icon-folder.gif   Conference Reports for NATAP  
  5th IAS Conference on HIV Pathogenesis, Treatment and Prevention
July 19th-22nd 2009
Capetown, South Africa
Back grey_arrow_rt.gif
HIV Prevention at IAS 2009
  5th IAS Conference on HIV Pathogenesis, Treatment and Prevention, Cape Town, South Africa, 19-22 July 2009

Jared Baeten, MD PhD
Connie Celum, MD MPH
University of Washington

HIV prevention occupied a central place at this year's IAS conference. Knowledge of HIV serostatus is key to prevention efforts, and there were several presentations about programmatic initiatives increasing uptake of HIV testing as individuals and couples. Dr. Ying Ru Lo of WHO presented on HIV testing in clinical and population-based settings (abstract TUSY303) and highlighted some of the issues in implementation and yield of provider-initiated HIV testing, district-wide household based testing (which has been successfully implemented in several countries in East Africa), and household testing of partners and families of HIV-infected persons. Specific abstracts related to these topics include one detailing high (95%) uptake and high yield (46% HIV seropositivity) with provider-initiated VCT of inpatients in Mulago Hospital, Kampala (Wanyenze, abstract MOPED075), and a pilot of family-centered testing in South Africa (Scorgie, abstract MOPED074).

The importance of high coverage of services, including knowledge of serostatus and targeted delivery of HIV prevention and care, was highlighted in a Tuesday plenary presentation by Dr. Stefano Bertozzi (abstract TUPL103). A newly-created operations research track provided an opportunity to hear about ongoing work implementing HIV prevention, including progress and challenges in scaling up male circumcision. A major theme of IAS was the concept of antiretrovirals for prevention, as summarized below, which provided an opportunity to bridge the too-often separate efforts for HIV prevention and treatment.

Dr. Ron Gray summarized biomedical prevention in a Tuesday plenary presentation (TUPL101), providing a summary and synthesis of the results of efficacy trials of microbicides, vaccines, circumcision, and treatment of sexually transmitted infections for the prevention of HIV, including his perspectives on possible reasons for some of the surprising or flat results of trials. A number of observational and interventional studies reporting data about potential novel prevention strategies are summarized below.

Antiretrovirals for HIV Prevention: Pre-Exposure Prophylaxis (PrEP), Post-Exposure Prophylaxis (PEP), and Treatment as Prevention

The use of antiretroviral therapy for HIV prevention - as PrEP or PEP to prevent HIV acquisition by HIV-uninfected persons and as treatment to decrease the infectiousness of HIV-uninfected persons - was a main theme of the conference, occupying two symposia (one prior to the conference), one oral abstract session, and a plenary session.

Treatment as prevention.

The rapid worldwide expansion of combination antiretroviral therapy during the past 15 years has revolutionized HIV-1 care. Several observational studies have reported that individuals initiated on antiretroviral therapy appear to be less infectious to their sexual partners, and the potential for antiretroviral therapy expansion to reduce the number of cases of new HIV-1 infections is a topic of discussion for both resource-rich and resource-limited settings. In a plenary talk, Dr. Reuben Granich from the World Health Organization presented mathematical modeling data suggesting that early HIV diagnosis and immediate ART initiation could markedly reduce the number of new cases of HIV infection in high-prevalence settings, potentially even to the point of shrinking the epidemic (abstract MOPL101). This strategy - often referred to as "test and treat" - would assume an intensive public health effort, including population-wide HIV testing (as frequently as yearly), antiretroviral therapy as soon as HIV is diagnosed (at much higher CD4 counts than currently recommended for clinical care), and life-long treatment. Whether test and treat can be evaluated as an HIV prevention intervention in communities with high HIV prevalences is under consideration. One late-breaker poster (Walensky, abstract LBPEC04) emphasized the challenges of test and treat for one community (Washington DC in the United States). Using mathematical modeling, the investigators found that a test and treat strategy would likely lengthen the average lifespan of HIV-infected persons (by engaging persons in care earlier and achieving awareness of HIV status), but it would also require very frequent testing (at least yearly), high acceptance of testing, and immediate ART initiation in order to have even a modest community-level impact on decreasing HIV transmission.

In another presentation, the benefits of antiretroviral therapy on reducing HIV infectiousness were reported from an observational database (Sullivan, abstract WEAC101). Among 2993 HIV serodiscordant couples from Rwanda and Zambia followed in a prospective cohort study, the rate of HIV transmission was 3.4 per 100 person-years in couples in which the HIV-infected partner was not taking antiretroviral therapy versus just 0.7 cases per 100 person-years among couples taking antiretroviral therapy (relative risk 0.21, 95% confidence interval 0.07-0.51). This abstract was a follow-up to data from this cohort presented at CROI earlier this year and updated the findings with stratification by gender of the HIV-infected partner. Although all 4 observed transmissions from individuals taking antiretroviral therapy were from women to men, there was not statistical evidence to suggest that antiretroviral therapy operates any more or less efficiently in reducing female-to-male versus male-to-female transmission of HIV. Samples were not available after ART initiation to test for viral load among those who transmitted after ART initiation. In sum, in this large observational cohort, in which antiretroviral therapy for HIV-infected persons was initiated for clinical care, antiretroviral treatment reduced the risk of HIV transmission by ∼80%.

In a poster presentation (Masese, abstract WEPDC106), data from a cohort of 899 Kenyan HIV-positive female sex workers were presented. During follow-up, 147 initiated antiretroviral therapy, according to Kenyan treatment guidelines. After starting antiretroviral therapy, women reported less unprotected sex (adjusted odds ratio 0.69, 95% confidence interval 0.48-0.98), which was corroborated by a decreased frequency of sexually transmitted infections (adjusted odds ratio 0.67, 95% confidence interval 0.44-1.02). For this core transmitter group, these findings are encouraging and emphasize an opportunity for engaging women at high-risk for transmitting HIV into HIV care and treatment.

Post-exposure prophylaxis (PEP).

One prominent challenge with PEP has been tolerability, with rates of completion of the usual 28-day course of therapy often <50%. Two presentations described recent experience with PEP regimens with better results. The first (Tosini, abstract WEAC102), a multi-center evaluation in France evaluated the combination of emtricitabine/tenofovir plus lopinavir/ritonavir among 222 persons presenting for PEP after a high-risk exposure. Two-thirds were male and most exposures were sexual (i.e., few were from injection drug use or associated with occupational injury, such as a needlestick to a healthcare worker). One hundred sixty-six (75%) completed the course of PEP, while 22 (10%) discontinued for side effects and 27 (12%) were lost to follow-up - a higher rate of completion than with historical PEP regimens. In the second study (Mayer, abstract WEAC104), 53 men presenting to a Boston clinic after high-risk sexual exposure were started on PEP using emtricitabine/tenofovir plus raltegravir. Significantly less diarrhea, nausea and vomiting, and fatigue were experienced, compared to historical controls (n=199) who had used zidovudine/lamivudine plus a protease inhibitor. The completion rate was 59% compared to 39% for historical controls. There were no HIV seroconversions in either study.

Pre-exposure prophylaxis (PrEP).

PrEP refers to HIV-uninfected persons using an antretroviral agent (e.g., oral or formulated as a topical microbicide) prior to high-risk exposures in an effort to block initial HIV infection, a situation analogous to malaria prophylaxis in travelers. There are currently seven efficacy trials of PrEP ongoing worldwide, and while the first results from these trials are not expected until at least 2010, there is high scientific enthusiasm for this novel potential HIV prevention strategy. One study (Mayer, abstract WEPEC080) interviewed 227 men in Boston about their knowledge and interest in PrEP, should it prove effective. Few (19%) had heard of PrEP, and only 1 had used it. Willingness to use PrEP was directly correlated with perceived efficacy: three-quarters felt PrEP would need to be >80% effective to be used. The possibility of side effects strongly influenced men's attitudes regarding PrEP.

HSV suppression

Multiple observational studies during the past two decades suggest strong epidemiologic synergy between the global epidemics of herpes simplex virus type 2 (HSV-2) and HIV, with HSV-2 increasing both HIV susceptibility and infectiousness. HSV-2 is the most common cause of genital ulcer disease worldwide and genital ulcers have been associated with enhanced sexual HIV transmission. Five proof-of-concept studies have demonstrated that daily suppressive therapy against HSV-2, using acyclovir or valacyclovir (a pro-drug of acyclovir), reduces plasma and genital HIV levels by 0.25-0.50 log10 copies/mL, suggesting that HSV-2 suppression could decrease HIV infectiousness.

Thus, the final results of the Partners in Prevention HSV/HIV Transmission Study, presented for the first time in Cape Town, were surprising and disappointing (Celum, abstract WELBC101). [Full disclosure: We are investigators on this study.] The design of this study was unique and complex: 3408 heterosexual HIV serodiscordant couples (67% in which the HIV-infected partner was female) were enrolled at 14 African sites. All HIV-infected partners also had HSV-2 and had CD4 counts that did not meet national guidelines for initiation of antiretroviral therapy. They were randomized to daily acyclovir or placebo and seen every month for clinical follow-up and adherence counseling. Their HIV-uninfected partners were seen every 3 months for HIV testing, for up to 24 months of follow-up. Thus, the intervention studied (daily acyclovir or placebo) was taken by the HIV-infected partners, while the outcome determining the efficacy of the intervention (HIV seroconversion) was measured as HIV acquisition rates among their partners.

Retention and follow-up were high. HIV seroincidence was 2.8 per 100 person-years (considerably lower than observational studies of HIV discordant couples, likely reflecting the monthly risk-reduction counseling). Acyclovir reduced both the frequency of genital ulcer disease due to HSV-2 (RR 0.27, 95% CI 0.20-0.36, p<0.001) and HIV plasma viral load (mean reduction 0.25 log10 copies/mL, 95% CI 0.22-0.29, p<0.001). Genital HIV levels were reduced by a similar amount in the acyclovir arm (Baeten, abstract LBPEA07). The relationship of genital HIV levels from the Partners in Prevention trial indicate a step-wise association with HIV transmission among the linked transmissions and that transmissions did occur at low viral loads. In spite of significant but modest reductions in plasma and genital HIV levels, there was no statistically significant effect on reducing sexual HIV transmission: of 84 HIV transmission events determined by viral sequencing to have occurred within the partnership, there were 41 and 43, respectively, on the acyclovir and placebo arms (HR 0.92, 95% CI 0.60-1.41, p=0.7). Notably, an accompanying presentation (Lingappa, abstract WELBC102) showed that acyclovir significantly slowed the rate of HIV disease progression (to CD4<200, need for antiretroviral therapy, or death) among the HIV-infected partners, although the magnitude of effect was modest (a 17% decrease in risk, translating to delaying antiretroviral therapy by 6-11 months) would be sufficient to implement. Ultimately, the results of this trial demonstrate that interventions aimed at reducing HIV transmission (and disease progression) may need to have very powerful effects to demonstrate true clinical benefit.

Male circumcision

Several sessions were devoted to scale-up of male circumcision services in high HIV prevalence areas. Three years ago, 3 large randomized trials conclusively demonstrated that circumcision reduces the risk of HIV acquisition in men. One large program, in the Nyanza Provence in western Kenya, reported high satisfaction (>99%) and a low rate of adverse events (<5%) among nearly 1200 men presenting for circumcision (Herman-Roloff, abstract LBPED01). Notably, <25% of men agreed to HIV testing prior to circumcision, emphasizing the continued need for voluntary counseling and testing services to integrate with circumcision services where possible. A qualitative study from the same region of 30 men who had undergone circumcision found that most adopted or maintained safer sex behaviors after circumcision and understood that the procedure offered only partial protection against HIV, suggesting productive synergy between the circumcision surgery and the pre-procedure counseling (Riess, abstract MOPEC023).

For HIV-infected men, the risks and benefits of circumcision are less clear than for HIV-uninfected men. A late-breaker from the Partners in Prevention HSV/HIV Transmission Study (Baeten, abstract LBPEC06) found that circumcised HIV-infected men were ∼40% less likely to transmit HIV to their female partners than uncircumcised men (p=0.1). These results suggest circumcision could carry some longer-term HIV transmission benefits for men who become HIV-infected in spite of being circumcised. However, the short-term risks are still a point of debate: a randomized trial of circumcision of HIV-infected men from Uganda was published in the Lancet the week prior to the IAS meeting (Wawer et al. 2009; 374:229). While the rate of adverse events was low (showing the procedure can be done safely in HIV-infected men), the rate of HIV transmission to female partners was higher among men who were circumcised (HR 1.49, p=0.4), particularly among men who initiated sex prior to wound healing.

For men who have sex with men (MSM), there is not yet consensus on whether circumcision offers protection against HIV. One study among 378 MSM from Soweto, South Africa who exclusively practiced insertive anal intercourse found those who were uncircumcised were at marked HIV risk (adjusted OR 4.5, 95% CI 3.1-6.7) compared with those who were circumcised (Lane, abstract MOPDC105). Most men also reported sex with female partners. Several studies continued to report high rates of high-risk behaviors and high prevalence and incidence of HIV among African MSM; in one study (Mwangome, abstract MOPEC013), HIV incidence among MSM sex workers in coastal Kenya was 9.9 per 100 person-years, compared with 3.2 per 100 person-years among female sex workers in the same area.

Topical Vaginal Microbicides

Relatively little new data were presented relating to topical vaginal microbicides. Two efficacy trials of tenofovir gel (CAPRISA 004 and the Microbicide Trials Network's VOICE study) are ongoing, and the field is awaiting the results of the Microbicide Development Program study 301, testing the non-specific agent PRO2000 (which showed some suggestion of modest efficacy in a smaller trial presented at CROI this year), which will report results late this year.

HPV infection and HIV risk

Three observational studies presented data suggesting human papilloma virus (HPV), which can cause genital warts and cervical cancer, may increase HIV risk. Two of the studies were done among participants in the randomized trials of male circumcision for prevention of HIV acquisition in men that were completed 3-4 years ago. Among Kenyan men (Smith, abstract WELBC104), 50% of men had HPV detected, and HPV on the glans (but not on the penile shaft) was associated with increased HIV risk (adjusted HR 1.8, 95% CI 1.1-2.9). This effect was not different for circumcised versus uncircumcised men or for men with HPV types that are high or low risk for causing cancer. Among South African men (Auvert, abstract TUAC202), 19% of men had HPV in a urethral swab, which was strongly associated with HIV (RR 4.6, 95% CI 2.2-9.3) but only for high-risk types. Generally similar findings were seen in a smaller study of South African female sex workers (Auvert, abstract TUPEC073). Since HPV and HIV are both transmitted, one concern with these findings is that they are confounded by sexual behaviors. Additionally, the temporal relationship of HPV prior to HIV acquisition is complicated by the sensitivity of assays, use of different sampling methods in the studies to date, and the natural history of HPV, as HPV can be cleared and reinfection occurs. Unfortunately, there has been limited success to date with interventions aimed at sexually transmitted infections as a strategy to prevent HIV.


Although antiretrovirals for prevention of sexual HIV transmission are just beginning to be discussed aggressively, these medications have been the mainstay for prevention of mother-to-child HIV transmission (PMTCT) for well over a decade. Several large important PMTCT studies were reported in Cape Town; all focused on more intensive antiretroviral therapy regimens - either maternal highly active antiretroviral therapy (HAART, for women with higher CD4 counts who would not yet require HAART for their own health) or infant antiretroviral prophylaxis. With WHO set to review PMTCT and infant breastfeeding guidelines later this year, the most up-to-date data available were presented.

Last year (Kumwenda, et al. The PEPI Study. New Engl J Med 2008; 359:119), researchers from Malawi demonstrated that 14 weeks of extended antiretroviral prophylaxis of HIV-exposed infants, using either nevirapine or nevirapine plus zidovudine, significantly reduced infant HIV infection during breastfeeding. In Cape Town, subgroup analyses from the PEPI Study were presented, demonstrating that antiretroviral prophylaxis of infants was effective at all maternal CD4 counts (Mofenson, abstract TUPEC053). Notably, for women with CD4 counts >350, transmission rates were very low in the context of prophylaxis (1.4%), similar to what has been reported with maternal HAART.

The Mma Bana Study from Botswana (Shapiro, abstract WELBB101) assessed virologic suppression in women with CD4 counts >200 who were started on HAART during late pregnancy. Among 560 women, plasma HIV RNA suppression (<400 copies/mL) was >90% at delivery and throughout 6 months of breastfeeding, using HAART regimens of abacavir/zidovudine/lamivudine or zidovudine/lamivudine/lopinavir/ritonavir. The overall HIV transmission rate was 1% and treatment-limiting adverse events were rare.

The Kesho Bora study (de Vincenzi, abstract LBPEC01) randomized 824 pregnant women with CD4 counts between 200 and 500 to HAART versus standard short-course antiretroviral therapy near the time of delivery. Cumulative HIV infection rates were 5.6% in the HAART arm versus 9.3% in the short-course arm at 12 months, a 40% reduction (p=0.05). Subgroups analyses suggested the benefit was greatest among women with CD4 counts between 200 and 350.

Finally, the BAN study (Chasela, abstract WELBC103) randomized 2637 pregnant women from Malawi with CD4 counts >250 to post-natal HAART or infant extended nevirapine prophylaxis. All received intrapartum short-course antiretrovirals, and a third control arm that included only the short-course regimen was stopped early. Infants were exclusively breastfed for 24 weeks then weaned over 4 weeks. At 28 weeks post-partum, HIV transmission rates were 6.4% for the control arm, 3.0% for the HAART arm (p=0.003 vs. control), and 1.8% for the infant prophylaxis arm (p<0.001 vs. control). The difference between the maternal HAART and infant prophylaxis arms did not reach statistical significance (p=0.1). Analysis of HIV-free survival showed similar relationships across the three arms. Thus, both maternal HAART, started after delivery, and infant prophylaxis are effective (essentially, more or less equally effective) in preventing HIV transmission to breastfed infants born to women with CD4 counts >250.

Distressingly, in spite of tremendous progress in discovery of novel strategies to minimize HIV acquisition in infants, getting antiretrovirals to women at risk of transmitting HIV to their infants remains a tremendous operational challenge. The PEARL study collected ∼29,000 anonymous cord blood specimens from women delivering at centers in Cameroon, Cote d'Ivoire, South Africa, and Zambia between 2007 and 2008 (Coetzee, abstract WELBD101). Samples were tested for HIV and for antiretrovirals. Overall, 12% were HIV seropositive. Overall, only ∼50% of births received single-dose nevirapine prophylaxis, and drop off in successful delivery of PMTCT medications occurred at each step leading up to delivery (e.g., offering of HIV testing, acceptance of testing, receipt of results, receipt of nevirapine, injestion of nevirapine). System-wide interventions must continue to target better performance of this proven prevention service.