Efficacy of Individual-Level Prevention Methods in Preventing Acquisition and Transmission of STDs and HIV Infection
Preexposure vaccination is one of the most effective methods for preventing transmission of 2 main STDs: HPV infection and hepatitis B (). In March 2007, the Advisory Committee on Immunization Practices (ACIP) issued guidelines for administration of the quadrivalent HPV vaccine to females aged ≤25 years [14
] (details at www.cdc.gov/std/hpv
). This vaccine confers protection against HPV types 6/11 (responsible for 90% of genital warts) and 16/18 (responsible for 70% of cervical cancers). In published clinical trials, the quadrivalent HPV vaccine has demonstrated efficacy for prevention of vaccine HPV type-related cervical, vaginal, and vulvar cancer precursor and dysplastic lesions, and external genital warts [15
]. Universal vaccination of girls aged 11–12 years is recommended, as is catch-up vaccination for girls and women aged 13–26 years. The vaccine is also efficacious in preventing infection in women aged 24–45 years not already infected with the relevant HPV types. [16
] Data on the efficacy of the quadrivalent HPV vaccines in protecting young men from vaccine-type HPV acquisition indicates similarly high levels of protection [17
], and the ACIP issued permissive guidance for immunization to prevent genital warts in young men in 2010. Both men and women are also likely to benefit from protection against anal intraepithelial neoplasia afforded by the quadrivalent vaccine. A bivalent vaccine that is effective in preventing cervical neoplasia associated with HPV types 16/18 has also been approved for use in the United States, and is recommended by ACIP [19
Individual-Level Biomedical Approaches to Prevention of Sexually Transmitted Infections, Including HIV Infection, by Level of Adherence Requireda
Immunization against hepatitis B has been routinely recommended for infants since 1991 and was subsequently recommended for adolescents. Although this has been temporally associated with marked declines in the incidence of hepatitis B virus infection in the United States [21
], sexual transmission still accounts for the majority of new infections, which are especially common among unvaccinated MSM. Consequently, hepatitis B vaccination is recommended for all adults who are at risk for sexual infection, including sex partners of persons positive for hepatitis B surface antigen, sexually active persons who are not in a long-term, mutually monogamous relationship, persons seeking evaluation or treatment for a STD, and MSM [22
]. Moreover, all HIV-infected persons should be immunized against hepatitis B, because the natural history of hepatitis B is accelerated in the setting of HIV, and coinfection imposes specific considerations in selection of antiretroviral agents. Hepatitis A vaccine is licensed and is recommended for MSM and illicit drug users (both injecting and noninjecting) [23
] (details available at http://www.cdc.gov/hepatitis
). Finally, new vaccine approaches aimed at hepatitis C, including peptide, recombinant protein, DNA and vector-based vaccines, have recently reached phase I/II human clinical trials, providing some promise for future control of this infection [24
Prospects for an effective HIV vaccine remain on the distant horizon. Recent disappointing results from human trials have stimulated a renewed focus on the basic biology of HIV pathogenesis. Two phase III trials of a vaccine aimed at eliciting neutralizing antibodies against the envelope glycoprotein 120 did not find protection against HIV infection [25
]. A phase IIB trial of the first T cell vaccine (Merck’s MRKAd5 HIV-1 gag/pol/nef trivalent product, using a replication-defective adenovirus type-5 vector with 3 HIV genes) was stopped in September 2007. Interim analysis revealed no protective effect against HIV acquisition and no reduction in initial viral loads among participants infected with HIV [27
]. Further analysis showed that preexisting immunity to adenovirus type 5 was directly associated with a significantly higher risk of acquiring HIV and that this untoward effect was further augmented among uncircumcised men. A community-based, randomized, double-blind, placebo-controlled trial performed in >16
000 Thai adults evaluated 4 priming injections of a recombinant canary pox vector vaccine (ALVAC-HIV) plus 2 booster injections of a recombinant glycoprotein 120 subunit vaccine (AIDSVAX B/E) [29
]. There was a trend toward prevention of HIV infection in the intention-to-treat analysis (vaccine efficacy, 26.4%; 95% confidence interval [CI], −4.0 to 47.9), but not in the per-protocol analysis (vaccine efficacy, 26.2%; 95% CI, −13.3 to 51.9). Vaccination did not affect HIV viral load or CD4 cell count in participants who acquired HIV infection during the trial.
The 2006 STD Treatment Guidelines noted that, when used consistently and correctly, male latex condoms are effective in preventing sexual transmission of HIV and other STDs, including chlamydia, gonorrhea, syphilis, genital HPV, and trichomoniasis [1
]. By limiting lower genital tract infections, male condoms might also reduce the risk of pelvic inflammatory disease in women [30
]. In heterosexual serodiscordant relationships in which condoms were consistently used, HIV-negative partners were 80% less likely to become HIV infected than persons in similar relationships in which condoms were not used [31
]. Condom use might also reduce the risk for transmission of HSV-2, although data for this effect are more limited [32
]. Finally, condom use reduces the risk of HPV infection [34
] and HPV-associated diseases (eg, genital warts and cervical cancer) [36
]. Use of condoms has been associated with regression of cervical intraepithelial neoplasia [37
] and clearance of HPV infection in women, and with regression of HPV-associated penile lesions in men [38
Since 2006, available data on male condom efficacy have emerged in several areas: (1) protection against infection with genital HPV, HSV-2, and C. trachomatis; (2) the methodology of self-reporting on consistent and correct condom use; and (3) interventions to reduce adolescents’ sexual risk behavior and absence of condom use during first sex on adolescents’ subsequent sexual risk behavior.
A prospective study among newly sexually active college women demonstrated that consistent condom use was associated with a 70% reduction in risk for genital HPV transmission; a large cross-sectional study supported slightly lower but significant efficacy in men. Investigators followed up 82 female university students who reported their first intercourse with a male partner either during the study period or within 2 weeks before enrollment [35
]. Cervical and vulvovaginal samples for HPV DNA testing and Pap smears were collected every 4 months. The incidence of genital HPV infection was 37.8/100 patient-years at risk among women whose partners used condoms for all instances of intercourse during the 8 months before testing, compared with 89.3/100 patient-years at risk in women whose partners used condoms <5% of the time (adjusted hazard ratio [AHR], 0.3; 95% CI, .1–.6). In participants in this study who reported 100% condom use by their partners, no cervical squamous intraepithelial lesions were detected in 32 patient-years at risk, whereas 14 incident lesions were detected during 97 patient-years at risk among women whose partners did not use condoms or used them less consistently. In a separate cross-sectional study from 2 cities in the United States, 393 men were assessed for 37 HPV types from 5 anogenital sites. Report of always using condoms was associated with lower odds of HPV detection (adjusted odds ratio [AOR], 0.50; 95% CI, .30–.83) [34
Prospective studies continue to support a protective effect of condoms against acquisition of genital herpes, chlamydia, and gonorrhea. In an analysis that pooled data from all published studies that prospectively assessed condom use and HSV-2 incidence, persons who always used condoms had a 30% decreased risk of acquiring HSV-2, compared with those who reported no condom use (P
= .01). [39
] Moreover, the risk of acquiring HSV-2 decreased by 7% for every additional 25% of the time that condoms were used (P
= .01). Conversely, HSV-2 acquisition rose steadily with report of increasing frequency of unprotected sex acts. These effects were consistent for men and women. Among men who participated in the Kenya circumcision trial, report of condom use at last vaginal intercourse was independently associated with lower rates of incident chlamydial or gonococcal infection detected during 2 years of follow-up (AHR, 0.64; 95% CI, .50-.82) [40
Cross-sectional data support the claim that male condoms protect against acquisition of C. trachomatis.
Three cross-sectional studies addressed risk of chlamydial infection as a function of various correlates of condom use. Data from a public STD clinic database examined chlamydia prevalence among known chlamydia contacts who were consistent versus inconsistent condom users. The AOR for chlamydial infection in consistent relative to inconsistent users was 0.10 (95% CI, .01–.83) [41
]. In a database from an Australian STD clinic, condom use was associated with a lower odds of rectal, but not urethral, chlamydia among MSM [42
]. In an urban adolescent healthcare clinic, the prevalence of chlamydia or gonorrhea was assessed as a function of reported condom use in 509 predominantly African American adolescent girls. Although consistent and correct use was reported uncommonly (in only 16% of subjects), it was associated with reduced odds of chlamydia (odds ratio [OR], 0.4; 95% CI, .2–1.0) and gonorrhea (OR, 0.1; 95% CI, 0–.7) [43
Data support the need for adolescents to receive comprehensive, current, and accessible information on prevention of STDs, HIV infection, and pregnancy, including information on condoms. Data from the 1994 to 2002 National Longitudinal Study of Adolescent Health (Add Health) compared subsequent sexual behaviors and risk of STI among adolescents who did and did not use a condom at their sexual debut [44
]. Adolescents who reported condom use at sexual debut were more likely to report condom use at most recent intercourse (on average, 6.8 years after sexual debut), and were half as likely to test positive for chlamydia or gonorrhea (AOR, 0.50; 95% CI, .26–.95). Reported number of lifetime sexual partners did not differ between the 2 groups. A separate analysis of Add Health data included teens enrolled in 2001 who were followed up 1 and 3 years later; teens who took a virginity pledge reported a longer time until sexual debut than teens who did not [45
]. However, overall sexual behaviors subsequent to pledging, including patterns of condom use, did not differ between these groups. A more recent analysis demonstrated that teens who took the pledge and who did have sex were less likely to use condoms at sexual debut [46
Estimating true condom efficacy requires that measures of consistent and correct use must be developed, understood, and used. Multiple problems with condoms occurred among 1152 participants who completed a supplemental questionnaire as part of Project RESPECT, a counseling intervention trial conducted at 5 publicly funded STD clinics in the 1990s [47
]. Nearly half (41%) of respondents reporting condom use indicated that condoms broke, slipped off, leaked, or were not used throughout intercourse in the previous 3 months. Nearly 9% of acts in which condoms were used resulted in potential STI exposure because of delayed application of condoms, breakage, early removal, slippage, or leakage. Critically, use problems were significantly associated with reporting inconsistent condom use, multiple partners, and other condom problems. Among 130 participants who were tested for gonorrhea and chlamydia at the time of the questionnaire completion and 3 months prior, no infections were detected among consistent users reporting no use problems. A significant dose-response relationship occurred between measures of increased protection from condom use and reduced gonorrhea and chlamydia rates.
Nonlatex condoms provide an acceptable alternative for persons unable to tolerate latex condoms. Two general categories of nonlatex condoms exist. The first type is made of polyurethane or other synthetic material and provides protection against STDs, HIV infection, and pregnancy equal to that of latex condoms. These condoms provide an acceptable alternative for persons unable to use latex condoms. A Cochrane review concluded that although one nonlatex condom (eZon) did not protect against pregnancy as well as its latex comparison condom, no differences were found in the typical use efficacy between the Avanti and the Standard Tactylon condoms and their latex counterparts [48
]. The nonlatex condoms had higher rates of clinical breakage than latex comparison condoms (OR for clinical breakage, 2.64 [95% CI, 1.63–4.28] to 4.95 [95% CI, 3.63–6.75]). The contraceptive efficacy of nonlatex condoms could not be estimated, and will require more research [48
]. The Food and Drug Administration (FDA) has published draft guidelines modifying the labeling on male latex condoms to reflect these findings [49
Laboratory studies indicated that the original version of the female condom (Reality), is an effective mechanical barrier to viruses and semen. If used consistently and correctly, the female condom might substantially reduce the risk for STI. Female condoms are safe to use repeatedly if proper care procedures are followed. Since the last guideline review, relatively few studies have been completed to evaluate the efficacy of female condoms in providing protection from STIs, including HIV infection. The new evidence uses postuse markers of semen to measure more precisely the mechanical barrier provided by this method. Other in vitro data assess a new formulation of the female condom.
Two systematic reviews supported the potential effectiveness of female condoms. The first reviewed 137 articles and abstracts on various aspects of the female condom and 5 randomized controlled trials on its effectiveness [50
]. The review concluded that although the evidence is limited, “the female condom is effective in increasing protected sex and decreasing STI incidence among women.” A second systematic review concluded that “randomised controlled trials provide evidence that female condoms confer as much protection from STIs as male condoms” [51
The comparative effectiveness of the male condom and female condom was assessed in a randomized controlled trial that assigned women to sequential use of 10 male latex condoms, then 10 female polyurethane condoms. [52
] The association between frequency and types of self-reported mechanical failure and semen exposure was measured based on prostate-specific antigen (PSA) levels. Moderate to high postcoital PSA levels were detected in 3.5% of male condom users and 4.5% of female condom users (difference, 1.4; 95% CI, 1.6–3.7). PSA levels were more frequent with mechanical problems and less frequent with other problems or correct use with no problems. Although mechanical problems were more common with the female condom, the risk of semen exposure was probably similar.
The FDA held an advisory meeting in December 2008 to review evidence in support of a new version of the female condom [53
]. The new version has a slightly modified shape and no seam and is made from nitrile (as opposed to polyurethane, the material used for the first version). Modifications to the manufacturing process as a result of this shift have resulted in considerable cost reductions to the product. The advisory panel voted to support FDA approval of the new female condom, and it became available in 2009. The new female condom is already in use in many countries outside the United States and has been endorsed by WHO after a similar review process. This new design should theoretically afford protection similar to the polyurethane female condom and allows for lower manufacturing cost. The female condom also has been evaluated for protection against HIV infection and other STIs protection during receptive anal intercourse [54
]. Although it might provide some protection, no new data are available to help to define its efficacy in this setting.
Three randomized controlled trials performed in healthy African men showed that male circumcision was effective in preventing HIV acquisition. In studies performed in Uganda, South Africa, and Kenya, men were randomized to be offered immediate or delayed (at 24 months) circumcision and followed up for 2 years for acquisition of HIV infection and other STDs [55
]. The summary rate ratio for reduction of HIV acquisition in the men who underwent immediate circumcision for the 3 trials was 0.42 (95% CI, .31–.57), identical to that obtained from observational studies, which translates into a protective effect of male circumcision of 58% [56
]. On the basis of these findings, a WHO and UNAIDS consultation in March 2007 recommended that circumcision be recognized as an effective intervention for preventing heterosexual HIV acquisition in men [59
]. WHO and UNAIDS also recommended that male circumcision be offered to HIV-negative men in addition, but not as a substitute, to other HIV risk-reduction strategies.
Circumcision also affords a similar level of protection against acquisition of other STIs, particularly nonulcerative pathogens, such high-risk genital HPV infection, and also against acquisition of genital herpes [60
]. In South Africa, after 21 months of follow-up, circumcision protected against high-risk HPV infection (OR, 0.57; 95% CI, 0.43–0.75), but not gonorrhea [60
]. The association between trichomoniasis and male circumcision remained borderline when age, ethnic group, number of lifetime partners, marital status, condom use, and HIV status were controlled for (AOR, 0.48; P
= .069). In the as-treated analysis, this association became significant (OR, 0.49 [P
= .030]; AOR, 0.41 [P
= .03]). The authors concluded that male circumcision reduces incident trichomoniasis among men. Men in Uganda were also followed up for acquisition of STDs for 2 years. At 24 months, the cumulative probability of HSV-2 seroconversion was 7.8% in men randomized to circumcision (1684 men who were initially HSV-2 seronegative) and 10.3% in the control group (1709 men who were initially HSV-2 seronegative) (AHR, 0.72; 95% CI, .56–.92; P
= .008) [62
]. The prevalence of high-risk HPV genotypes was 18.0% in the intervention group and 27.9% in the control group (adjusted risk ratio, 0.65; 95% CI, .46–.90; P
= .009). However, no significant difference between the 2 study groups was observed in the incidence of syphilis (AHR, 1.10; 95% CI, .75–1.65; P
= .44). Among men enrolled in the Kenya study, circumcision afforded no protection against incident gonorrhea, chlamydia, or trichomoniasis [40
No randomized controlled trials of circumcision have been performed among men in the United States However, a cross-sectional analysis reported that among 394 heterosexual African- American men attending a Baltimore STD clinic who reported known HIV exposure, circumcision was significantly associated with lower HIV prevalence (10.2% vs 22.0%; adjusted prevalence rate ratio [PRR], 0.49; 95% CI, .26–.93). No such association was seen for men with unknown HIV exposure [63
The benefits of circumcision to MSM are unproved. A meta-analysis of studies reported that overall, circumcised MSM had lower odds of being infected with HIV (OR, 0.86; 95% CI, .65–1.10), an association that did not reach statistical significance and that was similar among men who reported primarily engaging in insertive anal sex [64
]. Among the 4889 participants in the VaxGen rgp 120 HIV vaccine study (87.4% of whom were circumcised), 342 men (7.0% of all participants) acquired HIV while enrolled; being uncircumcised was not associated with incident HIV infection (AHR, 0.97; 95% CI, .56–1.68) among men who reported unprotected insertive anal sex with HIV-infected partners [65
]. However, studies conducted prior to the introduction of highly active antiretroviral therapy demonstrated a significant inverse association of circumcision with HIV infection (OR, 0.47; 95% CI, .32–.69) [64
]. No significant effects were seen for non-HIV STIs, and the investigators concluded that more data were needed.
Unfortunately, the benefits of male circumcision in reducing HIV acquisition in men do not extend to women; however, other benefits may occur. Female sex partners of men who participated in the Uganda circumcision trial were followed up to assess effects on their genital symptoms and vaginal infections [66
]. Among women with normal vaginal flora scores at enrollment, rates of bacterial vaginosis (BV) at follow up were significantly lower in wives of men who had been circumcised compared with men who had not (prevalence risk ratio [PRR], 0.80; 95% CI, .65–.97). In women with BV at enrollment, persistent BV at 1 year was significantly lower in the intervention arm than control arm women (PRR 0.83; 95% CI, .72–.96). The adjusted prevalence risk ratio of genital ulcer disease among wives of circumcised men compared with uncircumcised men was 0.78 (95% CI, .61–.99), consistent with circumcision efficacy of 22%. The adjusted prevalence risk ratio for trichomoniasis in intervention arm wives relative to controls was 0.55 (95% CI, .34–.89; efficacy 45%). The authors concluded that male circumcision may have direct benefits for prevention of genital ulceration, trichomoniasis, and BV in female partners and that this should be considered when plans are made to scale up of male circumcision programs to prevent HIV infection.
Implementation of male circumcision as a strategy for preventing HIV infection remains to be fully defined. Concerns include possible disinhibitory effects on sexual risk behaviors, complications from unsafe or inexperienced providers, and acceptability by substantial numbers of men at highest risk for HIV infection [67
]. Male circumcision is a complement to, not a substitute for, other HIV risk-reduction strategies. WHO and UNAIDS recommend that countries with hyperendemic and generalized HIV epidemics and low prevalence of male circumcision expand access to safe male circumcision services within the context of ensuring universal access to comprehensive HIV infection programs, including prevention, treatment, care, and support.
Nonspecific Topical Agents
In general, results of topical agents with nonspecific antimicrobial activity for the prevention of HIV and other STDs have been disappointing [67
]. A randomized controlled trial comparing vaginal application of 0.5% PRO 2000 (a synthetic polyanionic polymer that blocks attachment of HIV to the host cell) with BufferGel (a vaginal buffering agent), placebo gel, and condom use only found that PRO 2000 was associated with a 30% reduction in risk of HIV acquisition relative to no gel use (AHR, 0.70; 95% CI, .46–1.08; P
= .10) or placebo gel use (AHR, 0.67; 95% CI, .44–1.02; P
= .06) [69
]. However, a considerably larger study (the MDP301 trial, conducted in 4 sub-Saharan African countries) assessing 0.5% PRO2000 relative to placebo gel found no protective effect [70
Other topical polyanion agents have not fared well either. A randomized controlled trial compared coitally dependent use of Carraguard (a carrageenan derivative with in vitro activity against HIV) with methylcellulose gel placebo among South African women at high risk for HIV infection. After 2 years follow-up, the incidence of HIV infection in the Carraguard group (n = 3011) was 3.3/100 woman-years, compared with 3.8/100 woman-years in the placebo group (n = 2994) (AHR, 0.87; 95% CI, .69–1.09). Applicator dye testing—one means of measuring actual vaginal insertion of the product—indicated that adherence to product was low (42% of sex acts overall). Self-reported product use was substantially higher than the estimate obtained from applicator testing, and some investigators have reported low accuracy for applicator dye testing [71
Two randomized controlled trials compared daily 6% cellulose sulfate (an HIV entry inhibitor) vaginal gel with corresponding placebo. A multicountry trial enrolled 1398 African women at high risk for HIV infection. Twenty-five newly acquired HIV infections occurred in the cellulose sulfate group, and 16 in the placebo group, with an estimated hazard ratio (HR) of infection for the cellulose sulfate group of 1.61 (P
= .13). This result, which is not significant, is in contrast to the interim finding that led to the trial’s being stopped prematurely (HR, 2.23; P
= .02) and the suggestive result of a preplanned secondary (adherence-based) analysis (HR, 2.02; P
= .05). Compared with placebo, cellulose sulfate had no significant effect on the risk of gonorrhea (HR, 1.10; 95% CI, .74–1.62) or chlamydia (HR, 0.71; 95% CI, .47–1.08). The authors concluded that cellulose sulfate did not prevent and may have increased the risk of HIV acquisition [73
]. A second randomized, placebo-controlled trial of cellulose sulfate in Nigeria was stopped prematurely after the data safety monitoring board of the multicountry trial concluded that cellulose sulfate might be increasing the risk of HIV acquisition [73
]. With the limited data available, cellulose sulfate gel appeared not to prevent transmission of HIV, gonorrhea, or chlamydial infection.
Two trials of the effectiveness of 1.0% C31G (Savvy; a surfactant) in preventing HIV acquisition were similarly disappointing. In the first, more women in the SAVVY group than in the placebo group reported reproductive tract adverse events [75
]. In the second, 33 seroconversions (21 in the SAVVY group and 12 in the placebo group) occurred in the 2153 participants. The cumulative probability of HIV seroconversion was 2.8% in the SAVVY group and 1.5% in the placebo group (P
= .121), with an HR of 1.7 for SAVVY versus placebo (95% CI, .9–3.5). [76
]. The trials indicated that SAVVY did not reduce the incidence of HIV infection and may have been associated with increased risks. Taken together, these studies do not support further testing of polyanion-type compounds with nonspecific activity against STDs and HIV.
Observational studies demonstrate that diaphragm use protects against cervical gonorrhea, chlamydia, and trichomoniasis [51
]. The MIRA trial examined the effect of a diaphragm plus polycarbophil (Replens) lubricant on HIV acquisition in women in Zimbabwe and South Africa. The authors found no additional protective effect of latex diaphragm, lubricant gel, and condoms on HIV acquisition compared with condoms alone [77
]. A subsequent analysis of data from this study evaluated outcomes of chlamydia and gonorrhea. [78
] Median follow-up time was 21 months, and the retention rate was >93%. A total of 471 first chlamydia infections occurred, 247 in the intervention arm and 224 in the control arm, with an overall incidence of 6.2/100 woman-years (relative hazard [RH], 1.11; 95% CI, .93–1.33) and 192 first gonococcal infections, 95 in the intervention arm and 97 in the control arm with an overall incidence of 2.4/100 woman-years (RH, 0.98; 95% CI, .74–1.30). Per-protocol results indicated that when diaphragm adherence was defined as ‘‘always use’’ since the last visit, a significant reduction in gonorrhea incidence occurred among women randomized to the intervention (RH, 0.61; 95% CI, .41–.91). The authors concluded that, although no difference by study arm was found in the rate of acquisition of chlamydia or gonorrhea, per-protocol, results suggested that consistent use of the diaphragm may reduce acquisition of gonorrhea.
Another analysis from the MIRA trial estimated the diaphragm’s effect on the incidence and clearance of HPV infection in women in Zimbabwe [79
]. There was no overall difference in incidence at the first postenrollment visit or at 12 months, or in HPV clearance at 12 months among women who were HPV positive at enrollment. However, clearance of HPV type 18 was lower in the diaphragm group at the exit visit (relative risk[RR], 0.55; 95% CI, .33–.89) but not at 12 months. Women reporting diaphragm and gel use at 100% of prior sex acts had a lower likelihood of having ≥1 new HPV type detected at 12 months (RR, 0.75; 95% CI, .58–.96). The authors concluded that diaphragms did not reduce the incidence of HPV infection or increase clearance. Diaphragms should not be relied on as the sole source of protection against STIs or HIV infection. Diaphragms used with nonoxynol-9 (N-9) spermicides have been associated with an increased risk for bacterial urinary tract infections in women.