BOIN

BOIN optimizes dose-finding in early-phase clinical trials to identify the maximum tolerated dose (MTD) using Bayesian optimal interval decision rules.

Key Features:

• Versatility: Applicable to a wide range of early-phase clinical trials, including single-agent and drug-combination studies, and scenarios with late-onset toxicities.
• MTD and OBD determination: Identifies the maximum tolerated dose (MTD) and can determine an optimal biological dose (OBD) by considering both toxicity and efficacy.
• Interval-based Bayesian decision rules: Uses pre-specified optimal toxicity intervals to guide dose escalation and de-escalation decisions.
• Statistical operating characteristics: Provides robust and powerful operating characteristics for decision-making in dose-finding trials.
• Data-driven decisions: Bases escalation and de-escalation on observed toxicity outcomes within the Bayesian optimal-interval framework.

Scientific Applications:

• Early-phase oncology trials: Optimizes dose selection in phase I and phase I/II studies for anticancer agents.
• Combination therapy dose-finding: Supports identification of safe dose combinations for multi-agent regimens.
• Trials with late-onset toxicities: Adapts dose-finding decisions to settings where toxicities may present later than typical observation windows.
• Optimal biological dose selection: Balances toxicity and efficacy considerations to inform selection of biologically optimal doses.
• Primary brain tumor studies: Facilitates MTD/OBD identification in challenging indications such as primary brain tumors.

Methodology:

BOIN applies Bayesian principles to construct optimal toxicity intervals and uses interval-based decision rules for dose escalation and de-escalation based on observed toxicity outcomes.