Question 1

What is target-mediated drug disposition (TMDD)?

Accepted Answer

Target-mediated drug disposition (TMDD) is a pharmacokinetic phenomenon in which a drug binds with high affinity to its pharmacological target (receptor, enzyme, or other protein) to such an extent that the binding significantly affects the drug's distribution and/or elimination kinetics. TMDD was described by Levy (1994) to explain the nonlinear pharmacokinetics of drugs for which elimination is substantially driven by receptor-mediated endocytosis or target-mediated catabolism. TMDD is most clinically important for monoclonal antibodies and biologic therapeutics — where the target (e.g. a soluble cytokine, cell-surface receptor) can be saturated at pharmacologically relevant concentrations, causing the apparent clearance to decrease as dose increases (dose-dependent PK). Examples: anti-PD-L1 antibodies (atezolizumab), anti-VEGF (bevacizumab), anti-PCSK9 (evolocumab), anti-IgE (omalizumab).

Question 2

How does TMDD differ from standard linear pharmacokinetics?

Accepted Answer

In standard (linear) pharmacokinetics, clearance and half-life are constant regardless of dose or concentration — doubling the dose doubles the AUC, and Cmax scales proportionally. In TMDD, two elimination pathways operate simultaneously: a non-saturable (linear) elimination pathway (FcRn recycling and proteolytic catabolism for mAbs; renal/hepatic for small molecules) and a saturable (target-mediated) elimination pathway. At low doses, target-mediated clearance dominates — rapid, dose-disproportionate elimination. At high doses that saturate the target, only the non-saturable pathway remains — slower, more predictable elimination. The result is that PK changes with dose: AUC increases more than proportionally with dose (concave upward in log-log plot), apparent half-life increases with dose, and Cmax/AUC relationships are non-constant. TMDD complicates dose selection and PK/PD modelling — requiring TMDD-specific models (full, quasi-steady state, or Michaelis-Menten approximations).

Question 3

Which drugs exhibit clinically significant TMDD?

Accepted Answer

TMDD is most prominent in monoclonal antibodies and biologic therapeutics targeting proteins expressed at low concentrations relative to the drug dose. Examples with clinically significant TMDD: bevacizumab (anti-VEGF) — VEGF in circulation is at picomolar concentrations, well below bevacizumab concentrations at therapeutic doses, causing pronounced TMDD at lower dose levels; omalizumab (anti-IgE) — serum IgE is at nanomolar concentrations, causing significant TMDD below about 0.5 mg/kg; evolocumab and alirocumab (anti-PCSK9) — PCSK9 levels are at nanomolar concentrations, causing non-linear PK at lower doses (PCSK9 target saturation evident at therapeutic doses); cetuximab (anti-EGFR) — EGFR on tumour and normal cells contributes to TMDD. Some small molecules also exhibit TMDD: digoxin binding to Na+/K+-ATPase causes non-linear distribution at higher concentrations; some high-affinity enzyme inhibitors binding at catalytic sites.

Question 4

How is TMDD modelled in pharmacokinetic analysis?

Accepted Answer

The full mechanistic TMDD model (Mager & Jusko, 2001) explicitly describes the drug-target binding, complex formation, and complex elimination. Parameters: kon (association rate constant), koff (dissociation rate constant), Kd = koff/kon (equilibrium dissociation constant), Rmax (total receptor/target concentration), ksyn (target synthesis rate), kdeg (target degradation rate), and kint (internalization/elimination rate of drug-target complex). Solving the full TMDD model requires numerical ODE solvers. The quasi-steady-state (QSS) approximation assumes drug-target binding reaches equilibrium quickly compared to drug PK — allows algebraic solution of target occupancy, reducing model complexity. The Michaelis-Menten (MM) approximation further simplifies by treating target-mediated clearance as Vmax/(Km + C) — acceptable when target concentration is much lower than drug concentration (drug in large excess). Model selection depends on available data density and drug concentration relative to target level.

Frequently asked questions