Advancing Platform Trials in Early Oncology by Using MATS and EXNEX in Randomized Controlled Trials.

Area of Science:

• Clinical Trial Design
• Oncology Drug Development
• Biostatistics

Background:

• Randomized controlled trials (RCTs) are standard but challenging in early-phase oncology, especially with small, heterogeneous patient populations.
• Recruitment difficulties, timeline pressures, and evolving prognoses complicate traditional RCTs for rare cancers.
• FDA's Project Optimus necessitates earlier dose optimization, which may vary across patient groups and treatments.

Purpose of the Study:

• To develop an innovative platform trial design for proof-of-concept (POC) decision-making in early oncology.
• To integrate dose optimization strategies aligned with FDA's Project Optimus within an RCT framework.
• To enhance statistical efficiency in small, similar patient populations through advanced statistical methods.

Main Methods:

• Adapted the Multi-Arm Two-Stage (MATS) design to include randomized control arms for robust inference.
• Incorporated a Stage 1 futility stopping rule before proceeding to Stage 2 dose optimization.
• Utilized the exchangeability-nonexchangeability (EXNEX) framework for control borrowing across cohorts to improve statistical efficiency.

Main Results:

• The proposed platform trial design supports robust POC decisions within an RCT setting.
• The design facilitates efficient dose optimization, addressing FDA's Project Optimus requirements.
• Control borrowing via the EXNEX framework significantly enhances statistical efficiency in small populations.

Conclusions:

• This novel platform trial design effectively addresses challenges in early oncology development for small patient populations.
• The integrated approach promotes efficient drug development by combining RCT rigor, early dose optimization, and statistical efficiency.
• The methodology offers a practical solution for advancing oncology therapeutics in rare and similar patient groups.