A novel workflow integrating whole-body PET microdosing data and therapeutic-dose pharmacokinetics across species to inform first-in-human dose selection

Introduction: First-in-human studies require dose extrapolation from pharmacokinetic animal studies combined with safety assessments. Subtherapeutic doses of radiolabelled drugs can be administered in preclinical and early clinical development to gain a dynamic pharmacokinetic understanding, potentially informing pharmacologically-based pharmacokinetics (PBPK) models through whole-body PET data. Aims: to develop a PET-informed modelling framework for preclinical-to-human extrapolation, with dolutegravir as a case-study. Methods: We developed a structured workflow integrating micro- and conventional-dose data for interspecies and dose extrapolation. We built a whole-body PBPK model in PK-Sim/MoBi (v12.1) using non-human primate (NHP) data in 5 key organs incorporating dolutegravir physico-chemical properties, protein binding, metabolism and efflux. Sensitivity analyses and parameter estimation were performed sequentially first with PET microdosing organ data over 3h, then with fluid and tissue concentrations following a 2.5 mg/kg IV injection. Finally, 100 Caucasian healthy adults (50% male, 20-80 years) receiving 50 mg qd po after high-fat meals were simulated using the two sets of estimated parameters and physiology-related parameter distributions provided by PK-Sim. Results: Dolutegravir blood data were well described in NHPs over 3 hours, with parameters adjusted to handle the macrodose-related changes. While microdose-based parameter estimates systematically underpredicted exposure, combining NHP micro- and conventional dose data predicted steady-state geometric mean AUC0-24 and Cmax closely matching human reported profiles, although slightly underpredicting Ctrough. Conclusions: PET-PBPK modelling combining micro- and conventional doses in NHPs successfully predicted dolutegravir concentrations in healthy volunteers, additionally informing tissue distribution. This proof of concept study supports early PET data acquisition to build robust priors for first-in-human studies.