Ravi Kumar Jairam (1), Maria Franz (2), Anna-Kaisa Rimpelä (3), Guangda Ma (3), Nina Hanke (2), Lars Kuepfer (1)
(1) Institute for Systems Medicine with Focus on Organ Interaction, University Hospital RWTH Aachen, Aachen, Germany, (2) Translational Medicine & Clinical Pharmacology, Boehringer Ingelheim Pharma GmbH & Co. KG, Biberach, Germany, (3) Drug Discovery Sciences, Boehringer Ingelheim Pharma GmbH & Co. KG, Biberach, Germany
Objectives: Rabbits play a pivotal role as a highly valued small animal model, particularly in the field of ocular therapeutics, where they serve as a crucial link between preclinical research and clinical applications. In this context, we have developed a physiologically based pharmacokinetic (PBPK) model designed specifically for rabbits, with a focus on accurately predicting the pharmacokinetic profiles of protein therapeutics following intravenous administration. Our goal was to comprehend the influence of key physiological factors on systemic drug disposition. Such a comprehensive understanding of physiological processes governing the pharmacokinetics of therapeutic proteins at the whole-body level is important for further analyses such as the disposition of protein therapeutics across various ocular compartments following intravitreal dosing in rabbits in an ocular PBPK model.
Methods: We developed systemic PBPK models for rabbits using intravenous data of antibodies and their fragments. For non-cross-reactive therapeutics (no target binding in rabbits), we employed the generic PBPK model in PK-Sim® without adjustments. However, for selected cross-reactive proteins, we integrated a specific target-mediated drug disposition (TMDD) process using MoBi®[1, 2]. For lower molecular weight proteins, renal clearance was included as an additional clearance mechanism.
Subsequently, based on the groundwork established by the Hutton-Smith et al. model [3], we refined an ocular PBPK model in R. Initially designed as a 3-compartment semi-mechanistic model, our extension incorporated a plasma compartment. Subsequent qualification confirmed the model’s ability to predict the distribution of bevacizumab across multiple ocular compartments including vitreous, aqueous, retina, and plasma, after intravitreal injection. To further enhance the accuracy, strain-specific ocular anatomical differences in rabbits were incorporated during model development, utilizing datasets from various publications.
Results: In our intravenous PBPK model development, we conducted an in-depth analysis of non-cross-reactive anti-glycoprotein D (anti-gD) antibodies and their fragments, as well as cross-reactive anti-vascular endothelial growth factor (anti-VEGF) Fab and F(ab)2. We meticulously adjusted parameters to reflect their unique characteristics: for non-modified anti-gD IgG, the dissociation constant for FcRn binding (KD-FcRn) was set to 2.5 µmol/L, and the hydrodynamic radius (Rh) was optimized to be 1.2 times smaller than the reported value for IgGs. In contrast, for anti-gD null IgG (mutations that prevent binding to FcRn), KD-FcRn was determined to be relatively 5-fold higher. Parameters such as Rh and renal clearance (RCL) were fine-tuned for Fab fragments, with the RCL set close to the reported estimated values for anti-gD Fab and for anti-gD F(ab)2 it was approximately 10-fold lower than for anti-gD Fab. Moreover, for anti-gD Fc proteins, KD-FcRn was set to a comparatively lower value than for Fab, RCL was adjusted to a 16-fold lower level, and Rh was fitted to 1.3-fold lower than anti-gD Fab. Furthermore, we estimated target-mediated drug disposition values for anti-VEGF Fab and F(ab)2, resulting in simulation profiles that accurately represented the observed data. These adjustments provided valuable insights into protein disposition systemically for IgG, F(ab)2, Fab and Fc.
In our ocular simulations for bevacizumab, we estimated the vitreal half-life within a narrow range, falling within 2-fold of the reported values. Similarly, the plasma half-life was set to closely align with reported values, while ensuring precise alignment with observed data across all ocular compartments of the injected eye and the plasma compartment.
Conclusions: We developed a tailored PBPK platform for rabbits, accurately predicting protein therapeutics systemic disposition post intravenous administration. We qualified our models across various antibodies and their fragments. Incorporating strain-specific ocular anatomical differences for ocular PBPK modeling, we successfully predicted bevacizumab’s ocular disposition after intravitreal administration. Our goal is to merge these advancements to construct a comprehensive eye-whole-body PBPK model. The PBPK models we have developed enhances preclinical pharmacokinetic research, aiding in translating findings to clinical settings and streamlining the drug development processes.
References:
[1] Niederalt, C., et al., A generic whole body physiologically based pharmacokinetic model for therapeutic proteins in PK-Sim. J Pharmacokinet Pharmacodyn, 2018. 45(2): p. 235-257.
[2] Kuepfer, L., et al., Applied Concepts in PBPK Modeling: How to Build a PBPK/PD Model. CPT Pharmacometrics Syst Pharmacol, 2016. 5(10): p. 516-531.
[3] Hutton-Smith, L.A., et al., Ocular Pharmacokinetics of Therapeutic Antibodies Given by Intravitreal Injection: Estimation of Retinal Permeabilities Using a 3-Compartment Semi-Mechanistic Model. Mol Pharm, 2017. 14(8): p. 2690-2696.
Reference: PAGE 32 (2024) Abstr 10831 [www.page-meeting.org/?abstract=10831]
Poster: Drug/Disease Modelling - Absorption & PBPK