Robust Analysis of Biomarker Data with Informative Missingness Using a Two-Stage Hypothesis Test in an HIV Treatment Interruption Trial: AIEDRP AIN503 /ACTG A5217

• Activity Number: 255
• Type: Contributed
• Date/Time: Tuesday, August 8, 2006 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Karen Messer*+ and Florin Vaida and Christine Hogan
• Companies: University of California, San Diego and University of California, San Diego and Columbia University
• Keywords: missing data ; clinical trial ; robust analysis

Abstract

AIN503/A5217 studies whether a period of early treatment with antiretroviral therapy may lower the viral setpoint in subjects recently HIV-1 infected. However, even under the null hypothesis control arm subjects are more likely to be missing endpoint data because of disease progression, leading to bias. Also, two comparisons between arms are important: one at identical times post-randomization, and the other at the same amount of time off-therapy, as measured post-randomization. Both comparisons are important. The null hypothesis is exchangeability under a time-shift between arms, which captures clinically relevant information and allows us to formalize a two-stage hypothesis test using both comparisons. With this framework, we show that the two-stage test can be adjusted for the missing data using a simple worst-rank substitution.