P. Girard (1), A. Kovar (2), B. Brockhaus (2), M Zuehlsdorf (2), M Schlichting (2), A. Munafo (1).
(1) Merck Serono S.A., Geneva, Switzerland; (2) Merck KGaA, Darmstadt, Germany
Objectives Cetuximab (CET) is a monoclonal antibody targeting the epidermal growth factor receptor which is prescribed for different solid tumor cancers, in association with cytotoxics (CTX). The treatment has been shown to reduce tumor size (TS) and improve overall survival (OS) [1]. We developed an exposure-TS-OS model to explore potential differences between weekly (QW) and every-2-week (Q2W) CET regimen in metastatic colorectal cancer patients.
Methods Patients from 3 studies, receiving CET weekly (400 mg/m² wk1 then 250 mg/m2/wk) or Q2W (400 or 500 mg/m2 every 2 wk) and one CTX regimen (FOLFIRI, FOLFOX4 or irinotecan) were analyzed. A simplified 2-compartment PK model with linear elimination was used to estimate individual AUC. PK was missing in one study and AUC fixed to population value adjusted by body weight. TS was described by differential equation:
dTS/dt=[Kf – PCET •AUCb•PCTX •exp(-Kr•t)]•TS
where t is time; Kf and Kr rate constants of TS formation and resistance; PCET and PCTX the CET and CTX treatment effect parameters; AUCb the AUC in biophase compartment [2]. From this model, predictions of early tumor shrinkage at week 8 (ETS) and time to nadir (TN time at which individual predicted TS is minimal) were estimated [3]. For both TS and OS models, KRAS mutation, health status, dosage regimen and study were tested as potential covariates. For OS model only, ETS and TN were tested separately. All modeling was performed with NONMEM7.2 (FOCEI) and R.
Results 369 patients contributed to 3821 PK, 2053 TS and 233 death observations. For TS model, 3 significant covariates were identified: KRAS mutation decreased treatment effect PCET by 63% and CTX increased it by 50%. Q2W regimen decreased significantly and only Kr, but this effect disappeared after excluding 2nd-line treatment resistant patient data. TS model provided excellent prediction and was qualified by stratified visual predictive check. For OS model, TN was found as the best predictor of survival, better than ETS, followed by baseline TS, health status and KRAS mutation.
Conclusion This exposure-TS model is the first one developed for cetuximab. It confirmed that KRAS wild-type is a predictive marker for OS. No difference between the treatment regimens was observed on any parameter, including the development of resistance, in first-line setting patients. For overall survival, increase in time to TS nadir was found as the best predictive covariate, as suggested by a previous analysis [4].
References
[1] Liu L et al. Cetuximab-based therapy versus non-cetuximab therapy for advanced cancer: a meta-analysis of 17 randomized controlled trials. Cancer Chemother Pharmacol,65, 2010.
[2] Claret L et al. Model-based prediction of phase III overall survival in colorectal cancer on the basis of phase II tumor dynamics. J.Clin.Oncol. 27, 2009.
[3] Olofsen E. Calculation of time-to-nadir, NMusers, 2004.
[4] Claret L et al. Evaluation of Tumor-Size Response Metrics to Predict Survival and Progression Free Survival in First-Line Metastatic Colorectal Cancer. Abst 2328, PAGE, 2012.
Reference: PAGE 22 (2013) Abstr 2902 [www.page-meeting.org/?abstract=2902]
Poster: Oncology