Introduction Microdosing approaches allow the administration of infra-therapeutic doses of drug candidates to humans. This approach, carried out upstream of Phase I, is referred as “Phase 0” [1]. Decisive clinical PK can therefore be obtained at a early stage compared to the traditional process. A microdose PET (positron emission tomography) study was performed in subjects to assess the role of Organic Anion-Transporting Polypeptides (OATP) for liver uptake and their impact on whole-body pharmacokinetics (WBPK), using the substrate Glyburide as a model drug [2]. The aim of this work was to model the WBPK of radiolabeled glyburide (11C-glyburide) with or without pre-infusion of rifampicin, a potent inhibitor of OATPs [2]. Methods Whole-body dynamic acquisitions after injection of a microdose of 11C-glyburide were obtained in 16 healthy volunteers using a PET scanner, with repeated acquisition, over 40 minutes. PET images were reconstructed manually outlining volumes of interest (VOIs). Time-activity curves (TACs) generated by the VOIs (MBq/cc) were converted to concentrations (µg/mL). 10 subjects also underwent a second session, where injection of 11C-glyburide was preceded by an infusion of rifampicin. Using Nonlinear Mixed Effect models (NLMEM), we modelled the kinetics in 5 selected organs of importance for the distribution and elimination: arterial blood compartment (aorta + left ventricle of the heart), liver, kidneys, pancreas and spleen . The structural model was built using the data from 11C-glyburide administered alone. Each organ was first modelled separately using standard compartment models, to obtain initial parameter estimates. Models were then fitted simultaneously for all 5 organs, investigating for each organ the number of compartments, linear and non-linear rate constants for distribution and elimination, and finally the residual error model. Model selection was based on the Bayesian Information Criterion (BIC). For organs which showed rapid distribution, a simple ratio of organ to blood concentrations was tested. The model selected was applied to the full dataset with both occasions, introducing interoccasion variability (IOV) and testing for correlations for all variabilities. Body weight was included using an allometric model on volumes of distribution, and kept if the BICc did not degrade by more than 3 points. The other covariates were incorporated using the COSSAC algorithm [3]. The covariance structure was simplified by removing variabilities if the RSE exceeded 100%. The final model was evaluated using Visual Predictive Check (VPC). Python was used for data analysis, and Monolix - 2023R1 for NLMEM. Results The final model included 6 compartments. Liver and kidney required two compartments each, representing within-tissue exchanges, while pancreas concentrations were found to be in rapid equilibrium with blood. The effect of body weight was included using an allometric model. Correlation between IOV random effects were found between Blood, Liver and Spleen volumes. The COSSAC algorithm showed that Rifampicin decreased distribution to liver. The transfer rate constant from blood to liver decreased by 38% (RSE = 84%), and a 48 % decrease in liver volume of distribution (RSE = 67%). In addition, the transfer rate constants between blood and kidneys were lower (51%) in subjects over 50 years old compared to young men. VPC showed good fits for the typical profiles in the different organs, but IIV appear to be overestimated and would need to be refined in a larger sample of subjects. Conclusions We developed a model to describe the WBPK of 11C-glyburide in 5 organs. Rifampicin was found to modify liver distribution, consistent with previous results from a non-compartmental analysis [2]. PET studies represent an opportunity to safely and dynamically follow the kinetics of radiolabeled drugs in the entire human body. A drawback is the limited observation period, which makes it difficult to estimate elimination reliably. Indeed, we found a two-fold underestimation of the elimination rate when compared to previous studies [4], as well as a smaller central volume of distribution. A perspective of this work is to combine PET data with preclinical PK studies using Bayesian approaches to predict the WBPK of investigational drugs even before traditional Phase I studies.