|
Activity Number:
|
123
- Binary and Ordinal Outcome Regression
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 30, 2018 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Biometrics Section
|
|
Abstract #330897
|
Presentation
|
|
Title:
|
A Bayesian Logistic Model with Covariate to Identify Optimal Dose for Heterogeneous Population in Phase I Oncology Trial
|
|
Author(s):
|
Xin Wei* and Michael Branson
|
|
Companies:
|
Celgene Corporation and celgene corporation
|
|
Keywords:
|
Oncology Clinical Trial;
dose finding;
MTD;
Bayesian Logistic Regression;
adaptive design;
CRM
|
|
Abstract:
|
Phase I oncology FIH trial aims at identifying the maximum tolerated dose (MTD) for patients that can be used in later trial. Previous work developed a Bayesian logistic model based continuous reassessment method (BLRM) that improves the precision of MTD identification. Our work demonstrates the further improvement of optimal dose finding for a heterogeneous population by modeling the patient risk factor as covariate. In addition, we use simulation to show that the non-parsimonious model does not compromise the ability of model to pick the MTD and other operating characteristics in the context of 3-parameter logistic regression for dose/toxicity relationship.
|
Authors who are presenting talks have a * after their name.
