|
Activity Number:
|
123
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 3, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biopharmaceutical Section
|
| Abstract - #305047 |
|
Title:
|
A Hierarchical Bayesian Approach for Controlled Phase II Cancer Trials in Heterogeneous Diseases
|
|
Author(s):
|
Lanjia Lin*+ and Jyotirmoy Dey and Lumay Chiang
|
|
Companies:
|
Florida State University and Novartis Pharmaceuticals and Novartis Oncology
|
|
Address:
|
, , ,
|
|
Keywords:
|
Hierarchical Bayesian ; Randomized phase II ; Controlled experiment ; Heterogeneous cancer
|
|
Abstract:
|
Subtypes of heterogeneous cancers often respond variably to the same therapy. An adaptive hierarchical Bayesian approach for conducting uncontrolled phase II trials in such cancers was proposed by Thall et al. (2003) where treatment effects are allowed to vary across subtypes while assuming a priori that they are exchangeable and correlated. We explore an extension of the Bayesian logistic set-up which allows conducting a controlled experiment where the primary focus is on the efficacy of an experimental therapy relative to a control in each subtype. The extended model allows the control effects for the various subtypes to be different and independent, with the hierarchy set on the relative treatment benefit of the experimental therapy across subtypes. Simulation results show gain is efficiency as the model borrows strength across subtypes where the treatment effect is congruous.
|
