Using physiologically-based pharmacokinetics-quantitative systems pharmacology model to optimize peptide-drug conjugates design

Introduction: Peptide-drug conjugates (PDCs) are being developed as new cancer treatments. Compared to larger antibody-drug conjugates (ADCs), they retain the advantage of a targeted therapy, binding to tumor-associated antigens (TAA) to deliver payload, while having better tumor penetration than ADCs. PDCs typically exhibit faster plasma clearance than ADCs, and therefore potentially have less tumor exposure to the drug. The pharmacokinetics is influenced by the molecular weight (MW) of the peptide used in the PDC. A physiologically-based pharmacokinetics (PBPK)-quantitative systems pharmacology (QSP) model would be useful to understand the interplay of pharmacokinetics (PK), tissue disposition, and tumor drug exposure of PDCs. Methods: A platform PBPK-QSP model was developed based on a published PBPK model (Li et al., 2019) and a genetic tumor QSP model (Scheuher et al., 2024) by coupling these two models with a MW-based tumor penetration relationship. Through calibrated model simulations and sensitivity analysis, we interrogate how parameters of the peptide (e.g., molecular weight, binding affinity towards TAA) impact the PK, tissue disposition, and tumor drug exposure. Results: The model predicted higher tumoral PDC concentration when the peptide was small, corresponding to the knowledge that smaller peptides resulted in better tumor penetration. However, a bell-shaped relationship was predicted between peptide size and PDC or payload exposure. The model predicts that a peptide with MW of approximately 86 kDa results in maximum tumor PDC or payload exposure despite exhibiting more rapid PK clearance than a 150 kDa ADC. This is in contrast with tissue disposition, as larger peptides led to higher non-tumoral tissue exposure. Further sensitivity analysis indicated this relationship is insensitive to tumor characteristics, such as TAA expression, or PDC binding affinity towards the TAA, but is sensitive to the peptide size. Conclusions: This work demonstrated that this platform PBPK-QSP model can be a useful tool to guide PDC design and lead selection by optimizing peptide size for PDCs.