Anyue Yin (1)*, G. D. Marijn Veerman (2)*, Johan G.C. van Hasselt (3), Christi M. J. Steendam (4), Hendrikus Jan Dubbink (5), Lena E. Friberg (6), Henk-Jan Guchelaar(1), Anne-Marie C. Dingemans (4), Ron Matthijssen (2), Dirk Jan A.R. Moes(1) * These authors contributed equally
(1) Department of Clinical Pharmacy and Toxicology, Leiden University Medical Center, Leiden, the Netherlands, (2) Department of Medical Oncology, Erasmus MC Cancer Institute, Rotterdam, Netherlands, (3) Division of Systems Pharmacology and Pharmacy, Leiden Academic Centre for Drug Research (LACDR), Leiden University, Leiden, The Netherlands, (4) Department of Pulmonary Diseases, Erasmus MC Cancer Institute, Rotterdam, Netherlands, (5) Department of Pathology, Erasmus MC Cancer Institute, Rotterdam, Netherlands, (6) Department of Pharmacy, Uppsala University, Uppsala, Sweden
Introduction: Erlotinib is an effective targeted treatment of EGFR-mutated non-small-cell lung cancer (NSCLC) [1, 2]. However, the occurrence of acquired drug resistance and the presence of primary resistance can limit its efficacy [1-3]. Additionally, a decrease in erlotinib exposures also showed to correlate with shorter progression-free survival[3].
To improve the treatment outcome for NSCLC patients, it is essential to better understand tumor heterogeneity and the development of drug resistance in lung cancer patients. Modelling tumor size dynamics considering tumor heterogeneity and evolving drug resistance related to resistant clones can support the study of drug resistance and treatment optimization[4]. On the basis of such model, linking erlotinib exposure to the tumor growth inhibition may further improve the understanding of the evolving tumor dynamics under treatment in lung cancer patients and the pharmacokinetics (PK)-pharmacodynamics (PD) relationship of erlotinib.
Objectives: To quantitatively characterize the erlotinib exposure-tumor dynamics relationship and the development of drug resistance in NSCLC patients considering tumor heterogeneity.
Methods:
Data
A dataset containing intensively sampled erlotinib concentrations was obtained from NSCLC patients from two previous PK studies[5, 6]. The data from patients who received erlotinib with water and without concomitant esomeprazole were included. Additionally, longitudinal tumor size measurements and sparsely sampled erlotinib concentrations were obtained from patients from the START-TKI study [3]. The tumor size measurements were assessed as the sum of longest diameters (SLD, mm) of target lesions by Response Evaluation Criteria In Solid Tumors (RECIST).
PK modelling
Population modelling analysis was performed with NONMEM (version 7.4.4). Based on the obtained PK data, a population PK model was first developed. Patients’ demographic information and lab test results were investigated as covariates. The final model was evaluated with goodness-of-fit (GOF) plots and visual predicted checks (VPC).
PK-PD modelling
A PK-PD model was then developed to characterize the tumor size dynamics. The estimated individual PK parameters were applied to simulate drug exposure. The tumor heterogeneity was considered and models considering 1) acquired resistance and 2) primary and acquired resistance, with or without drug exposure (trough concentration) dependent decay were explored. Next to the GOF and VPC plots, the models were also evaluated by the Akaike information criterion (AIC), the percentage of predicted tumor sizes within ± 20% of the observations, and the root-mean-square deviation (RMSE) between predicted and observed values.
Results:
The intensively sampled erlotinib PK time curves (13 measurements per patient) were obtained from 29 NSCLC patients. The tumor size measurements (in median 7 measurements (range 2-19) per patient) and the sparsely sampled erlotinib concentrations (in median 8 measurements (range 1-21) per patient) were collected from 18 patients with EGFR-mutated NSCLC from the START-TKI study[3].
A two-compartment population PK model with first-order absorption with lag time and first-order elimination was demonstrated to best describe the obtained PK data. No significant covariate was identified. The predictability of the developed model was demonstrated to be sufficient with the GOF and VPC plots.
The results of tumor dynamics modelling showed that the model with only acquired resistance had better fitness than the model considering primary resistance (AIC decreased by 10.725). The percentage of predicted tumor sizes within ± 20% of the observations was 97.8%. Incorporating an exposure-dependent decay resulted in the same percentage but decreased RMSE from 3.78 to 3.74. The GOF and VPC plots also indicated sufficient predictability of the final model.
Conclusions: The erlotinib concentrations collected in NSCLC patients were well predicted by the developed two-compartment PK model. The tumor dynamics of the included NSCLC patients were well captured by the model with acquired resistance and exposure-dependent treatment effect. On the basis of the current results, the PK-PD relationship of erlotinib and the correlation between tumor dynamics and genetic biomarkers in NSCLC patients can be further explored.
References:
[1] Nagano, T., M. Tachihara, and Y. Nishimura, Mechanism of Resistance to Epidermal Growth Factor Receptor-Tyrosine Kinase Inhibitors and a Potential Treatment Strategy. Cells, 2018. 7(11).
[2] Oliveira, K.C.S., et al., Current Perspectives on Circulating Tumor DNA, Precision Medicine, and Personalized Clinical Management of Cancer. Mol Cancer Res, 2020. 18(4): p. 517-528.
[3] Steendam, C.M.J., et al., Plasma Predictive Features in Treating EGFR-Mutated Non-Small Cell Lung Cancer. Cancers (Basel), 2020. 12(11).
[4] Yin, A., et al., Anti-cancer treatment schedule optimization based on tumor dynamics modelling incorporating evolving resistance. Sci Rep, 2022. 12(1): p. 4206.
[5] van Leeuwen, R.W., et al., Influence of the Acidic Beverage Cola on the Absorption of Erlotinib in Patients With Non-Small-Cell Lung Cancer. J Clin Oncol, 2016. 34(12): p. 1309-14.
[6] Veerman, G.D.M., et al., Influence of Cow’s Milk and Esomeprazole on the Absorption of Erlotinib: A Randomized, Crossover Pharmacokinetic Study in Lung Cancer Patients. Clin Pharmacokinet, 2021. 60(1): p. 69-77.
Reference: PAGE 30 (2022) Abstr 10133 [www.page-meeting.org/?abstract=10133]
Poster: Drug/Disease Modelling - Oncology