Most HIV-1 infections worldwide are transmitted sexually, and the plasma viral load has commonly been used as a surrogate predictor of HIV-1 transmission risk.1 However, genital viral load has been shown to predict HIV-1 transmission risk independent of plasma viral levels,2 and genital shedding of HIV-1 can occur even in women with undetectable plasma levels of HIV-1.3 Lower genital inflammation could influence viral shedding, and possibly provide an explanation for the sometimes imperfect correlation between plasma and genital viral loads.
Generalized immune activation is a hallmark of HIV-1 pathogenesis.4 Whether this activation drives viral replication and shedding remains unknown, although expression of activation markers on lower genital T cells is associated with genital viral shedding and with mucosal CD4 depletion.5,6 Acute HIV-1 infection is associated with elevated levels of both proinflammatory (IL-6 and IL-12) and anti-inflammatory (IL-10) cytokines in the genital tract,7 and elevated proinflammatory cytokines in cervicovaginal lavage (CVL) fluid have been associated with HIV-1 shedding.8,9 Thus, local inflammation could have a substantial influence on genital viral shedding by driving recruitment and/or activation of HIV-1–infected cells in the lower genital tract. We set out to attain a better understanding of how the systemic and genital inflammatory environments modulate HIV-1 transmission risk by examining the relationship between systemic and genital cytokine levels and viral loads.
Plasma and vaginal samples collected on the same day were evaluated from 48 HIV-1 seropositive women enrolled in a prospective cohort study in Mombasa, Kenya. None of the women had previously received or were receiving antiretroviral therapy at the time of sample collection. Vaginal samples were collected using a Dacron swab immersed in 1 mL of freezing medium (70% RPMI 1640, 20% fetal calf serum, and 10% dimethyl sulfoxide). Full details of the participants and the sample collection procedures have been previously described.10–12 All participants provided informed consent, and the ethical review committees at the University of Washington, the Fred Hutchinson Cancer Research Center, and the University of Nairobi approved the study protocol.
Plasma and vaginal viral loads were determined by the Gen-Probe HIV-1 viral load assay.10 The limit of detection for HIV-1 viral load was 100 copies/μL; values that were below the level of detection were assigned a value of 50 copies/mL. Cytokine concentrations were determined using the Milliplex MAP Human Cytokine/Chemokine Pre-mixed 26 Plex (Millipore Corporation, Billerica, MA) on Luminex 200 (Luminex Corporation, Austin, TX). The upper limit of cytokine detection was 10,000 pg/mL and the lower limit was 1.52 pg/mL. Samples that were less than the lower limit of detection were assigned a value of 0.76 pg/mL. No analyses were performed on cytokines in which >95% of the observations were below the level of detection (IL-2, IL-3, IL-4, IL-5, and TNF-β in plasma). For sexually transmitted infection (STI) detection, a vaginal saline wet mount was examined for the presence of yeast and motile Trichomonas vaginalis parasites, and vaginal Gram stained slides were evaluated for the presence of bacterial vaginosis using Nugent criteria.13 Culture for Neisseria gonorrhoeae was performed on modified Thayer-Martin media. Genital ulcer disease was diagnosed based on clinician examination.
Statistical analyses were performed in SPSS version 20 (IBM Corp, Armonk, NY). Cytokine levels were log10 transformed; if more than 20% of the measurements were below the lower limit of detection, a dichotomous variable was used reflecting levels at or below versus above the lower limit of detection. Linear regression was used to assess the relationships between the log10 plasma HIV-1 RNA and plasma cytokine levels, log10 vaginal HIV-1 RNA and vaginal cytokine levels, log10 plasma HIV-1 RNA and vaginal cytokine levels, and log10 vaginal HIV-1 RNA and plasma cytokine levels. Regression diagnostics including p-plots and residual plots were performed to assess the fit of the models; the linear regression models fit well. Adjustment for multiple comparisons was performed by controlling for a false discovery rate of 5%. The relationship between plasma and vaginal cytokine levels was assessed using Spearman rank correlation coefficients.
The median plasma viral load was 4.69 log10 copies/mL [interquartile range (IQR): 4.14–5.12], and the median genital viral load was 2.47 log10 copies/mL with 21% below the limit of detection (IQR: 1.97–3.77). Plasma viral load was strongly correlated with vaginal viral load (Pearson correlation coefficient = 0.47, P = 0.001).
Vaginal cytokine concentrations were highest for IL-8, and more than 20% of observations were below the limit of detection for many cytokines (Table 1). Vaginal levels of several proinflammatory cytokines and chemokines (IP-10, MIP-1β, IL-7, IFN-α2, G-CSF, IL-6, IL-8, and TNF-α) were significantly associated with the vaginal viral load. Only levels of IP-10, a chemokine secreted by monocytes, endothelial cells, and fibroblasts in response to IFN-γ, remained significantly associated with shedding after controlling for a false discovery rate of 5% (Table 1). For every 1 log10 pg/mL increase in vaginal, there was a 0.744 log10 increase in vaginal HIV-1 RNA.
Because plasma HIV-1 RNA levels were significantly associated with vaginal HIV-1 RNA levels, we evaluated whether cytokine levels predicted vaginal HIV-1 RNA levels after adjusting for plasma HIV-1 RNA levels. Vaginal levels of IP-10, MIP-1β, IL-7, IFN-α2, IL-8, and TNF-α remained significantly associated with vaginal HIV-1 RNA levels after adjustment for plasma HIV-1 RNA levels, although only IP-10 remaining significant after controlling for a false discovery rate of 5% (Table 1). Finally, because hormonal contraceptive (HC) use and STI have been associated with viral shedding14,15 and may be associated with genital inflammation, we adjusted for plasma log10 HIV-1 RNA, HC use, and STIs. STI screening was limited to bacterial vaginosis, T. vaginalis, yeast, genital ulcer disease by clinician exam, and N. gonorrhoeae by culture. We found that vaginal levels of IP-10, MIP-1β, IL-7, IL-8, and TNF-α were significantly associated with vaginal HIV-1 RNA levels before controlling for a false discovery rate of 5%, although none of these associations were significant after adjusting for multiple comparisons (Table 1). There were no statistically significant associations between vaginal cytokine concentrations and plasma HIV-1 RNA levels (P > 0.08 for all cytokines).
With the exception of IP-10, plasma cytokines levels were lower than vaginal cytokine levels, and more than 20% of the values were below the limit of detection for the several cytokines (Table 2). The only plasma cytokine that was significantly associated with plasma HIV-1 RNA levels was IP-10 (regression coefficient = 0.728, 95% CI: 0.047 to 1.409, P = 0.04, Table 2), and this association was not significant when controlling for a false discovery rate of 5%. There were no statistically significant relationships between any of the plasma cytokines and vaginal HIV-1 RNA (P > 0.08 for all cytokines). In fact, plasma and vaginal cytokine levels were generally not statistically significantly associated. Only IL-10 (Spearman rho = 0.3449, P = 0.016) and IL-15 (Spearman rho = 0.386, P = 0.007) levels were significantly associated between the plasma and vaginal compartments, but these associations were driven entirely by the fact that only 3 subjects had levels of IL-10 and one individual had levels of IL-15 above the limit of detection in the plasma.
We report here that local inflammation is associated with increased genital HIV-1 RNA shedding, independent of plasma viral load. Thus, these findings indicate limitations of using plasma viral load to assess transmission risk and highlight the importance of considering the impact of the local inflammatory environment in modulating transmission. Local inflammation was associated with substantial alterations in the genital viral load, which could result in increased risk of transmission. For instance, in a recent study of serodiscordant couples in Africa, each 1 log10 increase in endocervical viral load was associated with a 1.67-fold relative risk of HIV-1 transmission to the uninfected partner (95% CI: 1.60 to 3.04).2
Chronic inflammation is a hallmark of HIV-1 disease progression,4 yet few prior studies have simultaneously evaluated the systemic and genital inflammatory environment in conjunction with HIV-1 viral load. Jaspan et al6 observed that the frequency of activated T cells in the blood correlated with that in lower genital tract of HIV-1 infected women, and that the latter predicted HIV-1 shedding. While we similarly found that concentrations of several proinflammatory cytokines predicted vaginal shedding (although only IP-10 remained significant after adjusting for multiple comparisons), we found that the systemic concentrations of proinflammatory cytokines were not predictive of genital concentrations nor were they associated with genital HIV-1 shedding. Together, these data indicate that measurements of systemic inflammation cannot necessarily be used to predict the genital tract inflammatory environment or shedding.
If the systemic inflammatory environment does not fully predict the genital environment, then it is important to understand what factors modulate genital inflammation and viral shedding. STIs are associated with genital inflammation and with increased HIV-1 shedding and transmission risk.15 Similarly, use of HCs are associated with genital inflammation,16,17 and several reports have now demonstrated that HCs increase HIV-1 transmission risk, although conflicting data exist.18,19 After adjusting for the plasma viral load, use of HCs, and the presence of several STIs, increased concentrations of several vaginal proinflammatory cytokines were associated with HIV-1 shedding, although none of these associations remained statistically significant after adjusting for multiple comparisons. Because inflammation can persist long after STI symptom resolution,20 post-STI inflammation and/or STIs that we did not adequately measure (such as subclinical HSV infection or reactivation) could explain the weak associations between inflammation and shedding in this model. Future studies will be necessary to more definitively establish whether even subclinical levels of genital inflammation increase transmission risk.
Significant strengths of this study include the use of matched plasma and vaginal collections from a well-characterized study of HIV-1 infection in women. Limitations of this analysis include the relatively small sample size, and the fact that some STIs that could have influenced genital inflammation were not assessed. In addition, variation in the volume of vaginal secretions on each swab could have accentuated the observed associations between concentrations of vaginal inflammatory markers and HIV-1 RNA. Detection of cytokines in vaginal swabs, which are reliable for detection of HIV-1 RNA,12 yielded lower cytokine concentrations than observed in some prior studies of CVL8,9 but were in agreement with other studies.7,21 Given the large number of analytes, the associations between inflammatory cytokines and shedding were significantly attenuated after controlling for a false discovery rate of 5%. However, this may be an overly conservative approach to analyzing these data, as many cytokines are positively correlated. Additional studies, with larger numbers of subjects from different risk groups and geographic regions, will be needed to confirm the observed associations between local inflammation and shedding.
These data bolster and extend previous reports that local inflammation is associated with HIV-1 shedding.5,6,8,9,21 Mitchell et al21 found that CVL IL-1β, a potent proinflammatory cytokine produced by macrophages, and IL-8, a chemokine produced by macrophages, epithelial cells, and other cells, were significantly associated with CVL HIV-1 RNA independent of plasma viral load. The relationship with IL-1β, but not IL-8, was attenuated after adjusting for STI, suggesting that HIV itself and/or noninfectious cases of inflammation may play a role in cervicovaginal HIV-1 shedding. Similarly, Spear et al and Makura et al8,9 found that several proinflammatory cytokines, including IL-6, were significantly associated with shedding. Genital proinflammatory cells were also associated with shedding in 2 studies,5,6 although in one, IL-6 and IL-8, which are secreted primarily by macrophages and epithelial cells, did not predict shedding.6 The differences between studies in precisely which cytokines or cells predict shedding could reflect the fact that the inflammatory mechanisms leading to T-cell activation versus IL-6/IL-8 production could be distinct and independently influence shedding. Furthermore, differences in cohorts and in sampling methods could account for some of these differences.
Together, these data highlight the fact that the genital environment may have a substantial influence of HIV-1 transmission. Genital inflammation is associated with CD4+ T-cell depletion both systemically and in the lower genital tract.6,7 Here, we show that genital inflammation is also associated with HIV-1 shedding in the genital tract, independent of systemic inflammation and viral load. Thus, efforts to decrease HIV-1 transmission must take into account the role of local inflammation and the fact that the local environment is not necessarily reflected in plasma measurements.
The authors would like to thank the clinical and laboratory staff in Mombasa for their tremendous efforts to recruit and retain the women in the cohort and for their collection and storage of samples. They gratefully acknowledge the women who participated in the study.
1. Quinn TC, Wawer MJ, Sewankambo N, et al.. Viral load and heterosexual transmission of human immunodeficiency virus type 1. Rakai Project Study Group. N Engl J Med. 2000;342:921–929.
2. Baeten JM, Kahle E, Lingappa JR, et al.. Genital HIV-1 RNA predicts risk of heterosexual HIV-1 transmission. Sci Transl Med. 2011;3:77ra29.
3. 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.
4. Giorgi JV, Hultin LE, McKeating JA, et al.. Shorter survival in advanced human immunodeficiency virus type 1 infection is more closely associated with T lymphocyte activation than with plasma virus burden or virus chemokine coreceptor usage. J Infect Dis. 1999;179:859–870.
5. Anderson Brenna L, Wang CC, Delong Allison K, et al.. Genital tract leukocytes and shedding of genital HIV type 1 RNA. Clin Infect Dis. 2008;47:1216–1221.
6. Jaspan HB, Liebenberg L, Hanekom W, et al.. Immune activation in the female genital tract during HIV infection predicts mucosal CD4 depletion and HIV shedding. J Infect Dis. 2011;204:1550–1556.
7. Bebell LM, Passmore J-A, Williamson C, et al.. Relationship between levels of inflammatory cytokines in the genital tract and CD4+ cell counts in women with acute HIV-1 infection. J Infect Dis. 2008;198(5):710–714.
8. Mukura LR, Ghosh M, Fahey JV, et al.. Genital tract viral load in HIV-1 positive women correlates with specific cytokine levels in cervical-vaginal secretions but is not a determinant of infectious virus or anti-HIV activity. AIDS Res Hum Retroviruses. 2012; [epub ahead of print, March 23, 2012].
9. Spear GT, Zariffard MR, Chen HY, et al.. Positive association between HIV RNA and IL-6 in the genital tract of Rwandan women. AIDS Res Hum Retroviruses. 2008;24:973–976.
10. Devange Panteleeff D, Emery S, Richardson BA, et al.. Validation of performance of the gen-probe human immunodeficiency virus type 1 viral load assay with genital swabs and breast milk samples. J Clin Microbiol. 2002;40:3929–3937.
11. 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.
12. John GC, Sheppard H, Mbori-Ngacha D, et al.. Comparison of techniques for HIV-1 RNA detection and quantitation in cervicovaginal secretions. J Acquir Immune Defic Syndr. 2001;26:170–175.
13. 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.
14. Wang CC, Mcclelland RS, Overbaugh J, et al.. The effect of hormonal contraception on genital tract shedding of HIV-1. AIDS. 2004;18:205–209.
15. Johnson LF, Lewis DA. The effect of genital tract infections on HIV-1 shedding in the genital tract: a systematic review and meta-analysis. Sex Transm Dis. 2008;35:946–959.
16. Ghanem KG, Shah N, Klein RS, et al.. Influence of sex hormones, HIV status, and concomitant sexually transmitted infection on cervicovaginal inflammation. J Infect Dis. 2005;191:358–366.
17. Baeten JM, Nyange PM, Richardson BA, et al.. Hormonal contraception and risk of sexually transmitted disease acquisition: results from a prospective study. Am J Obstet Gynecol. 2001;185:380–385.
18. Heffron R, Donnell D, Rees H, et al.. Use of hormonal contraceptives and risk of HIV-1 transmission: a prospective cohort study. Lancet Infect Dis. 2012;12:19–26.
19. Blish CA, Baeten JM. Hormonal contraception and HIV-1 transmission. Am J Reprod Immunol. 2010;65:302–307.
20. Zhu J, Hladik F, Woodward A, et al.. Persistence of HIV-1 receptor-positive cells after HSV-2 reactivation is a potential mechanism for increased HIV-1 acquisition. Nat Med. 2009;15:886–892.
21. 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.