Statistical design of noninferiority multiple region clinical trials to assess global and consistent treatment effects

J Biopharm Stat. 2017;27(6):933-944. doi: 10.1080/10543406.2017.1293075. Epub 2017 Mar 15.

Abstract

Noninferiority multiregional clinical trials (MRCTs) have recently received increasing attention in drug development. While a major goal in an MRCT is to estimate the global treatment effect, it is also important to assess the consistency of treatment effects across multiple regions. In this paper, we propose an intuitive definition of consistency of noninferior treatment effects across regions under the random-effects modeling framework. Specifically, we quantify the consistency of treatment effects by the percentage of regions that meet a predefined treatment margin. This new approach enables us to achieve both goals in one modeling framework. We propose to use a signed likelihood ratio test for testing the global treatment effect and the consistency of noninferior treatment effects. In addition, we provide guidelines for the allocation rule to achieve optimal power for testing consistency among multiple regions. Extensive simulation studies are conducted to examine the performance of the proposed methodology. An application to a real data example is provided.

Keywords: Consistency of treatment effects; global treatment effect; multiregional clinical trial; noninferiority clinical trial; random effects model; signed likelihood ratio test.

MeSH terms

  • Drug Approval / methods
  • Drug Approval / statistics & numerical data*
  • Global Health / statistics & numerical data*
  • Humans
  • Models, Statistical*
  • Multicenter Studies as Topic / methods
  • Multicenter Studies as Topic / statistics & numerical data*
  • Randomized Controlled Trials as Topic / methods
  • Randomized Controlled Trials as Topic / statistics & numerical data*
  • Sample Size
  • Treatment Outcome