Population PBPK Modelling of Trastuzumab: a Framework for Quantifying and Predicting Inter-individual Variability

In this work we proposed a population physiologically-based pharmacokinetic (popPBPK) framework for quantifying and predicting inter-individual pharmacokinetic variability using the anti-HER2 monoclonal antibody (mAb) trastuzumab as an example. First, a PBPK model was developed to account for the possible mechanistic sources of variability. Within the model, five key factors that contribute to variability were identified and the nature of their contribution was quantified with local and global sensitivity analyses. The five key factors were the concentration of membrane-bound HER2 ([Formula: see text]), the convective flow rate of mAb through vascular pores ([Formula: see text]), the endocytic transport rate of mAb through vascular endothelium ([Formula: see text]), the degradation rate of mAb-HER2 complexes ([Formula: see text]) and the concentration of shed HER2 extracellular domain in circulation ([Formula: see text]). [Formula: see text] was the most important parameter governing trastuzumab distribution into tissues and primarily affected variability in the first 500 h post-administration. [Formula: see text] was the most significant contributor to variability in clearance. These findings were used together with population generation methods to accurately predict the observed variability in four experimental trials with trastuzumab. To explore anthropometric sources of variability, virtual populations were created to represent participants in the four experimental trials. Using populations with only their expected anthropometric diversity resulted in under-prediction of the observed inter-individual variability. Adapting the populations to include literature-based variability around the five key parameters enabled accurate predictions of the variability in the four trials. The successful application of this framework demonstrates the utility of popPBPK methods to understand the mechanistic underpinnings of pharmacokinetic variability.