Drug and target parameters impacting target profiles for mono- and multi-specific NANOBODY® molecules – a model-based sensitivity analysis
Viktoria Stachanow (1), Lieselot Bontinck (1), Maria Laura Sargentini-Maier (1)
(1) Sanofi R&D, Ghent, Belgium
Objectives: NANOBODY® molecules are heavy chain only antibodies with molecular weights of 12-15 kDa, i.e., approximately a 10th of the size of common antibodies, and are currently under research and development (R&D) for various therapeutic areas [1,2]. In drug R&D, it is crucial to know which drug properties should be optimized to achieve the desired therapeutic outcome. The interplay between the drug and the target(s) can be complex for biologicals, such as NANOBODY® molecules. Thus, the aim of this analysis was to identify the drug and target parameters which impact the target suppression for mono- and multi-specific NANOBODY® molecules in a model-based sensitivity analysis.
Methods: A simple target-mediated drug disposition (TMDD) model of a bispecific NANOBODY® molecule with one soluble and one membrane-bound target was applied in deterministic simulations in Berkeley Madonna® [3] (Version 8.3.18). To also account for mono-specific NANOBODY® molecules targeting a soluble or membrane-bound target, the TMDD model was reduced accordingly. Tested scenarios included a 10-fold increase/decrease of the following parameters: drug elimination rate (kel), target binding affinity (KD), synthesis and degradation rate of the free target (ksyn and kdeg_f), complex on- and off-rate (kon and koff) and complex elimination rate (Kel_cpx). Furthermore, since literature values of target baseline concentrations can vary substantially, target baseline changes by factors up to 1000 were evaluated. All scenarios described single dose administrations of 10, 100, or 1000 mg NANOBODY® molecule in a typical individual of 70 kg. The resulting free target profiles were subsequently plotted in R/RStudio [4,5] (Version 4.2.0).
Results: The target profiles were impacted by changes of all tested drug and target parameters. For both mono- and bispecific NANOBODY® molecules, a 10-fold lower KD, i.e., higher affinity resulted in a longer and higher target suppression. 10-fold increased kon and koff rates, i.e., faster complex turnover with unchanged KD, lead to higher target suppression. A 10-fold higher kel, i.e., a shorter drug half-life, resulted in a shorter target suppression. The extent of the change varied between the mono-specifics with soluble and membrane-bound target since the kel was assumed to be equal to kel_cpx for soluble targets. Ten times higher ksyn and kdeg_f, i.e., faster free target turnover, lead to shorter and lower target suppression, again with a differently pronounced impact in the two mono-specifics with soluble or membrane-bound target since for membrane-bound targets kdeg_f was assumed to be equal to kel_cpx. In the bi-specific NANOBODY® molecule, a 1000-fold change of the membrane bound target baseline concentration resulted in a changed suppression duration for the soluble target, indicating cross-target influences. In this case, the shorter suppression of the soluble target due to a 1000-fold higher baseline of the membrane-bound target could not be substantially improved by 1000-fold changes in the affinities of either target.
Conclusions: The changes in the speed, onset and/or extent of target suppression observed in the simulated scenarios can differ between soluble- and membrane bound mono-specifics, depending on project-specific parameter values and assumptions. In bi-specifics, membrane-bound target parameters can impact soluble target profiles, most likely due to TMDD. Depending on the project and simulated scenario, optimization of KD does not always lead to a higher target engagement. In conclusion, simulations are crucial for each NANOBODY® molecule project to understand the complex dynamics between drug and target(s). Moreover, data is needed to confirm model assumptions and thus to increase the reliability of model simulations.
References:
[1] Harmsen M and De Haard H. Appl. Microbiol. Biotechnol. 77:13–22, 2007.
[2] https://www.sanofi.com/en/science-and-innovation/research-and-development/technology-platforms/nanobody-technology-platform
[3] University of California Berkeley. Berkeley Madonna. http://www.berkeleymadonna.com/
[4] The R Project for Statistical Computing. R. https://www.r-project.org/
[5] RStudio, Inc. RStudio. https://www.rstudio.com/