Saskia Fuhrmann (1,4), Charlotte Kloft (3) and Wilhelm Huisinga (2)
(1) Institute of Biochemistry, Universitaet Potsdam, Germany; (2) Institute of Mathematics, Universitaet Potsdam, Germany; (3) Institute of Pharmacy, Dept. Clinical Pharmacy & Biochemistry, Freie Universitaet Berlin; (4) Graduate Research Training Program PharMetrX: Pharmacometrics & Computational Disease Modelling, Freie Universitaet Berlin and Universitaet Potsdam, Germany
Objectives: One of the primary applications of PBPK modeling is to predict efficacious and non-toxic clinical dose and human PK for new chemical and biological entities (NCEs, NBEs). In silico methods to predict tissue-to-plasma partition coefficients have significantly advanced the use of PBPK models for NCEs. The routine application of PBPK models to predict PK of NBEs, in particular monoclonal antibodies (mAbs), however is still challenging [1]. Antibody biodistribution coefficients (ABCs)[2] strongly support extrapolation of mAb PK between species. The objective of our study was to evaluate the feasibility of a generic PBPK modeling approach incorporating ABCs as a general approach to study mAb PK in different preclinical species and humans. The focus was on mAbs binding to soluble targets.
Methods: PK data of 22 mAbs of IgG isotype from mice, rat and humans were obtained from literature and analyzed with the generic PBPK model parameterized by ABC values [3]. The only unknown parameter(s) relate to species-dependent clearance processes. In most cases, also in the presence of target, linear PK for mAbs was observed – depending on dose range and target concentrations. In the dose-linear range, only linear total clearance was estimated. For mAb PK data resulting from a large dose range, we estimated both, linear and nonlinear clearance, with the generic PBPK approach extended by a Michaelis-Menten TMDD model.
Results: Estimates of the total linear mAb clearance in humans were similar for different mAbs and ranged from 0.1 to 0.5 l/day, which is comparable to the clearance of endogenous IgG and within the reported range for mAbs [4]. The median clearance 0.223 l/day can serve as a default value for linear total clearance to obtain good predictions for PK of mAbs binding to soluble targets in humans. Often, also in the presence of a target, the unspecific clearance is the dominating clearance process; and only for lower concentrations, nonlinear PK was observed with target-mediated clearance dominating over the unspecific clearance.
Conclusions: We showed that a generic PBPK approach parameterized by ABC values allows to study mAb PK across species with only clearance as unknown parameter. For a given mAb the generic PBPK model with median linear plasma clearance can provide reference prediction, e.g. to study the extent of target-mediated elimination.
References:
[1] H. M Jones et al. The AAPS Journal, 15: 377-387, 2012.
[2] Shah et al. mAbs, Vol.5: 297-305, 2013.
[3] W.Huisinga, S.Fuhrmann, L.Fronton, B.F.Krippendorff Target-Driven Pharmacokinetics of Biotherapeutics in Application of ADME and Translational PK/PD in the development of therapeutic biologics (Eds. H. Zhou and F.-P. Theil) Wiley, 2015.
[4] N. Dirks et al. Clinical pharmacokinetics, 49: 633-659, 2010.
[5] L. Fronton et al. J Pharmacokinet Pharmacodyn, 41: 87-107, 2014.
[6] B. Kuang et al. Bioanalysis, 2: 1125-1140, 2010.
Reference: PAGE 25 () Abstr 5749 [www.page-meeting.org/?abstract=5749]
Poster: Drug/Disease modeling - Absorption & PBPK