Modeling of tumor size reduction patterns in advanced melanoma under treatment with MK-3475, a potent antibody against PD-1
Jeroen Elassaiss-Schaap (1), Andreas Lindauer (1), Alexandre Sostelly (1)#, Malidi Ahamadi (2), Kevin Gergich (3), Peter Kang (3), Dinesh de Alwis (3) and Rik de Greef (1)
(1) QP2, MSD, Oss, Netherlands; #currently employed at Merck Serono, Germany; (2) QP2, Merck, Whitehouse Station, NJ USA; (3) Clinical Oncology, Merck, Whitehouse Station, NJ USA
Objectives: Limited information is available on the dynamics of tumor size (TS) in advanced melanoma. In the current analysis a population model was developed based on TS data from patients treated with MK-3475, a potent immune modulating antibody against PD1, to characterize the patterns of TS dynamics and to investigate the influence of covariates.
Methods: In a phase I clinical trial, 411 patients with advanced melanoma received MK-3475 at different dose levels, see also . TS was quantifiable as the sum of the longest dimensions of all target lesions by RECIST 1.1 assessment in 364 patients. A mixed effects model describing TS over time was developed on the basis of the model by Claret et al.  in Nonmem7.1 . Model performance was assessed by bootstraps and VPCs using PsN .
Results: TS patterns over time after dosing of MK-3475 appeared to differentiate from available literature [2, 5]; especially little to no tumor regrowth was present. The Claret model was modified to accommodate this and other features of the data. Specifically, heterogeneities in the data reduced the adequacy of standard variability components as was particularly visible in VPCs. Four patterns of tumor growth and shrinkage were discernable in the data and characterized using four mixture groups in the $MIX functionality of NONMEM. Particular adjustments were made to also include patients without post-baseline measurements. Subsequent stepwise covariate modeling identified the baseline number of target lesions, number of lymph node lesions and ECOG scores as significant and influential covariates. Baseline TS was also an important correlate of tumor response and was incorporated into the model through the mixture component. When accounting for these covariates, pretreatment with ipilimumab was not a significant predictor of tumor size change despite a higher disease burden in these patients.
Conclusions: A mixed effects model capturing tumor size reduction in melanoma patients dosed with MK-3475 was successfully developed. In a new approach, patterns in tumor size reduction under treatment with this immune modulator were characterized using mixture components. Baseline tumor size, number of target lesions and number of lymph node lesions were identified as important covariates in the response to treatment.
 Topalian SL, Hodi FS, Brahmer JR, Gettinger SN, Smith DC,`McDermott DF, et al. Safety, activity, and immune correlates of Anti-PD-1 antibody in cancer. N Engl J Med 2012;366(26):2517-9.
 Claret L, Girard P, Hoff PM, Van Cutsem E, Zuideveld KP, Jorga K, et al. Model-based prediction of phase III overall survival in colorectal cancer on the basis of phase II tumor dynamics. J Clin Oncol 2009;27(25):4103-8.
 Beal SL, Sheiner LB, Boeckmann AJ & Bauer RJ (Eds.) NONMEM Users Guides. 1989-2011. Icon Development Solutions, Ellicott City, Maryland, USA.
 Pearl-speaks-NONMEM (psn.sourceforge.net)
 Chiu J, Ouellet D. Exposure-response analysis of the effect of trametinib, a MEK inhibitor, on tumor size in patients with V600 BRAF mutation positive melanoma. American Conference on Pharmacometrics; 2013 May 12-15. Fort Lauderdale, FL, 2013.