Negative Control Outcome Adjustment in Early-Phase Randomized Trials: Estimating Vaccine Effects on Immune Responses in HIV Exposed Uninfected Infants

Early-phase trials to evaluate vaccine immunogenicity are hampered by unique challenges, especially in HIV-exposed uninfected infants due to small trial sizes, imbalanced randomization ratios, and persistent non-vaccine-elicited immune responses inherited from the mother. We extend methods project for covariate adjustment in randomized trials to include post-baseline variables unaffected by treatment, or Negative Control Outcomes. In vaccine trials in HEU infants, two possibilities for NCOs are immune responses to HIV-1 antigens not included in the vaccine or immune responses to other routine early-childhood vaccinations. In numerical experiments and two reanalyses of early-phase HIV-1 vaccine trials in HEU infants, we demonstrate superior performance of our method. Our results suggest that off-target endpoints can enhance the efficiency of early-phase randomized trials conducted in populations with previous exposure to the pathogen of interest.

Our paper is published in Statistics in Medicine and can be found here.