Hippolyte Darre1, Arnaud Nativel, Perrine Masson1, Loïc Etheve1, Nicolas Ratto1, Adèle L’Hostis1, Claudio Monteiro1
1 Nova In Silico
Introduction Osimertinib, a third-generation EGFR tyrosine kinase inhibitor, is a standard of care for first-line treatment in advanced EGFR-mutated non-small cell lung cancer (NSCLC) with exon 19 deletions or exon 21 L858R mutations. In metastatic settings, osimertinib’s clinical efficacy relies partly on its penetration into multiple metastatic sites, including the brain, bone, and liver. Physiologically based pharmacokinetic (PBPK) modeling offers a mechanistic means of predicting plasma and tissue concentrations over extended treatment schedules, thereby supporting dose optimization and combination strategies. We developed a PBPK model that characterizes osimertinib distribution in both primary tumors and metastatic lesions, with the aim of enabling personalized dosing considerations and in silico trial simulations. Objectives To create and validate a whole-body PBPK model of osimertinib that predicts concentration–time profiles across key organs and metastatic sites, to integrate this model within a quantitative systems pharmacology (QSP) NSCLC framework on the Jinko platform, and to simulate multiple virtual Phase III trials in EGFR-mutated metastatic NSCLC. Methods A whole-body PBPK model was built using tissue-to-plasma partition coefficients derived from Rodgers and Rowland theory [1]. Compartments included lungs, brain, liver, and bone, each extended with metastatic subcompartments informed by tissue-specific perfusion rates. Osimertinib metabolism accounted for formation of AZ5104 and subsequent pharmacokinetics. Model calibration employed published clinical PK datasets capturing both steady-state measurements and reported accumulation over multiple cycles [2]. The calibrated PBPK model was then combined with a QSP-based disease model (ISELA-V2) [3] on the Jinko platform [4], enabling comprehensive three-year virtual trials in a diverse population of virtual patients presenting with metastatic EGFR-mutated NSCLC. Results The PBPK–QSP integrated simulation reproduced observed plasma exposures of osimertinib with <12% deviation from clinical data. Tissue-level concentrations predicted in metastatic sites (brain, bone, liver) aligned with the efficacy signals observed in clinical settings. In a FLAURA2-like virtual trial, the combined model accurately captured progression-free survival and hazard ratios, confirming robustness in linking estimated drug concentrations to long-term clinical endpoints. These findings underscore the model’s utility in predicting multi-year outcomes under various dose and combination scenarios, consistent with results seen in other in silico validations [5]. Conclusion A whole-body PBPK model incorporating metastatic compartments can reliably portray osimertinib’s extended tissue distribution in EGFR-mutated NSCLC. When implemented within the Jinko platform, this PBPK–QSP approach provides a powerful tool for guiding dose optimization and exploring combination regimens in multi-year trial simulations. By capturing both plasma and lesion-specific drug dynamics, it has the potential to accelerate clinical development and refine therapeutic strategies for patients with metastatic disease.
[1] Rodgers T, Rowland M. Physiologically based pharmacokinetic modelling 2: predicting the tissue distribution of acids, very weak bases, neutrals and zwitterions. J Pharm Sci. 2006 Jun;95(6):1238-57. doi: 10.1002/jps.20502. Erratum in: J Pharm Sci. 2007 Nov;96(11):3153-4. [2] Planchard D, et al. (2016). Osimertinib Western and Asian clinical pharmacokinetics in patients and healthy volunteers: implications for formulation, dose, and dosing frequency in pivotal clinical studies. Cancer Chemother Pharmacol. doi: 10.1007/s00280-016-2992-z [3] Darré H, Masson P, Nativel A, et al. (2024). Comparing the Efficacy of Two Generations of EGFR-TKIs: An Integrated Drug–Disease Mechanistic Model Approach in EGFR-Mutated Lung Adenocarcinoma. Biomedicines, 12(3):704 [4] Jinko modeling & simulation platform, Nova In Silico, www.jinko.ai, (2025, March) [5] Duruisseaux M, Masson P, Nativel A, et al. (2024). Prospective validation of in silico clinical trials prediction using an EGFR-mutated NSCLC mechanistic model (FLAURA2 and MARIPOSA). J Clin Oncol, 42(16_suppl):8614
Reference: PAGE 33 (2025) Abstr 11426 [www.page-meeting.org/?abstract=11426]
Poster: Drug/Disease Modelling - Oncology