Quantitative decision-making in randomized Phase II studies with a time-to-event endpoint.

Area of Science:

• Clinical trial design
• Biostatistics
• Drug development

Background:

• Traditional Go/No-Go decisions after proof-of-concept studies are limited by small, isolated trials.
• Hypothesis testing in traditional methods struggles to fully assess efficacy evidence strength.

Purpose of the Study:

• To propose a quantitative Bayesian/frequentist decision framework for Go/No-Go criteria and sample size evaluation in Phase II randomized trials.
• To develop an integrated quantitative approach for clinical development programs considering Phase II and III trials with common time-to-event endpoints.

Main Methods:

• Developed a quantitative Bayesian/frequentist decision framework for Go/No-Go criteria.
• Proposed an integrated approach for Phase II and III trials, allowing a discount of Phase II data.
• Utilized a time-to-event endpoint for analysis.

Main Results:

• The proposed framework quantitatively integrates treatment effect uncertainty for decision-making.
• Results confirm that increasing Phase II sample size yields a greater increase in Phase III probability of success compared to equal increases in Phase III sample size.
• Demonstrated quantitative decision-making with a real-world example in metastatic pancreatic cancer.

Conclusions:

• The quantitative framework offers improved Go/No-Go decision-making and sample size optimization in early-phase clinical trials.
• Strategic increases in Phase II sample size are more impactful for overall program success than solely increasing Phase III sample size.
• The approach provides a robust method for evaluating clinical development programs with shared endpoints.