Letizia Carrara1, Ludwig Vincent2, Alice Ke2, Adrien Tessier1, Laurence Launay3, Sandeepraj Pusalkar3, Tharin Limsakun4, Kate Wu4, Elise Lemasson5, Samer El Bawab1, Mohammad Hossain4, Sylvain Fouliard1
1Quantitative Pharmacology, Translational Medicine; Servier, 2Certara UK Limited, Simcyp Division, Sheffield, UK, 3DMPK, Translational Medicine; Servier, 4Clinical Pharmacology, Translational Medicine; Servier, 5CMC; Servier
OBJECTIVES: vorasidenib is an orally available brain-penetrant and potent inhibitor of isocitrate dehydrogenase-1 and 2 (IDH1 and IDH2) mutant proteins for the treatment of patients with grade 2 astrocytoma or oligodendroglioma with a susceptible IDH1 or IDH2 mutation following surgery [1]. In vitro, vorasidenib is a CYP1A2 substrate, induces CYP2B6, CYP2C8, CYP2C9, CYP2C19 and CYP3A4, and inhibits P-gp and BCRP transporters. The objective of this analysis was to develop a physiologically based pharmacokinetic (PBPK) model to predict the drug-drug interaction (DDI) potential of vorasidenib as a victim (CYP1A2 substrate) and as a perpetrator of DDI (CYPs inducer and transporters inhibitor). METHODS: The model was developed using SimCYP v21. In vitro particle size distribution, disintegration time and dissolution data were coupled with clinical pharmacokinetics (PK) data to parameterize a mechanistic absorption model and bridge the different formulations used in the clinical development program. The main drug metabolizing enzyme was identified based on in vitro (CYP1A2). The fraction metabolized by such enzyme (fmCYP1A2) was estimated based on the biotransformation pathways and metabolite profiling in human mass balance study and verified against the data obtained from a clinical DDI study with a strong CYP1A2 inhibitor. The in vitro interaction parameters measuring the interaction potential of vorasidenib towards the CYPs and transporters mentioned above were included into the model. Clinical PK data from Healthy Volunteers (HV) and patient studies following single and multiple doses (SD and MD) were incorporated into the model. Subsequently, the model was applied to simulate the perpetrator and victim DDI potential of the drug. DDI was evaluated as the geometric mean ratios (GMR) of AUC and Cmax of the victim drug in the presence/absence of the perpetrator. RESULTS: A full PBPK model was developed, with Advanced Dissolution Absorption and Metabolism (ADAM) model [2] for absorption, enzymatic clearance (fmCYP1A2=97%) and an additional organ to capture the terminal elimination phase. The model allowed to describe the concentration-time profile of vorasidenib observed in HV (dose range: 10 to 50 mg) and in subjects with glioma (dose range: 10 to 300 mg QD): prediction error (predicted/observed in mean data for AUC and Cmax) was within 2-fold in the majority of the cases [3]. The model also adequately described the DDI effect observed in clinical study with ciprofloxacin, a strong CYP1A2 inhibitor, on single dose PK of vorasidenib (prediction error within 1.32-fold). PBPK MODEL SIMULATIONS FOR VORASIDENIB AS A VICTIM OF DDI. The predicted changes in vorasidenib steady-state exposure in the presence of CYP1A2 modulators are: -a vorasidenib GMR of AUCtau and Cmax of 7.18 and 5.70, respectively, following treatment with fluvoxamine (50mg BID for 100 days) a strong CYP1A2 inhibitor. -a similar effect on vorasidenib exposure following treatment with phenytoin (300 mg QD for 30 days) or rifampicin (600 mg BID for 30 days), both moderate CYP1A2 inducers, with a vorasidenib GMR of AUCtau and Cmax of 0.610 and 0.703 (phenytoin) and 0.609 and 0.702 (rifampicin). PBPK MODEL SIMULATIONS FOR VORASIDENIB AS A PERPETRATOR OF DDI. The simulated changes in exposure of sensitive substrates for CYPs and transporters following MD of vorasidenib (40mg QD for 30 days) were as follows: •Midazolam (CYP3A4 substrate): AUCinf GMR=0.179 and Cmax GMR=0.208 •Bupropion (CYP2B6 substrate): AUCinf GMR=0.789 and Cmax GMR=0.818 •Repaglinide (as CYP2C8 substrate only): AUCinf GMR=0.912 and Cmax GMR=0.946 •S-Warfarin (CYP2C9 substrate): AUCinf GMR=0.816 and Cmax GMR=0.968 •S-Mephenytoin (CYP2C19 substrate): AUCtau GMR=0.650 and Cmax GMR=0.705 •Rosuvastatin (BCRP substrate): AUCinf GMR=1.05 and Cmax GMR=1.21 •Digoxin (P-gp substrate): AUCinf GMR=1.00 and Cmax GMR=1.02 CONCLUSIONS: The PBPK model predicted potentially clinically relevant PK interactions for vorasidenib as victim and as perpetrator of DDI, offering valuable insights for evaluating and managing DDI risk.
[1] Mellinghoff, Ingo K., et al. “Vorasidenib in IDH1-or IDH2-mutant low-grade glioma.” New England Journal of Medicine 389.7 (2023): 589-601. [2] Jamei, Masoud, et al. “Population-based mechanistic prediction of oral drug absorption.” The AAPS journal 11 (2009): 225-237. [3] Wagner, Christian, et al. “Predicting the effect of cytochrome P450 inhibitors on substrate drugs: analysis of physiologically based pharmacokinetic modeling submissions to the US Food and Drug Administration.” Clinical pharmacokinetics 54 (2015): 117-127.
Reference: PAGE 33 (2025) Abstr 11362 [www.page-meeting.org/?abstract=11362]
Poster: Drug/Disease Modelling - Absorption & PBPK