Sexual and mother-to-child transmission of HIV-1 are highly correlated with plasma viral RNA levels, and transmission likely occurs in most cases through contact with HIV-1 in genital secretions [1,2]. Plasma HIV-1 RNA levels are often used as a surrogate measure for genital tract HIV-1 RNA levels in terms of transmission risk . Studies of genital tract HIV-1 shedding have demonstrated a strong correlation between plasma HIV-1 RNA and genital tract HIV-1 RNA , with plasma HIV-1 RNA levels as the strongest predictor of genital tract HIV-1 RNA detection . However, HIV-1 RNA in the genital tract can be detected in women with plasma viral loads (PVLs) below 500 copies/ml , suggesting that viral replication in the genital tract and plasma can be divergent [7–10]. Several conditions can increase the risk for genital tract HIV-1 RNA shedding, including the pharmacology of specific antiretroviral agents , genital infections , oral contraceptive use , and pregnancy .
Increasingly, HIV-infected women are receiving highly active antiretroviral therapy (HAART) leading to plasma viral suppression, but data are limited regarding the temporal patterns of genital tract HIV-1 RNA shedding among women suppressed on HAART . Though studies have described residual genital tract viremia among patients on HAART [6,11,16–18], data still remain limited on longitudinal genital tract HIV-1 shedding patterns in these women.
The study aims to understand the effect of HAART on genital tract HIV-1 shedding in an effort to better understand possible continued risk of HIV transmission even with controlled plasma viremia [19,20]. The present study was undertaken among HIV-1-infected women receiving HAART with below-detectable PVL during the 6 months prior to study enrollment to describe the observed longitudinal patterns of genital tract HIV-1 RNA shedding by placing patients into defined shedding categories and examining the associations between shedding category and covariates such as age, race/ethnicity, HAART regimen, PVL, and CD4 cell count; compare concentrations of and visit-level patterns of HIV-1 RNA shedding within genital tract subcompartments; quantify rates of genital tract HIV-1 RNA shedding by genital tract subcompartment over all study visits and when PVL is undetectable; examine the relationships between genital tract infections [bacterial vaginosis, trichomoniasis, candidiasis, and Herpes simplex virus type 2 (HSV-2) PCR positivity] and CD4 and genital tract HIV-1 RNA shedding within each genital tract subcompartment; and describe the dynamics of HIV-1 RNA in the plasma and genital tract using visit-to-visit transitions from and to detectable and undetectable PVL or genital tract HIV-1 RNA.
Participants were recruited from the Miriam Hospital Immunology Center (Providence, Rhode Island, USA), and included HIV-1-infected women on HAART with below-detectable PVL at least 6 months prior to the screening visit. The Miriam Hospital Institutional Review Board approved the study. Participants were excluded if they were younger than 18 years, pregnant, or had pelvic surgery within 3 months. Baseline demographic, medical, sexual and reproductive history were collected. Participants were followed every 4 weeks for 1 year. Women were advised not to have sexual intercourse, douche, or insert intravaginal products at least 48 h prior to study visits. Pelvic examinations were performed and paired plasma and genital tract specimens were collected at each visit, scheduled to coincide with the mid-point of the menstrual cycle as much as possible. Participants were asked about medication adherence.
Participants with at least three consecutive study visits with undetectable PVL (≤80 copies/ml) were subdivided into three genital tract HIV-1 RNA shedding categories: persistent, intermittent, or nonshedder. Classifications were based upon observed longitudinal genital tract HIV-1 RNA viral load patterns during visits with undetectable PVL. A woman was classified as a persistent shedder if she had at least two consecutive monthly visits with detectable genital tract HIV-1 RNA and undetectable PVL. An intermittent shedder had detectable genital tract HIV-1 RNA and undetectable PVL on visits that were preceded and followed by visits with undetectable genital tract and plasma HIV-1 RNA. Nonshedders had no detectable genital tract HIV-1 RNA during visits with undetectable PVL. A woman was classified as indeterminate (or unknown) if she had fewer than three consecutive visits or too few consecutive visits with undetectable PVL to place her in one of the other three categories. Visits with detectable PVL were not used for classification.
Screening PVL was quantitated by bDNA (Chiron, Inc.) with a lower limit of detection of 75 copies/ml. After enrollment, PVL and genital tract viral loads were quantitated by NucliSens HIV-1 QT (bioMérieux, Inc, Durham, North Carolina, USA). The limit of detection was 80 copies/ml. For genital specimens the volume collected was 24 ul (three SnoStrips/ 2 TearFlo), and the limit of detection was 3300 copies/ml. Samples were invalid when there was not adequate amplification of the internal calibrators. Genital tract secretions were collected using Sno-strip filter paper from the endocervix, ectocervix, and vagina. After Sno-strip manufacture was discontinued in late 2005, TearFlo strips were used. A cervicovaginal lavage (CVL) using 10 ml of normal saline was also collected and processed for white blood cell counts. Women with hysterectomy contributed only vaginal samples. Tests for gonorrhea (culture), Chlamydia (culture), and syphilis (serology) were done at baseline and when clinically indicated. Tests for Trichomonas vaginalis, bacterial vaginosis, symptomatic candidiasis, and HSV-2 in the genital tract were done at every visit. Trichomoniasis was diagnosed by the presence of trichomonads on saline mount. Candida was diagnosed by the presence of hyphae/spores on wet mount. Bacterial vaginosis was diagnosed by the Amsel definition . HSV DNA was detected by real-time PCR. The presence of seminal fluid in genital tract secretions was detected using the ABA card P30 test. Blood testing included CD4 cell counts, PVL, HSV-2 antibody and syphilis serologies. Every attempt was made to procure a sample from each genital subcompartment; but at some visits, women did not have adequate secretions for all subcompartments. Every attempt was made to quantify HIV viral load and samples with invalid viral load measures were reanalyzed a second time.
The following outcome variables were analyzed: shedding category derived by aggregating patient-specific longitudinal data of genital tract HIV-1 RNA shedding; presence and concentration of HIV-1 RNA in blood plasma and in the vagina, endocervix, and ectocervix; presence of genital tract HIV-1 RNA in at least one subcompartment (yes/no); and concomitant shedding in the vagina, endocervix, and ectocervix.
Because only four participants were persistent shedders, the intermittent and persistent shedding groups were aggregated for comparison with the nonshedding group. Comparisons included age, race/ethnicity, baseline CD4 cell count, HAART regimen, taking a protease inhibitor (nucleoside/nucleotide reverse transcriptase inhibitors achieve higher genital tract concentrations compared to protease inhibitors ), having a hysterectomy, and taking hormonal contraceptives. The categorical covariates were compared using nonparametric Fisher's exact tests and risk ratios [95% confidence interval (CI)] and the continuous covariates were compared using Wilcoxon rank sum tests and nonparametric estimates of the difference in medians (95% CI).
Patterns of concomitant HIV-1 RNA shedding within genital tract subcompartments were described. The medians and ranges of HIV-1 RNA concentrations were calculated for each subcompartment and in the vagina of women with hysterectomies for visits with at least one detectable HIV-1 RNA.
For each subcompartment and for all subcompartments combined, estimates of the overall probability of having detectable genital tract HIV-1 RNA when PVL was detectable and undetectable were obtained using logistic regression for longitudinal binary data. These models, and the longitudinal models described below, were fitted using generalized estimating equations. Robust standard errors were used for significance testing and confidence interval estimation.
For each subcompartment and all subcompartments combined, the within-patient and between-patient effects of detectable PVL on the odds of detectable genital tract shedding were estimated with logistic regression having the following covariates: detectable PVL at baseline (yes/no) and the visit-specific difference in PVL detection status compared to baseline (taking values −1, 0, or 1). The PVL effects were adjusted for patient age, race/ethnicity, and taking a HAART regimen containing a protease inhibitor at baseline. Models were not fitted for women with hysterectomies because there were only 11 of them.
To explore potential associations between genital tract HIV-1 RNA shedding and other covariates [the presence of genital tract infections, genital tract white blood cell count (genital tract WBC), or CD4 cell count], two additional covariates were added to the PVL model: baseline presence/absence of the genital tract infection or value of genital tract WBC or CD4 cell count and for the dichotomous outcomes (presence/absence) change from baseline as described for PVL and for the continuous covariates (genital tract WBC count and CD4 cell count) difference between each visit's value and the baseline value. Each covariate was included one at a time in a separate model. Only those genital tract infections with overall prevalence greater than 5% were analyzed.
The temporal dynamics of HIV-1 RNA between the plasma and genital tract were evaluated using Markov transition models . We examined whether a visit with detectable genital tract HIV-1 RNA was more likely to be preceded by a visit with detectable PVL and undetectable genital tract HIV-1 RNA or a visit with undetectable PVL and genital tract HIV-1 RNA, and whether a visit with detectable PVL was more likely to be preceded by a visit with detectable genital tract HIV-1 RNA and undetectable PVL or a visit with undetectable PVL and genital tract HIV-1 RNA. The independent variables in the two models were the same: detectable PVL (Y/N) and genital tract viral load in any subcompartment (Y/N) at the previous visit and their interaction, with adjustments made for currently adhering with HAART (Y/N), adherence with HAART at the previous visit (Y/N), age, race/ethnicity group, and taking a protease inhibitor-containing HAART regimen (Y/N).
All statistical analyses were conducted using R (version 2.9.0; Vienna, Austria, 2009). A P value below 0.05 was considered statistically significant.
Sixty-two women completed the initial screening; two women dropped out before the first study visit and one woman had one incomplete study visit only. The demographics and baseline clinical measures of the 59 remaining women who contributed a total of 582 study visits (median 12 visits) are presented in Table 1. At baseline, the median CD4 cell count was 458 cells/μl (range 120–1346). Almost all participants had below detectable PVL (95%) and genital tract viral load (98%). Half of the participants (49%) were on non-nucleoside reverse transcriptase inhibitor (NNRTI)-based HAART and 42% were on protease inhibitor-based HAART. Bacterial vaginosis was the most common lower genital tract infection (20%). All women were HSV-2 IgG antibody seropositive. Thirty-two of 59 women (54%) had detectable HIV-1 RNA at least once in the genital tract during the study period. Twenty-two of 59 (37%) women had detectable genital tract HIV-1 RNA during a study visit when PVL was undetectable. Five hundred and sixty-five of 582 (97%) visits had valid PVL and PVL was undetectable at 78% of these visits.
Results of participant-level analysis
Four women (6.8%) were classified as persistent shedders, 18 (31%) as intermittent shedders, and 27 (46%) as nonshedders (Table 2). There were 10 women (16.9%) of unknown shedding status. Six of these women provided 63 visits (on average 10.5 visits/woman) but none had three consecutive study visits with undetectable PVL. The remaining four women of unknown shedding status attended fewer than three study visits.
Comparison of the persistent and intermittent shedders with the nonshedders suggested that women with hysterectomies were less likely to be a persistent or intermittent shedder than a nonshedder (risk ratio 0.14, 95% CI 0.02–0.99). The remaining covariates were not found to be significantly different between shedders and nonshedders, yet some of the estimated differences in medians and risk ratios and their CIs do not exclude possibly meaningful differences between the groups.
Results of visit-level analysis
Among women with a uterus, 403 of 464 (87%) visits had valid genital tract samples from all three subcompartments. Fifty of these 403 visits (12%) had detectable genital tract viral load in at least one subcompartment (Fig. 1). Sampling three subcompartments increased detection of HIV-1 genital tract viral load compared to sampling a single subcompartment. No genital tract subcompartment (or pattern of subcompartments) appeared more likely to shed virus. It was more likely (72% of the visits) for just one subcompartment to have HIV shedding compared to more than one subcompartment (P < 0.05). The majority of women on HAART had undetectable genital tract viral load (Table 3, Fig. 2). The maximum quantity of HIV-1 RNA in the genital tract when PVL was undetectable was 456 000 copies/ml in the endocervix, 648 000 copies/ml in the ectocervix, and 480 000 copies/ml in the vagina. Women with hysterectomies provided 118 follow-up visits, 104/118 (88%) with valid measures of vaginal genital tract viral load (Table 3). The PVL for the visit with genital tract viral load of 2 720 000 copies/ml was 4000 copies/ml. The maximum quantity of HIV in the vagina among women with hysterectomies when PVL was undetectable was 68 000 copies/ml.
The estimated overall probabilities of genital tract HIV shedding were between 6 and 8% of visits in each of the three subcompartments (Table 4); genital tract HIV shedding in at least one subcompartment was estimated to occur at 13% of visits (95% CI 9–18%). Shedding in at least one of the three subcompartments occurred at 9% of visits when PVL was undetectable (95% CI 6–14%). The odds of shedding in each genital tract subcompartment and in at least one subcompartment were significantly increased when PVL was detectable. Among women without hysterectomies, the between-patient (or cross-sectional) odds ratio (OR) of shedding ranged from 4.3 for shedding in at least one subcompartment (95% CI 1.8–10.2) to 15.7 for shedding in the endocervix (95% CI 5.6–43.9). The within-patient (or longitudinal) OR of shedding ranged from 2.8 in the vagina (95% CI 1.2–6.3) to 5.6 in the endocervix (95% CI 2.6–12.3). Three genital tract infections were diagnosed at more than 5% of visits – Candida, HSV-2 DNA, and BV – and were examined as possible predictors of genital tract shedding; having any sexually transmitted infection (STI) was also examined. None of these covariates was found to be significantly positively associated with genital tract shedding by subcompartment or in any subcompartment (data not shown).
Analysis of genital tract and plasma dynamics
Genital tract viral load was significantly more likely to be detectable when PVL in the previous visit was detectable (OR 2.15, 95% CI 1.1–4.3). However, PVL was not significantly more likely to be detectable when genital tract viral load was detectable at the previous visit (OR 0.91, 95% CI 0.27–3.1).
The current study found that HIV-1 RNA can be detected in the genital tract of women receiving HAART, even if PVL remains undetectable. Over half of the women at some time-point had documented transient genital HIV-1 RNA shedding, which occurred in all genital tract subcompartments. To the best of our knowledge, this is the first longitudinal study to categorize HIV-1 RNA genital subcompartment shedding patterns in women with below-detectable PVL. A previous cross-sectional study by Kovacs et al.  also found genital tract HIV-1 RNA in over half the women. Neely et al.  found detectable genital tract HIV-1 RNA in 15% of women. An earlier study among men found that HIV-1 is primarily shed in an intermittent manner and shedding patterns in semen are related to compartmentalization between blood and plasma . These findings have implications for the low but not absolute lack of risk of sexual transmission of HIV-1 with well controlled PVL . Our observation that HIV-1 RNA is still present in the genital secretions of women with undetectable PVL, at times at very high copy numbers, suggests that the potential risk for sexual transmission remains .
Over a third of the women demonstrated discordance between PVL and genital tract viral load. When PVL was undetectable, significant amounts of virus could still be found in the genital tract ranging from 480 000 to 648 000 copies/ml. This finding is similar to earlier studies and may demonstrate viral compartmentalization in the blood and genital tract [9,11,25,26]. Despite the evidence of discordance between plasma and the genital tract, PVL still drove the presence of genital tract virus. Earlier studies have found that there is a 2.6 times greater odds of genital tract HIV-1 RNA shedding for each log10 unit increase in plasma HIV-1 RNA level, and that rebounds in HIV-1 RNA levels occur first in plasma or concurrently in the genital tract .
We examined viral shedding in all three genital subcompartments among women with a uterus. A persistent question has been whether a subcompartment of the female genital tract preferentially sheds virus . It has been suggested that the ectocervix is more conducive to HIV replication than is the endometrium . We found no genital tract subcompartment was more likely to shed virus. In almost three-quarters of the women, only a single subcompartment had evidence of viral shedding. If a woman did shed, she was most likely to shed intermittently. Havlir et al.  found that intermittent plasma viremia occurred frequently among patients with PVL less that 200 copies/ml, but this was not associated with virologic failure. In our study, no woman with below-detectable PVL with detectable genital tract shedding experienced virologic failure. Further studies need to be done to assess whether intermittent viremia in the genital tract leads to subsequent virologic failure in patients receiving HAART. Studies also need to assess whether the quantifiable HIV-1 RNA during these shedding episodes are from replicating virus or nonreplicating viral particles and whether they are infectious.
We are unable to predict when a woman with below detectable PVL will have genital tract HIV-1 RNA shedding. Due to the episodic, unpredictable nature of genital tract shedding, it will be difficult to establish a threshold for HIV transmission based on genital tract viral load among HAART-experienced women with below-detectable PVL. If samples had only been collected in a single subcompartment, many cases of detectable genital tract HIV-1 RNA would have been missed, emphasizing possible localization of HIV-1 RNA replication within genital tract subcompartments. Currently, there is a lack of consensus whether collecting samples in different subcompartments is necessary as differing results from each subcompartment could be driven by variation in sample collection and contamination, rather than underlying HIV shedding patterns. The current study documented robust levels of virus in each subcompartment. Detectable genital tract HIV-1 RNA varied widely within a subcompartment ranging from 3300 to 2.8 million copies/ml or less. Even among women who had undergone hysterectomy, vaginal HIV-1 RNA was quantifiable. One of these women had a PVL of only 4000 copies/ml, though she had a genital tract viral load of 2.7 million copies/ml.
The prevalence of genital tract infections was low in this study. No significant associations were documented with any of the tested STIs possibly due to the low prevalence of STIs in this cohort. Although studies have shown genital tract infections to up-regulate HIV shedding [29,30], few studies have been conducted among women with suppressed PVL on HAART. Our group identified the presence of genital tract white blood cells as increasing the risk of detectable genital tract viral load among primarily ART-naive women, but individual infections were not independently associated with increased viral load shedding . Inflammatory processes rather than infections may drive HIV-1 viral shedding.
There are several limitations to the study. This study only collected cell-free rather than also cell-associated virus and analyzed only RNA, not proviral DNA. Also, the median age of the women was 45 years, and hence these women may not be a representative sample of HIV-infected women. The prevalence of STIs was low so it was not possible to evaluate temporal patterns of genital HIV-1 shedding with STIs. Despite the fact that all women were HSV-2 antibody-positive, few women (5%) had detectable HSV-2 DNA at baseline and between 4 and 9% of women had detectable HSV-2 DNA at each of the follow-up visits. It is possible that we may have missed low levels of HIV genital tract shedding (i.e. <3300 copy range) due to assay sensitivity and differences in cut-off between peripheral blood and genital viral loads. Strengths of the current study include longitudinal data collection with a median of 12 follow-up visits per woman over a 1-year period, and inclusion of women who had undergone hysterectomy.
The Swiss Federal Commission for HIV/AIDS has suggested that seropositive individuals with no STI and on suppressive ART for greater than 6 months do not transmit HIV . Stürmer et al.  reported HIV-1 transmission in a serodiscordant couple despite successful ART of the HIV-infected partner. Frozen blood samples were analyzed phylogenetically from the HIV-1-positive and the newly infected partner before treatment and shortly after seroconversion, respectively, and showed a true relationship . Despite the long-term immune restorative effects of HAART and the impact of HAART in markedly reducing viral load, the current study found that women could continue to shed virus in their genital tract secretions. Whereas genital tract shedding is primarily driven by plasma viremia, clinicians may not be able to solely rely on HAART to eradicate the potential for the sexual and perinatal transmission of HIV. Further studies are needed to examine whether a threshold of genital tract viral load exists for sexual transmission as well as possible surrogate markers for transmission potential. The findings of the present study add to the growing evidence that in the HAART era, women with below-detectable PVL may have less risk of HIV sexual transmission on a population level, but may continue to be possibly infectious on an individual level.
The work was supported in part by National Institutes of Health (NIH) RO1 AI40350: K24 AI066884; Lifespan/Tufts/Brown Center for AIDS Research (P30AI42853); and Emory Center for AIDS Research (P30 AI050409).
Each author was a co-investigator for the project and had intellectual input in the study design, data analyses and had significant input in the final manuscript. Jess Ingersoll and Jackie Kurpewski in particular processed specimens, run the assays, troubleshooted lab problems and contributed to the interpretation of the results.
1. Quinn TC, Wawer MJ, Sewankambo N, Serwadda D, Li C, Wabwire-Mangen F, 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. Mofenson L, Lambert JS, Stiehm ER, Bethel J, Meyer WA 3rd, Whitehouse J, et al
. Risk factors for perinatal transmission of human immunodeficiency virus type 1 in women treated with zidovudine. Pediatric AIDS
Clinical Trials Group Study 185 Team. N Engl J Med 1999; 341:385–393.
3. Kalichman S, Di Berto G, Eaton L. Human immunodeficiency virus viral load
in blood plasma and semen: review and implications of empirical findings. Sexually Transm Dis 2008; 35:55–60.
4. Cu-Uvin S, Caliendo AM, Reinert S, Chang A, Juliano-Remollino C, Flanigan TP, et al
. Effect of highly active antiretroviral therapy on cervicovaginal HIV
-1 RNA. AIDS
5. Cu-Uvin S, Snyder B, Harwell JI, Hogan J, Chibwesha C, Hanley D, et al
. Association between paired plasma and cervicovaginal lavage fluid HIV
-1 RNA levels during 36 months. J Acquir Immune Defic Syndr 2006; 42:584–587.
6. Dornadula G, Zhang H, VanUitert B, Stern J, Livornese L Jr, Ingerman MJ, et al
. Residual HIV
-1 RNA in blood plasma of patients taking suppressive highly active antiretroviral therapy. JAMA 1999; 282:1627–1632.
7. Fiore JR, Suligoi B, Saracino A, Di Stefano M, Bugarini R, Lepera A, et al
. Correlates of HIV
-1 shedding in cervicovaginal secretions and effects of antiretroviral therapies. AIDS
8. Sullivan S, Mandava U, Evans-Strickfaden T, Lennox JL, Ellerbrock TV, Hart CE. Diversity, divergence, and evolution of cell-free human immunodeficiency virus type 1 in vaginal secretions and blood of chronically infected women: associations with immune status. J Virol 2005; 79:9799–9809.
9. Kovacs A, Wasserman SS, Burns D, Wright DJ, Cohn J, Landay A, et al
, DATRI Study Group; WIHS Study Group. Determinants of HIV
-1 shedding in the genital tract
of women. Lancet 2001; 358:1593–1601.
10. Craigo J, Patterson BK, Paranjpe S, Kulka K, Ding M, Mellors J, et al
. Persistent HIV
type 1 infection in semen and blood compartments in patients after long-term potent antiretroviral therapy. AIDS
Res Hum Retrovirus 2004; 20:1196–1209.
11. Neely MN, Benning L, Xu J, Strickler HD, Greenblatt RM, Minkoff H, et al
. Cervical shedding of HIV
-1 RNA among women with low levels of viremia while receiving highly active antiretroviral therapy. J Acquir Immune Defic Syndr 2007; 44:38–42.
12. Cu-Uvin S, Hogan JW, Caliendo AM, Harwell J, Mayer KH. Carpenter CC; HIV
Epidemiology Research Study. Association between bacterial vaginosis and expression of human immunodeficiency virus type 1 RNA in the female genital tract
.1. Clin Infect Dis 2001; 33:894–896.
13. Wang C, McClelland RS, Overbaugh J, Reilly M, Panteleeff DD, Mandaliya K, et al
. The effect of hormonal contraception on genital tract
shedding of HIV
14. Clemetson D, Moss GB, Willerford DM, Hensel M, Emonyi W, Holmes KK, et al
. Detection of HIV
DNA in cervical and vaginal secretions. Prevalence and correlates among women in Nairobi, Kenya. JAMA 1993; 269:2860–2864.
15. Anderson B, Wang CC, Delong AK, Liu T, Kojic EM, Kurpewski J, et al
. Genital tract
leukocytes and shedding of genital HIV
type 1 RNA. Clin Infect Dis 2008; 47:1216–1221.
16. Pasquier C, Moinard N, Sauné K, Souyris C, Lavit M, Daudin M, et al
. Persistent differences in the antiviral effects of highly active antiretroviral therapy in the blood and male genital tract
17. Sungkanuparph S, Overton ET, Seyfried W, Groger RK, Fraser VJ, Powderly WG. Intermittent episodes of detectable HIV
viremia in patients receiving nonnucleoside reverse-transcriptase inhibitor-based or protease inhibitor-based highly active antiretroviral therapy regimens are equivalent in incidence and prognosis. Clin Infect Dis 2005; 41:1326–1332.
18. Leruez-Ville M, Dulioust E, Costabliola D, Salmon D, Tachet A, Finkielsztejn L, et al
. Decrease in HIV
-1 seminal shedding in men receiving highly active antiretroviral therapy: an 18 month longitudinal study (ANRS EP012). AIDS
19. Wilson DP,Law MG, Grulich AE, Cooper DA, Kaldor JM. Relation between HIV viral load and infectiousness: a model-based analysis. Lancet
20. Cohen M, Gay C, Kashuba AD, Blower S, Paxton L. Narrative review: antiretroviral therapy to prevent the sexual transmission of HIV
-1. Ann Internal Med 2007; 146:591–601.
21. Amsel R, Totten PA, Spiegel CA, Chen KC, Eschenbach D, Holmes KK. Nonspecific vaginitis. Diagnostic criteria and microbial and epidemiologic associations. Am J Med 1983; 74:14–22.
22. Diggle P, Heagerty P, Liang K, Zeger S. Analysis of longitudinal data
. 2nd ed. Oxford University Press; 2002.
23. Gupta P, Leroux C, Patterson BK, Kingsley L, Rinaldo C, Ding M, et al
. Human immunodeficiency virus type 1 shedding pattern in semen correlates with the compartmentalization of viral Quasi species between blood and semen. J Infect Dis 2000; 182:79–87.
24. Stürmer M, Doerr HW, Berger A, Gute P. Is transmission of HIV
-1 in nonviraemic serodiscordant couples possible? Antivir Ther 2008; 13:729–732.
25. Andreoletti L, Skrabal K, Perrin V, Chomont N, Saragosti S, Gresenguet G, et al
. Genetic and phenotypic features of blood and genital viral populations of clinically asymptomatic and antiretroviral-treatment-naive clade a human immunodeficiency virus type 1-infected women. J Clin Microbiol 2007; 45:1838–1842.
26. Chomont N, Hocini H, Grésenguet G, Brochier C, Bouhlal H, Andréoletti L, et al
. Early archives of genetically-restricted proviral DNA in the female genital tract
after heterosexual transmission of HIV
27. Asin S, Eszterhas SK, Rollenhagen C, Heimberg AM, Howell AL. HIV
type 1 infection in women: increased transcription of HIV
type 1 in ectocervical tissue explants. J Infect Dis 2009; 200:965–972.
28. Havlir DV, Bassett R, Levitan D, Gilbert P, Tebas P, Collier AC, et al
. Prevalence and predictive value of intermittent viremia with combination HIV
therapy. JAMA 2001; 286:171–179.
29. McClelland R, Wang CC, Overbaugh J, Richardson BA, Corey L, Ashley RL, et al
. Association between cervical shedding of herpes simplex virus and HIV
30. Cohen M, Hoffman IF, Royce RA, Kazembe P, Dyer JR, Daly CC, et al
. Reduction of concentration of HIV
-1 in semen after treatment of urethritis: implications for prevention of sexual transmission of HIV
-1. AIDSCAP Malawi Research Group. Lancet 1997; 349:1868–1873.
31. Vernazza P, Hirschel B, Bernasconi E, Flepp M. Les personnes seropositives ne souff rant d'aucune autre MST et suivant un traitment antiretroviral effi cie ne transmettent pas le VIH par voie sexuelle. Bull Med Suisses 2008; 89:165–169.