Online Program Home
My Program

Abstract Details

Activity Number: 260 - SPEED: Topics in Bayesian Analysis
Type: Contributed
Date/Time: Monday, July 30, 2018 : 3:05 PM to 3:50 PM
Sponsor: Section on Bayesian Statistical Science
Abstract #332888
Title: Dose Finding Methods Using Two Endpoints in Early Clinical Studies
Author(s): Kyounghwa Bae* and Zhentao Tong and Vladimir Dragalin
Companies: Janssen Research & Development, LLC and North Carolina State University and Janssen R&D
Keywords: Dose Finding; Phase I study; Bayesian; Optimal biological dose

For the new molecularly targeted agents and immune therapies, the goal of a dose finding study is to identify the optimal biological dose (OBD), defined as the lowest dose having the highest probability of efficacy, while also safeguarding patients against excessive dose-limiting toxicity. Therefore, the maximum tolerated dose may not be the most important target in such studies and alternative dose escalation procedures are needed that take into consideration both efficacy and toxicity endpoints. The most well-known methods, such as EffTox(Thall 2004) and bCRM(Braun 2002) utilize an association between the two endpoints but the results are sensitive to the specification of the association. However, it is difficult in practice to estimate the association between the efficacy and toxicity from pre-clinical data, especially in a first-in-human study. We developed a method that combines the probability of toxicity and the probability of efficacy(or surrogate) calculated using two independent models to select safe and efficacious doses for the next cohort in the dose escalation and finally to identify the OBD.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program