Ghazal Montaseri (1), Éric Fernandez (1), Irene Waizenegger (2), Gabriela Gremel (2), Melanie Wurm (2), Elisa Oquendo Cifuentes (3), Tammy Bigwarfe (3), Andrzej Bienczak (1,4), Matthias Freiwald (1), Valerie Nock (1).
(1) Boehringer Ingelheim Pharma GmbH & Co. KG, Biberach an der Riß, Germany. (2) Boehringer Ingelheim Pharma GmbH & Co. KG, Vienna, Austria. (3) Boehringer Ingelheim Pharma GmbH & Co. KG, Ridgefield, USA. (4) Novartis Pharma AG, Basel, Switzerland.
Introduction: A previously published semi-mechanistic tumor biodistribution – target engagement – response model for monoclonal antibodies (mAb) constitutes a suitable framework to implement model-informed drug development across projects in oncology to assess distribution to, target binding in, and growth of tumors [1]. The model consists of a two-compartmental pharmacokinetic (PK) model to describe plasma PK and a physiologically-based tumor compartment, itself containing vascular, endosomal and interstitial sub-compartments. Dynamic receptor binding in the central and tumor interstitial compartments, dynamics FcRn bindings in the endosomal space and time-varying tumor volume are among eminent features of the model.
One of the important processes is the extravasation of mAb from vasculature to interstitial space. In the original model, the extravasation process was modeled assuming that the dominant driving mechanism is convection. Whereas this assumption holds true for large molecules such as IgGs, for small biological molecules such as nanobodies the diffusion mechanism also plays a role in the extravasation and has to be included in the model.
Objectives: In this work, we aimed at setting up a framework to model tumor distribution of small and large biologics by enriching a previously published model employing the two-pore theory [2-4]. This allows to physiologically model the extravasation pathway considering both, diffusion and convection, processes.
Methods: Initial investigations showed that the tumor biodistribution model had to be revised to address issues identified with respect to mass-balance, scaling of tumor distribution rate constants and derivative of receptor occupancy (RO) as a target engagement marker. After these revisions, the part of the structural model to represent the extravasation of a compound from vasculature to the interstitial space was modified according to the two-pore theory. The theory assumes that the capillary walls consist of small and large pores and a molecule moves between the two spaces via the pore passages driven by convection and diffusion mechanisms. Following the laws of fluid mechanics, the two-pore theory mathematically describes clearance of antibodies due to the two mentioned processes. Whereas for small antibodies diffusion (due to concentration gradients) is the prevailing driving force for compound transport, large antibodies extravasate mainly due to the convection driven by the lymph flow. Profiting from the two-pore theory, we are able to model the extravasation for any antibody with a pre-specified molecular weight. To investigate the impact of the two-pore theory, we compared the tumor biodistribution model with the original and the two-pore theory-based extravasation models. All modelling and simulation efforts were done using NONMEM version 7.4.3 and MATLAB 2019a, data processing and plotting was done using R version 3.5.2.
Results: Two key parameters in the two-pore model are the molecular weight of the antibody and permeability surface area of the tumor. Knowing the molecular weight one can derive the permeability surface area for the tumor in mice [3,4] and hence no additional parameters need to be estimated compared to the original model. For IgGs including the model corrections and employing the two-pore theory the interstitial influx and therefore interstitial concentrations were increased. As a result, the internalization half-life of antibodies was impacted and still in the reported range for clinically applicable antibodies (i.e., minutes to hours).
Conclusions: Following the two-pore theory, we were able to extend the potential of the previously published model to represent not only large molecule data, but also small biologics. Furthermore, the two-pore theory paves the way for application of the model for a wider range of antibodies from nanobodies to IgGs without increasing the parameter estimation burden.
References:
[1] Lindauer A et al. CPT Pharmacometrics Syst. Pharmacol. (2017) 6, 11-20.
[2] Niederalt C et al. J Pharmacokinet Pharmacodyn (2018) 45, 235-257.
[3] Sepp A et al. J Pharmacokinet Pharmacodyn (2019) 46, 339-359.
[4] Li Z and Shah D.K. J Pharmacokinet Pharmacodyn (2019) 46, 305–318.
Reference: PAGE () Abstr 9487 [www.page-meeting.org/?abstract=9487]
Poster: Drug/Disease Modelling - Oncology