Post-randomization Biomarker Effect Modification Analysis in an HIV Vaccine Clinical Trial

Area of Science:

• Immunology
• Biostatistics
• Epidemiology

Background:

• The HVTN 505 trial found no overall efficacy for an HIV vaccine.
• Immune response markers in vaccinated individuals correlated with infection risk, suggesting subgroup effects.

Purpose of the Study:

• To develop robust statistical methods for analyzing treatment effect modification by intermediate response variables.
• To re-evaluate the HVTN 505 trial data using these new methods to identify potential protective subgroups.

Main Methods:

• Adapted methods from Survivor Average Causal Effect (SACE) analysis for Principal Stratification (PS) analysis.
• Utilized a binary intermediate response variable for analysis.
• Applied the adapted methods to the HVTN 505 trial data.

Main Results:

• The adapted PS analysis revealed that the vaccine partially protected individuals with specific vaccine-induced T-cell functions.
• This contrasts with the trial's overall null finding, highlighting the importance of subgroup analysis.

Conclusions:

• New, robust PS methods can be developed by adapting SACE analysis techniques.
• The HVTN 505 vaccine may offer partial protection to specific subgroups defined by immune responses.
• Statistical frameworks are crucial for uncovering nuanced treatment effects in clinical trials.