Plasma EC 90 multiples are drug, target site and species specific—A physiologically based pharmacokinetic modelling example of nirmatrelvir central nervous system distribution

Background and Purpose Achieving minimum unbound drug concentration in plasma above in vitro EC 90 is a common antiviral pharmacokinetic/pharmacodynamic (PK/PD) target. However, it often overlooks restricted distribution into target sites like the central nervous system (CNS), where a trough concentration above 1 × EC 90 is needed. Currently, the required EC 90 multiplier ( n ) is defined empirically without mechanistic evidence. This study utilizes physiologically based pharmacokinetic (PBPK) modelling of nirmatrelvir in rats to bridge the gap between plasma and target‐site exposure, providing a mechanistic basis for the minimal EC 90 multiplier in antiviral dose selection. Experimental Approach A nirmatrelvir population PK (popPK) model developed from rat plasma data was integrated into the LeiCNS PBPK platform. Evaluated against rat cerebrospinal fluid (CSF) concentration, the PBPK model was used to simulate drug distribution across various CNS compartments, to assess the plasma‐to‐periphery relationship of antivirals with varying blood–brain barrier (BBB) permeability. Key Results The simulation results highlight how both unbound fraction in plasma ( f u ) and unbound tissue‐to‐plasma partition coefficient ( K p,uu ) are the primary determinants of the minimal EC 90 multiplier. We demonstrated that aiming to reach plasma C min above an arbitrary multiple of the EC 90 without accounting for drug and species‐dependent parameters, and related body distribution, can lead to insufficient drug exposure at peripheral target sites. Conclusion and Implications Target‐site concentration, rather than total plasma concentration, is the mechanistically correct effect site for linking to in vitro EC 90 . By incorporating both f u and K p,uu , a more mechanistically informed EC 90 multiplier can be predicted a priori.