Design and statistical principles of the SHIVA trial

Area of Science:

• Oncology
• Genomics
• Clinical Trial Design

Background:

• Molecularly targeted agents (MTAs) efficacy is linked to specific tumor molecular alterations.
• Current cancer drug development primarily relies on tumor type, not molecular profile.
• The SHIVA trial investigated alternative treatment selection strategies for refractory cancers.

Purpose of the Study:

• To evaluate whether tumor biology or tumor location/histology is superior for selecting MTAs in refractory cancer patients.
• To review the statistical principles, strengths, and limitations of the SHIVA randomized trial.
• To analyze an omic-based algorithm for guiding targeted agent selection.

Main Methods:

• The SHIVA trial is a randomized study comparing treatment selection based on tumor biology (omic-based algorithm) versus standard histology/location.
• Statistical principles for trial design and algorithm development are reviewed.
• The impact of potential algorithm misspecification on treatment selection was quantified.

Main Results:

• The SHIVA trial aimed to determine the optimal approach for selecting molecularly targeted agents.
• The study emphasized the need for standardized, transparent methods in developing treatment selection algorithms.
• Quantification of the impact of partially incorrect algorithms was performed.

Conclusions:

• The SHIVA trial provides insights into optimizing treatment selection for cancer patients using molecular profiling.
• Standardization and clear methodology are crucial for omic-based treatment algorithms.
• Understanding algorithm limitations is key to their effective clinical implementation.