|
Activity Number:
|
145
- Leveraging External Data in Clinical Trials
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 8, 2022 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Biopharmaceutical Section
|
|
Abstract #323139
|
|
|
Title:
|
BOREC: A Bayesian Optimal Design for Randomized Dose Expansion Cohorts in Oncology Trials
|
|
Author(s):
|
Jinjie Chen* and Rong Liu and Ruitao Lin
|
|
Companies:
|
BMS and BMS and MD Anderson
|
|
Keywords:
|
Bayesian ;
Dose expansion;
Oncology;
Monotherapy;
Combination therapy
|
|
Abstract:
|
Nowadays, as dose expansion cohorts are getting increasing popularity in practice because of their attractive and versatile utilities in further assessing the safety and efficacy, especially in combination therapy development. We propose a novel Bayesian method to design the combo-versus-mono expansion phase where multiple novel combinations are compared with the backbone monotherapy. The monotherapy can be an ongoing agent under investigation or an already approved agent. We incorporates the information from the dose-escalation phase through the use of robust meta-analytic prior. In the combo-versus-mono phase, we use dual-criterion Bayesian posterior probabilities to determine the go/no-go decisions for each combination arm. The cutoff values of the posterior probabilities are optimized such that the resulting design controls both the agent-level and family-level false-positive rates and maximizes the true positive rate. The design can be further optimized by minimizing the average expected sample size such that the design has savings in the sample size. The proposed design is evaluated by extensive simulation studies and exemplified by a planning trial at BMS.
|
Authors who are presenting talks have a * after their name.
