Utilization of Brain Physiologically Based Pharmacokinetic Modeling in Support of CNS Drug Development: Case Example of Fenebrutinib.

While accurately predicting the human brain penetration of drugs has been notoriously challenging, recent physiologically based pharmacokinetic (PBPK) modeling work has attempted to integrate passive permeability, active transport, and physiological parameters to support central nervous system (CNS) exposure predictions. In this study, a bottom-up PBPK approach was utilized prospectively to predict human cerebrospinal fluid (CSF) concentrations of fenebrutinib. In vitro permeability, efflux transporter, and brain binding data were generated and extrapolated in vitro to in vivo for required inputs in a five-compartment mechanistic brain PBPK model. The approach was first assessed using a compound with publicly available clinical CSF data, which informed the use of empirical scalars to address possible in vitro to in vivo translation gaps. The refined bottom-up PBPK approach was then used to predict the CSF concentrations of fenebrutinib. These predictions aligned well with observed CSF data as predicted values fell within 1.5 fold of the observed measurements, indicating the fenebrutinib has good brain penetration in multiple sclerosis (MS) patients, which were available afterward. This work demonstrates the utility of PBPK modeling to enable prospective CNS exposure prediction and support early clinical decision-making.