Adaptive Dose Finding Under Model Uncertainty Using MCP-Mod

• Activity Number: 30
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Tobias Mielke*+
• Companies: Aptiv Solutions
• Keywords: Dose Finding ; MCP-Mod ; Adaptive designs ; Optimal Designs

Abstract

The determination of the optimal dose and the characterization of the dose-response relation are key objectives in drug development. Uncertainties on the underlying dose-response relation may result in inefficient and unsuccessful studies. Adaptive design and analysis methods may take these uncertainties into account in the planning stage. The MCP-Mod approach is an adaptive analysis method targeting the uncertainty on the dose-response model. A candidate set of dose response shapes is used to define a set of optimized trend tests. Based on the results of these trend tests, a subset of the candidate shapes is selected and fit to the data. The EMA draft qualification opinion on MCP-Mod states this approach as an efficient methodology for the design and analysis of dose-finding studies under model uncertainty. Adaptive designs additionally integrate updated information on the dose-response into optimized randomizations to dose-groups. After a short introduction on the MCP-Mod methodology, different adaptive dose-finding approaches will be presented. Limitations and benefits of standard and innovative dose-finding approaches will be discussed based on a practical simulation example.