|
Activity Number:
|
488
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Government Statistics
|
| Abstract - #307856 |
|
Title:
|
Adaptive Dose-Finding Designs for Single and Dual Agent Combinations Using Accumulated Data from All Patients
|
|
Author(s):
|
Sumithra J. Mandrekar and Daniel J. Sargent*+
|
|
Companies:
|
Mayo Clinic and Mayo Clinic
|
|
Address:
|
200 1st St SW, Rochester, MN, 55901,
|
|
Keywords:
|
Phase I ; Trinomial Outcome ; CR Model ; Efficacy ; Toxicity ; CRM
|
|
Abstract:
|
Historically, designs for dose seeking trials have been geared towards finding the maximum tolerated dose, with safety as the primary outcome. Agents with specific biologic targets with unknown dose efficacy curves and limited toxicity profile require novel designs. We present designs for identifying an optimal dose or combination region utilizing both toxicity and efficacy data. A continual reassessment method with straightforward dose selection criteria using accumulated data from all patients treated thus far is employed based on the flexible continuation-ratio model. A generalization of the single agent model allowing for separate toxicity and efficacy curves for each agent to generate a dose outcome surface for the two agent combination is used. Our simulation studies demonstrated that the proposed designs have favorable operating characteristics under a variety of scenarios.
|
