Objective: This study used a whole-body physiologically based pharmacokinetic (WB-PBPK) model for moxifloxacin, plus in vitro and in vivo literature data on its interaction with macrophages, to interpret biopsy results generated from patients undergoing primarily colorectal surgery.
Methods: A WB-PBPK model was developed using PK-Sim® software and refined using observed plasma profiles. The model was assessed by comparing predictions of unbound interstitial concentrations with in vivo data from a microdialysis study.
Results: Incorporating in vitro data on the percentage volume of macrophages in a colorectal resection (8.1%) plus the in vivo kinetic and accumulation potential of moxifloxacin in macrophages into the WB-PBPK model, biopsy concentrations and kinetics were predicted and compared with observed data. The WB-PBPK model accurately described adipose and muscle interstitial unbound concentrations. The predicted biopsy concentrations (including interstitial, intracellular, vascular space and macrophages) were slightly greater than the observed values, although the kinetic (i.e. observed biopsy half-life = 21 hours) was similar to that of moxifloxacin in macrophages (20.8 hours) and thus similar to the predicted biopsy half-life. A reduction in the predicted biopsy concentrations to match the observed data required a decrease in the volume fraction of macrophages from 8.1% to 3.6%.
Conclusion: When plasma concentrations are known, WB-PBPK is a method to determine interstitial and intracellular concentrations. In this study, integration of biopsy data with WB-PBPK allowed for generation and testing of hypotheses to determine the reason for the observed biopsy kinetics. This type of translational modelling may lead to a better understanding of the anti-infective pharmacokinetic/pharmacodynamic relationship.