Cervical HIV-1 DNA shedding is frequently detected among HIV-1–infected women (26%–86% of specimens)1–3 despite potent suppressive antiretroviral therapy (ART) and may be a contributor to sexual HIV-1 transmission. The mechanisms responsible for HIV-1 DNA shedding are not well understood; elucidation of these mechanisms could be useful for prevention of sexual and perinatal HIV-1 transmission.
We hypothesized that proinflammatory cytokines and coinfections in the female genital tract are associated with cervical HIV-1 DNA shedding. To test this hypothesis, we evaluated associations between vaginal proinflammatory cytokines, ART, vaginal coinfections, and HIV-1 DNA shedding in cervical cytobrush samples from HIV-1–infected women in the United States and in Kenya.
Specimens and clinical data were collected during prospective observational studies of HIV-1–infected women in Seattle, WA; Rochester, NY; and Nairobi, Kenya between 2002 and 2009. Study participants were 18–50 years of age, nonpregnant, and had no symptomatic genital infections at study entry. The University of Washington, University of Rochester, and University of Nairobi Institutional Review Boards approved the studies, and all subjects provided written informed consent.
Participants had 3–4 study visits per year over 1–5 years, as described previously.4 HIV-1 RNA from plasma and cervicovaginal lavage (CVL) specimens were quantified by real-time PCR assay.5 HIV-1 DNA was quantified in cervical cytobrush specimens by quantitative polymerase chain reaction (PCR) for the gag region of HIV-1.6 Samples with 1–4 copies of HIV-1 DNA/10 μL extracted DNA were reported as <5 copies/cytobrush. The human betaglobin gene7 was amplified to evaluate the number of cells; samples containing fewer than 100,000 cells were considered inadequate. CVL was tested for interleukins (IL)-1β, -6, -8, and secretory leukocyte protease inhibitor (SLPI) as previously described.8 Urine was collected for Neisseria gonorrhoeae and Chlamydia trachomatis detection by nucleic acid amplification (COBAS Amplicor PCR, Roche, Pleasanton, CA). Trichomonas vaginalis was detected by the InPouch culture system (Hardy Diagnostics, Santa Maria, CA); testing for trichomoniasis was not performed routinely in Kenyan women and was excluded from analysis in this group. Genital shedding of herpes simplex virus-1 and -2 and cytomegalovirus (CMV) was evaluated using PCR of CVL.9,10 Bacterial vaginosis (BV) was diagnosed using Nugent criteria.11 Vaginal yeast and Lactobacillus were detected by culture.12 Detection of an infectious agent in <1% of the genital specimens was considered too sparse for further analysis.
Demographic data at enrollment were compared between US and Kenyan women using Student t test and χ2 test. HIV-1 DNA concentrations were grouped for analysis in 4 categories: undetectable, very low (<5 copies/cytobrush), lesser than the median (5–100 copies/cytobrush), and more than the median (>100 copies/cytobrush). We used multinomial logistic regression and linear regression with generalized estimating equations specifying an independent correlation structure and robust standard errors to assess associations between genital tract infections, cytokines, and HIV-1 DNA shedding. Based on substantial differences in several predictors between US and Kenyan women, these 2 cohorts of women were analyzed separately. All analyses were conducted using Stata version 10 (StataCorp, Inc, College Station, TX).
Data from 136 women were included: 56 from the United States and 80 from Kenya. US women contributed 316 visits with complete data [median, 4/woman; interquartile range (IQR) = 2–6], and Kenyan women 259 visits (median, 3/woman; IQR = 2–4). At enrollment, mean age of US participants was 39 ± 6 years compared with 32 ± 5 years for Kenyan women (P < 0.01). US women were primarily African American (58%), 20% white, and 22% other ethnicities. In US women, mean CD4+ T-cell count at study entry was 436 ± 258 cells/mL, and mean plasma HIV-1 RNA concentration was 2.9 ± 1.4 log10 copies/mL, whereas Kenyan women had both higher CD4+ T-cell count (582 ± 232 cells/mL; P < 0.001) and plasma HIV-1 RNA concentration (3.4 ± 1.3 log10 copies/mL; P = 0.004). ART use was reported by 38 of 56 US women at 179 (57%) of 316 visits, in contrast to 8 (10%) of 80 Kenyan women taking ART at 13 (5%) of 259 visits (P < 0.01). When reporting use of ART, US women showed viral suppression (plasma HIV-1 RNA <30 copies/mL) at 104 (58%)/179 visits, whereas Kenyan women were suppressed at only 2 (15%)/13 visits.
HIV-1 DNA was detected in cervical cytobrush specimens collected at 250 visits from US women (80%) and 207 visits from Kenyan women (79%; P = 0.57). Of 139 specimens with undetectable HIV-1 DNA, 21 (11 from the United States and 10 from Kenya) had inadequate sample and were excluded, leaving 118 (20%) cervical specimens with undetectable HIV-1 DNA. In women with quantifiable HIV-1 DNA, median quantity of cervical HIV-1 DNA differed by nationality: 25 copies/cytobrush (IQR = 5–91) in US women versus 69 copies/cytobrush (IQR = 12–210) in Kenyan women (P < 0.01). However, after adjustment for HIV-1 plasma RNA concentration, this difference was no longer significant (P = 0.37). HIV-1 RNA was detected in CVL collected at 59 (19%) visits by US women and at 148 (57%) visits by Kenyan women (P < 0.01). This difference remained significant after controlling for plasma viral load (P < 0.01). Quantification of the betaglobin gene detected a median of 640,000 cells/specimen (IQR = 321,000–1,266,700) in US women, higher than the median of 459,000 cells/specimen in Kenyan women (IQR = 228,300–780,000; P = 0.03).
US women had significantly lower genital cytokine concentrations compared with Kenyan women. Among US subjects, median IL-1β was 13.7 pg/mL (IQR = 3.6–52.5), compared with 64.9 pg/mL in Kenyan women (IQR = 21.2–264.4) (P < 0.01), with significant differences also seen between the median IL-6 in US women of 4.8 pg/mL (IQR = 2.5–16.2) versus 25.5 pg/mL (IQR = 9.6–72.3) in Kenyan women (P < 0.01); IL-8, with US median of 245 pg/mL (IQR = 114–635) versus Kenyan median of 1273 pg/mL (IQR = 506–3080; P < 0.01); and a median SLPI in US women of 74,559 pg/mL (IQR = 21,171–184,854) versus 222,776 pg/mL (IQR = 101,821–485,043) in Kenyan women (P < 0.01). These differences remained significant even after controlling for differences in prevalence of genital tract infections (data not shown).
Women with higher plasma viral load and lower CD4+ T-cell count were more likely to have higher quantities of HIV-1 DNA detected in cervical specimens in both US and Kenyan populations (Table 1). In US women, none of the studied genital infections were associated with higher levels of cervical HIV-1 DNA, whereas in Kenyan women yeast vaginitis was positively and cervicitis negatively associated with cervical HIV-1 DNA shedding. BV and abnormal vaginal flora were associated with increased concentrations of IL-1β in both US and Kenyan women, but other associations differed between women from the 2 countries (Table 2). Yeast and BV were more common among US women, whereas cervicitis, CMV, and herpes simplex virus shedding were more common among Kenyan women. Gonorrhea and Chlamydia were present in <1% of samples, thus, were not included in analysis. ART was not associated with HIV-1 DNA shedding in univariate analysis, but when the model was controlled for plasma HIV-1 concentration, US women on ART had a higher risk of shedding. We performed a stratified analysis in this group and found that the increased risk of HIV-1 DNA shedding was largely due to the 42% of women who were on ART but not suppressed (data not shown). Too few Kenyan women were on ART to perform the same analysis.
Among US women, after controlling for plasma HIV-1 RNA, the concentration of IL-1β in CVL was significantly higher in women with >100 copies of HIV-1 DNA in cervical secretions than those with no HIV-1 DNA shedding (median 47 pg/mL vs 7 pg/mL; P = 0.02). This was also true for IL-6 (median 9 pg/mL vs 4 pg/mL; P = 0.05) and lower concentrations of SLPI (median 31,119 pg/mL vs 99,175 pg/mL; P = 0.008). IL-8 was not significantly different between woman with ≥100 copies/cytobrush of HIV-1 DNA detected and those with no detectable cervical HIV-1 DNA (median, 211 pg/mL vs 245 pg/mL; P = 0.53). In Kenyan women, no differences in cytokine concentrations were seen between specimens with ≥100 copies/cytobrush HIV-1 DNA (n = 88) and those with no HIV-1 DNA detected (n = 52) (data not shown).
This study of HIV-1 infected Kenyan and US women found higher levels of proinflammatory cytokines in the CVL of Kenyan compared with US HIV-1–infected women. Kenyan women had higher rates of some genital infections, such as cervicitis and CMV, but US women were more likely to have yeast. Differences in cytokines persisted after controlling for genital infections, suggesting a larger difference in immune milieu between US and Kenyan women. Cohen et al13 found that HIV-1–uninfected Kenyan adolescents had a higher number of activated CD4+ T cells in the genital tract than American adolescents even after controlling for genital infections. The genital inflammation in these Kenyan populations may stem from systemic infections with parasites, malaria, or TB, from vaginal washing practices, or from unmeasured genital tract infections.
In the US cohort, women with the highest quantities of cervical HIV-1 DNA had higher genital IL-1β and IL-6 concentrations compared with women with no HIV-1 DNA detected. Both cytokines are common to many immune response pathways in the innate and adaptive immune response, thus, may implicate activation of any of several pathways and do not point to a specific mechanistic connection.14–16 Interestingly, no genital infections were positively associated with the highest levels of HIV-1 DNA in either population. Together, these results suggest that there is not a direct causal pathway between the genital infections we studied, these classic proinflammatory cytokines, and HIV-1 DNA genital shedding. The stimuli for HIV-1 DNA genital shedding may be from infections that were not measured in this cohort, from systemic factors, or from noninfectious local stimuli.
An association between genital infections and cytokine concentrations in genital secretions has been reported across multiple studies, although the cytokines and infections studied have varied. Across studies, BV is associated with IL-1β in both US17,18 and Kenyan19 women, but few reports have examined IL-1β and other infections. IL-6 seems to have little relationship with any genital infection,17,20–22 whereas increased IL-8 has been associated with trichomoniasis,23 cervicitis,24 and yeast25 in many but not all17,25 studies. Few of these studies enrolled African women, thus, we are unable to assess whether the differences we observed in associations between US and Kenyan women are an anomaly or reflect regional differences. Even fewer studies have assessed the link between genital infections and genital tract HIV-1 DNA; one showed no association between yeast vulvovaginitis and cervical HIV-1 DNA detection,2 whereas a Kenyan study found an association with mucopurulent cervical discharge, but did not assess specific pathogens.26
HIV-infected cells may pose a risk for sexual transmission,27 similar to the association between HIV-1 DNA in blood and breast milk and mother-to-child transmission.28 Plasma HIV-1 RNA concentration is the factor most commonly associated with detection of HIV-1 DNA in cervicovaginal secretions.2,29,30 Although previous studies have reported lower rates of HIV-1 DNA shedding in women on ART31–33 compared with those who are not,1,10,30 detection of HIV-1 DNA is much more common than HIV-1 RNA when women have a suppressed plasma viral load.5 In this study, US women on ART but not suppressed in the plasma had a higher risk of HIV-1 DNA shedding than women not on ART. This is likely because, in our US cohort, women on ART were likely to be sicker (as evidenced by the low rate of plasma viral suppression), creating a more inflammatory environment.
Our study only assessed cervical HIV-1 DNA shedding and not detection in vaginal secretions. Other investigators have found a higher prevalence of HIV-1 DNA shedding in cervical compared with vaginal samples.26 Our evaluation of cervical cytobrush samples detected HIV-1 DNA at higher rates than studies which have used cervical Dacron swabs or cervicovaginal lavage1,29,30 but similar to rates reported by another study using cytobrush samples.3 This is likely because the cytobrush picks up more cellular material than swabs or lavage. However, our cytokine measurements were made on CVL fluid, an indirect reflection of cervical inflammation. Our participants contributed visits both on and off ART, and with and without virologic suppression, which increases the heterogeneity of the data and may mask differences in determinants of genital shedding.
Our study did not detect a direct causal pathway linking genital infections, the classic proinflammatory cytokines IL-1β, IL-8, and IL-6 and HIV-1 DNA shedding. Additionally, local effects of inflammation or infection on cervical HIV-1 DNA shedding are eclipsed by systemic factors such as plasma HIV-1 RNA concentration.
1. Graham SM, Holte SE, Peshu NM, et al.. Initiation of antiretroviral therapy leads to a rapid decline in cervical and vaginal HIV-1 shedding. AIDS. 2007;21:501–507.
2. Spinillo A, Zara F, Gardella B, et al.. The effect of vaginal candidiasis on the shedding of human immunodeficiency virus in cervicovaginal secretions. Am J Obstet Gynecol. 2005;192:774–779.
3. Iversen AK, Attermann J, Gerstoft J, et al.. Longitudinal and cross-sectional studies of HIV-1 RNA and DNA loads in blood and the female genital tract. Eur J Obstet Gynecol Reprod Biol. 2004;117:227–235.
4. Mitchell C, Hitti J, Paul K, et al.. Cervicovaginal shedding of HIV type 1 is related to genital tract inflammation independent of changes in vaginal microbiota. AIDS Res Hum Retroviruses. 2011;27:35–39.
5. Zuckerman RA, Lucchetti A, Whittington WL, et al.. Herpes simplex virus (HSV) suppression with valacyclovir reduces rectal and blood plasma HIV-1 levels in HIV-1/HSV-2-seropositive men: a randomized, double-blind, placebo-controlled crossover trial. J Infect Dis. 2007;196:1500–1508.
6. Micek MA, Blanco AJ, Beck IA, et al.. Nevirapine resistance by timing of HIV type 1 infection in infants treated with single-dose nevirapine. Clin Infect Dis. 2010;50:1405–1414.
7. Coutlee F, He Y, Saint-Antoine P, et al.. Coamplification of HIV type 1 and beta-globin gene DNA sequences in a nonisotopic polymerase chain reaction assay to control for amplification efficiency. AIDS Res Hum Retroviruses. 1995;11:363–371.
8. Mitchell CM, Balkus J, Agnew KJ, et al.. Bacterial vaginosis, not HIV, is primarily responsible for increased vaginal concentrations of proinflammatory cytokines. AIDS Res Hum Retroviruses. 2008;24:667–671.
9. Corey L, Huang ML, Selke S, et al.. Differentiation of herpes simplex virus types 1 and 2 in clinical samples by a real-time taqman PCR assay. J Med Virol. 2005;76:350–355.
10. Mostad SB, Kreiss JK, Ryncarz AJ, et al.. Cervical shedding of cytomegalovirus in human immunodeficiency virus type 1-infected women. J Med Virol. 1999;59:469–473.
11. Nugent RP, Krohn MA, Hillier SL. Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation. J Clin Microbiol. 1991;29:297–301.
12. Eschenbach DA, Davick PR, Williams BL, et al.. Prevalence of hydrogen peroxide-producing Lactobacillus species in normal women and women with bacterial vaginosis. J Clin Microbiol. 1989;27:251–256.
13. Cohen CR, Moscicki AB, Scott ME, et al.. Increased levels of immune activation in the genital tract of healthy young women from sub-Saharan Africa. AIDS. 2010;24:2069–2074.
14. Barker BR, Taxman DJ, Ting JP. Cross-regulation between the IL-1beta/IL-18 processing inflammasome and other inflammatory cytokines. Curr Opin Immunol. 2011;23:591–597.
15. Ben-Sasson SZ, Caucheteux S, Crank M, et al.. IL-1 acts on T cells to enhance the magnitude of in vivo immune responses. Cytokine. 2011;56:122–125.
16. Dinarello CA. A clinical perspective of IL-1beta as the gatekeeper of inflammation. Eur J Immunol. 2011;41:1203–1217.
17. Spear GT, Kendrick SR, Chen HY, et al.. Multiplex immunoassay of lower genital tract mucosal fluid from women attending an urban STD clinic shows broadly increased IL1β and lactoferrin. PLoS One. 2011;6:e19560.
18. Anton G, Rid J, Mylonas I, et al.. Evidence of a TH1-shift of local vaginal inflammatory response during bacterial vaginosis. Infection. 2008;36:147–152.
19. Rebbapragada A, Howe K, Wachihi C, et al.. Bacterial vaginosis in HIV-infected women induces reversible alterations in the cervical immune environment. J Acquir Immune Defic Syndr. 2008;49:520–522.
20. Hemalatha R, Ramalaxmi BA, Krishnaswetha G, et al.. Cervicovaginal inflammatory cytokines and sphingomyelinase in women with and without bacterial vaginosis. Am J Med Sci. 2012;344:35–39.
21. Mattsby-Baltzer I, Platz-Christensen JJ, Hosseini N, et al.. IL-1beta, IL-6, TNFalpha, fetal fibronectin, and endotoxin in the lower genital tract of pregnant women with bacterial vaginosis. Acta Obstet Gynecol Scand. 1998;77:701–706.
22. Weissenbacher TM, Witkin SS, Gingelmaier A, et al.. Relationship between recurrent vulvovaginal candidosis and immune mediators in vaginal fluid. Eur J Obstet Gynecol Reprod Biol. 2009;144:59–63.
23. Simhan HN, Anderson BL, Krohn MA, et al.. Host immune consequences of asymptomatic Trichomonas vaginalis infection in pregnancy. Am J Obstet Gynecol. 2007;196:59.e1–5.
24. Sawada M, Otsuki K, Mitsukawa K, et al.. Cervical inflammatory cytokines and other markers in the cervical mucus of pregnant women with lower genital tract infection. Int J Gynaecol Obstet. 2006;92:117–121.
25. Spear GT, Zariffard MR, Cohen MH, et al.. Vaginal IL-8 levels are positively associated with Candida albicans and inversely with lactobacilli in HIV-infected women. J Reprod Immunol. 2008;78:76–79.
26. Clemetson DB, Moss GB, Willerford DM, et al.. Detection of HIV DNA in cervical and vaginal secretions. Prevalence and correlates among women in Nairobi, Kenya. JAMA. 1993;269:2860–2864.
27. Salle B, Brochard P, Bourry O, et al.. Infection of macaques after vaginal exposure to cell-associated simian immunodeficiency virus. J Infect Dis. 2010;202:337–344.
28. Arvold ND, Ngo-Giang-Huong N, McIntosh K, et al.. Maternal HIV-1 DNA load and mother-to-child transmission. AIDS Patient Care STDS. 2007;21:638–643.
29. Benki S, McClelland RS, Emery S, et al.. Quantification of genital human immunodeficiency virus type 1 (HIV-1) DNA in specimens from women with low plasma HIV-1 RNA levels typical of HIV-1 nontransmitters. J Clin Microbiol. 2006;44:4357–4362.
30. Andreoletti L, Chomont N, Gresenguet G, et al.. Independent levels of cell-free and cell-associated human immunodeficiency virus-1 in genital-tract secretions of clinically asymptomatic, treatment-naive African women. J Infect Dis. 2003;188:549–554.
31. Cu-Uvin S, DeLong AK, Venkatesh KK, et al.. Genital tract HIV-1 RNA shedding among women with below detectable plasma viral load. AIDS. 2010;24:2489–2497.
32. Graham SM, Masese L, Gitau R, et al.. Genital ulceration does not increase HIV-1 shedding in cervical or vaginal secretions of women taking antiretroviral therapy. Sex Transm Infect. 2011;87:114–117.
33. Henning TR, Kissinger P, Lacour N, et al.. Elevated cervical white blood cell infiltrate is associated with genital HIV detection in a longitudinal cohort of antiretroviral therapy-adherent women. J Infect Dis. 2010;202:1543–1552.