Tomoko Freshwater (1), Pravin R Jadhav (1), Lee Hodge (2), William D. Hanley (1), Bharath Kumar (1)
(1) Merck & Co., Inc., Kenilworth, NJ, USA (2) qPharmetra LLC, Andover, Massachusetts, USA
Objectives: 1) To assess a competitive landscape for MRL-1 by comparing with market leaders and key competitors (disease-modifying antirheumatic drugs [DMARDs] and anti-tumour necrosis factor [TNF]) and define relative advantages of MRL-1. 2) To compare different study designs and optimize sample size through a simulation-based approach to provide confidence that the Phase IIb design offers high probability of demonstrating a robust dose-response relationship.
Methods: 1) benchmark efficacy database for the MRL-1 and competitor was created through literature on compounds approved or evaluated for treatment of rheumatoid arthritis. The database included longitudinal summary-level information for efficacy endpoints at different dose levels and patient characteristics (covariates) for 31 drugs (12 used) tested in 88 controlled clinical trials (85 were used) in more than 200,000 patients. The model-based meta-analyses including time course observations for efficacy endpoints, ACR20 and other key endpoints, were conducted using NONMEM and R to characterize the probability that MRL-1 would be greater, similar or worse on efficacy endpoints given uncertainty in the literature by establishing the time course of efficacy endpoints which enabled comparison of onset of action relative to competitors. 2) To optimize Phase IIb design, the evaluation of different sample size and dose options was conducted by integrating the MRL-1 model with key trial design attributes in a clinical trial simulation (CTS).
Results: (1) Based on the competitive landscape modeling, there was a 50 to 70% probability that MRL-1 would be similar or better on the ACR20 efficacy end point vs. all competitors. (2) The exposure-efficacy relationship allowed exploration of alternative dosing regimens and helped us to better characterize different dosing (QD/BID) regimens. The simulations allowed us to define the most efficient sample size and dose arms. CTS was used by MRL-1’s project team to guide its design of the next trial.
Conclusions: A comparator model supported the decision to continue with clinical development of MRL-1 based on the efficacy comparison. Further, the modeling efforts were pivotal to select Phase IIb doses and the sample size needed to fully characterize the exposure-response for efficacy parameters.
Reference: PAGE 23 () Abstr 3253 [www.page-meeting.org/?abstract=3253]
Poster: Methodology - Other topics