Abstract Body

HIV-exposed uninfected (HEU) infants have increased rates of severe lower respiratory tract infection (LRTI), sepsis, hospitalization and death. We examined the incidence of LRTI in HEU according to maternal antibody transfer and infant antibody production.

We enrolled 247 HEU and 88 HIV-unexposed uninfected (HUU) Brazilian infant/mother pairs, including 107 HEU and 16 HUU with LRTI in the first 6 months of life. Antibodies to the following agents were measured by ELISA: respiratory syncytial virus (RSV) and pneumococcus (PNC) 1, 5, 6, 14 in mothers (delivery) and infants (0, 6 months); influenza A (Flu) and parainfluenza viruses (PIV) 1, 2, 3 (infants 0, 6 months); tetanus toxoid (infants 6 months).

Compared to HUU, HEU infants had lower antibody levels at birth for all respiratory agents (p<0.0001), although maternal antibodies to PNC and RSV did not differ by HIV status. Transplacental transfer of maternal antibodies was lower for RSV in HEU vs. HUU (mean±SD ratios=1.3±3.5 vs. 1.8±0.8; p=0.05). Infant: mother PNC antibody ratios were <1 in both HEU and HUU, but the differences between HEU and HUU were  not statistically significant. Compared to mothers of LRTI-, those of LRTI+ HEU had higher antibody levels to PNC 1 and 6 and those of LRTI+ HUU to PNC 5 and 14 (p≤0.04). Flu, PIV, RSV and PNC antibodies at birth were similar in LRTI+ vs. LRTI- HEU or HUU, except for higher PNC 5 and 14 levels in LRTI+ vs. LRTI- HUU (p≤0.05). At 6 months, HEU and HUU had similar antibody responses to tetanus vaccine regardless of LRTI status. After controlling for birth levels, HEU had lower RSV (p<0.001), higher PIV 1, 2, 3 (p≤0.001) and similar Flu antibodies (p=0.11) compared with HUU at 6 months. At 6 months, LRTI+ HEU had higher anti-RSV antibody levels (p=0.08) and rates of seroconversion to ≥1, 2 or 3 paramyxoviruses (p=0.05, 0.02 and 0.06, respectively) than LRTI- HEU.

The incidence of LRTI in HEU infants correlated with the frequency of paramyxovirus infections, but not with low levels of transferred maternal antibodies or with infant failure to make antibodies in response to infections or vaccines. The higher maternal PNC antibody levels in LRTI+ vs. LRTI- infants suggested that mothers of LRTI+ infants had higher rates of PNC infection and/or carriage increasing infant exposure to PNC and possibly contributing to LRTI morbidity. Collectively, our data suggest that environmental factors and innate and/or cell-mediated immune defects predispose HEU to LRTI.