Comparison of Cervicovaginal Lavage, Cervicovaginal Lavage Enriched With Cervical Swab, and Vaginal Tampon for the Detection of HIV-1 RNA and HSV-2 DNA in Genital Secretions

Delany, Sinéad MD, PhD*; Rosas, Raul MSc†; Mlaba, Nonkululeko MD*; Clayton, Tim MSc†; Akpomiemie, Godspower MPH*; LeGoff, Jérôme MD, PhD‡; Capovilla, Alexio PhD§; Bélec, Laurent MD, PhD‡; Stevens, Wendy PhD§; Mayaud, Philippe MD, MSc†

JAIDS Journal of Acquired Immune Deficiency Syndromes: 1 December 2008 - Volume 49 - Issue 4 - pp 406-409
doi: 10.1097/QAI.0b013e31818c7f75
Brief Report: Clinical Science

Methods: We compared the performance of 3 collection methods for cervicovaginal secretions [cervicovaginal lavage (CVL), CVL enriched with a cervical swab (eCVL), and vaginal tampon (VT)] to identify the most reliable method for detection of cervicovaginal HIV-1 and herpes simplex virus type 2 (HSV-2). HIV-1 RNA (Nuclisens EasyQ; BioMerieux, Marcy-l'Etoile, France), HSV-2 DNA (real-time polymerase chain reaction), and microscopic blood and semen traces were detected in samples from 19 HIV-1-HSV-2-coinfected women seen at 4 weekly visits.

Results: HIV-1 RNA was detected in 49 (79%) of 62 eCVLs, 41 (61%) of 67 CVLs, and 27 (57%) of 47 VTs. Detection of HIV-1 RNA was higher in eCVL compared with CVL [45/58 (78%) vs. 32/58 (55%); risk ratio 1.41, 95% confidence interval 1.05 to 1.88].

Conclusions: Although more eCVLs were contaminated with microscopic blood (29%) than CVLs (22%) or VTs (7%), detection of HIV-1 RNA remained higher using eCVL compared with CVL (risk ratio 1.43, 95% confidence interval 1.02 to 2.02) in uncontaminated samples. HSV-2 DNA was detected in less than 10% of samples by each method but in 7 (37%) of 19 women overall by 1 or more methods.

From the *Reproductive Health and HIV Research Unit, University of the Witwatersrand, Johannesburg, South Africa; †London School of Hygiene & Tropical Medicine, London, United Kingdom; ‡Université Paris V, Equipe Immunité et Biothérapie Muqueuse, Unité INSERM Internationale U743 (Immunologie Humaine), Centres de Recherches Biomédicales des Cordeliers and Laboratoire de Virologie, Hôpital Européen Georges Pompidou, Paris, France; and §Department of Haematology and Molecular Biology, University of the Witwatersrand, Johannesburg, South Africa.

Received for publication March 13, 2008; accepted August 29, 2008.

The authors do not report any conflict of interest.

Correspondence to: Sinéad Delany-Moretlwe, MD, PhD, Reproductive Health and HIV Research Unit, University of the Witwatersrand, PO Box 18512, Hillbrow, Johannesburg 2038, South Africa (e-mail:

Article Outline

Previous studies examining the role of HIV-1 or herpes simplex virus type 2 (HSV-2) dynamics using cervicovaginal secretions (CVS) have shown great variability in the detection and quantification of both HIV-1 RNA and HSV-2 DNA, depending on anatomic site of collection, method of specimen collection, frequency of collection, specimen storage, and specimen testing. Comparisons between studies are therefore difficult because these measurements have not been well standardized.1,2 Sampling methods for the collection of CVS have involved endocervical swabs or cervicovaginal lavage (CVL). CVL may have the advantage of increased sampling surface area and collection of a large sample volume, which can be fractionated for various analyses.3,4 However, accurate quantification of virus may be more problematic in CVL samples because of the inherent difficulties with volume standardization coupled with the possible dilution of virus levels.5 Cervical swabs, by contrast, allow localized sampling of HIV-1 shed from the cervix but may lead to blood contamination of samples because of mucosal disruption, resulting in possible misclassification of shedding results.6,7 Recently, these 2 sample collection modalities were combined (CVL followed by endocervical swab) by investigators in Burkina Faso in a technique they called “enriched CVL” (eCVL) to detect and quantify both HIV-1 RNA and HSV-2 DNA in a single specimen with seemingly high sensitivity.8 Self-administered vaginal tampons (VTs) have previously been shown to be useful for the detection of nonulcerative sexually transmitted infections9,10 and HIV-1 RNA11 and may provide a useful alternative to clinician-administered methods in studies that would require frequent sampling.

Paired comparisons of some of these CVS collection methods have been made previously,4,7,12-14 but the methods have not been compared for the purpose of simultaneous detection of genital HIV-1 and HSV-2 shedding, nor did the comparison studies use a longitudinal design, which has the advantage of allowing for assessment of the reliability of a method and investigation of the intermittent nature of genital viral shedding. We compared the performance of different methods of CVS collection to identify the most reliable and practical method for detection of both genital HIV-1 RNA and HSV-2 DNA with the lowest risk of blood contamination, which could be used for the assessment of viral outcomes in a clinical trial.

Back to Top | Article Outline


Nonpregnant, HIV-1-seropositive women aged 18 years and above attending an HIV support group in Johannesburg, South Africa, were invited to participate in the study, which took place between August and October 2005. Women providing signed informed consent were interviewed using a questionnaire, and physical examination was performed for clinical staging of HIV/AIDS.15 HIV serostatus was confirmed using a rapid test on fingerprick blood (Abbott Determine HIV 1/2; Abbott Diagnostics, Wiesbaden, Germany). A venous blood sample was drawn for HSV-2 serologic testing (HerpeSelect; Focus Diagnostics, Cypress, CA), defining antibody index values ≥3.5 as positive,16 and CD4 count (FlowCARE PLG CD4; Beckman Coulter, Hialeah, FL). Eligible participants, with antibodies to HSV-2 and HIV-1 and classified in World Health Organization HIV/AIDS stages I or II, were seen at consecutive weekly visits for a total of 4 visits, and genital samples were collected at each visit. Visits, which fell during the menstrual period, were deferred by a few days until the completion of menses. Participants were counseled to avoid vaginal cleansing on the day of their clinic visits and unprotected sexual intercourse for 72 hours before (or to use condoms) to avoid contamination of samples with semen. Participants with sexually transmitted infection symptoms and signs at enrollment visit were treated using national syndromic management guidelines, which did not include acyclovir in the case of genital ulcer disease (GUD).

One sterile Dacron endocervical swab was applied to the cervical os and rotated 360 degrees, avoiding any mucosal trauma. A lavage was performed using 6 mL of phosphate-buffered saline (PBS) for 60 seconds, using a technique that has been previously described.5 Half of the reaspirated volume was collected in a sterile tube for CVL and half in a separate tube for the preparation of eCVL into which the cervical swab was agitated, then discarded. The order of CVS sample collection was alternated. At 2 of the visits, the swabs were collected first, whereas at the remaining visits, the lavage was performed first to assess whether swab collection resulted in additional trauma and bleeding that might impair shedding measurement. In addition, at the enrollment visit, participants were educated to insert a VT for 15-30 minutes up to 12 hours before their clinic visits. At each visit, participants were given a collection tube filled with 10 mL of PBS to transport their VT specimens to the clinic at their subsequent visit. All collected specimens were kept on ice until transferred to the laboratory. CVL specimens were centrifuged at 1000g for 10 minutes. The supernatant was aliquoted, whereas the cellular pellet was resuspended in 200 μL of PBS (pH 7.2) and aliquoted. The whole eCVL fluid was aliquoted without centrifugation. VTs were removed from their collection tubes, squeezed, and eluted; and a 1000-μL aliquot was preserved. All samples were stored at −70°C until testing.

CVL supernatant, eCVL whole fluid, and VT eluants were ultracentrifuged at 4°C for 1 hour. In each case, the sample volume was reduced to 200 μL and the excess supernatant was removed immediately. Nucleic acid extraction was performed by the Boom extraction method using the Nuclisens EasyMAG system (BioMerieux, Marcy l'Etoite, France). Specimens were tested using the Nuclisens EasyQ Version 1.2 HIV-1 assay (BioMerieux), following the manufacturer's protocol, with a lower limit of detection of 125 copies per milliliter. This method has been validated in our laboratory for detection of subtype C in plasma, the predominant HIV-1 subtype in southern Africa.17 We used an in-house real-time polymerase chain reaction (PCR) for the detection of HSV-2 DNA. This assay was developed using the Roche Hybridization FastStart PCR kit performed on the Roche LightCycler PCR system (Roche Molecular System, Basel, Switzerland). The procedures for differentiation of HSV-1 and HSV-2 were adapted from validated methods.2,18 The lower limit of quantification was 10 copies per reaction or approximately 60 copies per milliliter of CVS fluid.

The presence of microscopic blood in CVL, eCVL, and VT eluants was determined using Combur 3 Test E dipstick (Roche Diagnostics, Mannheim, Germany). Samples were considered positive for values ≥250 erythrocytes per milliliter. Detection of prostate-specific antigen, used as a proxy marker of semen contamination, was performed using the ADVIA Centaur immunoassay (Bayer Diagnostics, Tarrytown, NY). The working range of this assay is 0.01-100 μg/L. Samples were considered positive for levels >0.4 μg/L.19

The proportion of samples with detectable HIV-1 RNA, HSV-2 DNA, and microscopic blood or semen traces was calculated by collection method. Viral shedding using the different specimen collection methods was compared using risk ratios (RRs) and 95% confidence intervals (CIs) to assess whether 1 method systematically detected a higher prevalence of shedding. Multilevel regression models were used to allow for the paired samples at each visit and also for the repeated visits for each woman. Because the order of the collection method might lead to different levels of contamination, samples were also compared after excluding those in which blood contamination was detected. Finally, we calculated the proportion of women with HIV-1 RNA, HSV-2 DNA, microscopic blood, and prostate-specific antigen detected by any method at each visit and over all visits.

Back to Top | Article Outline


In total, 19 women were enrolled with a median age of 31 years (range 18-43 years) and a median CD4 count of 506 cells per cubic millimeter (interquartile range 300-754 cells/mm3). Seventeen (89%) and 2 (11%) women completed 4 and 3 visits, respectively. A high proportion of women reported having ever experienced GUD (68%), with 8 women (42%) found to have a GUD on clinical examination at enrollment. A total of 74 CVLs and 74 eCVLs were collected (96% of expected samples), and 52 (91%) of 57 expected VTs were brought to the clinic.

For HIV-1 RNA testing, results were available for 67 CVLs, 62 eCVLs, and 47 VTs. Most (n = 17) of the remaining samples had invalid results due to poor or no amplification of the internal control in PCRs. HIV-1 RNA was detected in 61%, 79%, and 57% of CVL, eCVL, and VT specimens, respectively (Table 1). Overall, rates of microscopic blood detection were highest in the eCVL and very low in VT. However, in samples uncontaminated with blood, eCVL still maintained the highest detection rate for HIV-1 RNA (77%) compared with CVL (58%) and VT (55%). By comparison, rates of semen detection seemed to be similar for both eCVL and CVL specimens.

In paired comparisons, detection of HIV-1 RNA was higher in eCVL compared with CVL [45/58 (78%) vs. 32/58 (55%); RR 1.41, 95% CI 1.05 to 1.88; P = 0.022]. HIV-1 RNA was also detected more frequently in eCVL irrespective of whether the swab was taken first [25/29 (86%) vs. 17/29 (59%); RR 1.47, 95% CI 0.98 to 2.21; P = 0.065] or lavage was taken first [20/29 (69%) vs. 15/29 (52%); RR 1.33, 95% CI 0.90 to 1.96; P = 0.14]. When blood-contaminated samples were excluded from paired comparisons, eCVL was still likely to detect HIV-1 RNA in more samples than CVL [30/40 (75%) vs. 21/40 (53%); RR 1.43, 95% CI 1.01 to 2.02; P = 0.043].

Both eCVL and CVL detected HIV-1 RNA in more samples than VT, although the observed increased detection with CVL was lower [eCVL vs. VT: 34/41 (83%) vs. 23/41 (56%); RR 1.48, 95% CI 1.08 to 2.01; CVL vs. VT: 27/42 (64%) vs. 22/42 (52%); RR 1.23, 95% CI 0.84 to 1.80].

We also considered whether the order of specimen collection influenced the rate of microscopic blood detection, and therefore perhaps HIV-1 RNA detection. When swabs were collected first, 9 (32%) of 28 HIV-1-positive eCVL samples and 6 (29%) of 21 HIV-1-positive CVL samples were found to have blood contamination. When lavage was done first, 6 (29%) of 21 HIV-1-positive eCVL samples and 5 (25%) of 20 HIV-1-positive CVL samples were found to have blood contamination. Comparisons between methods for HSV-2 detection were difficult because of the low number of shedding episodes. HSV-2 DNA was detected in less than 10% of samples by each method (Table 1).

HIV-1 RNA was detected at least once in all women by 1 or more specimens collected over the 4 time points. The proportion of women who shed HIV-1 RNA at each of the 4 visits was high and consistent with at least 84% with HIV-1 RNA detected by at least 1 method. Eleven women (58%) shed HIV-1 at each of the 4 visits (persistent shedders). There was no difference in mean CD4 count between persistent and intermittent HIV-1 shedders (514 vs. 434 cells/mm3, P = 0.58).

HSV-2 DNA was detected at least once in 7 (37%) women by 1 or more collection methods over the follow-up period. The proportion of women with detectable HSV-2 DNA at any given visit was low (fewer than 20% at any visit). The cumulative prevalence of HSV-2 genital shedding over the 4 visits was 1 (5%), 2 (11%), 5 (26%), and 7 (37%), respectively.

Although 16 (84%) of 19 women had microscopic blood traces in their samples, only 6 (32%) of 19 women had cervical contact bleeding during clinical procedures, suggesting that microscopic blood detection may be too sensitive a method and not well correlated with a clinically significant blood contamination. Semen traces were found in samples of 9 (47%) of the women at least once.

Back to Top | Article Outline


The main purpose of this study was the comparison of methods to collect CVS, which can be used as study outcomes and surrogate markers of HIV-1 and HSV-2 sexual transmissibility. Enriched CVL, a previously unevaluated technique which combines cervical swabbing and lavage,8 seemed to be the best collection method for detection of both HIV-1 RNA and HSV-2 DNA by detecting the greatest proportion of shedders. This method was associated with higher rates of microscopic blood contamination. However, when contaminated specimens were excluded from the analyses, rates of detection of HIV-1 RNA remained highest in eCVL compared with both CVL and VT. There was an indication that HIV-1 RNA detection was higher if the swab was taken before lavage, but there was little suggestion that the order of sample collection influenced blood detection. However, numbers involved are small, and caution should be exercised in interpreting these results. In studies where blood contamination may be considered an issue, CVL and VT may offer some advantages over eCVL, in that they are less traumatic methods of specimen collection and therefore less likely to induce bleeding.

VTs were a useful additional method of CVS collection. Although detection rates for HIV-1 RNA were lower than for eCVL, they were similar to rates observed with CVL. Higher rates of HIV-1 genital shedding were detected in VT than have been observed previously.11 VT performed as well as CVL for the detection of HSV-2 DNA, although the number of HSV-2-positive samples was small. VT may be useful for increasing the number of measurement points in intervention trials, where repeated sampling is required. Generally, VTs have much less stringent transport criteria than swabs, making them more suitable for field-based trials in developing country settings.

The frequency of HIV-1 RNA detection in CVS was high in this study, as in other studies of HIV-infected women in Africa.8,13,20 High plasma viral load and HIV subtype have been associated with increased genital HIV-1 in other studies,1,21,22 but these samples were not collected in this study and so correlations cannot be made.

The frequency of HSV-2 DNA detection in CVS was lower than that reported in other studies8,23 and may be due to limitations in the sampling method or processing. However, through repeated sampling, we were able to demonstrate that in this cohort of women with relatively high CD4+ counts, 37% of them shed HSV-2 DNA over 4 visits. HSV-2 viral shedding is intermittent, and as many as 50% of shedding episodes may last <6 hours.24 Timing of sample collection combined with our small sample size may have precluded detection of some HSV shedding episodes.

In conclusion, this study demonstrated a high frequency of genital HIV-1 shedding, lower rates of HSV-2 shedding, and the need for repeated sampling to identify women with intermittent viral shedding. Enriched CVL may be a more sensitive collection method for the detection of HIV-1 and HSV-2 but may be more likely to cause bleeding which could contaminate CVS specimens and complicate the interpretation of results. Tampons were an acceptable self-administered method and can potentially be used to increase the frequency of sampling in intervention trials with viral shedding outcomes.

Back to Top | Article Outline


The authors are grateful to the volunteers who participated in this study; to the Reproductive Health and HIV Research Unit (Johannesburg) staff who provided support for project implementation; to Grant Napier (G.N.), Wouter Le Roux (W.L.R.), and staff at Contract Laboratory Services (CLS) who assisted with testing of samples; and to Helen Weiss (London School of Hygiene and Tropical Medicine) for initial advice on study design. The authors acknowledge the support of the City of Johannesburg-Department of Health. Contributions of authors: S.D., R.R., and P.M. designed the study and obtained funding. The study was conducted by S.D. and N.M. Laboratory testing was conducted by W.L.R. under supervision of G.N., A.C., and W.S., with external quality control by J.L. and L.B. G.A. provided data management support. S.D., T.C., and P.M. conducted the analyses and wrote the first draft of the article, which was reviewed and approved by all authors.

Back to Top | Article Outline


1. Coombs RW, Reichelderfer PS, Landay AL. Recent observations on HIV type-1 infection in the genital tract of men and women. AIDS. 2003;17:455-480.
2. Legoff J, Bouhlal H, Gresenguet G, et al. Real-time PCR quantification of genital shedding of herpes simplex virus (HSV) and human immunodeficiency virus (HIV) in women coinfected with HSV and HIV. J Clin Microbiol. 2006;44:423-432.
3. Ndjoyi-Mbiguino A, Ozouaki F, Legoff J, et al. Comparison of washing and swabbing procedures for collecting genital fluids to assess cervicovaginal shedding of herpes simplex virus type 2 DNA. J Clin Microbiol. 2003;41:2662-2664.
4. Andreoletti L, Gresenguet G, Chomont N, et al. Comparison of washing and swabbing procedures for collecting genital fluids to assess shedding of human immunodeficiency virus type 1 (HIV-1) RNA in asymptomatic HIV-1-infected women. J Clin Microbiol. 2003;41:449-452.
5. Belec L, Meillet D, Levy M, et al. Dilution assessment of cervicovaginal secretions obtained by vaginal washing for immunological assays. Clin Diagn Lab Immunol. 1995;2:57-61.
6. 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.
7. Baron P, Bremer J, Wasserman SS, et al. Detection and quantitation of human immunodeficiency virus type 1 in the female genital tract. The Division of AIDS Treatment Research Initiative 009 Study Group. J Clin Microbiol. 2000;38:3822-3824.
8. Nagot N, Foulongne V, Becquart P, et al. Longitudinal assessment of HIV-1 and HSV-2 shedding in the genital tract of West African women. J Acquir Immune Defic Syndr. 2005;39:632-634.
9. Sturm PD, Connolly C, Khan N, et al. Vaginal tampons as specimen collection device for the molecular diagnosis of non-ulcerative sexually transmitted infections in antenatal clinic attendees. Int J STD AIDS. 2004;15:94-98.
10. Tabrizi SN, Paterson BA, Fairley CK, et al. Comparison of tampon and urine as self-administered methods of specimen collection in the detection of Chlamydia trachomatis, Neisseria gonorrhoeae and Trichomonas vaginalis in women. Int J STD AIDS. 1998;9:347-349.
11. Webber MP, Schoenbaum EE, Farzadegan H, et al. Tampons as a self-administered collection method for the detection and quantification of genital HIV-1. AIDS. 2001;15:1417-1420.
12. Coombs RW, Wright DJ, Reichelderfer PS, et al. Variation of human immunodeficiency virus type 1 viral RNA levels in the female genital tract: implications for applying measurements to individual women. J Infect Dis. 2001;184:1187-1191.
13. 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.
14. Ndjoyi-Mbiguino A, Ozouaki F, Legoff J, et al. Comparison of washing and swabbing procedures for collecting genital fluids to assess cervicovaginal shedding of herpes simplex virus type 2 DNA. J Clin Microbiol. 2003;41:2662-2664.
15. WHO. Interim WHO Clinical Staging of HIV/AIDS and HIV/AIDS Case Definitions for Surveillance (African Region). Geneva, Switzerland: World Health Organisation; 2005. WHO/HIV/2005.02.
16. Ashley-Morrow R, Nollkamper J, Robinson NJ, et al. Performance of focus ELISA tests for herpes simplex virus type 1 (HSV-1) and HSV-2 antibodies among women in ten diverse geographical locations. Clin Microbiol Infect. 2004;10:530-536.
17. Stevens W, Horsfield P, Scott LE. Evaluation of the performance of the automated NucliSENS easyMAG and EasyQ systems versus the Roche AmpliPrep-AMPLICOR combination for high-throughput monitoring of human immunodeficiency virus load. J Clin Microbiol. 2007;45:1244-1249.
18. Burrows J, Nitsche A, Bayly B, et al. Detection and subtyping of Herpes simplex virus in clinical samples by LightCycler PCR, enzyme immunoassay and cell culture. BMC Microbiol. 2002;2:12.
19. Pepin J, Fink GD, Khonde N, et al. Improving second-generation surveillance: the biological measure of unprotected intercourse using prostate-specific antigen in vaginal secretions of West African women. J Acquir Immune Defic Syndr. 2006;42:490-493.
20. Cowan FF, Pascoe SJ, Barlow KL, et al. Association of genital shedding of herpes simplex virus type 2 and HIV-1 among sex workers in rural Zimbabwe. AIDS. 2006;20:261-267.
21. Kovacs A, Wasserman SS, Burns D, et al. Determinants of HIV-1 shedding in the genital tract of women. Lancet. 2001;358:1593-1601.
22. John-Stewart GC, Nduati RW, Rousseau CM, et al. Subtype C is associated with increased vaginal shedding of HIV-1. J Infect Dis. 2005;192:492-496.
23. Mbopi-Keou FX, Gresenguet G, Mayaud P, et al. Interactions between herpes simplex virus type 2 and human immunodeficiency virus type 1 infection in African women: opportunities for intervention. J Infect Dis. 2000;182:1090-1096.
24. Mark KE, Wald As, Magaret S, et al. Rapid onset and clearance of genital HSV reactivations in immunocompetent adults: The virus is usually “on”. Presented at: 17th International Society for STD Research and 10th International Union Against Sexually Transmitted Infections; 2007; Seattle, WA.

African women; cervicovaginal lavage; genital shedding; HIV-1 RNA; HSV-2 DNA

© 2008 Lippincott Williams & Wilkins, Inc.