POOR ANTIBODY RESPONSE TO YELLOW FEVER VACCINATION IN CHILDREN INFECTED WITH HUMAN IMMUNODEFICIENCY VIRUS TYPE 1

Abstract

The AIDS epidemic in children has raised concern regarding the safety and effectiveness of vaccines used in the routine immunization of children infected with HIV-1. Studies have established that most routine childhood vaccines can be safely administered to HIV-infected children but that some antigens are less immunogenic in this population.1 Since the mid-1980s, AIDS has emerged as a significant public health problem among children in Abidjan, the principal city of Côte d'Ivoire, where 0.8% of healthy children and 8.2% of hospitalized children were documented to be HIV-seropositive in 1989 and 1992, respectively.2, 3 Because Côte d'Ivoire lies in the yellow fever (YF) endemic zone, YF vaccination is part of the routine childhood immunization schedule. YF vaccine is a live attenuated virus vaccine that is generally well-tolerated, produces an immune response in 95% of vaccinees and affords excellent protection from YF infection.4. However, there are no published data on the immunogenicity of YF vaccination in HIV-infected children. We assessed this issue in a cohort of children followed in Abidjan. Methods. Data for this analysis were collected during a prospective cohort study of mother-to-child transmission of HIV that was conducted from September, 1990, to October, 1994, in Abidjan.5 Children born to 138 HIV-1-seropositive women, 122 HIV-2-seropositive women, 69 women seroreactive to both HIV-1 and HIV-2 and 266 HIV-seronegative women were enrolled at birth and followed for up to 48 months. Each child was examined at 1, 2 and 3 months of age and every 3 months thereafter, and a blood sample was obtained at each visit. Serum specimens were screened by HIV-1 and HIV-2 enzyme immunoassays (either Genetic Systems, Seattle, WA, or Genelavia Mixte, Diagnostics Pasteur, Paris, France), and HIV infection was confirmed and the virus type specified by a synthetic peptide-based assay (Pepti-LAV 1-2®, Diagnostics Pasteur) and virus-specific Western blot (Dupont de Nemours, Geneva, Switzerland, for HIV-1; Diagnostic Pasteur for HIV-2).6 Children who had a positive HIV-1 Western blot at ≥15 months of age were considered HIV-infected, whereas children with a negative HIV-1/2 enzyme immunoassay at ≥15 months of age were considered HIV-uninfected. Between June, 1991, and June, 1993, as part of the routine immunization schedule, all children received a single 0.5-ml dose of 17D YF vaccine (Pasteur Mérieux Sérums et Vaccin, Daccar, Senegal) and measles vaccine at a mean of 10 months of age (range, 7 to 14 months). All children were vaccinated in one clinic, and the staff administering the vaccine were unaware of the child's HIV infection status. For this analysis we selected all HIV-infected children with available pre- and postvaccination serum samples and approximately 3 times as many randomly selected HIV-uninfected children. For all children YF neutralization antibody titers were compared between a prevaccination serum sample obtained on the day of vaccination or on the previous visit and a postvaccination sample obtained during a routine follow-up visit at a mean of 5 months after vaccination (range, 2 to 10 months). To assess the influence of maternal YF antibody on a child's response to YF vaccine, YF titer was measured on a maternal serum sample obtained at delivery. Neutralizing antibody response to YF vaccination was assessed by plaque reduction neutralization assay, with 2-fold dilutions starting at a serum dilution of 1:10.7 An adequate antibody response to YF was defined as an increase in YF titers from <10 on the prevaccination to ≥10 on the postvaccination sample. Results. Of the 33 HIV-1-infected children identified in the study, 26 had sufficient follow-up to have had a postvaccination clinic visit, and 18 of these had available pre- and postvaccination serum samples. Sixteen of these children were born to HIV-1-seropositive mothers and 2 to HIV-1/2 dually reactive mothers. Sixty-one HIV-uninfected children were selected as a comparison group, but 4 of these children were excluded from the analysis because their prevaccination serum sample had a YF titer of ≥10. The remaining 57 HIV-uninfected children included 38 born to HIV-negative mothers and 19 to HIV-positive mothers (4 HIV-1, 10 HIV-2 and 5 HIV-1/2 dually reactive) (Table 1). The 2 groups of children had a similar sex ratio and similar nutritional profile as measured by the mean weight-for-height and height-for-age percentiles. HIV-infected children were slightly older than HIV-uninfected children at the time of vaccination (10.5 vs. 9.7 months, P = 0.02), but the intervals from vaccination to postvaccination serology were similar in the two groups. None of the HIV-infected children met the pediatric clinical AIDS case definition.8TABLE 1: Characteristics of children evaluated to assess response to yellow fever vaccination, Abidjan, Côte d'Ivoire Only 3 (17%) of the 18 HIV-infected children had an adequate YF antibody response compared with 42 (74%) of the 57 HIV-uninfected children, (relative risk 0.2; 95% confidence interval 0.1 to 0.6; P < 0.0001) (Fig. 1). Among the 45 children with an adequate antibody response, the geometric mean YF titers at the postvaccination visit were 32 for HIV-infected and 33 for uninfected children. Among HIV-uninfected children, 68% of children born to HIV-positive mothers had an adequate antibody response compared with 76% of children born to HIV-negative mothers (relative risk 0.8; 95% confidence interval 0.3 to 1.8; P = 0.75). The presence of maternal YF antibody titer at delivery did not significantly alter the child's YF vaccination response. In fact children born to mothers with a YF titer of ≥10 at delivery had a slightly better YF response than did children born to mothers with a lower titer (HIV-infected children, 29% vs. 10%, P = 0.54; HIV-uninfected children, 88% vs. 65%, P = 0.16). Children were followed for a median of 29 months after YF vaccination, and no child developed YF. The frequency of adverse reactions could not be determined given that these were not systematically ascertained.Fig. 1: Postvaccination yellow fever neutralizing antibody titer in HIV-infected and uninfected children (bar, median). All prevaccination titers were <10.Discussion. Our results indicate that HIV-infected children may respond less well to YF vaccination than do HIV-uninfected children and that despite vaccination some HIV-infected children may be at risk of acquiring YF. These results need to be interpreted with some caution because of several limitations. First this study was not specifically designed to describe the response to YF vaccination. Second the proportion of HIV-uninfected children with an adequate YF antibody response was lower than the 95% response that has been reported in other studies,9 suggesting that the antigenicity of the vaccine used, vaccine storage or administration were less than optimal. Finally because lymphocyte subtyping was not performed in these children, we cannot assess the relationship between response to YF vaccine and level of HIV-related immunosuppression. To our knowledge no other data exist regarding the immunogenicity of YF vaccine in HIV-infected children. However, poor immunogenic response to other vaccines has been documented among children infected with HIV. For example only 11 (55%) of 20 HIV-infected children who had received measles vaccine showed detectable measles antibody compared with 12 (92%) of 13 HIV-uninfected children.10 Similarly only 35% of HIV-infected children receiving hepatitis B vaccine had a measurable antibody response compared with 92% of HIV-uninfected children.11 The reason for the poor immune response to YF vaccine is unknown but may be related to immunologic dysfunction associated with HIV infection. Alternatively concurrent HIV infection may interfere with normal YF vaccine replication in mononuclear cells. Further studies are required to confirm this finding and, if confirmed, to determine the biologic mechanism of this poor immunologic response and whether this poor response results in a higher risk for YF disease. Toussaint S. Sibailly, M.D. Stefan Z. Wiktor, M.D., M.P.H. Theodore F. Tsai, M.D., M.P.H. Bruce C. Cropp Ehounou R. Ekpini, M.D. Georgette Adjorlolo-Johnson, M.D., M.P.H. Emannuel Gnaore, M.D., M.P.H. Kevin M. DeCock, M.D., F.R.C.P., D.T.M.H. Alan E. Greenberg, M.D. Projet RETRO-CI (TSS, SZW, ERE, GAJ, AEG); National AIDS Control Program (EG); Abidjan, Côte d'Ivoire Division of HIV/AIDS Prevention; Centers for Disease Control and Prevention; Atlanta, GA (SZW, KMD, AEG) Division of Vector-Borne Infectious Diseases; Centers for Disease Control and Prevention; Ft. Collins, CO (TFT, BCC)