Subgroup-Specific Treatment Effect Assessment with Randomized Trials

• Activity Number: 547
• Type: Invited
• Date/Time: Wednesday, August 1, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Social Statistics Section
• Author(s): Tianxi Cai*+ and Lihui Zhao and Lu Tian and Brian Claggett and Lee-Jen Wei
• Companies: Harvard University and Northwestern University and Stanford University and Harvard University and Harvard University
• Keywords: Subgroup Analysis ; Biomarker ; Robust Inference ; Resampling ; Randomized Clinical Trial

Abstract

Clinical trials that evaluate treatment benefit (TB) focus primarily on estimating the average benefit. However, a treatment reported to be effective may not be beneficial to all patients. For example, the benefit of giving chemotherapy prior to hormone therapy with Tamoxifen in the adjuvant treatment of postmenopausal women with lymph node negative breast cancer depends on the ER-status. Due to the toxicity of chemotherapy, it is crucial to identify patients who will and will not benefit from chemotherapy. This gives rise to the need of accurately predicting TB based on important markers.

Traditional approach to evaluating TB is to fit a regression model and assess the interactions between treatment assignment and covariates. Such methods, while useful for hypothesis testing, may have limited ability in quantifying TB due to the restrictive model assumptions. When there is only a single marker, We propose two-step robust procedures for evaluating the performance of multiple markers in predicting subgroup specific TB. We also develop procedures for comparing markers in such predictions.