INHANCE: An Adaptive Confirmatory Study with Dose Selection at Interim

Abstract

Adaptive designs use accumulating data to modify aspects of the study without undermining its validity and integrity. There are key practical and statistical issues that must be considered when planning an adaptive design, for example, controlling the Type I error rate, prespecifying the adaptation criteria, and setting up well-organised procedures to maintain data confidentiality at interim analyses. The number of adaptations should be limited, preferably to just one in the confirmatory setting. There are several possible types of adaptation; this chapter focuses on a pivotal confirmatory two-stage adaptive design with dose selection at interim, illustrated by the INHANCE study. This comprised a dose-finding stage with dose selection after 14 days of treatment, and a second stage evaluating efficacy and safety during 26 weeks of treatment. An independent data monitoring committee (DMC) selected two indacaterol doses (out of four) based on predefined decision rules and review of unblinded results from an interim analysis. The decision rules were based on trough (24 h post-dose) and early (1–4 h post-dose) bronchodilator effect after 14 days, and safety data. Selected doses were continued into the second, 26 weeks, stage. The INHANCE study provides a successful example of the use of an adaptive design in the confirmatory setting.