“Affinity” is defined as, “a spontaneous or natural liking or sympathy for someone or or something.” This concept applies also to the biologics (large molecules) we help to develop. Drugs like monoclonal antibodies (mAb) or bispecific antibodies are ideal drug candidates since they have very high affinity to bind to their target substance or site. Given the variability of the targets, safety profiles, and therapeutic windows, it is important to understand the characteristics of the affinity of the target and how to translate phenomena such as target mediated drug disposition (TMDD).
What is TMDD?
TMDD is the phenomenon in which a drug binds with high affinity to its pharmacological target site (such as a receptor) to such an extent that this affects its pharmacokinetic (PK) characteristics. The concept was first formulated by Gerhard Levy in 1994 and has been the focus of extensive research to improve the understanding and application of TMDD (1). The target binding and subsequent elimination of the drug-target complexes could affect both drug distribution and elimination, and result in nonlinearity of PK in a dose-dependent manner. This is most commonly observed as linear PK at high dose levels or high concentrations and nonlinear PK at low dose levels or low concentrations.
TMDD PK Models & Characteristics
Successful development of biologics requires accurate prediction of human exposure. This is more challenging for biologics demonstrating TMDD. Due to the nonlinearity of exposure, simple allometric scaling models cannot be applied. Mechanistic-based models that quantitatively describe the drug–targeting interactions are needed for more accurate simulation and prediction. Over time various TMDD model structures have been proposed and tested against observed data, and / or used for prediction. A representative model structure is shown below.
The structure of these model is commonly determined by factors such as drug dosing route (intravenous or not), drug distribution (number of compartments), location of the target (tissue or blood), and binding kinetics (fast or slow, high or low affinity, reversible binding or not, single target or not, elimination mechanisms, and others).
A key characteristic of TMDD is the dose-dependent pharmacokinetic behavior. The figure below shows representative concentration–time profiles at different dose levels after bolus intravenous administration and application of the model structure from the above figure. The concentration–time profiles can be divided into four phases according to the concentrations. The first phase shows quick decrease in concentration, corresponding to initial binding to target and also distribution into the peripheral compartment. The second phase shows linear elimination, where the target is saturated with drug. At this phase, elimination is mainly by non-target related routes, together with fixed rate of target mediated elimination, which is negligible. The PK is largely linear. At the third phase, the concentration becomes lower so the targets are not all saturated, and both non-target mediated and target mediated elimination routes are important. Nonlinear PK is observed at this phase. At the last phase, the concentration is so low that targets are not saturated, the target mediated elimination becomes the mainly route of elimination, and the PK becomes linear again.
The concentration–time profiles of a given drug may not show all the phases, depending on the model structure and other factors including binding kinetics, target turnover rates, elimination of drug, and drug-target complex. Another common factor is the bioanalysis detection sensitivity. High detection limits may limit detection at the lower end of the profiles. Anti-drug antibodies (ADA) are also common for biologics and can impact the shapes of the concentration–time profiles. Understanding the characteristics of TMDD PK profiles can be helpful in identifying whether the nonlinearity is due to TMDD or ADA.
When to use a TMDD model and model structure determination
A TMDD model may not be needed in all cases. Determining when to use a TMDD model is based on factors such as the type of molecule, the shape of the concentration–time profiles, and the noncompartmental pharmacokinetic analysis (NCA) results. Biologics are more likely to show TMDD in their PK profiles because they are designed to bind to their target with high affinity. However, small molecules can also have TMDD kinetics. The shape of the concentration – time profiles and NCA analysis results can indicate if non-linear clearance (CL) and volume of distribution (V) are indicators of TMDD kinetics. TMDD model structure should be determined based on information on the drugs and the pharmacological mechanisms. These could include the properties of the targets (soluble or not, locations, concentrations, and binding kinetics with the drugs). Such information or parameters may be measured in vitro, or estimated by data-fitting if the concentrations can be measured where targets are not saturated.
Applying TMDD model for human PK prediction
Multiple studies have shown that, in general, PK data from nonhuman primates (NHP) are more preferable and recommended for predicting human mAb PK (2). This is based on the observation that most therapeutic mAbs bind to NHP antigens more often than to rodent antigens, due to the greater sequence homology observed between NHP and humans. PK parameters (CL and V) can be scaled allometrically to human, or estimated by transforming the concentration–time profiles from monkey to human using the species-invariant time method. The targetrelated parameters are generally kept the same between monkeys and humans, or use experimentally determined values, if available. A sensitivity analysis may be performed to estimate the impact of the individual parameters. Caution should be taken when predicting PK in patients, because parameters at disease state may be significantly different from those observed at healthy state or in NHP, thus need to be evaluated or justified based on information available.
Each biologic drug candidate is unique in mechanism and performance, but also possesses some similar properties and the pursuit of common strategies to develop. It takes both experience and careful investigation to develop successful TMDD models and a high affinity for specialized science.