Yan Feng (1), Xiaoning Wang (1), Dave Clawson (1), Satyendra Suryawanshi (1), Manish Gupta (1), and Amit Roy (1)
(1) Clinical Pharmacology and Pharmacometrics, Bristol-Myers Squibb, Princeton, NJ
Introduction: Ipilimumab (IPI) was the first immune-checkpoint inhibitor (ICE) to demonstrate an improvement in overall survival (OS) of advanced melanoma (AdvMel) patients [1], and the approved dosing regimen for AdvMel in the US and EU is 4 doses of 3 mg/kg, given once every 3-weeks (Q3W). The benefit-risk of 10 mg/kg vs 3 mg/kg IPI in AdvMel was evaluated in a post-approval randomized phase 3 study (CA184169) [2], as objective tumor response was higher with 10 mg/kg in a phase 2 dose-ranging study [3]. The overall survival (OS) of AdvMel patients in CA184169 treated with IPI 10 mg/kg Q3W (4 doses) was significantly better than that of the 3 mg/kg dose. However, the objective response rate (ORR) by RECIST criteria, duration of response, and progression free survival (PFS) were similar in both dose arms. Characterization of the tumor burden (TB) time-profile may enable identification of one or more features of tumor response that are associated with OS following treatment with ICE agents.
Objectives:
- Characterize the TB time-profile of patients treated with ipilimumab treatment, by a model-based description of longitudinal tumor burden data
- Utilize the model to assess differences in patterns of TB time-profiles
Methods: The TB time-profiles of AdvMel patients in CA184169 who received ipilimumab, and for whom tumor burden data were available (N=345 for 3 mg/kg; N=344 for 10 mg/kg) were described by a nonlinear mixed-effects TGD model. The sum of the longest diameters of target lesions was used as a surrogate for TB. The TB at a given time (t) =baseline tumor burden, first-order shrinkage rate, linear growth rate, and the tumor burden at steady-state, respectively. A unimodal model and mixture models with 2 or 3 subpopulations were evaluated. TBSS was only defined for mixture models with a no growth subpopulation (TG fixed to 0). Bayesian information criterion (BIC) was used to guide model selection.
Results: The TGD model with 3 subpopulations provided an adequate description of the observed data, and had the lowest BIC value among tested unimodal and mixture models. The 3 (no growth, intermediate, and fast tumor growth) subpopulations determined by the mixture model identified subjects with qualitatively different tumor growth dynamics, as evidenced by differences in the distributions of TG and TS values. TB0 was higher in subjects in fast tumor growth subpopulation than the other two subpopulations. The TS was similar in both dose arms, but the TG of patients in the 10 mg/kg arm was ~30% slower than that of patients in the 3mg/kg arm (0.020 vs. 0.029 cm/week). The fractions of patients in the 3 and 10 mg/kg dose arms who were categorized into the no-growth subpopulation (representing durable response) were approximately 15.1% and 21.7%, respectively.
Conclusions: The TGD profiles of AdvMel patients treated with IPI were adequately described by a mixture TGD model, with 3-subpopulations (fast, intermediate, and no-growth). Even though PFSwas similar between 3 and 10 mg/kg, the proportion of subjects achieving durable responses were higher, and tumor progression rate was lower in patients who received 10 mg/kg relative to those received 3 mg/kg.
References:
[1]. Hodi FS, O’Day SJ, McDermott DF, et al. Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med. 2010;363(8):711-23.
[2]. Ascierto PA, Vecchio M Del, Robert C, et al. Ipilimumab 10 mg / kg versus ipilimumab 3 mg / kg in patients with unresectable or metastatic melanoma : a randomised , double-blind , multicentre , phase 3 trial. Lancet Oncol. 2017;18:611-622.
[3]. Wolchok JD, Neyns B, Linette G, et al. Ipilimumab monotherapy in patients with pretreated advanced melanoma: a randomised, double-blind, multicentre, phase 2, dose-ranging study. Lancet Oncol. 2010;11(2):155-64.
Reference: PAGE 27 (2018) Abstr 8687 [www.page-meeting.org/?abstract=8687]
Poster: Drug/Disease Modelling - Oncology