Subcutaneous Absorption Coupled to TMDD

Why Subcutaneous Input Changes the TMDD Profile

After subcutaneous administration, drug does not enter the systemic circulation all at once. It first occupies an absorption depot, and its rate of arrival depends on the depot amount and the absorption rate constant. Once in the central compartment, the drug can bind a target, dissociate from it, and be removed through target-mediated processes as well as nonspecific clearance. The observed concentration-time curve therefore reflects both the timing of drug input and the concentration-dependent disposition.

This interaction matters because target-mediated elimination is often most influential when free drug concentrations are not high enough to saturate the target-mediated pathway. A slowly absorbed dose may spend longer in that concentration range than a rapidly absorbed dose. Conversely, rapid input can raise free concentrations enough to reduce the fraction of target-mediated elimination operating per unit drug concentration. These are mechanistic possibilities, not universal predictions: the outcome depends on binding, target turnover, internalization, and absorption parameters together.

A Full Coupled Model

A standard model assumes first-order absorption from one depot, a well-mixed central compartment, and mass-action binding. Free drug undergoes first-order nonspecific elimination at rate constant \(k_{el}\). Free target is synthesized at rate \(k_{syn}\) and degraded at rate constant \(k_{deg}\). Complex dissociates at rate constant \(k_{off}\) and is internalized at rate constant \(k_{int}\). The equations are:

Here \(k_{on}\) is the association rate constant. The binding terms remove free drug and free target together and add complex; dissociation reverses that exchange. Internalization removes complex and its drug from the modeled system. The depot begins at the administered dose, \(A(0)=D\), while the initial concentrations \(C(0)\), \(R(0)\), and \(RC(0)\) are set by the dosing history. In a drug-free baseline with no complex, the free-target steady state is \(R(0)=k_{syn}/k_{deg}\).

Adding the free-drug and complex equations shows why binding itself is not drug elimination: the association and dissociation terms cancel. With a common central volume and compatible concentration units, the modeled drug balance is

Thus, complex formation can change the free concentration immediately, but net loss through the target pathway occurs through complex internalization in this formulation. An assay that measures free drug reports \(C\); a total-drug assay may reflect both free drug and complex, subject to the assay’s measurement properties.

The Linear Absorption Benchmark

Before adding target-mediated disposition, it is useful to establish the concentration profile under first-order absorption and linear elimination alone. Assume one depot, constant \(F\) and \(V\), first-order absorption, and first-order elimination with rate constant \(k\). There is no target binding in this benchmark. Solving the depot equation gives \(A(t)=D e^{-k_a t}\). Substitution into the central-compartment equation and integration gives:

The peak occurs where \(dC/dt=0\). Differentiating and setting the derivative to zero gives \(k_a e^{-k_a t}=k e^{-kt}\), so the positive peak time is \(t_{\max}=\ln(k_a/k)/(k_a-k)\). This peak-time formula is a useful benchmark, but it does not apply unchanged when TMDD makes elimination concentration-dependent.

How Target Saturation Enters the Coupled Profile

A useful reduced description is available when complex adjusts quickly relative to the changes in free drug and the total target pool is approximately constant over the interval of interest. Under a quasi-steady-state approximation, complex concentration is approximated by

where \(R_{tot}=R+RC\). The target-mediated removal rate is then \(k_{int}RC\), which has the saturable form \(V_{\max}C/(K_m+C)\), with \(V_{\max}=k_{int}R_{tot}\). The reduced subcutaneous model becomes

This approximation requires rapid complex adjustment and a sufficiently stable total target pool; it is not simply a new exact form of the full equations. When \(C\ll K_m\), target-mediated removal is approximately \((V_{\max}/K_m)C\), so it behaves nearly linearly. When \(C\gg K_m\), it approaches \(V_{\max}\), and its contribution per unit concentration becomes smaller as concentration rises. At a concentration peak in this reduced model, instantaneous input equals instantaneous removal:

Because both \(A\) and \(C\) change over time, this balance generally must be evaluated from the model trajectory. It is not, by itself, a formula for the peak time.

Interpreting Profiles, Diagnostics, and Alternatives

The absorption rate controls how quickly drug becomes available systemically; the target mechanism controls how disposition changes as free drug and target interact. Slower absorption can delay and flatten input, while target saturation can reduce the incremental effect of target-mediated removal at high concentrations. Their combination may alter peak height, peak time, and the post-peak decline. A linear absorption model remains a useful comparison, but its Bateman curve assumes linear disposition and should not be used to explain a concentration-dependent TMDD profile.

When model predictions or data appear inconsistent, check whether the assumed mechanism is plausible before interpreting parameter estimates. Useful diagnostics include plotting observed and predicted concentrations over time, comparing early absorption-phase and later decline patterns, and checking whether free-drug or total-drug measurements are being compared with the corresponding model quantity. Look for systematic time-dependent deviations, implausible parameter values, and predictions that change markedly when the model is initialized or simplified. Such patterns can signal a misspecified absorption process, inadequate target turnover assumptions, or a reduced approximation being used outside its valid range.

The full ODE model explicitly represents free target and complex and is appropriate when their dynamics matter. A quasi-steady-state or rapid-equilibrium reduction can be more parsimonious when its timescale and target-pool assumptions are defensible. A linear absorption model is simpler still, but omits target-mediated nonlinearity. These alternatives answer different mechanistic questions; a simpler model should not be selected merely because its profile looks smooth.