qTPI: A quasi-toxicity probability interval design for phase I trials with multiple-grade toxicities

Area of Science:

• Clinical trial design
• Biostatistics
• Pharmacology

Background:

• The National Cancer Institute's Common Terminology Criteria for Adverse Events has advanced drug development.
• Phase I trials increasingly collect multiple-grade toxicity data, necessitating advanced statistical designs.
• Existing designs often lack transparency and appropriate handling of nuanced toxicity data.

Purpose of the Study:

• To propose a novel, transparent, and statistically sound Phase I clinical trial design for multiple-grade toxicities.
• To introduce the quasi-toxicity probability interval (qTPI) design, integrating a quasi-continuous toxicity measure.
• To enhance the safety, accuracy, and reliability of dose selection in early-phase drug development.

Main Methods:

• Developed the quasi-toxicity probability interval (qTPI) design within a Bayesian framework.
• Incorporated a quasi-continuous toxicity probability (qTP) measure, mapping multi-grade toxicities via a severity weight matrix.
• Continuously updated the dose-toxicity curve using accumulating patient data.

Main Results:

• Numerical simulations demonstrated that qTPI designs offer superior safety, accuracy, and reliability compared to binary toxicity designs.
• The qTPI design simplifies parameter elicitation, avoiding the need for multiple hypothetical cohorts.
• Illustrative application to a soft tissue sarcoma trial showed effective patient-by-patient dose allocation.

Conclusions:

• The qTPI design provides a robust and efficient statistical framework for Phase I trials with multiple-grade toxicities.
• This approach enhances the precision and safety of dose-finding studies in oncology.
• qTPI represents a significant advancement in clinical trial methodology for evaluating drug safety and efficacy.