Pre-clinical and clinical applications of PBPK models for monoclonal antibodies and other large molecule modalities.
Iain Gardner
Translational DMPK Science, Certara UK LTD –Simcyp Division, Sheffield, UK
Physiologically based pharmacokinetic (PBPK) models for large molecules combine information on the biological system together with information specific to the protein of interest to allow simulation of the concentrations of the protein in blood and tissues following dosing to the animal or human subject of interest.
In this presentation the physiological processes that need to be considered in PBPK models of large molecules will be discussed. In addition to the typical physiological information need to construct PBPK models for small molecule drugs (e.g. tissue volumes and blood flows), there are a number of additional considerations that need to be taken into account when constructing PBPK models for therapeutic proteins and other large modalities (eg convection of proteins from the blood to the interstitial space, binding of some proteins to FcRn with subsequent recycling within the endosomal space of the vasculature and the effect of target binding and dynamics on the pharmacokinetics of the protein of interest). In addition, information is becoming available to enable the contribution of different tissues to the elimination of therapeutic proteins to be included mechanistically within PBPK models for proteins.
Despite the inherent challenges in constructing PBPK models for therapeutic proteins, much progress has been made over the last 40 years. For instance, approaches to link distribution of proteins from the vascular to interstitial space of different tissues dependent on the molecular size of the therapeutic protein have been described (eg Gill et al, 2016, AAPS J, 18, 156). This allows PBPK models to be used to describe the concentration of therapeutic proteins in different human tissues (eg synovial fluid, lung etc). In vitro-in vivo extrapolation (IVIVE) approaches to link in vitro data on non-specific binding and FcRn transport/affinity to make predictions of human pharmacokinetics of specific monoclonal antibody proteins have been described and approaches to use this information in PBPK models have been described in the literature (Jain et al., PNAS 2017, www.pnas.org/cgi/doi/10.1073/pnas.1616408114; Kelly et al., mAbs 7,770,2015, Avery et al., Mabs, 10, 244, 2018, Chung et al., Mabs, 2019).
PBPK models for proteins have also been linked to target mediated disposition (TMD) models and to TMD models adapted to account for the binding of bispecific antibodies to two independent molecular targets or to account for shedding of target (eg Li et al., CPT: PSP, 2014, 3, e96). Different approaches to obtain estimates of target protein concentrations to allow inclusion of TMD in PBPK models will be exemplified. Lastly, due to the separation of physiological system information from the inherent properties of the compounds, PBPK models for proteins have the ability to account for differences in physiology changes with age and disease allowing the disposition of proteins to be simulated in children and renally and hepatically impaired subjects for example. Examples of these different applications will be highlighted in the presentation.