Abstract
AbstractThe Multiple Comparison Procedure and Modelling (MCPMod) approach has been shown to be a powerful statistical tool that can significantly improve the design and analysis of dose finding studies under model uncertainty. Due to its frequentist nature, however, it is difficult to incorporate into MCPMod information from historical trials on the same drug. Recently, a Bayesian version of MCPMod has been introduced by Fleischer et al. (2022) to resolve this issue, which is particularly tailored to the situation where there is information about the placebo dose group from historical trials. In practice, information may also be available on active dose groups from early phase trials, e.g., a dose escalation trial and a preceding small Proof of Concept trial with only a placebo and a high dose. To address this issue, we introduce a Bayesian hierarchical framework capable of incorporating historical information about both placebo and active dose groups with the flexibility of modelling both prognostic and predictive between-trial heterogeneity. Our method is particularly useful in the situation where the effect sizes of two trials are different. Our goal is to reduce the necessary sample size in the dose finding trial while maintaining its target power.
Publisher
Cold Spring Harbor Laboratory
Reference17 articles.
1. Bayesian clinical trials
2. The use of the area under the ROC curve in the evaluation of machine learning algorithms;Pattern Recognition Letters,1997
3. Combining Multiple Comparisons and Modeling Techniques in Dose-Response Studies
4. European Medicines Agency (EMA) (2014). Qualification Opinion of MCP-Mod as an efficient statistical methodology for model-based design and analysis of Phase II dose finding studies under model uncertainty. https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/qualification-opinion-mcp-mod-efficient-statistical-methodology-model-based-designen.pdf.
5. An introduction to ROC analysis;Pattern Recognition Letters,2006