Mathilde Marchand (1), Rong Zhang (2), Benjamin Wu (2), Ya-Chi Chen (2), Sandhya Girish (2), Jin Jin (2), Rene Bruno (3)
(1) Certara Strategic Consulting, France, (2) Genentech/Roche, USA (3) Genentech/Roche, France.
Introduction: The anti–PD-L1 (programmed death-Ligand1) antibody atezolizumab (Tecentriq®) targets human PD-L1 on tumor-infiltrating immune cells and tumor cells, and inhibits PD-L1 interaction with programmed death 1 (PD-1) and B7.1 receptors, both of which can provide inhibitory signals to T cells. Atezolizumab is being investigated as a potential therapy against various hematologic malignancies and solid tumors and is approved for the treatment of patients with metastatic non-small cell lung cancer (NSCLC), small cell lung cancer, triple-negative breast cancer and locally advanced or metastatic urothelial carcinoma (mUC) in a number of countries. The population pharmacokinetics (popPK) of atezolizumab has been described in patients with various types of cancers collected during phase 1 studies using a time-stationary linear two-compartment model [1,2].
Objectives: The purpose of this work was to update the initial popPK model of atezolizumab and quantify the impact of time-varying covariates on atezolizumab clearance (CL) based on a larger population of cancer patients.
Methods: PK data in cancer patients mainly NSCLC and mUC from 3 clinical studies: PCD4989g (Phase I), OAK and IMvigor211 (Phase III), were pooled (1519 patients with 9165 PK samples) and analyzed with NONMEM® [3]. The structural PK model was kept and the impact of covariates on atezolizumab CL was reassessed on this population. Demographical, physiological covariates (incl. inflammatory markers) were tested. The first step was to identify the baseline covariates affecting atezolizumab CL (at p < 0.01), all significant covariates (univariate assessment (p < 0.01) were included in a full model and subjected to a backward elimination at p < 0.001) (baseline covariate model). The second step was to re-estimate the model allowing baseline covariates to vary over time (time-varying covariate model). Finally, a previously proposed empirical time-varying function for CL [4, 5] was added baseline covariate model. Those three models were evaluated and compared using classical tools (i.e., objective functions (OFV), goodness of fit plots and prediction-corrected visual predictive checks, pcVPCs).
Results: The final time-varying covariate PK model included the effects of body weight, albumin levels, tumor size, anti-drug antibodies (ADA), gender, neutrophil count, alkaline phosphatase and bilirubin levels on atezolizumab CL. The comparison of the three models: baseline covariate, time-varying covariate, and empirical time-varying models, demonstrated that both time-varying models resulted in a clear improvement of the fit and pcVPCs; the best model being the time-varying covariate model (OFV improved by 643 points compared with the baseline model vs 308 points for the empirical model). In this model, the main driver for change in CL over-time was albumin level variations with decrease in CL paralleling increase in albumin (improvement in patient’s status). Time-varying ADAs had a small impact (9% increase in CL). As in the previous time-stationary model [1, 2], covariates did not impact atezolizumab CL by more ± 20% threshold except +27% in low albumin patients.
Conclusion: Time-varying CL driven by time-varying patient status is confirmed for atezolizumab. The impact of covariates on atezolizumab CL is similar in magnitude to the one found with the stationary model [1, 2] with no further impact on atezolizumab dosing recommendations. The results support the hypothesis that CL variation over time is associated with patients’ prognostic factors and disease status as shown with other checkpoint inhibitors [4, 6, 7].
References:
[1] M. Marchand et al. Population pharmacokinetics for atezolizumab in cancer patient. ACOP7 2016, Seattle, (T-70)
[2] Stroh M et al. Clinical pharmacokinetics and pharmacodynamics of atezolizumab in metastatic urothelial carcinoma. Clin Pharmacol Ther. 2017 Aug;102(2)
[3] Beal SL et al. 1989-2011. NONMEM Users Guides. Icon Development Solutions, Ellicott City, Maryland, USA
[4] Liu C et al. Association of time-varying clearance of nivolumab with disease dynamics and its implications on exposure-response analysis. Clin Pharmacol Ther. 2017 May;101(5)
[5] Li H et al. Time dependent pharmacokinetics of pembrolizumab in patients with solid tumor and its correlation with best overall response. J Pharmacokinet Pharmacodyn. 2017 Oct;44(5)
[6] Baverel PG et al. Population pharmacokinetics of durvalumab in cancer patients and association with longitudinal biomarkers of disease status.
Clin Pharmacol Ther. 2018 Apr;103(4)
[7] Turner DC et al. Pembrolizumab exposure-response assessments challenged by association of cancer cachexia and catabolic clearance. Clin Cancer Res. 2018 Dec;24(23)
Reference: PAGE () Abstr 9257 [www.page-meeting.org/?abstract=9257]
Poster: Drug/Disease Modelling - Oncology