Pharmacokinetics of paclitaxel and its metabolites using a mechanism-based model
Martin Fransson(1), Henrik Gréen(2), Jan-Eric Litton(1), Lena E. Friberg(3)
(1) Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden; (2) Division of Drug Research/Clinical Pharmacology, Department of Medical and Health Sciences, Faculty of Health Sciences, Linköping University, Linköping, Sweden; (3) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden
Objectives: The influence of the genotypes of the metabolizing enzymes, CYP2C8 and CYP3A4 on the clearance of paclitaxel are still not fully established. To further investigate the impact of these enzymes on the metabolic pattern of paclitaxel in vivo this study aimed to expand a previously developed mechanism-based model for population pharmacokinetics of paclitaxel, where the solvent Cremophor EL explains the non-linear disposition , to also include the kinetics of its primary metabolites; 6α-hydroxypaclitaxel (6α) and p-3'-hydroxypaclitaxel (p3), and its secondary metabolite; 6α-, p-3'-dihydroxypaclitaxel (6α-p3), which is formed by further oxidization of the primary metabolites.
Methods: 33 women diagnosed with gynaecological cancer were treated with paclitaxel in combination with carboplatin during a 3-h infusion at a dose of 175 mg/m2 (n=30) or 135 mg/m2 (n=3). Genotypes were determined for CYP2C8, CYP3A4 and ABCB1/mdr-1 variants along with CYP3A4 activity. Population pharmacokinetic analysis of plasma samples was performed using NONMEM. The PRIOR subroutine with prior information from literature was used to support lack of data for unbound concentrations of paclitaxel  and concentrations of Cremophor EL .
Results: Model building was based on 1156 samples; 345 from paclitaxel, 332 from 6α, 336 from p3 and 143 from 6α-p3. Parameters for paclitaxel were close to prior values. Estimated unbound metabolite concentrations were best fitted using a one compartment model. Total 6α and p3 concentrations were both found to be dependent on Cremophor EL concentrations, and were best fitted using a Hill equation with an additive Cremophor EL component. No association between total 6α-p3 and Cremophor EL was found. Clearance/fm of 6α was significantly bidirectional correlated with the mdr-1 tri allele G2677T/A on a level of P < 0.05.
Conclusion: Paclitaxel metabolite kinetics seems to be highly influenced by Cremophor EL concentrations. The mdr-1 tri allele G2677T/A may affect clearance of the paclitaxel metabolite 6α-hydroxypaclitaxel. No correlation between population variability in the metabolism of paclitaxel and genotype variants of metabolizing enzymes could be identified.
 Henningsson, A., Karlsson, M.O., Gianni, L., Vigano, L., Sparrebom, A., 2001. Mechanism-based pharmacokinetic model for paclitaxel. J. Clin. Oncol. 19, 4065-4073.
 Henningsson, A., Sparrebom, A., Loos, W.J., Verweij, J., Silvander, M., Karlsson, M.O., 2005. Population pharmacokinetic model for Cremophor EL. PAGE 14, 770, Abstract.