Design of paediatric trials with benefit-risk endpoints using a composite score of adverse events of interest (AEI) and win-statistics

ABSTRACT

A fundamental problem in the regulatory evaluation of a therapy is assessing whether the benefit outweighs the associated risks. This work proposes designing a trial that assesses a composite endpoint consisting of benefit and risk, hence, making the core of the design of the study, to assess benefit and risk. The proposed benefit risk measure consists of efficacy measure(s) and a risk measure that is based on a composite score obtained from pre-defined adverse events of interest (AEI). This composite score incorporates full aspects of adverse events of interest (i.e. the incidence, severity, and duration of the events). We call this newly proposed score the AEI composite score. After specifying the priorities between the components of the composite endpoint, a win-statistic (i.e. win ratio, win odds, or net benefit) is used to assess the difference between treatments in this composite endpoint. The power and sample size requirements of such a trial design are explored via simulation. Finally, using Dupixent published adult study results, we show how we can design a paediatric trial where the primary outcome is a composite of prioritized outcomes consisting of efficacy endpoints and the AEI composite score endpoint. The resulting trial design can potentially substantially reduce sample size compared to a trial designed to assess the co-primary efficacy endpoints, therefore it may address the challenge of slow enrollment and patient availability for paediatric studies.

KEYWORDS:

• Benefit-risk assessment
• AEI composite score
• win-statistics

Acknowledgment

The authors would like to thank Dr Jeff Maca for his insights on the framework of incorporating severity and duration into the assessment of adverse events.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.