Assurance methods for designing a clinical trial with a delayed treatment effect

Area of Science:

• Biostatistics
• Clinical Trial Design
• Immuno-oncology

Background:

• Assurance calculations offer a Bayesian alternative to traditional power calculations for clinical trial planning.
• Immuno-oncology drug development often features delayed treatment effects, where survival curves diverge later.
• Ignoring parameter uncertainty in assurance calculations can lead to insufficient sample sizes and increased trial failure risk.

Purpose of the Study:

• To present a novel elicitation technique for estimating parameters in trials with delayed treatment effects.
• To demonstrate how to compute assurance using elicited prior distributions.
• To improve the success rate and efficiency of Phase III immuno-oncology trials.

Main Methods:

• Developed a new elicitation technique for delayed treatment effects.
• Computed assurance using elicited prior distributions.
• Illustrated the methodology with a practical example and developed open-source software.

Main Results:

• The proposed method accounts for parameter uncertainty crucial for delayed treatment effects.
• Demonstrated practical application and provided open-source software for assurance calculation.
• The methodology can enhance statistical power and reduce failure risk in relevant trials.

Conclusions:

• The new assurance calculation methodology is suitable for immuno-oncology trials with delayed treatment effects.
• This approach helps ensure adequate sample sizes and increases the likelihood of detecting clinically relevant treatment effects.
• The methods have the potential to optimize Phase III trial efficiency and success rates.