SPIRIT: A seamless phase I/II randomized design for immunotherapy trials

Area of Science:

• Oncology
• Clinical Trial Design
• Immunotherapy

Background:

• Immunotherapy harnesses the immune system to combat cancer, necessitating innovative clinical trial designs.
• Current trial designs may not optimally capture immunotherapy's unique mechanisms.
• Identifying the optimal biological dose (OBD) is crucial for effective immunotherapy treatment.

Purpose of the Study:

• To propose a Bayesian seamless phase I/II randomized design for immunotherapy trials, named SPIRIT.
• To define the OBD based on restricted mean survival time (RMST).
• To integrate progression-free survival (PFS) and immune response for dose optimization and efficiency.

Main Methods:

• The SPIRIT design features two seamlessly integrated stages: dose-finding and dose-confirmation.
• Stage 1 identifies safe doses through toxicity monitoring.
• Stage 2 adaptively randomizes patients to safe doses, jointly modeling PFS and immune response to determine the OBD.

Main Results:

• Simulation studies demonstrate that SPIRIT possesses desirable operating characteristics.
• The proposed SPIRIT design outperforms conventional clinical trial designs in identifying the OBD.
• SPIRIT effectively utilizes PFS as the primary endpoint and immune response as an ancillary endpoint.

Conclusions:

• The SPIRIT design offers an efficient and robust approach for optimizing immunotherapy clinical trials.
• This novel design facilitates the identification of the optimal biological dose for cancer immunotherapy.
• SPIRIT enhances clinical trial efficiency by integrating multiple endpoints and adaptive randomization.