Unravelling the pharmacokinetic interaction of ticagrelor and MEDI2452 (ticagrelor-antidote) using mouse data
Peter Gennemark (1), Mark Penney (2), Susanne Pehrsson (1), Ann-Sofie Sandinge (1), Annika Janefeldt (1), Sufyan Maqbool (2), Shimona Madalli (3), Joanne Goodman (2), Sven Nylander (1), Joachim Almquist (4, 5)
(1) Cardiovascular and Metabolic Diseases, Innovative Medicines and Early Development Biotech Unit, AstraZeneca, M÷lndal, Sweden; (2) Clinical Pharmacology and DMPK, MedImmune, Cambridge, UK; (3) Cardiovascular and Metabolic Diseases Research, MedImmune, Cambridge, UK; (4) Fraunhofer-Chalmers Centre, Chalmers Science Park, Gothenburg, Sweden; (5) Systems and Synthetic Biology, Department of Biology and Biological Engineering, Chalmers University of Technology, Gothenburg, Sweden.
Objectives: The investigational ticagrelor-neutralizing antibody fragment MEDI2452 has been developed to rapidly and specifically reverse the antiplatelet effects of ticagrelor . Due to the dynamic interaction of ticagrelor, the ticagrelor active metabolite (TAM), and MEDI2452, analysis of pharmacokinetic (PK) data becomes non-trivial. The main objective of the present work is to better understand the PK of ticagrelor, TAM, and MEDI2452, and in particular to be able to predict free plasma concentrations that can bind to and inhibit platelet P2Y12 of ticagrelor and TAM.
Methods: The modelling process consists of three main steps: i) set-up a mathematical model of the combined ticagrelor-MEDI2452 PK in the mouse based on data of separately administered ticagrelor and MEDI2452, and on assumptions supported by literature; ii) validate and refine the model on several different combined ticagrelor-MEDI2452 PK data sets not used for setting up the model; and iii) use the model to understand the complex PK resulting from the ticagrelor-MEDI2452 interaction, and to predict free levels of ticagrelor and TAM and let these predictions drive a pharmacodynamic (PD) turn-over model under different experimental designs.
Results: We propose a mechanistic PK model, including a special observation model for post-sampling equilibration, which is validated and refined using four different combined ticagrelor-MEDI2452 mouse in vivo treatment data sets. A comparison against alternative models strengthen our a priori belief that MEDI2452-bound ticagrelor is primarily eliminated together with MEDI2452 in the kidney, and not recycled to the plasma, thereby providing a key assumption for the extrapolation to humans. The model predicts free ticagrelor and TAM plasma concentrations, which, in turn, drive a PD model that successfully predicts platelet inhibition level.
Conclusion: The proposed PK model of ticagrelor, TAM, and MEDI2452 has been validated on several mouse data sets and is useful for predicting free plasma concentrations of ticagrelor and TAM. The model significantly improves PK and PD understanding, experimental design, and translational confidence.
 Buchanan, A. et al. Structural and functional characterization of a specific antidote for ticagrelor. Blood 125, 3484-90 (2015).