Marion Kerioui(1,2,3), Solène Desmée(2), François Mercier(4), Alyse Lin(5), Ben Wu(5), Jin Y Jin(5), René Bruno(6), Jérémie Guedj(1)
(1)Université de Paris, INSERM, IAME, F-75018 Paris, France (2)Université de Tours, Université de Nantes, INSERM SPHERE, UMR 1246, Tours, France (3)Clinical Pharmacology, Roche/Genentech, Paris, France (4)Biostatistics – Roche Innovation Center Basel, Basel Switzerland (5)Clinical Pharmacology, Genentech Inc., South San Francisco, CA, USA (6)Clinical Pharmacology, Roche/Genentech, Marseille, France
Context: Immunotherapy showed impressive results in oncology in the past few years, but novel patterns of response complicate the association between treatment and survival. Studies report a higher inter-patients variability in response pattern under immunotherapy than chemotherapy, with some patients experimenting an unexpected progression of the tumor size (hyperprogression)[1], some others responding after a progression (pseudoprogression)[2], and some with a very stable and durable response over time. Moreover, the intra-patients variability could also be larger, with lesions responding and other progressing within one patient [3]. The lesion location and its interaction with the treatment could affect the kinetics of the tumor and the risk of death. Thus, the use of the RECIST criterion, relying on the sum of the target lesions as a marker of the tumor size should be further evaluated. In previous work in metastatic bladder cancer patients, joint modelling allowed to characterize the association between tumor size and survival, and to develop efficient tools to predict individual survival under immunotherapy treatment[4]. Considering the kinetics of each target lesions specific to a patient and integrating the impact of the location of each lesion to capture the tumor kinetics in different organs could increase the model predictability of the survival.
Objectives: We aimed to develop a multilevel nonlinear joint model to describe the target lesions kinetics and their association with patient survival, depending on the treatment. The model included impact from the lesion location to investigate the tumor kinetics in different organs and their possibly different degrees of association with survival.
Methods: We investigated data from a phase 3 clinical trial (IMVigor211 [5]) of 900 advanced urothelial carcinoma patients randomized between an immunotherapy treatment arm (Atezolizumab) and a chemotherapy control arm. We built a multilevel nonlinear joint model, exploring effect from the treatment arms and the location sites on patient response. For that, we used the simplified Tumor Growth Inhibition (sTGI) model [6] to describe the lesions’ kinetics and the survival model was a Weibull proportional hazard model with a link between current tumor kinetics in different organs and survival. We selected baseline covariates on the longitudinal and survival parts by forward procedure, and then immunotherapy effect on the different parameters following the same strategy. We provided diagnostic plots at individual and population levels.
Results: Chemotherapy treatments lead to a rapid decrease of tumor size. Conversely, immunotherapy showed a modest decrease in tumor size initially, but a longer decrease and a more durable response compared to chemotherapy. We found significantly different tumor dynamics from one organ to another. Notably, we noticed a faster growth of the lesions in the liver, and a stronger treatment effect in the lymph nodes. The association between tumor kinetics and survival turned out to be sensitive to treatment arm and location of the lesion.
Conclusions: This work provides a framework to evaluate the source of variability in response to immunotherapy compared to chemotherapy.
References:
[1] Maxime Frelaut, Christophe Le Tourneau, and Edith Borcoman. Hyperprogression under immunotherapy. International journal of molecular sciences, 20(11):2674, 2019.
[2] Edith Borcoman, Amara Nandikolla, Georgina Long, Sanjay Goel, and Christophe Le Tourneau. Patterns of response and progression to immunotherapy. American Society of clinical oncology educational book, 38:169–178, 2018.
[3] E Borcoman, Y Kanjanapan, S Champiat, S Kato, V Servois, R Kurzrock, Sanjay Goel, P Bedard, and C Le Tourneau. Novel patterns of response under immunotherapy. Annals of Oncology, 30(3):385–396, 2019.
[4] Coralie Tardivon, Solène Desmée, Marion Kerioui, René Bruno, Benjamin Wu, France Mentré, François Mercier, and Jérémie Guedj. Association between tumor size kinetics and survival in patients with urothelial carcinoma treated with atezolizumab: Implication for patient follow-up. Clinical Pharmacology & Therapeutics, 106(4):810– 820, 2019.
[5] Thomas Powles, Ignacio Duràn, Michiel S Van Der Heijden, Yohann Loriot, Nicholas J Vogelzang, Ugo De Giorgi, St´ephane Oudard, Margitta M Retz, Daniel Castellano, Aristotelis Bamias, et al. Atezolizumab versus chemotherapy in patients with platinum-treated locally advanced or metastatic urothelial carcinoma (imvigor211): a multicentre, open-label, phase 3 randomised controlled trial. The Lancet, 391(10122):748–757, 2018.
[6] Laurent Claret, Manish Gupta, Kelong Han, Amita Joshi, Nenad Sarapa, Jing He, Bob Powell, and René Bruno. Evaluation of tumor-size response metrics to predict overall survival in western and chinese patients with first-line metastatic colorectal cancer. Journal of clinical oncology, 31(17):2110–2114, 2013.
Reference: PAGE () Abstr 9259 [www.page-meeting.org/?abstract=9259]
Poster: Oral: Methodology - New Modelling Approaches