Dose Finding in Oncology Trials Guided by Ordinal Toxicity Grades Using Continuous Dose Levels

Area of Science:

• Oncology
• Clinical Trial Design
• Biostatistics

Background:

• Traditional dose-finding trials often rely solely on binary dose-limiting toxicity (DLT) data.
• This binary approach may not fully capture the spectrum of adverse events, potentially impacting patient safety during early-phase drug development.

Purpose of the Study:

• To present a novel Bayesian adaptive design for dose-finding in early-phase oncology trials.
• To enhance patient safety by incorporating intermediate toxicity grades (Grade 2) alongside DLTs.
• To improve the efficiency of estimating the maximum tolerated dose (MTD).

Main Methods:

• The design employs an escalation with overdose control principle.
• It utilizes a proportional odds model to characterize the dose-toxicity relationship.
• Intermediate Grade 2 toxicity data is integrated to guide dose adjustments, ensuring more cautious escalation.

Main Results:

• The Bayesian adaptive design demonstrated safety and acceptable efficiency in estimating the MTD.
• Operating characteristics were evaluated under various scenarios for true MTD and Grade 2 toxicity rates.
• Simulations confirmed the design's robustness for a planned sample size of twenty patients.

Conclusions:

• The proposed Bayesian adaptive design offers a safer and more refined approach to dose escalation in oncology.
• Integrating intermediate toxicity data improves ethical considerations in first-in-human trials.
• The design provides efficient MTD estimation, crucial for advancing novel cancer therapeutics.