Physiologically-Based Pharmacokinetic Modeling to Guide the Development of Therapeutic Antibodies
Glassman, Patrick M.
MetadataShow full item record
The use of model-based approaches for a priori prediction of monoclonal antibody (mAb) pharmacokinetics (PK) and pharmacodynamics (PD) has had mixed results, particularly in a translational setting where a model developed in preclinical species is scaled to predict clinical behavior of a drug candidate. Nearly all model-based approaches for interspecies scaling of antibody PK utilize compartmental models that rely on fitting of parameters that may not be readily scaled for model development. Physiologically-based pharmacokinetic (PBPK) models present an attractive platform for mechanistic integration of knowledge describing physiology, disease pathology, and molecular interactions to facilitate prediction of the biodistribution of therapeutic molecules. In the first section of this dissertation, we have scaled and extended a PBPK model that was previously developed for description of mAb disposition in mice to predict the non-linear, target-mediated PK of antibodies administered to monkeys and to patients. In order to describe interactions between mAbs and target molecules appropriately, we have proposed an algorithm to convert semi-quantitative immunohistochemical (IHC) scores into cellular receptor number, based on data from cells in culture. Additionally, in order to capture inter-antibody variability due to poorly understood factors (e.g. molecular charge and glycosylation), we have obtained estimates of variability in parameters related to non-specific endocytosis of mAbs into endothelial cells and convective uptake into tissue via paracellular pores. The models proposed in this work were able to generate reasonable predictions of the plasma PK of mAbs directed against epithelial antigens in both monkeys and in man, and the model was used to predict the plasma PK of mAbs that bind to lymphocyte targets and to tumor shed antigens (e.g. antigens that are present in both a membrane-bound and circulating form). In the second section of the dissertation, model-based approaches were used to understand the impact of molecular engineering of mAbs on PK, in an effort to design molecules with optimal PK/PD properties. PBPK modeling was used to understand the impact of “catch and release” (or pH-sensitive) antigen binding on the plasma PK of antibodies directed against a soluble target, proprotein convertase subtilisin kexin type 9 (PCSK9), and the proposed model was used to generate the hypothesis that engineering “catch and release” binding would be useful for mAbs against targets that are rapidly turned over and are expressed at concentrations exceeding targeted mAb concentrations. Additionally, a PBPK model was proposed to predict the influence of binding to the transferrin receptor (TfR) on the brain uptake of mAbs, which was able to characterize the improved brain exposure of lower affinity anti-TfR mAbs relative to high affinity mAbs. In addition to work with PBPK, the peripheral model of target-mediated disposition was used to propose a Target Classification System that could be used by groups seeking to determine whether affinity engineering of therapeutic mAbs would be lead to improvements in dose-potency. Finally, histidine derivatives were incorporated into the primary structure of mAbs in an effort to shift the pH-sensitivity in binding to FcRn. Inclusion of 2-iodo-L-histidine, which would be expected to lead to decreased ionization in the early endosome led to decreased exposure of a murine mAb following IV administration, relative to the parental mAb. In summary, the body of work presented in this dissertation evaluates the use of physiologically-based pharmacokinetic models at two stages of drug development, (i) in early discovery where PBPK could be used to guide molecular engineering and lead optimization, and (ii) in a translational setting where PBPK could be used to make a priori predictions of mAb disposition in patients. The results presented here highlight the potential that routine implementation of PBPK in antibody drug development could have in selection of antibodies and targets with the potential for optimal in vivo pharmacological properties (e.g. pharmacokinetics and pharmacodynamics).