Using Multiple Biomarkers as a Bridge to Examine the Efficacy of Cancer Lifestyle Intervention Trials

• Activity Number: 341
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Hongying Li*+ and Loki Natarajan
• Companies: University of California at San Diego and University of California at San Diego
• Keywords: lifestyle intervention trial ; multiple biomarkers ; composite risk score

Abstract

In cancer lifestyle intervention trials, the ultimate goal is to improve survival. But this usually requires a very long-term follow-up and is often not feasible. Instead, intervention effects on intermediate outcomes, such as biomarkers, may be more easily obtained. Of note, lifestyle interventions frequently impact multiple pathways and biomarkers. Rarely there is an existing model that considers all the relevant biomarkers together. Thus how to translate intervention related changes in multiple, possibly correlated biomarkers into survival benefits raises analytic challenges. To address this, we will develop a method that yields appropriate weights for different biomarkers, which are then used to combine them into a composite risk score. The efficacy of the intervention trial is assessed by the change of the risk score. The appropriate relative weights are derived by comparing a disease model conditioning on the joint distribution of the markers to the corresponding marginal models and incorporating the correlations of the markers. We describe the proposed analytic method in a logistic regression setting and will present simulation studies to evaluate our proposed method.