HIV Articles  
HIV-Specific IgG in Cervicovaginal Secretions of Exposed HIV-Uninfected Female Sexual Partners of HIV-Infected Men
  KATE BUCHACZ,1 BHARAT S. PAREKH,2 NANCY S. PADIAN,3 ARIANE VAN DER STRATEN,3 SUSAN PHILLIPS,2 JANET JONTE,3 and SCOTT D. HOLMBERG2 1Department of Epidemiology and Biostatistics, University of California, Berkeley, Berkeley, California 94720. 2Centers for Disease Control and Prevention, Atlanta, Georgia 30333. 3Department of Obstetrics, Gynecology and Reproductive Sciences, University of California San Francisco, San Francisco, California 94105.
Volume 17, Number 18, 2001, pp. 1689-1693
The presence of human immunodeficiency virus (HIV)-specific antibodies was examined in plasma and cervicovaginal (mucosal) samples of 24 HIV-exposed uninfected (EU) female sexual partners of HIV-infected men, and compared with findings in 18 HIV-infected and 15 low-risk HIV-uninfected women. Only HIV-infected women had detectable HIV-specific immunoglobulin G (IgG) (18 of 18) or HIV-IgA (6 of 18) in cervicovaginal samples by enzyme immunoassay (EIA). However, 3 of 24 EU women had positive Western blot (WB) for HIV-IgG in cervicovaginal secretions, while 2 of 24 EU women and 1 of 15 low-risk controls had indeterminate IgG-WB. EU women with positive or indeterminate IgG-WB in the cervicovaginal samples were similar in risk to the remaining EU women. None of the HIV-uninfected women had mucosal HIV-IgA. The findings suggest that some sexually or parenterally exposed HIV-uninfected women might develop low-level mucosal IgG responses. However, it appears unlikely that HIV-specific cervicovaginal antibodies play a major role in protection from HIV infection in this EU population.

THERE HAS BEEN a continued interest in the study of epidemiologic, genetic, and immune characteristics of individuals such as long-term sexual partners of HIV-infected persons, commercial sex workers, and injection drug users, who have remained HIV uninfected despite repeated exposures to HIV.1-5 Some previous studies have suggested that exposure to HIV-1 might induce development of HIV-specific IgA and/or IgG antibodies in the genital mucosa or serum that could be protective against systemic HIV infection, but the findings to date have been variable and inconsistent.1,2,6-9
Three (13%) of 24 EU women had positive IgG-WB results in their cervicovaginal secretions, suggesting that sexual and/or parenteral (IDU) exposure to HIV may induce development of local HIV-specific mucosal antibodies in the female genital tract without causing systemic seroconversion. However, the repeatedly negative EIA results indicate that mucosal antibody levels were low in these women. Only immunoglobulins of the IgG class were detected in the mucosal samples and no local antigen-specific IgA. It remains to be determined whether IgG in the cervicovaginal secretions of the EU women are only markers of past exposure to HIV, or whether they may also contribute to protection from HIV infection.12,13
HIV-specific IgG in EU women could result from prior exposure to noninfectious quantities or subunits of HIV or clearance of a primary HIV infection. Indeterminate IgG-WB results in plasma or cervicovaginal secretions of low-risk and EU women may, however, be due to exposure, possibly during pregnancy, to other molecules that are antigenically cross-reactive with HIV proteins,14 and therefore may reflect nonspecific reactivity. Indeterminate IgG-WB patterns-bands p24 and p17 being most common-are observed in 10-15% of serum samples in the uninfected (EIA-negative) populations.14 Because total IgG is found in greater concentrations than IgA in the cervicovaginal secretions of uninfected women,15 there might also be a greater likelihood of detecting cross-reactive non-HIV-specific IgG than IgA in the mucosal samples from EU and low-risk women in our study. Finally, it is possible that negative EIA but positive or indeterminate IgG-WB findings in mucosal samples may result from contamination with immunoglobulins from the seminal fluid of the HIV-infected partner, because the self-report of sexual abstinence for 72 hr before the study visit may be unreliable.
The EU participants in our study have remained HIV uninfected despite a moderate to high degree of exposure through unprotected vaginal sex, and in some cases anal sex and needle sharing, with HIV-infected male partners for extended periods of time. Differences in reported frequencies of HIV-specific IgA and IgG between our study and some previous reports may be related to laboratory methods and/or various sociodemographic and risk characteristics of the study populations. Although the level of exposure was substantial for most EU women, it is possible that some women had too few exposures to come in contact with HIV and to develop local antibody responses. The majority of EU women in our study have been exposed to one HIV-infected male partner, and EU women with positive or indeterminate mucosal IgG-WB had no other STIs. By contrast, in some previous studies of commercial sex workers, factors such as high rates of partner exchange, frequent STIs, and specific sexual or hygienic practices may be associated with higher observed frequencies of HIV-specific immunoglobulins in the genital secretions of HIV-uninfected women.1,6 Consistent with previous reports,15 all HIV-infected women in our study had HIV-specific IgG in both plasma and genital fluids, and a substantial proportion also had HIV-IgA in all specimens assayed.
Our study was exploratory in nature and had a modest sample size. Nonetheless, the low frequency of IgG-WB positive results and negative EIA findings suggest that HIV-specific cervicovaginal antibodies do not play a major role in protection from HIV infection in our EU population. Further studies are being conducted to evaluate possible role of HLA (our unpublished data) and other genetic factors and cellular immune responses16 in reducing susceptibility to HIV infection in this group of women. Continued investigation of the immune and molecular mechanisms that are activated in the genital mucosa of exposed but uninfected individuals, which might confer immunity to HIV infection, is important for the development of an effective vaccine for prevention of sexual HIV transmission. 17 Longitudinal studies of exposed uninfected individuals (counseled to reduce their risk) may be best able both to measure local humoral immune responses over time, and to investigate their potential role in modifying susceptibility to sexual acquisition of HIV.
Of 24 HIV-exposed but uninfected (EU) women, 46% were white, 33% were black, and the remaining 21% were Asian, Hispanic, Native American, or of other/mixed race or ethnicity. The mean age of EU women was 40 years (range, 20-50 years). The median duration of the sexual relationship with HIV-infected male partners among EU women was 29 months (range, 2-352 months). EU women reported a median of 46 (range, 5-388) unprotected vaginal sex episodes and a median of 0 (range, 0-12) unprotected anal sex episodes with their HIVinfected partners within the last 2 years. In addition to sexual risk, 15 EU women (63%) had a history of injection drug use, and 9 (38%) reported sharing needles with an HIV-infected person in the previous 2 years. Twenty-two EU women (92%) had been pregnant in the past. None of the EU women had syphilis or current gonorrhea or chlamydia infections, but 12 (50%) were found to have bacterial vaginosis, trichomoniasis, or candidiasis at the time of the pelvic exam and 11 (46%) reported a lifetime history of chlamydia, gonorrhea, syphilis, or herpes. HIV-infected controls (n =18) were similar to EU women with respect to sociodemographic and risk characteristics, including race, age, and parity. However, the low-risk control women (n 5 15) were on average younger, fewer had been pregnant, and fewer were black, as compared with EU and HIV-infected women (data not shown).
HIV-IgG and -IgA findings
Two EU women (8%) and one low-risk control (7%) had plasma samples greactiveh to selected HIV proteins on the IgG Western blot (Table 1) and were classified as IgG-WB indeterminate, whereas all remaining HIV-uninfected women had no HIV-specific IgG detectable in their plasma by EIA and confirmatory Western blot. All EU and low-risk women also had negative EIA for HIV-IgG or -IgA in their cervicovaginal secretions. However, four EU women (17%) and one low-risk control (7%) had IgG-WB bands in the CVL samples (Fisher exact test, p =0.63). In addition, among the 20 EU and 15 lowrisk controls who had cervical wicks available for testing, two EU women (10%) and one low-risk control (7%) had IgG-WB bands in the cervical wick samples. Of all mucosal (CVL and wick) samples tested, three EU women (13%) had positive IgGWB results, whereas two EU women and one low-risk control had indeterminate IgG-WB (Table 1). Only one EU woman (4%) and none of the low-risk controls had IgA-WB bands in the cervical wick sample (indeterminate result).
All HIV-infected women had positive EIA and WB for HIVIgG in their plasma, cervical wick, and CVL, with the exception of one woman who had IgG in plasma and CVL only. The proportions of HIV-infected women who had IgA by EIA and/or who had IgA-WB bands were, respectively, 50 and 100% in plasma, and 33 and 88% in the mucosal samples. The levels of HIV-IgG measured by optical density (OD) in EU women with positive or indeterminate WB were low in comparison with those observed in HIV-infected subjects. For EU subjects included in Table 1, the mean level of IgG in the CVL of four EU women was 0.037 (range, 0.014 to 0.102) and the levels in the cervical wicks of 2 EU women were 0.010 and 0.014. By contrast, the mean levels of HIV-specific IgG in HIVinfected women were 2.3 (range, 0.3 to 3.5) and 2.6 (range, 0.6 to 3.5) in cervical wicks and CVL, respectively.
Sociodemographic and risk correlates of
cervicovaginal IgG among EU women

The 5 EU women who had positive or indeterminate IgGWB results in their CVL and/or wick samples had a somewhat higher estimated number of unprotected vaginal sex exposures with HIV-infected male partners in the previous 2 years than the 19 EU women who had no IgG-WB bands in cervicovaginal specimens (median, 83 vs. 43 times, p =0.57). Unexpectedly, EU women with positive or indeterminate IgG-WB results were significantly less likely to have had STIs at the time of study examination (Fisher exact test, p =0.04), but this could be a chance finding. There were no significant differences with respect to any other sociodemographic, clinical, or risk characteristics between EU women with IgG-WB bands and the remaining EU women.
Factors potentially associated with false-positive or
indeterminate WB results

The majority of women in the study had been pregnant, and of six women with positive or indeterminate IgG-WB in their CVL and/or wick (five EU women and one low-risk control), all but one had a history of pregnancy. None of these six women had any STIs at the clinical examination or occult blood in their CVL. All EU women with mucosal IgG-WB bands reported no unprotected vaginal sex with their HIV-infected male partners within 72 hr of the study visit.
  iconpaperstack view older Articles   Back to Top