Invasive bacterial infections in children born to women infected with human immunodeficiency virus type 1.

To determine the rates and characteristics of invasive bacterial infections in children infected with the human immunodeficiency virus type 1 (HIV-1), we conducted a prospective, longitudinal, observational cohort study of infants born to HIV-1-infected mothers between Dec. 1, 1985, and Sept. 30, 1989. Of 104 subjects whose HIV-1 infection status could be definitively determined, 21 were infected with HIV-1 and 83 were not. In all, 11 (48%) of 23 invasive infections occurred among 10 HIV-1-infected patients and 12 (52%) of 23 occurred among 11 uninfected subjects. Infections with Streptococcus pneumoniae (n = 8), all of which were community acquired, accounted for the greatest proportion (35%) of the organisms isolated from either the blood or the cerebrospinal fluid. Five episodes of pneumococcal bacteremia occurred in the HIV-infected patients; all resolved promptly after treatment was begun, and no serious focal infections developed. Of 13 instances of bacteremia with an organism other than S. pneumoniae, seven were nosocomial. The rate of community-acquired invasive bacterial infections among the HIV-infected children was nearly three times higher than the rate in the non-HIV-infected children (overall, 1.02 infections per 100 person-months vs 0.37 infection per 100 person-months; rate ratio, 2.8; p = 0.05). Most of the increased risk occurred in children > 1 year of age. In contrast, the difference in the rates of infection between those patients in the two groups who were less than 12 months of age was not significant (1.3 infections per 100 person-months vs 0.81 infection per 100 person-months; rate ratio, 1.6; p = 0.47). We conclude that the rate of invasive bacterial infection is higher in HIV-infected children than in their peers, especially after 1 year of age.

Quantitative leukoviremia and immune complex-dissociated antigenemia as predictors of infection status in children born to mothers infected with human immunodeficiency virus type 1.

Four methods of culturing human immunodeficiency virus type 1 (HIV-1) from peripheral blood mononuclear cells and two serum antigen tests were assessed as predictors of infection status in children born to HIV-1-infected mothers. Of 36 infants whose cocultures were quantitative, all 15 who were deemed to be infected with HIV-1 (nine with symptoms, six without symptoms) by clinical criteria or persistence of Western blot reactive antibody had positive culture results, and all 21 uninfected seroreverters had negative culture results (sensitivity = 100%; specificity = 100%). Quantitative coculture was more sensitive than a technique in which cells were counted and stimulated with phytohemagglutin but not cocultivated with cells from seronegative donors, and more sensitive than two other qualitative techniques evaluated in samples from 80 children, in which cells were not enumerated before culture. The level of leukoviremia in children with symptoms did not differ appreciably from the level of leukoviremia in symptom-free infected children. Among those with positive results on quantitative coculture, only 40% also had free HIV-1 antigen in serum, whereas 86% had antigen in immune complexes. Among the methods evaluated, quantitative HIV-1 coculture was the best indicator of infection status in children.