Christine Xu, Vanaja Kanamaluru
Pharmacokinetics and Pharmacometrics, Drug Disposition, Safety and Animal Research, Sanofi Bridgewater, NJ, USA
Objectives: The objective of this analysis was to predict CYP3A based drug-drug interactions for fedratinib, a JAK2-selective inhibitor.
Methods: Fedratinib is a substrate of CYP3A4 with dose- and time-dependent nonlinear pharmacokinetics (PK). In vitro, fedratinib demonstrated time-dependent inhibition of CYP3A4 and a concentration-dependent increase in CYP3A4 gene expression (induction). A PBPK model was developed to predict complex drug-drug interactions involving induction and inhibition of CYP3A4[1]. SimCYP simulator was used with inputs of physico-chemical parameters and in vitro and in vivo PK data of fedratinib. After verification, the model was applied to predict fedratinib drug interactions that had not been assessed clinically, namely with ketoconazole at a different fedratinib dosing regimen or other perpetrators such as erythromycin or rifampicin.
Results: The PBPK model simulated PK profiles of fedratinib following a single 500 mg dose in healthy subjects or repeated 500 mg once daily (QD) doses in myelofibrosis patients were consistent with the observed profiles in vivo. The PBPK model predicted increase in fedratinib Cmax and AUC following co-administration of 200 mg twice daily (BID) ketoconazole with a single 300 mg dose of fedratinib in healthy subjects was 2.07 fold and 3.02 fold, respectively. The corresponding observed increase in the clincal study was 1.93 and 3.06 for respectively. These results confirmed the performance of the PBPK model for subsequent application to untested clinical situations. Co-administration of 200 mg BID ketoconazole or 500 mg three times daily erythromicin with a single 400 mg dose of fedratinib predicted to an increase in AUC by 2.7-fold and 2.2-fold, respectively. However, repeated administration of 400 mg QD fedratinib with the CYP3A4 inhibitors predicted a smaller increase (1.7-fold and 1.2-fold, respectively). Co-administration of 600 mg QD doses of rifampin with a single 400 mg dose and repeated 400 mg QD doses of fedratinib predicted to a decrease in AUC by 5.6-fold and 1.8-fold respectively.
Conclusions: After a single dose of fedratinib, the PBPK model predicted up to a 3-fold increase or a 6-fold decrease in exposure of CYP3A4 inhibition or induction, respectively. The magnitude of effect was smaller after repeated doses and is likely due to an interplay of auto inhibition and induction of CYP3A4 by fedratinib. The PBPK model could be used to support dose recommendations when fedratinib is co-administered with CYP3A4 inhibitors or inducers[2].
References:
[1] Xu C., Djebli N., Kanamaluru V. Physiologically based pharmacokinetic (PBPK) approach to discern potential population differences in patients with refractory solid tumors and healthy subjects: the effects of fedratinib on CYP3A4 substrate midazolam 5th American Conference on Pharmacometrics (ACOP) 2014 [2] After the completion of this study, cases consistent with Wernicke’s encephalopathy were reported in patients. All clinical trials involving fedratinib were halted.
Reference: PAGE 24 (2015) Abstr 3545 [www.page-meeting.org/?abstract=3545]
Poster: Drug/Disease modeling - Absorption & PBPK