Toxicity burden score: a novel approach to summarize multiple toxic effects

Area of Science:

• Pharmacology and Toxicology
• Clinical Trial Design
• Biostatistics

Background:

• Current cancer trial toxicity assessment relies on dose-limiting toxicity (DLT), which oversimplifies complex safety data.
• DLT fails to capture the nuances of multiple, lower-grade toxic effects or differentiate toxicity types and severity.
• This limitation hinders comprehensive safety evaluation during drug development.

Purpose of the Study:

• To introduce and validate a novel Toxicity Burden Score (TBS) for summarizing cancer trial toxicity.
• To demonstrate how TBS can provide a more detailed and accurate representation of patient safety compared to traditional DLT.
• To explore the application of TBS in improving dose-finding trial designs.

Main Methods:

• Developed a weighted sum approach to calculate the Toxicity Burden Score (TBS).
• Estimated severity weights for different toxicity grades and types using regression analysis on historical data.
• Validated the TBS method using historical bortezomib trial data and simulated dose-finding scenarios.

Main Results:

• Derived specific weights for various toxicities, including hematologic (platelets) and neurologic (neuropathy), and non-hematologic effects.
• Demonstrated that defining dose-limiting toxicity (DLT) based on TBS significantly reduced the probability of selecting doses exceeding the maximum tolerated dose in simulated trials.
• The TBS effectively incorporates information on the grade and type of multiple toxicities.

Conclusions:

• The Toxicity Burden Score (TBS) presents a feasible and informative method for summarizing complex toxicity data in cancer clinical trials.
• TBS offers advantages over traditional DLT by integrating information on the severity and type of multiple adverse events.
• Further validation in large prospective studies is recommended before widespread clinical application.