Background: Human immunodeficiency virus (HIV) exposed infants are at high risk of Mycobacterium tuberculosis exposure, have high rates of progression to tuberculosis (TB) disease and are at significant risk of bacille Calmette-Guérin (BCG) induced adverse events.
Objective: To evaluate a delayed BCG vaccination strategy in HIV-exposed infants.
Design: A randomised trial of routine BCG vaccination given at birth compared to 14 weeks of age in HIV-exposed non-infected and non-HIV-exposed infants to investigate longitudinal BCG-induced immune responses using a 7-day whole blood interferon-gamma (IFN-γ) enzyme-linked immunosorbent assay.
Results: A significantly higher proportion of infants had positive responses to M. tuberculosis purified protein derivative (PPD) and BCG at 14 weeks in the birth vs. delayed vaccination groups (P = 0.001 for both). This difference was no longer apparent at weeks 24 or 52. Among infants vaccinated at birth, the 14-week IFN-γ response to M. tuberculosis PPD was lower among HIV-exposed than non-exposed infants (276.5 pg/ml vs. 790.2, P = 0.048). Among all infants, there were significant correlations between the magnitude of IFN-γ responses to BCG, M. tuberculosis PPD, TB 10.4 and culture filtrate protein 10/early secreted antigenic target 6.
Conclusions: The timing of vaccination had limited effect on BCG-induced IFN-γ responses, which waned considerably over 1 year despite initial vigorous responses in both vaccination groups. The lower responses in HIV-exposed non-infected infants suggest potentially altered mycobacterial immunity early in life.