Novel physiologically based pharmacokinetic modeling of patupilone for human pharmacokinetic predictions

Patupilone (EPO906) is a novel potent microtubule stabilizer, which has been evaluated for cancer treatment. A novel physiologically based pharmacokinetics (PBPK) model was developed based on nonclinical data to predict the disposition of patupilone in cancer patients. After a single intravenous dose (1.2 mg/kg) in male Han-Wistar rats, the tissue distribution of C-14-patupilone was investigated by quantitative whole-body autoradiography (QWBA). The blood radioactivity and patupilone concentration were determined by LC-MS/MS and liquid scintillation counting. A novel PBPK model was developed based on rat tissue concentration data to predict blood concentration-time profiles of patupilone in cancer patients. PBPK parameters derived from the rat were applied to a human PBPK model. Phase I clinical pharmacokinetic data in Caucasian and Japanese cancer patients at various doses ranging from 0.75 to 10 mg/m(2) were successfully described using the PBPK approach. Patupilone dispositions in lung, heart, muscle, spleen, liver, brain, adipose, and testes of rats were well described using the PBPK model developed assuming a perfusion rate-limited distribution between different compartments. For skin and bone marrow, concentration-time profiles were modeled assuming a permeability-limited distribution between different compartments. The simulated human pharmacokinetic profiles from the PBPK model showed good agreement with observed clinical pharmacokinetic data, where the model predicted AUC, t (1/2), V (ss), and CL values were within approximately twofold of the observed values for all dose groups. The distribution of patupilone in rats was well described by a PBPK model based on measured tissue distribution profiles generated by QWBA combined with metabolism data. The human PBPK model adequately predicted blood pharmacokinetics of patupilone in cancer patients. The PBPK model based upon preclinical tissue distribution data can aid in successful prediction of pharmacokinetics in humans.