Physiologically based pharmacokinetic modeling to assess metabolic drug-drug interaction risks and inform the drug label for fedratinib

Abstract

Purpose: Fedratinib (INREBIC^®), a Janus kinase 2 inhibitor, is approved in the United States to treat patients with myelofibrosis. Fedratinib is not only a substrate of cytochrome P450 (CYP) enzymes, but also exhibits complex auto-inhibition, time-dependent inhibition, or mixed inhibition/induction of CYP enzymes including CYP3A. Therefore, a mechanistic modeling approach was used to characterize pharmacokinetic (PK) properties and assess drug-drug interaction (DDI) potentials for fedratinib under clinical scenarios.

Methods: The physiologically based pharmacokinetic (PBPK) model of fedratinib was constructed in Simcyp^® (V17R1) by integrating available in vitro and in vivo information and was further parameterized and validated by using clinical PK data.

Results: The validated PBPK model was applied to predict DDIs between fedratinib and CYP modulators or substrates. The model simulations indicated that the fedratinib-as-victim DDI extent in terms of geometric mean area under curve (AUC) at steady state is about twofold or 1.2-fold when strong or moderate CYP3A4 inhibitors, respectively, are co-administered with repeated doses of fedratinib. In addition, the PBPK model successfully captured the perpetrator DDI effect of fedratinib on a sensitive CY3A4 substrate midazolam and predicted minor effects of fedratinib on CYP2C8/9 substrates.

Conclusions: The PBPK-DDI model of fedratinib facilitated drug development by identifying DDI potential, optimizing clinical study designs, supporting waivers for clinical studies, and informing drug label claims. Fedratinib dose should be reduced to 200 mg QD when a strong CYP3A4 inhibitor is co-administered and then re-escalated to 400 mg in a stepwise manner as tolerated after the strong CYP3A4 inhibitor is discontinued.

Keywords: Drug–drug interaction; Fedratinib; Metabolic inhibition; PBPK; Simcyp.

Fig. 1

Modeling Procedures for the PBPK Model of Fedratinib

Fig. 2

Model-simulated and Clinically Observed Mean Plasma Concentration Profiles of Fedratinib in Healthy Subjects Following Single Doses of 300 mg (Subplot a) or 500 mg (Subplot b) Fedratinib and in Myelofibrosis (MF) Patients Following Repeated Doses (QD) of 300 mg (Subplot c), 400 mg (Subplot d), and 500 mg (Subplot e) Fedratinib. The clinically observed mean plasma concentrations are represented as the open symbols. The mean and individual trial simulations using the default Simcyp^® “Healthy Volunteers” are represented as solid black and gray curves (Number of virtual trials = 10), respectively, in Subplot A and B. The mean simulations using the default Simcyp^® “Healthy Volunteers” and “Cancer” populations are presented and solid and dashed black curves in Subplot c–e

Fig. 3

Summary of Model-predicted Drug-drug Interactions between Fedratinib (as the Victim) and CYP3A4 Modulators in Single Dose (Subplot a) and Repeated Dose (Subplot b) Scenarios. Subplot a: In the DDI simulations under the single-dose scenario, fedratinib was administered to healthy subjects as a single dose of either 300 mg (in the “Observation” and “Verification” categories) or 400 mg (in the “Prediction” category), and AUC refers to AUC_inf. Subplot b: In the DDI simulations under the repeated-dose scenario, fedratinib was administered to healthy subjects as 400 mg QD (in the “Prediction” category), and AUC refers to AUC_τ at steady state. In both subplots, the symbols represent geometric means of the estimates and error bars represent the 90% confidence intervals of the estimates

Fig. 4

Comparison between Model-Predicted and Clinically Observed Cmax (Subplot a) and AUC_τ (Subplot b) in Myelofibrosis Patients

Fig. 5

Model-simulated Fractions of Metabolism or Excretion (Fm/Fe) in Healthy Subjects and MF patients Following Repeated Doses of 400 mg QD Fedratinib. D1 Day 1, SS Steady state

Fig. 6

PBPK Simulation Design for Fedratinib Dose Re-Escalation after Discontinuation of Ketoconazole and the Simulated PK Profiles of Fedratinib in Cancer Patients