Highly active antiretroviral therapy (HAART) has resulted in a significant decrease in HIV replication in the blood, and has significantly improved the clinical outcome among HIV-infected patients [1,2]. Effective antiretroviral therapy has also reduced HIV-1 RNA levels and culturable virus in the semen [3–5].
HIV-1 RNA or proviral HIV-1 DNA have been isolated from the female genital tract [6–10]. Those studies have identified significant correlates of genital tract shedding, including plasma viral load, CD4 cell count, pregnancy, hormonal contraception, genital tract infections/inflammation, cervical ectopy and vitamin A deficiency [6–11].
More recent studies have shown a positive correlation between female genital tract HIV shedding and plasma viral load, and have demonstrated that the initiation of antiretroviral therapy reduces the amount of HIV-1 RNA in the female genital tract [12,13]. Most of the studies involved a limited number of women.
Understanding the effect of antiretroviral therapy on viral shedding in the female genital tract is important for the development of strategies to prevent heterosexual as well as mother-to-infant transmission of HIV.
The relationship of cervicovaginal lavage HIV-1 RNA level to plasma viral load and CD4 cell count was investigated. The frequency of cervicovaginal lavage and plasma HIV-1 RNA levels below detectable levels (< 400 copies/ml) were determined in women on no therapy, on non-HAART therapy and on HAART regimens. The effect of initiating HAART on the timing of HIV-1 RNA suppression in the blood plasma and genital tract was also compared in antiretroviral-naïve women.
The study population consisted of 205 HIV-positive women who receive medical care at the Miriam Hospital Immunology Center, Providence, RI, USA. The study was approved by the Miriam Hospital Institutional Review Board. All the women gave written informed consent. Seventy-nine of the women were participants in a longitudinal study of HIV disease progression in women, the HIV Epidemiology Research Study (HERS). The women were interviewed about their demographic characteristics, medical history, and sexual and reproductive history. They were advised not to have sexual intercourse, to douche, or to insert any vaginal products at least 48 h before sample collection. All the women underwent a microscopic examination of cervicovaginal secretions to diagnose Candida infection, trichomoniasis and bacterial vaginosis. Paired plasma and cervicovaginal lavage specimens were collected from each woman. None of the cervicovaginal lavage specimens were collected during menses. Cervicovaginal lavage was done by instilling 10 cc of normal saline solution directed at the cervical os and then aspirating after a few seconds.
Treatment with two nucleosides with one or more protease inhibitors or a non-nucleoside reverse transcriptase inhibitor was defined as HAART. Treatment with nucleosides alone was considered to be non-HAART therapy.
Seven antiretroviral-naïve women starting HAART had paired plasma and cervicovaginal lavage viral load measured daily for one week, at 2 weeks and again at 1 month after initiating therapy. CD4 cell counts were performed using standard techniques; all were within 6 months of the plasma and cervicovaginal lavage collection (49% were performed on the same day).
A commercially available technique for quantification was used to measure HIV-1 RNA in both the plasma and cervicovaginal lavage (nucleic acid sequence-based amplification assay, NASBA; Organon Teknika Corp., Durham, NC, USA). This method is isothermal and amplifies only HIV-1 RNA and not proviral DNA. Blood specimens were anticoagulated in ethylenediamine tetraacetic acid, and plasma was separated within 1 h of collection. An aliquot of 200 μl or 1 ml of unfractionated cervicovaginal lavage was used for HIV-1 RNA quantification. The plasma HIV-1 RNA viral load was measured using 100 μl or 1 ml of plasma. Cervicovaginal lavage and plasma were added to lysis buffer within 3 h of collection and stored at −70°C. Specimens were shipped in dry ice to Massachusetts General Hospital, and stored at −70°C until they were run for HIV-1 RNA testing.
Results were expressed as copies/ml of plasma or cervicovaginal lavage. The lower limit of detection was 400 copies/ml.
Both univariate and multivariate logistical models were used to assess the contribution of different factors on the detection of HIV-1 RNA in cervicovaginal secretions. The factors included plasma viral load, CD4 cell count and antiretroviral therapy. The association between plasma viral load levels and genital tract viral load levels was assessed by the Spearman rank correlation test.
The demographic information from 205 HIV-infected women in this cross-sectional analysis is summarized in Table 1. The majority of the women were white and had a history of injection drug use. The median CD4 cell count was 354/mm3 with a range of 4–1862/mm3. One third of the women were on HAART.
Plasma HIV-1 RNA was detectable in 146 out of 205 women (71%). The median plasma viral load was 19 000 copies/ml, with a range of 420–3 000 000 copies/ml. Fifty-three of the subjects (26%) had detectable HIV-1 RNA in cervicovaginal lavage with a median of 11 000 copies/ml and a range of 450–1 200 000 copies/ml.
Two women (1%) had detectable HIV-1 RNA in cervicovaginal lavage but not in the plasma. One woman had a cervicovaginal viral load of 15 000 copies/ml while on HAART, and the other had a genital tract viral load of 3200 copies/ml on didanosine monotherapy.
Fifty-one of the 205 women (25%) had HIV-1 RNA detected in both plasma and genital tract secretions, and 95 out of 205 (46%) had virus detected in the plasma but not in the genital tract. Fifty-seven out of 205 (28%) had no detectable HIV-1 RNA in either the plasma or genital tract. Among women with detectable plasma HIV-1 RNA, 35% had detectable cervicovaginal lavage HIV-1 RNA. Among women with no detectable plasma HIV-1 RNA, 97% had no detectable genital tract HIV-1 RNA. A Spearman rank correlation test showed a relationship between the levels of HIV-1 RNA in cervicovaginal lavage and the levels of HIV-1 RNA in the plasma (r = 0.42;P < 0.001).
Cervicovaginal HIV-1 RNA detection was significantly correlated with plasma HIV-1 RNA levels and negatively with CD4 cell counts (Fig. 1a and b). Among women with plasma viral loads of over 10 000 copies/ml, 48% had detectable HIV-1 RNA in the genital tract (P < 0.001). Fifty-one per cent of the women with CD4 cell counts of less than 200/mm3 had detectable HIV-1 RNA in the genital tract (P < 0.001).
A significant correlation was found between the detection of HIV-1 RNA in both blood and cervicovaginal lavage and the type of antiretroviral therapy (Fig. 2). Among those on HAART, the plasma viral load was less than 400 copies/ml in 35 out of 67 (52%) women compared with 13 out of 47 (18%) and 11 out of 64 (17%) women on no therapy and on non-HAART therapy, respectively (P < 0.001). Cervicovaginal HIV-1 RNA was less than 400 copies/ml in 51 out of 67 (85%) women on HAART compared with 51 out of 74 (69%) on no therapy and 44 out of 64 (69%) on non-HAART therapy (P < 0.045).
Twenty-two out of 205 (9%) women had cervicovaginal lavage HIV-1 RNA greater than the plasma viral load. For these women, the genital tract HIV-1 RNA level was 1.2 to 75-fold higher than plasma HIV-1 RNA (10/22 were at least five times higher). Cervicovaginal viral load ranged from 3900 to 1 200 000 copies/ml. Of these women, eight were infected heterosexually and 14 through injection drug use.
There were two cases of Candida vaginitis (culture positive for Candida species with abnormal vaginal discharge and either erythema, edema of the vulva or vagina, or pruritus) with cervicovaginal viral loads of less than 400 copies/ml. There were 12 cases of bacterial vaginosis (Amsel's criteria) but only two had detectable cervicovaginal viral loads. These women had high levels of plasma viral load (110 000 and 32 000 copies/ml). No patient was positive for trichomoniasis.
On multivariate analysis, the strongest predictor for the detection of HIV-1 RNA in the genital tract was a plasma viral load of over 10 000 copies/ml [odds ratio (OR) 5.8, 95% confidence interval (CI) 2.7–12.5]. A low CD4 cell count (< 200/mm3) was also associated with genital tract HIV shedding (OR 3.5, 95% CI 1.6–7.8). Age, race, and mode of transmission did not correlate with the detection of HIV-1 RNA in the genital tract.
Seven antiretroviral-naïve women who initiated therapy with two nucleoside reverse transcriptase inhibitors plus a protease inhibitor or a non-nucleoside reverse transcriptase inhibitor had intensive monitoring of plasma and genital tract HIV-1 RNA suppression. Three women started on zidovudine or stavudine with lamivudine and indinavir; three started on zidovudine or stavudine with lamivudine and nelfinavir; and one woman initiated therapy with zidovudine, lamivudine and efavirenz. Their CD4 cell counts ranged from 5 to 476/mm3 (median 65/mm3). Baseline plasma viral loads ranged from 25 000 to 780 000 copies/ml, and cervicovaginal viral loads ranged from less than 400 to 51 000 copies/ml. Three women had cervicovaginal viral loads of over 400 copies/ml at baseline. For these seven women, there was a 1.8–3.3 log10 decrease in the plasma viral load after initiating therapy. In all three women with detectable genital tract HIV-1 RNA, there was a 0.7–2.1 log10 drop in genital tract viral loads. Cervicovaginal HIV-1 RNA levels dropped to below detectable levels within 1–14 days of initiating HAART. The decrease in genital tract viral load may occur before the drop in plasma viral load (Fig. 3).
In this study, we determined that genital tract HIV-1 RNA was correlated positively with plasma HIV-1 RNA and negatively with the CD4 cell count. The use of HAART was significantly associated with below detectable levels of HIV-1 RNA in both plasma and cervicovaginal secretions. HIV-1 RNA suppression in the genital tract occurred rapidly within 2 weeks of initiating HAART among antiretroviral-naïve women. HIV-1 suppression in the genital tract may occur before suppression in blood plasma. These results indicated that the presence of higher levels of HIV-1 RNA in the genital tract compared with blood plasma had no correlation with the mode of HIV acquisition.
The study has several limitations. The use of cervicovaginal lavage does not allow for analysis by cell-associated versus cell-free virus. Proviral DNA in genital secretions was not measured. The quantity of infected cells could be an important factor of variation in cervicovaginal lavage HIV-1 RNA levels. Shaheen et al.  detected HIV-1 in cellular and acellular fractions of cervicovaginal lavage specimens. In their study, the plasma viral load showed significant correlation with the detection of HIV-1 RNA in cervicovaginal lavage and a less apparent correlation with viral DNA in cervical cells. Another study by Loussert-Ajaka et al. , however, found that HIV-1 RNA detected in cervicovaginal secretions was correlated with DNA detection in the same sample and with HIV-1 RNA detection in the plasma. The dilution caused by the lavage technique may affect the ability to detect virus compared with other techniques such as cervical canal fluid adsorbed by Sno-strip. None of the specimens were collected during menses and blood-contaminated samples were excluded; occult blood contamination was not, however, measured by quantitative techniques. HIV-1 RNA was performed with a lower limit of detection of 400 copies/ml. The results obtained may be different with the use of an ultrasensitive assay (50 copies/ml).
Several studies [14,15] have shown a correlation between female genital tract HIV viral load and blood plasma viral load, except for the study by Rasheed et al. . Recently, Hart et al.  analysed results from 52 women and demonstrated significant correlations of HIV-1 RNA levels in blood plasma with those in vaginal secretions and cervical mucus. Their study showed that the amount of cell-free HIV RNA in blood plasma was correlated with that in vaginal secretions (r = 0.64;P < 0.001). We found similar results among 205 women (r = 0.42;P < 0.001). This correlation could be interpreted as evidence that blood and the female genital tract are not separate compartments. However, the detection of HIV-1 RNA in the genital tract in women with below-detectable levels of the virus in the blood argues against a single compartment theory. Shaheen et al.  showed that in most patients there is a high degree of similarity observed between viral sequences derived from the blood and cervicovaginal lavage. However, two patients had closely related but somewhat distinct genotype variants in the blood and cervicovaginal lavage. One subject showed clear compartmentalization in which distinct viral genotypes were observed in the blood and cervicovaginal lavage. In men, there is evidence for compartmentalization. Differences have been observed in phenotype and genotype between blood and semen . Coombs et al.  found support for viral compartmentalization between semen and blood on the basis of a lack of association between the culturability of virus in semen and the viral RNA level in the blood, discordant distribution of viral phenotypes, discordant viral RNA levels, and differences in the biological variability of viral RNA levels. Differences in HIV resistance patterns between blood and semen have also been noted . Only the analysis of the homology of HIV species present in different compartments such as blood, semen, female genital tract secretions and cerebrospinal fluid (performed on a clonal basis either on env or pol genes) will allow researchers to establish whether these are truly separate compartments.
Combination antiretroviral therapy has produced marked and durable suppression of viral replication in the blood , and this is a major advance in the treatment of HIV infection . If effective suppression of viral load occurs in both the plasma and genital tract secretions, this may have an impact on the sexual spread of HIV as well as in mother-to-infant transmission. Recent data from the Women and Infants Transmission Study and the AIDS Clinical Trials Group (ACTG) 185 confirmed that the level of plasma HIV-1 RNA was the best predictor of the risk of perinatal transmission [20,21]. Although transmission was zero in those with less than 1000 copies/ml of plasma viral load, other studies [22,23] have reported perinatal transmission among women with below-detectable plasma viral loads. There is no defined threshold for plasma viral load below which there is no risk of transmission. Those studies did not assess the level of virus in the genital tract during pregnancy.
The plasma viral load can be reduced by more than 2 logs by effective combination antiretroviral therapy . Vernazza et al.  noted that antiviral treatment resulted in a significant fall in the viral load in semen that paralleled the viral reduction in the blood. The study by Hart et al.  reported that new treatment regimens resulted in a 0.5–2.1 log10 decrease in the viral load in vaginal lavage samples and on follow-up, 71% of their patients had undetectable vaginal lavage virus. Our data showed that 85% of women on HAART had undetectable cervicovaginal viral load compared with 69% on non-HAART and no therapy. Among antiretroviral-naïve women who started HAART and had daily paired plasma and cervicovaginal viral loads monitored, there was a 1.8–3.3 log10 decrease in the plasma viral load paralleled by a 0.7–2.1 log10 decrease in the genital tract viral load. In three women, the fall to undetectable levels in the cervicovaginal secretions preceded that in the plasma.
Several studies have reported the presence of virus in seminal cells despite receiving HAART [25,26]. HIV variants with genotypic resistance markers have been found in semen. Mayer et al. recently reported on HIV-1 protease inhibitor resistance mutants in the semen of patients receiving combination therapy. Similar studies need to be performed in women. The transmission of resistant HIV variants is an important public health concern, because it can undermine the advances that have been possible with effective antiretroviral therapy in the prevention of the spread of HIV infection.
1. Hammer SM, Squires KE, Hughes MD. et al. A controlled trial of two nucleoside analogues plus indinavir in persons with human immunodeficiency virus infection and CD4 counts of 200 per cubic millimeter or less.
:AIDS Clinical Trials Group 320 Study Team.
N Engl J Med 1997, 337: 725 –733.
2. Carpenter CCJ, Fischl MA, Hammer S. et al. Antiretroviral therapy for HIV infection in 1997: recommendations of an international panel.
JAMA 1997, 277: 1962 –1969.
3. Vernazza PL, Gilliam BL, Flepp M. et al. Effect of antiviral treatment on the shedding of HIV-1 in semen.
AIDS 1997, 11: 1249 –1254.
4. Gupta P, Mellors J, Kingsley L. et al. High viral load in semen of human immunodeficiency virus type 1 infected men at all stages of disease and its reduction by therapy with protease and nonnucleoside reverse transcriptase inhibitors.
J Virol 1997, 71: 1 –5.
5. Gilliam BL, Dyer JR, Fiscus SA. et al. Effects of reverse transcriptase inhibitor therapy on the HIV-1 burden in semen.
J Acquired Immune Defic Syndr Hum Retrovirol 1997, 15: 54 –60.
6. Kreiss JK, Coombs R, Plummer F. et al. Isolation of human immunodeficiency virus from genital ulcers in Nairobi prostitutes.
J Infect Dis 1989, 160: 380 –384.
7. Henin Y, Mandelbrot L, Henrion R, Pradinaud R, Coulaud JP, Montagnier L. Virus excretion on cervicovaginal secretions of pregnant and non-pregnant HIV-infected women.
J Acquired Immune Defic Syndr 1993, 6: 72 –75.
8. Goulston C, McFarland W, Katzenstein D. Human immunodeficiency viruses type 1 RNA in the female genital tract.
J Infect Dis 1998, 177: 110 –113.
9. Kreiss J, Willerford DM, Hensel M. et al. Association between cervical inflammation and cervical shedding of human immunodeficiency virus DNA.
J Infect Dis 1994, 170: 1597 –1601.
10. John GC, Nduati RW, Mbori-Ngacha D. et al. Genital shedding of human immunodeficiency virus type 1 DNA during pregnancy: association with immunosuppression, abnormal cervical or vaginal discharge, and severe vitamin A deficiency.
J Infect Dis 1997, 175: 57 –62.
11. Iversen AKN, Larsen AR, Jensen T. et al. Distinct determinants of human immunodeficiency virus type 1 RNA and DNA loads in vaginal and cervical secretions.
J Infect Dis 1998, 177: 1214 –1220.
12. Shaheen F, Sison AV, McIntosh L, Mukhtar M, Pomerantz RJ. Analysis of HIV-1 in the cervicovaginal secretions and blood of pregnant and nonpregnant women.
J Hum Virol 1999, 2: 154 –166.
13. Loussert-Ajaka I, Mandelbrot L, Delmas MC. et al. HIV-1 detection in cervicovaginal secretions during pregnancy.
AIDS 1997, 22: 1575 –1581.
14. Cu-Uvin S, Caliendo AM. Cervicovaginal human immunodeficiency virus secretion and plasma viral load in human immunodeficiency virus-seropositive women.
Obstet Gynecol 1997, 90: 739 –743.
15. Hart CE, Lennox JL, Pratt-Palmore M. et al. Correlation of human immunodeficiency virus type 1 RNA levels in blood and the female genital tract.
J Infect Dis 1999, 179: 871 –882.
16. Rasheed S, Zhiliang L, Dong X, Kovac A. Presence of cell-free human immunodeficiency virus in cervicovaginal secretions is independent of viral load in the blood of human immunodeficiency virus-infected women.
Am J Obstet Gynecol 1996, 175: 122 –130.
17. Zhu T, Wang N, Carr A. et al. Genetic characterization of human immunodeficiency virus type 1 in blood and genital secretions: evidence for viral compartmentalization and selection during sexual transmission.
J Virol 1996, 70: 3098 –3107.
18. Coombs RW, Speck CE, Hughes JP. et al. Association between culturable human immunodeficiency virus type 1 (HIV-1) in semen and HIV-1RNA levels in semen and blood: evidence for compartmentalization of HIV-1 between semen and blood.
J Infect Dis 1998, 177: 320 –330.
19. Eron JJ, Vernazza PL, Johnston DM. et al. Resistance to HIV-1 to antiretroviral agents in blood and seminal plasma: implications for transmission.
AIDS 1998, 12: F181 –F189.
20. Garcia PM, Kalishi LA, Pitt J. et al. Maternal levels of plasma human immunodeficiency virus type 1 RNA and the risk of perinatal transmission.
N Engl J Med 1999, 341: 394 –402.
21. Mofenson LM, Lambert JS, Stiehm ER. et al. Risk factors for perinatal transmission of human immunodeficiency virus type 1 in women treated with zidovudine.
N Engl J Med 1999, 341: 385 –393.
22. Cao Y, Krogstad P, Korber BT. et al. Maternal HIV-1 viral load and vertical transmission of infection: the Ariel Project for the prevention of HIV transmission from mother to infant.
Nat Med 1997, 3: 549 –552.
23. Sperling RS, Shapiro DE, Cooms RW. et al. Maternal viral load, zidovudine treatment, and the risk of transmission of human immunodeficiency virus type 1 from mother to infant.
N Engl J Med 1996, 335: 1621 –1629.
24. Gulick RM, Mellors JW, Havlir D. et al. Treatment with indinavir, zidovudine, and lamivudine in adults with human immunodeficiency virus infection and prior antiretroviral therapy.
N Engl J Med 1997, 337: 734 –739.
25. Zhang H, Dornadulla G, Beumont M. et al. Human immunodeficiency virus type 1 in the semen of men receiving highly active antiretroviral therapy.
N Engl J Med 1998, 339: 1803 –1809.
26. Mayer KH, Boswell S, Goldstein R. et al. Persistence of human immunodeficiency virus in semen after adding indinavir to combination antiretroviral therapy.
Clin Infect Dis 1999, 28: 1252 –1259.