Optimal biological dose selection in dose-finding trials with model-assisted designs based on efficacy and toxicity:...

Area of Science:

• Oncology
• Clinical Trial Design
• Pharmacology

Background:

• The optimal biological dose (OBD) concept is crucial for phase I oncology trials, balancing anti-cancer efficacy and toxicity.
• Model-assisted designs are increasingly used for dose-escalation, incorporating both toxicity and efficacy data.
• Numerous methods exist for OBD selection, but their comparative performance remains unclear.

Purpose of the Study:

• To conduct a comprehensive simulation study evaluating the operating characteristics of different optimal biological dose selection approaches.
• To compare the performance of various methods for estimating efficacy probability in OBD selection.
• To provide guidance on selecting appropriate OBD selection measures based on dose-escalation procedures.

Main Methods:

• A comprehensive simulation study was performed to assess the performance of various optimal biological dose (OBD) selection approaches.
• Operating characteristics of different utility functions for toxicity-efficacy trade-offs were analyzed.
• The impact of different dose-escalation procedures on OBD selection was investigated.

Main Results:

• The choice of the optimal biological dose (OBD) selection measure is influenced by the specific dose-escalation procedure employed.
• Simulation results highlighted key features of utility functions used to balance toxicity and efficacy.
• Modeling efficacy probability may offer limited advantages in optimal biological dose selection.

Conclusions:

• Practitioners must carefully consider the interplay between dose-escalation methods and OBD selection measures.
• The optimal biological dose (OBD) selection strategy should be tailored to the specific clinical trial design.
• Further research may be needed to refine efficacy modeling for improved optimal biological dose determination.