Thomas Rodier (1), David Balakirouchenane (1), Céline Narjoz (2), Jennifer Arrondeau (3), Nihel Khoudour (1), Michel VIDAL (1,4), Elisabeth Fabre (5), Marie Wislez (6), François Goldwasser (3), Benoit Blanchet (1), Alicja Puszkiel (1,4)
(1) Department of Pharmacokinetics and Pharmacochemistry, Cochin Hospital, AP-HP, CARPEM, 75014 Paris, France (2) Department of Clinical Chemistry, Hôpital Européen Georges Pompidou, Assistance Publique Hôpitaux de Paris, University of Paris, Paris, France. (3) Department of Medical Oncology, Hôpital Cochin, AP-HP, 75014 Paris, France. (4) UMR8038 CNRS, U1268 INSERM, Faculty of Pharmacy, University of Paris, PRES Sorbonne Paris Cité, CARPEM, 75006 Paris, France (5) Department of Oncology, Hôpital Européen Georges Pompidou, Paris, France. (6) Department of Thoracic Oncology, AP-HP, Groupe Hospitalier HUPC, Hôpital Cochin, Paris, France; Centre de Recherche des Cordeliers, Université Paris Descartes, Complement, Inflammation and Cancer, Paris, France.
Objectives: Osimertinib is an oral third-generation epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) approved for the treatment of metastatic non-small cell lung cancer (NSCLC) patients harboring EGFR mutation. High inter-individual variability (IIV) in osimertinib pharmacokinetics (PK) could be one of the factors contributing to large variability in clinical response among patients. Osimertinib is metabolized via CYP3A4/5 and is a substrate of drug transporters P-glycoprotein (P-gp) and Breast Cancer Resistance Protein (BCRP). Therefore, polymorphisms in genes encoding for these proteins could impact osimertinib PK. The aims of this study were to develop a population PK model of osimertinib in unselected NSCLC patients and to identify covariates contributing to the IIV in its PK.
Methods: Plasma samples were collected from unselected NSCLC patients in three Parisian hospitals as part of routine care and the quantification of osimertinib concentrations was performed by a validated liquid chromatography–mass spectrometry (LC-MS/MS) method [1]. The PK data were analyzed using the nonlinear mixed effects modeling program MONOLIX® (version 2020R1). One- and two-compartment models with first-order absorption and elimination and a delay in absorption (lag time or transit compartments) were explored. The covariates included genetic polymorphisms (CYP3A4*22, CYP3A5*3, CYP1A2*1F , ABCB1 3435C>T, ABCB1 2677G>T), clinical factors (age, sex, ethnicity, Body Mass Index (BMI), Eastern Cooperative Oncology Group (ECOG) performance status, smoking status, CYP3A, P-gp, ABCG2, CYP1A2 inhibitors or inductors) and biological variables (ASAT, ALAT, CRP, PAL, LDH, bilirubinemia). The log-likelihood ratio test was used to discriminate between models and a decrease of at least 3.84 (p < 0.05) in objective function value (OFV) was considered statistically significant for one additional parameter in the model-building process or in the forward insertion of a covariate.
Results: A total of 433 osimertinib steady-state plasma concentrations were available for 87 patients treated at doses ranging from 40 mg to 160 mg once daily. Median number of samples per patient was 3 (range 1-21) and median sampling time was 14.25 h (range 0.5-28.3) after last dose intake. Eighty patients (92%) had homozygous wild-type genotype for CYP3A4*22 and 7 patients (8%) had CYP3A4*1/*22 genotype associated with decreased CYP3A4 activity. CYP3A5 non-expressors (i.e. CYP3A5*3/*3 genotype) constituted 69% of the study population. All genotypes were in the Hardy-Weinberg Equilibrium except for CYP3A5*3 . However, this deviation was not observed in Caucasian patients (75% of our population). A one-compartment model with first order absorption and elimination provided the best fit. The first-order absorption rate constant (ka) was fixed to 0.24 h-1 according to a previously published value [2] to allow for adequate estimation of all PK parameters. The mean estimates (RSE%) of apparent clearance (CL/F) and apparent distribution volume (V/F) were 10.4 L/h (10%) and 651 L (13%), respectively. The IIV in CL/F and V/F were 46% (RSE) and 55% (RSE) respectively. Residual unexplained variability was described by a proportional error model and was estimated to 36% (4.1%). The vpc of the base model showed good agreement between observed and simulated data. In the univariate analysis, sex (?OFV = −4.85, p = 0.03), CYP3A5*3 (?OFV = −3.91, p = 0.048), ABCB1 2677G>T (?OFV = −4.17, p = 0.04) and CYP1A2 inductor (?OFV = −6.7, p = 0.01) were significantly associated with CL/F. Furthermore, concomitant intake of a P-gp inhibitor and ethnicity were significantly associated with CL/F (?OFVP-gp = −8.24, p = 0.004, ?OFVEthnicity = −4.01, p = 0.04) and V/F (?OFVP-gp = −5.3, p = 0.02, ?OFVEthnicity = −3.96, p = 0.05).
Conclusions: To the best of our knowledge, this is the first study describing osimertinib PK in unselected “real-life” patients and evaluating the impact of genetic polymorphisms on its PK parameters. The mean estimates of PK parameters and IIV are in accordance with a previously published popPK model based on data from clinical trials [2]. The multivariate covariate analysis is ongoing and will help to identify factors explaining high IIV in osimertinib PK.
References:
[1] Reis R, Labat L, Allard M, Boudou-Rouquette P, Chapron J, Bellesoeur A, et al. Liquid chromatographytandem mass spectrometric assay for therapeutic drug monitoring of the EGFR inhibitors afatinib, erlotinib and osimertinib, the ALK inhibitor crizotinib and the VEGFR inhibitor nintedanib in human plasma from non-small cell lung cancer patients. J Pharm Biomed Anal. 5 sept 2018;158:174-83.
[2] Brown K, Comisar C, Witjes H, Maringwa J, de Greef R, Vishwanathan K, et al. Population pharmacokinetics and exposure-response of osimertinib in patients with non-small cell lung cancer. Br J Clin Pharmacol. juin 2017;83(6):1216-26.
Reference: PAGE 29 (2021) Abstr 9784 [www.page-meeting.org/?abstract=9784]
Poster: Drug/Disease Modelling - Oncology