Integrating Target Mediated Drug Disposition (TMDD) into a minimal physiologically based modelling framework: evaluation of different quasi-steady-state approximations
Mezzalana Enrica (1), Lavezzi Silvia Maria (1), Zamuner Stefano (2), De Nicolao Giuseppe (1), Ma Peiming (3), Simeoni Monica (2)
(1) Department of Electrical, Computer and Biomedical Engineering, University of Pavia, Pavia, Italy, (2) Clinical Pharmacology Modelling and Simulation, GlaxoSmithKline, UK, (3) Clinical Pharmacology Modelling and Simulation, GlaxoSmithKline, China
Objectives: Target-mediated drug disposition (TMDD) often occurs with monoclonal antibodies (mAbs); hence various models were developed to describe this phenomenon [1]. Cao and Jusko [2] incorporated a quasi-steady-state (QSS) TMDD component into the minimal physiologically-based PK (mPBPK) model [3]. Assuming peripheral target binding, we explored the mPBPK model combined with the full TMDD model, denoted model I, and three approximated models [4], denoted model II, III and IV, to assess the feasibility of these approximations with respect to identifiability issues. The aim is to understand how informative is plasma PK about binding processes occurring in periphery.
Methods: With the PK parameters estimated in [2] for romosozumab (a mAb for the cure of osteoporosis), models I, II, III, IV were simulated using R and SimulX and sensitivity analysis performed. In particular, mAb PK data were simulated according to model I. Then, the full and approximated models were fitted with NONMEM 7.3 in different situations (without binding, with binding only in leaky tissues, and only in tight tissues) and with different types of data (drug concentration in plasma alone, with the addition of total drug concentration in the binding site, with the addition of total receptor concentration in the binding site). Drug related parameters were estimated while volumes and flow rate parameters were fixed to typical physiological values.
Results: For all models (I-IV), the contribution of elimination through internalization of the complex can be evaluated from drug concentration profiles in plasma: when binding is simulated in tight tissues, this contribution results negligible if compared to the situation without binding. Instead, if binding is simulated in leaky tissues, or both leaky and tight, the nonlinear elimination due to TMDD prevails. Identifiability issues arose almost under any condition, especially when attempting to partition the plasma and the non-linear peripheral clearance. In simulation, model I and II are almost equivalent, while model II yields different profiles primarily for receptor concentration in binding sites, while model IV exhibits a substantially different behavior.
Conclusions: : The full TMDD model and three QSS approximations [4] were plugged into a mPBPK model developed for mAbs [3]. Simulated profiles were compared and identification issues evaluated in the most significant contexts relative to binding site and measurements availability.
References:
[1] Theoretical Considerations of Target-Mediated Drug Disposition Models: Simplifications and Approximations, P. Ma. Pharm Res. 2012, vol 29, 866-882.
[2] Incorporating target-mediated drug disposition in a minimal physiologically-based pharmacokinetic model for monoclonal antibodies. Y. Cao, W.J. Jusko. J Pharmacokinet Pharmacodyn. 2014, vol 41, 375-387.
[3] Second-generation minimal physiologically-based pharmacokinetic model for monoclonal antibodies. Y. Cao, J.P. Balthasar, W.J. Jusko. J Pharmacokinet Pharmacodyn. 2013, vol 40(5).
[4] Gaining insights into the consequences of target-mediated drug disposition of monoclonal antibodies using quasi-steady-state approximations. H.P. Grimm. J Pharmacokinet Pharmacodyn. 2009, vol 39, 407-420.
