ACCEPTED. Objective: We derived an estimate of male condom effectiveness during anal sex among men

Transcription

1 JAIDS Journal of Acquired Immune Deficiency Syndromes Publish Ahead of Print DOI: /QAI Title: Condom Effectiveness for HIV Prevention by Consistency of Use among Men Who Have Sex with Men (MSM) in the U.S. Authors: Dawn K. Smith 1, Jeffrey H. Herbst 1, Xinjiang Zhang 1, Charles E. Rose 1 Affiliations: 1 Division of HIV/AIDS Prevention, NCHHSTP, Centers for Disease Control and Prevention, Atlanta, GA Running Head: MSM Condom Effectiveness Disclaimer: The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention. Conflict of Interest: None declared Sources of Funding: None declared Corresponding Author and Reprint Requests: Dr. Dawn K. Smith Centers for Disease Control and Prevention 1600 Clifton Rd, MS E-45 Atlanta, GA Dsmith1@cdc.gov Abstract Objective: We derived an estimate of male condom effectiveness during anal sex among men who have sex with men (MSM) because the most widely used estimate of condom effectiveness (80%) was based on studies of persons during heterosexual sex with an HIV-positive partner. Design: Assessed male condom effectiveness during anal sex between MSM in two prospective cohort studies of HIV incidence by self-reported consistency of use. Methods: Analyzed data combined from US participants in the EXPLORE trial ( ) public use dataset and in the VAX 004 trial ( ) dataset. Initially HIV-uninfected MSM enrolled in these trials completed baseline and semiannual interviews about their sexual

2 2 behaviors with male partners and underwent HIV testing. Using a time-to-event model, effectiveness of consistent condom use in preventing HIV infection was estimated among men reporting receptive and/or insertive anal sex with an HIV-positive partner, and consistency of condom use. Results: Among MSM reporting any anal sex with an HIV-positive male partner, we found 70% effectiveness with reported consistent condom use (compared to never use) and no significant protection when comparing sometimes use to never use. This point estimate for MSM was less than the 80% effectiveness estimate reported for heterosexuals in HIV discordant couples reporting consistent condom use. However, the point estimates in the two populations are not statistically different. Only 16% of MSM reported consistent condom use during anal sex with male partners of any HIV status over the entire observation period. Conclusions: These estimates are useful for counseling efforts, and for modeling the impact and comparative effectiveness of condoms and other prevention methods used by MSM. Keywords: MSM, HIV, condom, effectiveness

3 3 Introduction Gay, bisexual, and other men who have sex with men (collectively referred to as MSM) continue to be disproportionately affected by HIV in the US. While MSM represent approximately 2% of the US adult population 1, they accounted for 62% of all newly diagnosed HIV infections and 79% of newly diagnosed HIV infections among adolescent and adult males in Sexual risk behaviors account for most HIV infections in MSM, and receptive anal intercourse without condom use has the highest risk for HIV acquisition among them 3-5. Common HIV-related risk factors among MSM include multiple concurrent sex partners 6,7, condomless anal intercourse with unknown serostatus or presumed HIV-negative partners 8,9, use of alcohol or drugs before or during sex 10, and inconsistent or incorrect condom use 11. Reducing the number of new HIV infections among MSM is a public health priority and the use of male latex condoms for the prevention of sexually acquired HIV infection is a critical component of prevention efforts in the US 12 and globally 13. Latex condoms provide a nearly impermeable barrier to HIV 14, and epidemiological studies of HIV-discordant heterosexual couples, in which one partner has HIV infection and one does not, indicate that correct and consistent condom use is highly effective in preventing HIV infection 15,16. Estimates of condom effectiveness have been reported in the literature, but mainly among heterosexual HIVdiscordant couples 15 and in cohort studies of clinic patients with diagnosed sexually transmitted infections (STIs) 17,18 that sometimes included an undefined minority of MSM 19. An analysis of HIV seroconversion studies published from found condoms to be 90 to 95% effective when used consistently during penile-vaginal intercourse 20. A large randomized community prospective study of 17,264 sexually active men and women in rural Uganda found

4 4 consistent condom use during penile-vaginal intercourse reduced HIV incidence by 63% (95% CI, 12%-85%), whereas inconsistent condom use offered no protection 21. A Cochrane review and meta-analysis of 14 cohort studies found consistent use of condoms accounted for an estimated 80% reduction in HIV seroconversion among heterosexuals with HIV-positive partners 15. In addition, recent studies show that consistent and correct use of condoms is highly effective in reducing the risk of acquiring other STIs among men and women 18. In contrast to studies of heterosexuals, a review published in yielded only one study that reported an estimate of the effectiveness of condom use during penile-anal intercourse among MSM. In a cohort of 2915 HIV-seronegative MSM from four centers of the Multicenter AIDS Cohort Study followed every 6 months for 24 months, the HIV seroconversion rate was significantly lower among men who reported using condoms with all of their sex partners compared to men who reported using condoms with none of their sex partners (0.7% vs. 2.4%, respectively) 23. Based on these incidence estimates, condom effectiveness for always use versus never use is computed to be 70.8%. A recent meta-analysis by the World Health Organization of 5 observational studies assessing condom effectiveness among MSM in developing countries, estimated a 64% reduction in HIV acquisition risk (95% CI, 32%-80%) 24. There is a critical need for studies to investigate condom effectiveness according to consistency of use during penile-anal intercourse among MSM. This paper reports analyses to estimate condom effectiveness among MSM based on data from two large prospective HIV prevention trials that assessed condom use and HIV serostatus at standard intervals over multiple years. The paper also compares the condom effectiveness estimate for MSM based on the current analyses

5 5 to the published estimate of condom effectiveness (80%) from a meta-analysis of cohort studies of heterosexuals 15. Methods We identified existing datasets from two HIV prevention trials that enrolled HIV-uninfected MSM in the US, which collected both risk behavior data and HIV test results at standard intervals (every 6 months), and had at least 100 HIV seroconversions over the course of the trial. With data from the VAX and Project EXPLORE trials 8, in each dataset we modeled condom effectiveness in preventing HIV seroconversion among MSM engaging in receptive or insertive anal intercourse with at least one HIV-positive partner. Our analyses focused on estimating the effectiveness of condoms in reducing HIV incidence among MSM who reported they always used condoms (i.e., consistent users) compared both to those who reported they never used condoms and to those who reported sometimes using condoms. We estimated condom effectiveness in each dataset separately and in both datasets combined. The VAX 004 trial enrolled HIV-uninfected MSM from 57 US sites during into a randomized placebo-controlled trial of a recombinant HIV-1 envelope glycoprotein subunit (rgp120) vaccine that proved ineffective at 36 months of follow-up 25. We excluded participants who reported injection drug use at any point before or during the trial, non-us participants, participants who were found to have HIV infection at baseline, and participants with no study visit after enrollment, as well as visit records during any interval during which participants reported having sex with women. Self-reported condom use was assessed by use of standard interviewer-administered questionnaires at baseline and every 6 months thereafter (see

6 6 Supplemental Digital Content, Table 1). HIV-1 status was determined at each study visit by a standard HIV-1 enzyme-linked immunosorbent assay; enzyme immunoassay-positive tests underwent confirmatory immunoblot assay; and for those confirmed antibody positive at any visit, a stored serum specimen from the prior study visit was tested by a highly sensitive and specific nucleic acid-based amplification test. After applying the exclusion criteria, 4,492 MSM with 23,567 visit records were eligible for analyses, of whom 315 (7.0%) acquired HIV infection. Among all men in the analyses, 2,167 MSM (48% of the included sample) reported any anal sex with an HIV-positive partner at any time during the study, of whom 154 (7.1%) acquired HIV infection. Of men reporting any anal sex with an HIV-positive partner, 1,343 (62%) reported any receptive anal sex with an HIV-positive partner, and 1,930 (89%) reported any insertive anal sex with an HIV-positive partner (see Supplementary Digital Content, Figure S1). Project EXPLORE enrolled HIV-uninfected MSM from six US cities during into a randomized, controlled behavioral intervention trial in which the active intervention arm did not significantly reduce HIV incidence compared with the control intervention arm 8. We excluded participants who reported injection drug use at any point before or during the trial, participants who were found to have HIV infection at baseline, participants with no study visit after enrollment, as well as visit records for any intervals during which participants reported having sex with women. At baseline and follow-up visits every 6 months, audio computer-assisted selfinterviewing was used to collect sexual behaviors (Table S1). At each visit, blood samples were tested for HIV antibodies by enzyme-linked immunosorbent assay; if positive, specimens were

7 7 retested in duplicate, and repeatedly reactive samples were confirmed by Western blot assay or immunofluorescence assay. After applying the exclusion criteria, 3,233 MSM with 18,884 visit records were eligible for analyses; of whom 195 (6.0%) acquired HIV infection. Among all men in the analyses, 1,323 MSM (41% of the included sample) reported any anal sex with an HIV-positive partner at any time during the study, of whom 71 (5.4%) acquired HIV infection. Of men reporting any anal sex with an HIV-positive partner, 871 (66%) reported any receptive anal sex with an HIVpositive partner, and 1095 (83%) reported any insertive anal sex with an HIV-positive partner (see Supplementary Digital Content, Figure S2). In analyses that used the combined trial datasets, after exclusions above, we included 7,725 MSM with 42,451 visits before selecting only those who reported sex with any HIV-positive partner at any study visit. There are 3,490 MSM who reported sex with an HIV-positive partner during 9,246 visits, and among these men 225 acquired HIV infection (6.4%) (see Supplementary Digital Content, Figure S3). We included MSM in both the active and control arms of the trials. We believe this is appropriate as: (1) our analysis was designed to determine the effectiveness of self-reported condom use on HIV incidence regardless of which intervention was received, and (2) in both trials, statistically significant differences in HIV incidence were not found when comparing the intervention and control arms. The VAX 004 and Project EXPLORE questionnaires included

8 8 comparable variables to assess condom use by consistency (always, sometimes and never) and by partner serostatus (see Supplementary Digital Content, Table S1). All analyses were conducted using three analysis datasets: EXPLORE, VAX 004, and EXPLORE and VAX 004 combined. We compared the distribution of demographic characteristics for the EXPLORE and VAX 004 datasets using chi-square test statistics. The demographic factors common to both datasets are age (categorized as 25, 26 30, 31 35, 36 years), race/ethnicity (categorized as non-hispanic white, non-hispanic black, Hispanic, Asian/Pacific Islander, other/unknown), education (categorized as < high school, high school, college, graduate school), and region (categorized as Northeast, Midwest, South, and West). We calculated the percentage of condom use during anal sex acts for each 6-month interval and the total number of HIV-positive partners by receptive, insertive, and the combined receptive and insertive anal sex categories for both the EXPLORE and VAX 004 data. We computed the HIV incidence per 100 person years (PY), where PY was calculated by summing over all intervals for all subjects in the condom risk group categories of never, sometimes, and always. In addition, we calculated the total number of anal sex acts per condom risk group category (always, sometimes, or never use) among men reporting an HIV-positive sex partner. The incidence and corresponding 95% confidence interval (CI) were calculated using a Poisson regression model with the PY as the offset in the Poisson model. The risk per sex act and 95% CI were estimated using a logistic regression model. In addition, the condom effectiveness based on risk per sex act and 95% CI for always versus never condom use was calculated using a log-binomial model.

9 9 We estimated condom effectiveness using receptive anal sex only, insertive anal sex only, and combined receptive and anal sex. We categorized condom use consistency for each subject for each 6 month interval as either, always, sometimes, or never based on the percentage of time they used a condom during an anal insertive or receptive sex act. To be categorized as always / never condom use the subject must have reported condom use during all or none of the anal sex acts (respectively) with the HIV-positive partner during the 6-month interval. Subjects were categorized as sometimes condom use if the proportion of time they used a condom during a 6 month interval was between 0 and 1. For each 6-month interval, a subject was coded as HIV-positive or HIV-negative based on serological test result at the visit marking the end of the prior study interval and analyzed using an interval censored time to event model. A complementary log-log (CLL) model was used to model the interval censored time to HIV seroconversion. The hazard ratio (HR) was estimated for comparison of the different categories of condom use, and effectiveness was defined as 100*(1 HR). We estimated condom effectiveness using both univariable and multivariable models. The univariable models only included the condom consistency group and interval. The multivariable model additionally controlled for age, race/ethnicity, education, and region regardless of statistical significance in the univariable models. All analyses were conducted using SAS version 9.2, and all p-values < 0.05 were considered statistically significant. A secondary analysis was conducted to compare the condom effectiveness estimates identified in this study for MSM during anal intercourse with HIV-positive partners to published condom effectiveness estimates based on 14 prospective studies for heterosexual men and women in HIV

10 10 discordant couples reported in the Cochrane review 15. The Cochrane review reported two estimates of condom effectiveness: one based on a meta-analysis of all studies reporting effectiveness results of always use versus never use, and a second based on a subset of studies with longer than average follow-up times. We compared our unadjusted estimate of condom effectiveness (i.e., not adjusting for study factors of age, race/ethnicity, education and region) to the two Cochrane estimates using a Poisson model with PY as the offset and study, condom use ( always, never ), and the interaction of study and condom use in both models. Results The demographic characteristics of the combined trial populations indicated the majority of participants were aged 31 years or older at enrollment, non-hispanic white, and reported having a college diploma or more advanced degree (Table 1). However, participants in Project EXPLORE were significantly more likely to be younger, non-white, report a higher education level, and reside in the Northeast or West than participants in the VAX 004 trial. Condom Effectiveness Per-Act Transmission Risk Findings The risk of HIV infection per act of anal intercourse reported with an HIV-positive partner was significantly higher for never condom use than always condom use in both trials for reported receptive anal sex (EXPLORE: vs p-value<0.001; VAX 004: vs p-value=0.0032) and for any anal sex (receptive or insertive) (EXPLORE: vs p- value<0.0001; VAX 004: vs p-value=0.0043) (Table 2). The per sex act risk of HIV infection was significantly higher for never condom use than for always condom use in

11 11 Project EXPLORE participants during reported insertive anal sex ( vs p- value=0.025) but not for VAX 004 participants ( vs p-value=0.051). Condom Effectiveness Estimate In the multivariable analysis of data combined from the two trials, controlling for study, age, race/ethnicity, education, and region of residence, the effectiveness of condom use for men reporting anal sex with an HIV-positive male partner was estimated comparing self-reported always use to never use. Estimated condom effectiveness was 72.3% (95% CI, 60.7%- 80.5%) for receptive anal sex, 62.9% (95% CI, 46.3%-74.3%) for insertive anal sex, and 70.5% (95% CI, 58.2%-79.2%) for any receptive or insertive anal sex (Table 3). The estimated effectiveness for condom use during insertive anal sex was not statistically different from that for receptive anal sex (p-value=0.17). Self-reported always use to sometimes use was significantly protective in multivariable analyses in each trial separately and when combined for any anal sex reported with HIV-positive partners. Condom effectiveness comparing sometimes use and never use was not significantly protective in either trial or both combined for any anal sex reported. All estimates of condom effectiveness were lower in VAX 004 than in EXPLORE but the pattern across sexual activity was the same (Table 3). While all analyses reported here included only US MSM, in a sensitivity analysis, when data from men at non-us sites in the VAX 004 trial were included, there was no statistically

12 12 significant difference in estimated condom efficacy for men reporting any receptive or insertive anal sex (data not shown). Consistency of Condom Use Table 4 presents the results of analyses of condom use consistency for reported partners of any HIV status (positive, negative or unknown) by trial and anal sex act. Among MSM in the two trials combined with male partners of any HIV status, for at least one 6-month interval, 65.0% reported always condom use, 73.8% sometimes condom use, and 40.2% never condom use for receptive or insertive anal sex. Across all their study visits, 16.4% reported always use, 14.3% sometimes use, and 4.9% never use. These rates were similar for MSM in each of the two trials (Table 4). As shown in Figure 1, for each 6-month study interval, the proportion of MSM reporting condom use during any anal sex with HIV-positive or unknown status partners was consistently higher than that reported with partners believe to be HIV-negative. Comparison of Findings to Cochrane Review Estimated condom effectiveness estimates for consistent use based on HIV incidence rates were reported in the Cochrane review 15 as 82.9% (95% CI 66.7 to 91.2%) for the entire set of included studies and 80.2% (95% CI 56.3 to 91.0%) for the subset of studies reporting longer follow-up times. [Note that the 95% CIs were not published but were calculated from the data provided in the Cochrane review]. The adjusted (Table 3) estimated effectiveness of condom use among MSM with HIV-positive partners based on estimated HIV incidence in the combined EXPLORE

13 13 and VAX 004 samples is 69.2% (95% CI, 58.0 to 80.3%). Our estimated effectiveness did not differ significantly from the Cochrane review estimate for all studies (estimated percentage difference is , 95% CI , p-value=0.10) or the subset of studies reporting longer follow-up times (estimated percentage difference is , 95% CI , p-value=0.27). Discussion Given that the proportion of new HIV infections in the US annually is greatest for MSM, it is important to have robust and accurate estimates of the effectiveness of condoms and other prevention methods that are specific to this population. This study was able to assess condom effectiveness among MSM who reported having HIV-positive partners in multiple sites in two large prospective cohorts. The point estimate in our analysis of condom effectiveness when always used by MSM during anal sex with any HIV-positive male partners is 70%, modestly less than the 80% estimate for condoms when always used by heterosexual HIV-discordant couples. Although these point estimates do not differ by tests of statistical significance, it is more appropriate to use the MSMspecific point estimate of 70% effectiveness for discussions and models involving anal sex among MSM than to continue use of the heterosexual 80% effectiveness point estimate for MSM. This study, found that inconsistent ( sometimes ) condom use with HIV-positive male partners over months to years offers minimal or no protection, underscoring the importance of the inclusion of messaging by their HIV prevention providers that supports the adoption and

14 14 continuation of consistent condom use for MSM. In these analyses and in many prior epidemiologic studies, inconsistent condom use with partners of any HIV status is more frequently reported by MSM than consistent condom use 6,26. While condom use is often measured over short time frames (e.g., at last sex, in past 3 months), rates of always use fall significantly when measured over longer time frames 7,19. HIV prevention providers should inform MSM who are not using condoms consistently of the low protection offered by inconsistent use and additional prevention options should be considered. There are limitations in the data available that affect this analysis. Self-reports of condom use are not validated, nor is self-report of partner HIV status. It is likely that some partners believed to have been HIV-negative may have unrecognized HIV infection. No data are available about the treatment status of sex partners reported to HIV infection. The analysis therefore cannot account for HIV acquisition through sex with partners who had unrecognized HIV infection (i.e., were reported as negative or unknown HIV status) or those whose partners had suppressed viral load. In addition, because most MSM engaged in multiple sexual behaviors, it is not possible to assess condom efficacy among MSM who had only one type of sexual behavior (e.g., oral sex but no anal sex, insertive anal sex but no oral or receptive anal sex). Nonetheless, we provide a robust estimate of 70% effectiveness of condoms for prevention of HIV acquisition among MSM reporting anal sex with HIV-positive male partners (with or without other sexual behaviors). This finding is consistent with the prior understanding that for all sexual exposures, among both MSM and heterosexuals, condoms offer substantial but partial protection against HIV acquisition when they are consistently used.

15 15 Our analysis also found that among MSM in the two trials, all of whom received sustained behavioral interventions, a low proportion (16.4%) reported always (consistent) condom use for all sexual events over a year or longer regardless of partner HIV status. MSM rarely reported anal sex with only HIV-positive partners. However, the analysis found that the frequency of condom use is greater with partners reported to have HIV infection or unknown HIV status and lower with partners of negative HIV status. The substantial protective effectiveness of condoms cannot be fully realized without their consistent use over prolonged periods of sexual activity and potential HIV exposure. Our findings of strong but partial effectiveness for consistent condom use, minimal effectiveness for inconsistent condom use, and low rates of consistent condom use over 1-2 years, even among persons receiving high quality risk-reduction counseling, may contribute to a better understanding of the persistent rates of HIV infection among MSM in the US, despite current levels of condom use promotion and provision. A recent analysis of data from MSM who were interviewed in the 20 cities participating in the National HIV Behavioral Surveillance System, found that reported anal sex without condoms anytime during the prior 12 months increased from 45% in 2005, to 54% in 2008, and 57% in 2011 (Chi-square test for trend, p-value <0.001) 27. From a public health perspective, these data suggest the need to further intensify efforts to educate HIV-uninfected MSM and improve their ability to accurately assess both their risk of HIV acquisition and the effectiveness of their current use of condoms (consistently or inconsistently) during anal sex. These data also suggest a need to provide access to additional highly-effective HIV prevention tools and strategies, including more frequent HIV testing (including testing as couples) and daily oral antiretroviral preexposure prophylaxis (PrEP) when

16 16 indicated 28. In addition, these findings may be useful to researchers modeling the impact and cost-effectiveness of behavioral interventions whose outcomes are measured by changes in reported condom use, and to those modeling combinations of biomedical and condom-based behavioral prevention methods to reduce the rate of new HIV infections among MSM. Acknowledgements We thank the study teams and participants of the VAX 004 and EXPLORE trials,

17 17 References 1. P DW, Johnson CH, Lansky A, et al. Estimating the Population Size of Men Who Have Sex with Men in the United States to Obtain HIV and Syphilis Rates. The Open AIDS Journal. 2012;6(Suppl 1: M6): Centers for Disease Control and Prevention. HIV Surveillance Report, ;23: Accessed 12 May Buchbinder SP, Glidden DV, Liu AY, et al. HIV pre-exposure prophylaxis in men who have sex with men and transgender women: a secondary analysis of a phase 3 randomised controlled efficacy trial. The Lancet infectious diseases. 2014;14(6): Varghese B, Maher JE, Peterman TA, Branson BM, Steketee RW. Reducing the risk of sexual HIV transmission: quantifying the per-act risk for HIV on the basis of choice of partner, sex act, and condom use. Sexually transmitted diseases. 2002;29(1): Patel P, Borkowf CB, Brooks JT, Lasry A, Lansky A, Mermin J. Estimating per-act HIV transmission risk: a systematic review. AIDS. 2014;28(10): Rosenberg ES, Khosropour CM, Sullivan PS. High Prevalence of Sexual Concurrency and Concurrent Unprotected Anal Intercourse Across Racial/Ethnic Groups Among a National, Web- Based Study of Men Who Have Sex With Men in the United States. Sexually Transmitted Diseases. Oct 2012;39(10): Crosby R, Shrier LA, Charnigo RJ, et al. A Prospective Event-Level Analysis of Condom Use Experiences Following STI Testing Among Patients in Three US Cities. Sexually Transmitted Diseases. 2012;39(10): Koblin B, Chesney M, Coates T. Effects of a behavioural intervention to reduce acquisition of HIV infection among men who have sex with men: the EXPLORE randomised controlled study. Lancet. Jul ;364(9428): Baggaley RF, White RG, Boily MC. HIV transmission risk through anal intercourse: systematic review, meta-analysis and implications for HIV prevention. Int J Epidemiol. Aug 2010;39(4): Carey JW, Mejia R, Bingham T, et al. Drug Use, High-Risk Sex Behaviors, and Increased Risk for Recent HIV Infection among Men who Have Sex with Men in Chicago and Los Angeles. Aids Behav. Dec 2009;13(6): Smith DK, Pals SL, Herbst JH, Shinde S, Carey JW. Development of a clinical screening index predictive of incident HIV infection among men who have sex with men in the United States. J Acquir Immune Defic Syndr. Aug 12012;60(4): Centers for Disease Control and Prevention. Condoms and STDs: Fact Sheet for Public Health Personnel. 2013; Accessed 12 September World Health Organization. Condoms for HIV prevention. 2010; Accessed 12 September Carey RF, Lytle CD, Cyr WH. Implications of laboratory tests of condom integrity. Sexually Transmitted Diseases. Apr 1999;26(4): Weller S, Davis K. Condom effectiveness in reducing heterosexual HIV transmission. Cochrane Database Syst Rev. 2002(1):CD Holmes KK, Levine R, Weaver M. Effectiveness of condoms in preventing sexually transmitted infections. Bulletin of the World Health Organization. 2004;82(6): Shlay JC, McClung MW, Patnaik JL, Douglas JMJ. Comparison of Sexually Transmitted Disease Prevalence by Reported Level of Condom Use Among Patients Attending an Urban Sexually Transmitted Disease Clinic. Sexually Transmitted Diseases. 2004;31(3):

18 Warner L, Stone KM, Macaluso M, Buehler JW, Austin HD. Condom Use and Risk of Gonorrhea and Chlamydia: A Systematic Review of Design and Measurement Factors Assessed in Epidemiologic Studies. Sexually Transmitted Diseases. 2006;33(1): Peterman TA, Tian LH, Warner L, et al. Condom use in the year following a sexually transmitted disease clinic visit. Int J STD AIDS. Jan 2009;20(1): Pinkerton SD, Abramson PR. Effectiveness of condoms in preventing HIV transmission. Soc Sci Med. May 1997;44(9): Ahmed S, Lutalo T, Wawer M, et al. HIV incidence and sexually transmitted disease prevalence associated with condom use: a population study in Rakai, Uganda. AIDS. Nov ;15(16): Silverman BG, Gross TP. Use and effectiveness of condoms during anal intercourse. A review. Sex Transm Dis. Jan 1997;24(1): Detels R, English P, Visscher BR, et al. Seroconversion, sexual activity, and condom use among 2915 HIV seronegative men followed for up to 2 years. Journal of acquired immune deficiency syndromes. 1989;2(1): World Health Organization. Prevention and treatment of HIV and other sexually transmitted infections among men who have sex with men and transgender people. 2011; Pitisuttithum P, Gilbert P, Gurwith M, et al. Randomized, Double-Blind, Placebo-Controlled Efficacy Trial of a Bivalent Recombinant Glycoprotein 120 HIV-1 Vaccine among Injection Drug Users in Bangkok, Thailand. Journal of Infectious Diseases,. 2006;194(12): Hakre S, Armstrong AW, O'Connell RJ, Michael NL, Scott PT, Brett-Major DM. A Pilot Online Survey Assessing Risk Factors for HIV Acquisition in the Navy and Marine Corps, JAIDS Journal of Acquired Immune Deficiency Syndromes. 2012;61(2): /QAI.1090b1013e31826a31815db. 27. Paz-Bailey G, Hall HI, Wolitski RJ, et al. HIV Testing and Risk Behaviors Among Gay, Bisexual, and Other Men Who Have Sex with Men - United States. Mmwr-Morbid Mortal W. Nov ;62(47): Smith DK, Grant RM, Weidle PJ, Lansky A, Mermin J, Fenton KA. Interim Guidance: Preexposure Prophylaxis for the Prevention of HIV Infection in Men Who Have Sex With Men (Reprinted from MMWR, vol 60, pg 65-68, 2011). Jama-J Am Med Assoc. Mar ;305(11):

19 1 Table 1. Demographic characteristics of men who have sex with men (MSM) in the EXPLORE and VAX 004 analysis datasets EXPLORE VAX 004 Combined Characteristic (N = 1323) (N = 2167) (N = 3490) Age Subjects Percent Subjects Percent P-value Subjects Percent < Race/Ethnicity Non-Hispanic White Non-Hispanic Black < Hispanic Asian / Pacific Islander Other / Unknown Education < High School High School < College Graduate School June 2014

20 2 Region Northeast Midwest < South West June 2014

21 3 Table 2. HIV incidence by condom risk group for men who have sex with men (MSM) with HIV-positive sex partners in the EXPLORE and VAX 004 studies. Receptive Anal Sex EXPLORE HIV+ 1 HIV- 2 Person 3 Years Incidence 4 (100 PY) and 95% CI Sex Acts Risk per sex act and 95% CI 5 Never Condom use Sometimes Condom use Always Condom use Total Insertive Anal Sex Never Condom use Sometimes Condom use Always Condom use Total Inserive+Receptive Anal Sex 6 Never Condom use June 2014

22 4 Sometimes Condom use Always Condom use Total VAX 004 Receptive Anal Sex Never Condom use Sometimes Condom use Always Condom use Total Insertive Anal Sex Never Condom use Sometimes Condom use Always Condom Total Insertive+Receptive Anal Sex 6 Never Condom use Sometimes Condom use June 2014

23 5 Always Condom use Total HIV+ is the number of subjects that seroconverted during an interval. 2 HIV- is the total number of subjects per interval per condom use group that did not seroconvert during an interval. 3 Person years are the total number of person years observed for all HIV+ and HIV- subjects. 4 Incidence is the rate of HIV+ per 100 person years. 5 Risk /sex act is the probability of HIV seroconversion per sex event. 6 Combined is the receptive plus insertive analysis. 30 June 2014

24 6 Table 3. Multivariable estimated condom use effectiveness among men reporting sex with an HIV-positive male partner, by levels of reported condom use, using an interval censored time to event model for EXPLORE, VAX 004, and combined EXPLORE and VAX 004 analyses. Factors in the model include: study, age, race/ethnicity, education, and region. Receptive Anal Sex Always versus Never Condom use Always versus Sometimes Condom use Sometimes versus Never Condom use Insertive Anal Sex Always versus Never Condom use Always versus Sometimes Condom use Sometimes versus Never Condom use Receptive + Insertive Anal Sex Always versus Never Condom use Always versus Sometimes Condom use Sometimes versus Never Condom use EXPLORE VAX 004 Combined EXPLORE and VAX 004* Effectiveness (95% CI) P-value Effectiveness (95% CI) P-value Effectiveness (95% CI) P-value 86.8 (73.1, 93.5) < (45.4, 76.4) < (60.7, 80.5) < (63.8, 91.6) < (45.4, 77.4) < (56.5, 79.1) < (-35.1, 57.4) (-61.7, 35.5) (-31.5, 35.8) (53.9, 90.3) < (33.3, 71.6) (46.3, 74.3) < (53.6, 90.2) < (32.9, 73.1) (47.7, 75.8) < (-80.1, 45.4) (-59.5, 34.0) (-48.8, 26.7) (71.6, 93.3) < (41.6, 74.4) < (58.2, 79.2) < (60.7, 90.6) < (44.7, 74.4) < (55.2, 77.0) < (-19.3, 56.8) (-50.9, 30.0) (-24.8, 32.1) June 2014

25 7 Table 4. Analysis of condom use consistency for reported sex partners of any HIV status (positive, negative, unknown) by study and anal sex act. All Intervals EXPLORE Receptive Anal Sex (N = 2810) Insertive Anal Sex (N = 2900) Insertive+Receptive Anal Sex (N = 3102) Frequency Percent Frequency Percent Frequency Percent Always Condom Use Sometimes Condom Use Never Condom Use Any Single Interval Always Condom Use Sometimes Condom Use Never Condom Use June 2014

26 1 Figure 1. Condom use percentage for subjects having HIV positive partners by interval for EXPLORE and VAX 004 VAX 004 VAX 004 VAX 004 Revised 3 October 2014