THE BASIS OF MOST HUMAN immunodeficiency virus (HIV) and sexually transmitted diseases (STD) behavioral interventions is reducing high-risk sexual behaviors. However, self-reported sexual behaviors are prone to reporting bias. Early studies by our group suggested that condom use is subject to overreporting, especially in settings such as clinics or public health venues, where they are aggressively promoted.1,2 Condoms, when used correctly, have been shown to be effective against transmission of most STD, including HIV, chlamydia, gonorrhea, syphilis, and herpes simplex virus type 2.3 As an effective intervention, correct and consistent condom use has been widely promoted as a critical component of HIV/STD prevention programs.4,5
The 2 approaches taken to reduce reporting bias for sexual behaviors have been improved interviewing techniques and biomarkers. Interview techniques that incorporate self-administered computerized questionnaires, such as AUDIO-computer assisted self-interview (A-CASI) and telephone-CASI (T-CASI), have been shown to reduce bias in reporting drug use, sexual behavior, same-sex contact, and other sensitive behaviors.6–10 Biomarkers that validate self-reports would help diminish the impact of reporting biases.
Initial biomarker approaches used incident STD as a biomarker of correct and consistent condom use. However, even in populations with high STD rates, most individuals having unprotected sex will not get infected. This paradox results in significant misclassification of risk behaviors.11,12 Non-STD semen biomarkers that have been evaluated include: acid phosphatase, prostate specific antigen (PSA), and human seminal plasma antigen, of which PSA was the most sensitive.13 However, PSA is detectable in vaginal secretions for only 24–48 hours and is also subject to low levels of cross-reactivity with other antigens.14–16
We previously reported initial work on a genetic marker Y-chromosome sequence (YCS) PCR.17,18 YCS that are unique to men were found in self-collected vaginal swabs up to 15 days following unprotected intercourse with a mean clearance half-life of 3.83 days and a lower detection limit of 5 chromosome copies. In practice, this offers the potential for a 1- to 2-week retrospective biomarker for sexual behavior. The original proof-of-concept study17 was validated using archived vaginal fluid samples of subjects who had well-characterized sexual histories.18 In this study, we evaluate the efficacy of condoms in preventing YCS transmission and validate that when correctly used, condoms do not result in a positive vaginal YCS signal.
Materials and Methods
Patient Population and Study Design
Women in monogamous heterosexual relationships were recruited by advertisements in local newspapers and at nearby university campuses. Informed consent was obtained from all participants. Eligible women were screened by a trained clinician before enrollment. Inclusion criteria included women who were over 18 years of age, sexually active, using either hormonal contraception or surgical sterilization (tubal ligation), monogamous, willing to abstain from intercourse for a “clearance period” of 2 weeks and then have protected intercourse with partner, willing to obtain SCVS, and not pregnant. Exclusion criteria included women who were unable to complete a daily coital diary or whose sexual partner had a vasectomy. In addition, subjects were counseled that if the abstinence period resulted in either relationship discord or potential vulnerability to emotional or physical abuse, they should then opt out of the study and notify the investigators. Subjects were notified that to protect them and maximize objectivity, they would be reimbursed for their time in the cycle, even if they withdrew early. Compensation for participation in the study was $100.00. The technique to obtain SCVS was explained to all participants, and the appropriate use of condoms was demonstrated using a penis model. All participants were provided with latex condoms. This study was granted approval from The Johns Hopkins University School of Medicine Institutional Review Board.
The study design is summarized in Figure 1. Women were asked to abstain from sexual intercourse for 2 weeks. This was based on our previous data that YCS may persist in vaginal secretions for up to 2 weeks after intercourse.17 On day 14, two “baseline” SCVS were obtained to document lack of YCS detection. On day 15, women were instructed to have protected sexual intercourse with their partner. On days 16 (“postcoital day 1”) and 17 (“postcoital day 2”), 2 additional SCVS were collected each day (total number of study swabs = 6). During the entire study period, women were instructed to keep a daily check-off diary, which included reports of any vaginal intercourse, condom use, condom malfunction, diaphragm use, spermicide use, lubricant use, hormonal contraceptive use, douching, menses, tampon or pad use, oral sex received, and any digital vaginal penetration. To minimize reporting bias, we made it clear to all participants that they would receive compensation despite breaches of study protocol as long as those breaches were documented in the daily diaries.
Y-Chromosome Detection Protocol and PSA Assay
SCVS were extracted for total DNA as described in the accompanying paper (Melendez J et al. “Detection and quantification of Y chromosomal sequences by real-time PCR using the LightCycler system”). Briefly, vaginal fluids were recovered from the SCVS by placing the swab in 0.5 mL of saline for 25 minutes, the sample was centrifuged for 3 minutes, and 100 μL of the supernatant was removed and stored at −80°C for PSA testing. The resulting sample was extracted for total DNA by a Proteinase K/dithiothreitol extraction protocol. Detection of Yc sequences was performed on the LightCycler system with primers and probes specific to a fragment of the SRY gene with positive and negative controls. PCR products were quantified with the absolute quantification module of the LightCycler system with the assistance of a standard concentration curve. Each assay was run with positive and negative controls.
Samples for PSA testing were sent to the University of Alabama at Birmingham for testing. PSA was measured with an enzyme-linked immunoassay as previously described.13
Because all subjects were instructed to use condoms, we expected all of the samples to be undetectable for YCS, both at “baseline” (since they were asked to abstain) and at postcoital days 1 and 2. Therefore, our main outcome was to evaluate the specificity of the assay being the probability of a negative assay result when condoms were used perfectly. A sample size of 50 for a hypothesized specificity of 95% allowed us to estimate this probability with a precision of ±3%. Specificity was calculated using the standard formula (the number of true negative results divided by the sum of true negative and false positive results) and binomial 95% confidence limits for specificity measures were reported. Independent proportions were compared using the χ2 test. P values <0.05 were assumed to represent statistical significance. All analyses were conducted using STATA v.8.2 (STATA Corp, College Station, TX).
Fifty-six women were eligible and were recruited. Forty-four of the 56 women (79%) completed the study. Of the 12 women who did not complete the study, 1 was disqualified because she mistakenly obtained the baseline swab after intercourse on day 15, and the remaining 11 decided not to participate and voluntarily withdrew from the protocol after enrollment. Fifteen percent of subjects were black and 85% white. 65% used hormonal contraception as a second contraceptive method. None of the women reported condom malfunction (breakage, slippage, etc.) on day 15.
There were 5 women who had detectable YCS on one or more of the swabs (Fig. 1). One of the 5 had a positive “baseline” swab with 2 subsequent positive postcoital swabs (patient 1, Fig. 1). Because of the baseline positive swab, the participant was excluded from further analyses. The remaining 4 patients with negative baseline swabs and positive postcoital swabs are also listed in Figure 1 (patients 2–5). The overall specificity of the Y-chromosome assay for detecting YCS following protected heterosexual intercourse was 92% (95%CI: 80%–98%). The PSA assay detected 2 subjects with positive baseline swabs; both subjects had a negative YCS assay and both subsequently had negative postcoital swabs. The 2 participants denied any sexual exposures (vaginal, oral, or digital) in the preceding 2 weeks. The PSA assay also detected 1 subject with a positive day 1 postcoital swab, which was negative based on YCS. That participant did not report any condom malfunction. The day 2 postcoital swab was subsequently negative by PSA. The overall agreement between the 2 assays was 82%. The overall specificity of the PSA assay following protected heterosexual intercourse was 98% (95%CI: 87%–100%).
Table 1 summarizes the prevalence of oral sex, digital penetration, menstrual bleeding, douching, hormonal contraceptive use, lubricant use, and spermicidal use 48 hours preceding postcoital swab collection. Except for patient 5, all the participants with positive postcoital swabs reported receptive digital penetration and/or receptive oral sex within 48 hours preceding swab collection.
Our study demonstrates that the assay is 92% specific for YCS detection in the setting of condom use during heterosexual intercourse. If we exclude patients who had detectable YCS and who reported receptive digital penetration and/or receptive oral sex 48 hours preceding swab collection (as both behaviors may be theoretically associated with YCS deposition), the specificity of the assay increases to 98%, a value equivalent to the specificity of PSA.13
The sensitivity of the assay for detecting Y-chromosome in the setting of oral sex or digital penetration in the preceding 48 hours is 19% (95%CI: 4%–46%) and 27% (95%CI: 6%–61%), respectively. The concentration of YCS ranged from 0.6 to 10.1 ng/mL following digital and/or oral sex (Fig. 1), much less than the 66.7 ng/mL mean concentration of YCS following unprotected vaginal intercourse.17 This suggests that the amplitude of the signal is lower when YCS is deposited as a result of oral or digital contact when compared with vaginal intercourse. A potential approach to minimize the effects of nonvaginal intercourse-related YCS deposition is to explore chemical means to destroy nonsperm cell DNA. For now, if the YCS detection assay is to be used, behaviors pertaining to digital penetration and oral sex should be documented.
PSA is not associated with oral sex or digital penetration, which increases its specificity relative to YCS detection. However, the shorter half-life of PSA limits its utility to self-reported behaviors preceding PSA detection by 48 hours. The advantage of YCS is its slower decay, making it useful to validate self-reported behaviors for 7 to 10 days following heterosexual exposure. Of the 44 women evaluated in this study, 7 reported oral and/or digital sex between 3 and 9 days preceding swab collection. Of those, only 1 (patient 5, Fig. 1) was found to have detectable YCS on her postcoital swabs. However, as can be seen in Figure 1, that patient’s baseline swab was negative despite her having had the oral exposure 24 hours preceding swab collection and no reported history of additional oral exposures. This suggests an inconsistency due to either laboratory error during processing of the baseline swab, or inaccurate self-report on the timing of oral sex. Thus, it appears that oral or digital sex preceding swab collection by more than 48 hours should not have a significant impact on the YCS biomarker. We believe that if these findings are confirmed, then it would be possible to recommend the use of YCS with the addition of PSA as a confirmatory testing if oral or digital exposure 48 hours preceding swab collection is reported by patients. This would maximize specificity of the YCS biomarker while minimizing the costs of frequent PSA testing.
Our finding of limited YCS detection following protected vaginal intercourse, even with the use of a highly sensitive method that allows the detection of as few as 5 YCS copies, lends further support to earlier studies demonstrating limited permeability of intact condoms to microparticles.19–21
This study has several limitations. The gold standard used to validate our assay was individual self-reports, which may be subject to information biases. To try and minimize bias, we designed the study so that participants understood that compensation was not linked to the 2-week abstinence period. We also provided daily diaries to enhance accuracy of self-reports, and we validated the YCS assay using PSA testing. Self-collected vaginal swabs are dependent on intrasubject and sampling variability. To minimize such an effect, 2 swabs per time period were collected. In addition, all swabs were tested for the presence of X-chromosome to ensure that the specimen was adequate. Finally, condom malfunction may be responsible for all or part of the positive YCS specimens rather than oral or digital exposures. We encouraged all participants to report any perceived condom malfunction during intercourse using the daily diaries. None of the women reported problems with condom use. This may be a result of the one-on-one teaching on proper use of condoms that all subjects participated in at the time of enrollment, or simply a result of the relatively small sample size coupled with the relatively infrequent occurrence of condom malfunction.22,23 It may also be that condoms may not be 100% effective at preventing YCS vaginal deposition.24 This latter explanation is less likely in view of the negative PSA confirmatory testing on the positive YCS specimens.
There are several additional characteristics of the YCS assay that need to be evaluated. In addition to the effects of receptive oral sex and digital penetration, the effects of menses and vaginal hygiene habits (e.g., douching) need to be assessed to define the optimal use of this test in large trials. In conclusion, the YCS assay appears to be a potentially useful biomarker to validate self-reported sexual behaviors in women. Even when using highly sensitive methods of detection, condoms appear to be effective at preventing the deposition of YCS when used in a correct and consistent manner.
1. Zenilman JM, Weisman CS, Rompalo AM, et al. Condom use to prevent incident STDs: The validity of self-reported condom use. Sex Transm Dis 1995; 22:15–21.
2. Turner CF, Miller HG. Zenilman’s anomaly reconsidered: Fallible reports, ceteris paribus, and other hypotheses. Sex Transm Dis 1997; 24:522–527.
3. Holmes KK, Levine R, Weaver M. Effectiveness of condoms in preventing sexually transmitted infections. Bull World Health Organ 2004; 82:454–461.
4. Eng TR, Butler WJ. The Hidden Epidemic. Washington, DC: National Academy Press; 1997.
5. Coates TJ, Aggleton P, Gutzwiller F, et al. HIV prevention in developed countries. Lancet 1996; 348:1143–1148.
6. Des Jarlais DC, Paone D, Milliken J, et al. Audio-computer interviewing to measure risk behaviour for HIV among injecting drug users: A quasi-randomised trial. Lancet 1999; 353:1657–1661.
7. Ghanem KG, Hutton HE, Zenilman JM, et al. Audio computer assisted self interview and face to face interview modes in assessing response bias among STD clinic patients. Sex Transm Infect 2005; 81:421–425.
8. Gribble JN, Miller HG, Cooley PC, et al. The impact of T-ACASI interviewing on reported drug use among men who have sex with men. Subst Use Misuse 2000; 35:869–890.
9. Rogers SM, Willis G, Al-Tayyib A, et al. Audio computer assisted interviewing to measure HIV risk behaviours in a clinic population. Sex Transm Infect 2005; 81:501–507.
10. Newman JC, Des Jarlais DC, Turner CF, et al. The differential effects of face-to-face and computer interview modes. Am J Public Health 2002; 92:294–297.
11. Pequegnat W, Fishbein M, Celentano D, et al. NIMH/APPC workgroup on behavioral and biological outcomes in HIV/STD prevention studies: A position statement. Sex Transm Dis 2000; 27:127–132.
12. Fishbein M, Pequegnat W. Evaluating AIDS prevention interventions using behavioral and biological outcome measures. Sex Transm Dis 2000; 27:101–110.
13. Lawson ML, Maculuso M, Bloom A, et al. Objective markers of condom failure. Sex Transm Dis 1998; 25:427–432.
14. Macaluso M, Lawson L, Akers R, et al. Prostate-specific antigen in vaginal fluid as a biologic marker of condom failure. Contraception 1999; 59:195–201.
15. Walsh TL, Frezieres RG, Nelson AL, et al. Evaluation of prostate-specific antigen as a quantifiable indicator of condom failure in clinical trials. Contraception 1999; 60:289–298.
16. Walsh TL, Frezieres RG, Peacock K, et al. Use of prostate-specific antigen (PSA) to measure semen exposure resulting from male condom failures: Implications for contraceptive efficacy and the prevention of sexually transmitted disease. Contraception 2003; 67:139–150.
17. Zenilman JM, Yuenger J, Galai N, et al. Polymerase chain reaction detection of Y chromosome sequences in vaginal fluid: Preliminary studies of a potential biomarker for sexual behavior. Sex Transm Dis 2005; 32:90–94.
18. Jadack RA, Yuenger J, Ghanem KG, et al. Polymerase chain reaction detection of Y-Chromosome sequences in vaginal fluid of women accessing a sexually transmitted disease clinic. Sex Transm Dis 2006; 33:22–25.
19. Jay GD, Drummond E, Lane B. Altered surface character of stretched condom latex. Contraception 1992; 45:105–110.
20. Lytle CD, Routson LB, Seaborn GB, et al. An in vitro evaluation of condoms as barriers to a small virus. Sex Transm Dis 1997; 24:161–164.
21. Retta SM, Herman WA, Rinaldi JE, et al. Test method for evaluating the permeability of intact prophylactics to viral-size microspheres under simulated physiologic conditions. Sex Transm Dis 1991; 18:111–118.
22. Rosenberg MJ, Waugh MS. Latex condom breakage and slippage in a controlled clinical trial. Contraception 1997; 56:17–21.
23. Trussell J, Warner DL, Hatcher RA. Condom slippage and breakage rates. Fam Plann Perspect 1992; 24:20–23.
24. Galvao LW, Oliveira LC, Diaz J, et al. Effectiveness of female and male condoms in preventing exposure to semen during vaginal intercourse: A randomized trial. Contraception 2005; 71:130–136.