max}\) because drug input occurs at different rates. An intravenous bolus enters rapidly; oral absorption is spread over time. Thus, multiplying a dose by the inverse of \(F\) can match AUC under the stated assumptions while failing to match a \(C_{\max}\)-based exposure limit.

A simple one-compartment model with first-order oral absorption illustrates the distinction. Assume linear disposition, a single dose, first-order absorption rate constant \(k_a\), elimination rate constant \(k_e\), volume of distribution \(V\), and no lag time. For \(k_a \ne k_e\), the concentration after oral dosing is:

Differentiating this expression with respect to time and setting the derivative to zero gives the time of peak concentration:

Substituting \(t_{\max}\) into \(C(t)\) gives the oral \(C_{\max}\). Unlike AUC, this peak depends on \(k_a\) as well as \(F\), dose, and disposition. This model is a useful illustration, not a universal description: delayed absorption, multiple absorption phases, nonlinear kinetics, or distributional complexity may require a different model or measured concentration-time data.

Assumptions, Diagnostics, and Limitations

Before using an AUC-based correction, verify that the model is appropriate. The dose-exposure relationship should be approximately linear over the relevant range: doubling dose should approximately double AUC, and clearance and bioavailability should not change materially with dose. AUC should be measured or predicted over a sufficiently complete interval. If the drug has nonlinear absorption, saturable first-pass metabolism, dose-dependent clearance, or time-varying bioavailability, the simple proportional correction may fail.

Route-specific bioavailability is often uncertain, especially early in development. Assess whether the estimate comes from a comparable formulation, dose, fed or fasted condition, and population. Sensitivity calculations or a conservative plausible range are more transparent than treating a weakly supported point estimate as exact. For an oral formulation with lower \(F\), AUC-based dose matching increases the administered dose; if that dose is impractical or unsafe to administer, the proposed route or formulation may need reconsideration.

Several observations can signal that the adjustment is not working as assumed:

• Dose-normalized AUC changes across dose levels, suggesting nonlinearity or dose-dependent bioavailability or clearance.
• Exposure differs substantially between formulations, food conditions, or batches, indicating that the selected \(F\) may not represent the planned clinical product.
• Measured \(C_{\max}\) or time to peak differs from predictions even when AUC is similar, indicating that absorption-rate assumptions need review.
• Early samples are sparse, late samples are missing, or the concentration-time profile is poorly characterized, making AUC or peak estimates unreliable.
• Local effects at the administration site or in the gut are relevant; systemic AUC matching cannot rule out route-specific local toxicity.

The main alternatives are complementary rather than interchangeable. Direct comparison of measured animal and human-route exposures is preferable when suitable data exist. A population PK or physiologically based PK model can represent absorption, first-pass effects, and route-specific kinetics when the simple equations are inadequate, but its assumptions and inputs need scrutiny. A dose-only conversion is less informative about systemic exposure when routes differ. Each method remains limited by the relevance and uncertainty of its data; no conversion removes uncertainty about species differences or the safety of the proposed exposure.

Applying the Route Adjustment Carefully