[Population pharmacokinetics of gemcitabine applied to personalize the dosage used in cancer patients]

Abstract

Objective: To develop and internally validate a population pharmacokinetic model for gemcitabine and its metabolite 2',2'-difluorodeoxyuridine (dFdU); and to evaluate its predictive perfomance for personalizing the dosage used in cancer patients.

Methods: Gemcitabine and dFdU plasma concentrations were determined in 18 cancer patients. A 2-compartment pharmacokinetic model was implemented in the NONMEN VI program to determine the appropriate pharmacokinetic parameters. The power to identify the parameters was assessed by parametric bootstrap, and the internal model validation was performed using nonparametric bootstrap and visual and numerical predictive check methods. The final predictive performance of the model was assessed for accuracy and precision during the first (a priori) and second (a posteriori) chemotherapy cycles.

Results: The mean and interpatient variability of gemcitabine and dFdU clearance was 2.70 L/min (31.0%) and 0.0515 L/min (35.8%), respectively. The estimated distribution volume at steady state was 30 L for gemcitabine and 238 L for dFdU. Internal validation confirmed that the population pharmacokinetic model was appropriate for describing the plasma concentrations of gemcitabine and dFdU over time, as well as its variability in the study population. The accuracy and precision of a posteriori gemcitabine plasma concentrations improved by 67% and 46%, respectively, compared to the a priori prediction.

Conclusion: The population pharmacokinetic model adequately characterised the gemcitabine and dFdU plasma concentrations in the study population over time, and can be used to accurately and precisely optimise gemcitabine dosing regimens in cancer patients.