|
Activity Number:
|
38
- Study Design: Historical Controls, Endpoint Timing, and Regulatory Considerations
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, July 29, 2018 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Medical Devices and Diagnostics
|
|
Abstract #330045
|
Presentation
|
|
Title:
|
Strategies for Validating Biomarkers Using Data from a Reference Set
|
|
Author(s):
|
Lu Wang* and Ying Huang
|
|
Companies:
|
Fred Hutchinson Cancer Research Center and Fred Hutchinson Cancer Research Center
|
|
Keywords:
|
Biomarker validation;
Group sequential methods
|
|
Abstract:
|
Biomarkers identified in discovery process need to be further validated. The evaluation of diagnostic biomarker, which is based on biobanks of stored specimens, is often expensive and time-consuming. Group sequential designs that allow early termination for futility and/or efficacy have been proposed to conserve resources and minimize cost. In this paper, we propose a rotation testing procedure based on the group sequential design for validating performance of candidate biomarkers. Given the amount of specimens from each individual is pre-determined, we develop analytical formulas for the expected numbers of biomarkers evaluated and discovered to have adequate performance using the proposed strategy. The performance of our proposed method for validating biomarkers with respect to the points on the receiver operating characteristic (ROC) curve are evaluated via extensive simulation studies. Different types of stopping boundaries and rules of sequential study designs are considered. Compared with the fixed-sample design, our proposed method is illustrated to discover more biomarkers with strong performance.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2018 program
|
