Evaluation of Biomarkers for Treatment Selection Using Individual Participant Data Meta-Analysis

• Activity Number: 420
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: ENAR
• Author(s): Chaeryon Kang* and Holly Janes
• Companies: University of Pittsburgh and Fred Hutchinson Cancer Research Center
• Keywords: Individual participant data meta-analysis ; Randomized clinical trial ; Treatment selection biomarker ; Personalized medicine

Abstract

Biomarkers that predict treatment effect heterogeneity across patient subgroups are important for treatment selection and have a potential to improve patient outcomes. A meta-analysis of individual participant data (IPD) is potentially more powerful than a single-study data analysis in evaluating markers for treatment selection. Our study was motivated by the IPD that were collected from three randomized control trials of hypertension and pre-eclampsia among pregnant women to evaluate the effect of labor induction over expectant management of the pregnancy in preventing maternal complications. In this study, we propose a statistical framework for evaluating markers for treatment selection given IPD meta-analytic data from small number of individual studies. We derive marker-based treatment rules by minimizing the average expected outcome across studies. We measure the performance of the marker-based rules by aggregating the study-specific performance, which is a clinically relevant and interpretable. Simulation studies and the application of the proposed methods to IPD meta-analytic data from the high risk pregnancy study are presented.