Zhiyuan Tan (1), Swantje Völler (1,2), Anyue Yin (3), Kaj van Schie (2), Hans Gelderblom (4), Amy Rieborn (3,4), Tom van der Hulle (4), Catherijne Knibbe (1,5), Dirk Jan Moes (3)
(1) Division of Systems Biomedicine and Pharmacology, Leiden Academic Centre for Drug Research, Leiden University, Leiden, The Netherlands. (2) Pharmacy, Leiden Academic Centre for Drug Research, Leiden University, Leiden, The Netherlands. (3) Department of Clinical Pharmacy & Toxicology, Leiden University Medical Center, Leiden, The Netherlands. (4) Department of Medical Oncology, Leiden University Medical Center, Leiden, The Netherlands. (5) Department of Clinical Pharmacy, St Antonius Hospital, Nieuwegein, The Netherlands.
Introduction: Cabozantinib, a second-generation receptor tyrosine kinase inhibitor, is approved for the treatment of metastatic renal cell carcinoma (mRCC) at a 60-mg oral daily dose1. It has been recommended as the preferred treatment option in latest mRCC guidelines2-4. The approved dose is supported by several population pharmacokinetics (PopPK)/pharmacodynamics (PD) studies with data from clinical trials5. In these studies, the predicted average steady-state cabozantinib concentration for 60 mg QD dose was 1125 ng/mL. However, a considerable gap between the real-world and clinical trial results has been observed. Cerbone et al. recently demonstrated that the median trough concentration was only 634 ng/mL for the patients who have disease control6. Furthermore, a recent study from our group showed that increased cabozantinib exposure was not associated with improved outcome in mRCC patients7. Hence, a PK/PD study is needed to study the exposure-response relationship in real-world patients.
Objectives: To evaluate the PK of cabozantinib using real-world patients’ therapeutic drug monitoring (TDM) data. A previously published PopPK model for cabozantinib reported in FDA registration documents (FDA model)8 was used as a basis for the analysis.
Methods: mRCC patients with cabozantinib TDM samples at the Leiden University Medical Center (LUMC) between August 2018 and December 2021 were identified. Patient data were analyzed using the FDA model consisting of a two-compartment disposition model with dual (fast and slow) lagged first-order absorption process 8. Initially, parameters were fixed to the reported values, except for the parameters of the statistical model, and covariates (race and gender) were removed. Then, the parameters clearance (CL) and inter-individual variability (IIV) on CL were estimated and compared to the corresponding estimates of the FDA model. The analysis was performed using NONMEM software (version 7.4.4). Goodness-of-fit (GOF) plots were used to evaluate the fit of the studied models.
Results: Twenty-seven patients (aged 65 (39-85) and 70.3% male) with 75 TDM observations were included in the current study. The patients were treated for a median 75 days with a median dose of 40 mg. Demographics were similar to those of the patients involved in the study of the FDA model. TDM observations mainly comprised trough samples (48%). The external model evaluation results showed that the cabozantinib TDM concentrations (DV) were generally adequately predicted by the FDA model with structural parameters fixed to reported values. The GOF plots indicated acceptable correspondence of observations (DV) – individual predictions (IPRED) and DV – population predictions (PRED). After estimating CL and IIV on CL, the coefficient of variation (CV%) of IIV on CL was 14.6%, which was lower than FDA report (46.2%). This may due to the sparse and less informative data as a high shrinkage was also observed (39%). The estimated CL was 2.84 L/h which is, slightly higher than the reported value (2.23 L/h).
Conclusions: In this model evaluation study with real-world patient data, a PK model that was based on patients included in studies for registration of cabozantinb (FDA model) resulted generally in adequate prediction for the real-world cabozantinib TDM data although covariates in the FDA model were not adopted. Including more data and/or denser data may help to better capture and predict the real-world cabozantinib PK data. Ultimately, the PK observations will be combined with PD outcomes in order to explore the PK/PD relation of cabozantinib in real world patients.
References:
[1] Lacy SA et al. Clin Pharmacokinet (2017) 56 (5), 477-491.
[2] Escudier B et al. Ann Oncol (2019) 30 (5), 706-720.
[3] NCCN Clinical Practice Guideline in Oncology. Kidney Cancer (Version 1.2021). 2020.
[4] Powles T. Ann Oncol (2021) 32 (3), 422-423.
[5] Nguyen L et al. J Clin Pharmacol (2019) 59 (11), 1551-1561.
[6] Cerbone L et al. ESMO Open (2021) 6 (6), 100312.
[7] Krens SD et al. BMC Cancer (2022) 22 (1), 228.
[8] Center for Drug Evaluation and Research FDA. CLINICAL PHARMACOLOGY AND BIOPHARMACEUTICS REVIEW(S) OF CABOMETYX (2015).
Reference: PAGE 30 (2022) Abstr 10056 [www.page-meeting.org/?abstract=10056]
Poster: Drug/Disease Modelling - Oncology