icon-    folder.gif   Conference Reports for NATAP  
  IAS 2015: 8th IAS Conference on
HIV Pathogenesis Treatment and Prevention
Vancouver, Canada
18-22 July 2015
Back grey_arrow_rt.gif
HIV Prevention at IAS 2015 Vancouver
  8th IAS Conference on HIV Pathogenesis, Treatment, & Prevention Vancouver, Canada 19-22 July 2015
Jared Baeten, MD PhD
Connie Celum, MD MPH
University of Washington
HIV prevention is central every international HIV conference, and this year's 8th IAS Conference an HIV Pathogenesis, Treatment, and Prevention continued that trend. Prevention was not always at the forefront at every meeting - for example, it was not until the 4th version of this conference that "and Prevention" was added to the title, which had previously focused solely on basic science to define viral pathogenesis and on evolutions in HIV treatment. At the 2015 meeting, however, prevention "competed" strongly with pathogenesis and treatment as a topical area of interest - with as many or more sessions devoted to prevention questions as to those other areas.
This year's meeting also emphasized the important interfaces between pathogenesis, treatment, and prevention. An important example was one of the most well-attended sessions of the meeting - the presentation of the START results, reporting the findings of that multicenter randomized trial to assess the clinical benefits of antiretroviral therapy (ART) initiation at the time of HIV diagnosis, compared to waiting for modest CD4 decline. Those results, which demonstrated that starting earlier has clinical benefits and were made public by press release in May 2015, were presented in a dedicated session at IAS 2015 and simultaneously published in the New England Journal of Medicine (http://www.nejm.org/doi/pdf/10.1056/NEJMoa1506816 ) [http://www.natap.org/2015/IAS/IAS_36.htm] Antiretroviral treatment is an important prevention tool, on top of its clinical benefits.
During the START session, the World Health Organization announced that its updated Consolidated Guidelines on the Use of Antiretroviral Drugs for Treating and Preventing HIV Infection, which are pending publication in the next few months, will include both recommendations that speak to the START results as well as a global recommendation that PrEP (pre-exposure prophylaxis) is recommended for all persons at substantial risk of HIV infection, regardless of geography or source of HIV risk.
As has been the case at recent International AIDS Conferences, much of the scientific program is available online, including abstracts, copies of slides, and webcasts of plenary and oral abstract sessions as well as a number of special session and symposia (searchable program available at http://www.ias2015.org/).
Antiretroviral-based HIV prevention strategies - including antiretroviral treatment (ART, often called treatment as prevention [TasP]) to reduce the infectiousness of HIV infected persons and pre-exposure prophylaxis (PrEP) for uninfected persons to prevent HIV acquisition - are powerful approaches for decreasing HIV spread. A Monday morning plenary by Dr. Francois Venter spoke to delivery of TasP and PrEP, both challenges and opportunities, and was an energetic start to the meeting (abstract MOPL0106 [https://www.youtube.com/watch?v=TiMK2Jmme10])..
PrEP (primarily oral tenofovir-based pills, the only approach to date to have received considerable attention at this year's IAS meeting - with studies ranging from pharmacology evaluations to better understand how much and when PrEP has to be taken to have its best benefits to a number of implementation science projects showing that PrEP is deliverable in a variety of settings worldwide. A recurring theme was that, when PrEP is available, those who start take it well.
A Monday afternoon symposium (MOSY01) was devoted to understanding daily and intermittent dosing strategies for PrEP, in a variety of populations. The impetus for the session was the completion of the HPTN 067 study, a multisite randomized trial of PrEP dosing - daily versus time-driven (twice weekly with a single pill boost after sex) versus event-driven (one dose prior to and one dose post-sex) for 24 weeks, as well as a 4-week observed dosing schedule for pharmacokinetics assessments. Three sites participated: Harlem, New York in the US and Bangkok, Thailand enrolled men who have sex with men (MSM) and transgender women (TGW) and Cape Town, South Africa enrolled women. Each site was independently powered, since local context could influence PrEP taking. The primary outcome was coverage for each sex act that occurred during the study period - with coverage defined as at minimum 1 pill taken within the 4 days prior to sex and 1 pill taken within the 1 day after sex (speaking to the idea that PrEP's efficacy requires at least some pre-exposure dosing as well as some post-exposure dosing - understanding that perhaps more than a single dose pre- and post-sex could be required in some populations/instances). In Cape Town, 179 women participated. Coverage was achieved for 75% of sex acts for women assigned to daily PrEP, 56% for those using time-driven PrEP, and 52% for those assigned to event-driven - surprisingly, demonstrating better performance for daily PrEP pill taking than intermittent taking in this population (perhaps because building a routine was important to do). The post-exposure dose was more often missed than the pre-exposure dose, particularly for the non-daily regimens, and overall adherence to the assigned PrEP schedule was highest for the daily dosing. There were 5 seroconversions during the study: 1 in the daily PrEP arm, 2 in the time-driven arm, and 2 in the event-driven arm; all had no evidence of medication use - emphasizing first that HIV risk was very high (5 seroconversions in 179 women followed for just a half a year) and second that HIV infections seem to occur principally when individuals are not taking PrEP. In summary, the HPTN 067 Cape Town results are very exciting - they show that a majority of women at real HIV risk can take PrEP, and those who took PrEP took enough to achieve HIV protection, particularly with daily-dosing. In Bangkok, 178 MSM/TGW were following. Like for the women in Cape Town, coverage of sex events was high and event-driven use had the lowest coverage: 85%, 84%, and 74%, and the missed dose was more often the post-exposure dose. Blood tenofovir concentrations consistent with HIV protection were high in all arms - essentially >90%. There were no seroconversions during the randomized phase of the study, in any of the arms. In Harlem, 179 MSM/TWGW were randomized. Coverage was 66% daily, 47% time-driven, 52% event-driven; 1 seroconversion occurred, in a daily arm participant who had very low concentrations of TFV. In all three sites, side effects were mild and no different than expected based on the known safety profile of FTC/TDF. Qualitative work that accompanied the study described how some participants felt it was easier to take a pill daily than to remember to take a pill on a non-daily schedule, how PrEP decreased anxiety and was highly valued as a result, but also how some of those taking PrEP battled stigma (to pill taking, to HIV risk, and to impressions of promiscuity). In sum, HPTN 067 shows that at-risk populations in diverse settings can take PrEP, and do so well enough to be protected against HIV. The symposium session also included several complementary talks -- addressing the pharmacology of HIV protection from PrEP (take away = both pre- and post-exposure dosing seems to be important) and how pharmacology may need to be adapted to different populations (for example, in HPTN 067, the women in Cape Town had better adherence to daily dosing than intermittent dosing, which would notably be in agreement with pharmacology data that vaginal levels of tenofovir may need to be dosed daily to achieve HIV protection). Finally, a PrEP advocate and user, Damon Jacobs, from the US, reminded the audience of the individual self-efficacy benefits of PrEP - renaming to Proactive Responsible Empowered Pleasure. Several individual abstracts from HPTN 067 were presented in full at IAS 2015, including: MOAC0305LB, MOAC0306LB, WELBPE23, TULBPE21.
A Tuesday oral abstract session was devoted to PrEP implementation projects (TUAC02). In the multisite US study entitled The Demo Project (Liu, abstract TUAC0202), PrEP was provided to MSM and TGW in STD clinics in San Franscisco and Miami and a community health center in Washington DC. Annual HIV incidence was estimated to be >2% in each site. Participants were enrolled from 2012 to 2014, and follow-up was quarterly for 48 weeks. The median age of subjects was 35 years and 20% were <25 years; 48% were white, 35% black, 7% Latino, 10% other. One-quarter had a baseline STI and all had behavioral risk for HIV.
Retention at 12 months was 78%, higher in those with prior PrEP knowledge and reporting condomless anal sex at baseline. Throughout follow-up 80-86% achieved protective levels of PrEP in dried blood spots - other adherence measures were consistent with the blood levels; 63% of participants had protective levels at all visits. PrEP interruptions occurred in 15% - mostly due to perceived side effects or changes in risk (to lower risk - and thus arguably less need for PrEP). Adherence was lower in Miami and among black race participants; it was not related to age, education, or drug use. Sexual behavior risk did not increase over follow-up - number of partners declined and condomless sex did not change; STI rates were high but not increasing over follow-up. Only 2 incident HIV infections happened during follow-up (at an incidence of 0.43 per 100 person-years) - both in individuals with low/no PrEP use. These results strongly indicate that at-risk MSM in the US uptake and adhere well to PrEP, with those with higher risk behaviors taking PrEP very well. In spite of ongoing condomless sex exposures (although not increasing over time) and high STI rates, HIV incidence was very low.
The TDF2 clinical trial, which demonstrated that PrEP was efficacious for HIV protection compared to placebo in a study among young heterosexuals in Botswana that ended in 2010, reported on its open-label extension study (Henderson, abstract TUAC0203). The open-label extension began in 2013 and offered open-label PrEP to those who had participated in TDF2 (and who remained HIV negative). Of 1219 TDF2 participants, only 335 were able to be screened for the open-label study, due to relocations and other factors given the gap between the end of TDF2 and the open-label; 229 started PrEP (86% of those eligible within the 335 screened). Number of sexual partners and frequency of sex without condoms decreased during the course of the study. A random subset of 120 participants had samples tested for TFV in blood spots - 93% had drug detected. Women were slightly less likely to have drug detected but detection in women was still quite high (>80%). No infections were observed - versus 5-6 cases expected based on the HIV incidence during the TDF2 study. Thus, PrEP adherence was high in Botswana in heterosexual young men and women.
ATN110 was a PrEP demonstration project for young MSM in Chicago (Hosek, abstract MOAC0204LB - http://www.natap.org/2015/IAS/IAS_76.htm). Young MSM carry the greatest risk of new infections in the US, but few young MSM participated in PrEP clinical trials. In ATN 110 young men interested in PrEP participated in a baseline behavioral intervention and were followed for 48 weeks with opportunity to take PrEP. 200 were enrolled (of 277 screened, with 11 excluded for new diagnosis of HIV infection). Mean age was 20 years, 78% identified as gay and 14% bisexual; they reported an average of 5 partners in the past month, 58% had condomless receptive anal intercourse with most recent partner, 22% had a bacterial STI at baseline. Adherence was good overall - by blood levels a majority achieved enough PrEP to provide HIV protection; adherence was lower in black compared to white participants. 4 became HIV infected (incidence 3.29 per 100 person-years) - none had tenofovir detected at the time of seroconversion. Sexual behavior risks did not increase over time, and those reporting condomless sex had higher PrEP levels. Overall the results demonstrate that young MSM can be identified who are interested in and appropriate for PrEP; adherence was good overall but had significant differences by race/ethnicity, although men having unprotected sex were more likely to take PrEP. Notably, adherence declined over time, particularly as visits spread out from monthly to quarterly - indicating that youth may need some more frequent contact (such as might be provided through in-person or mobile check-ins).
In another PrEP demonstration project, among MSM and TGW at multiple clinical sites in Brazil (Grinsztein, abstract TUAC0205 - http://www.natap.org/2015/IAS/IAS_76.htm), PrEP interest and uptake were high, with just over 50% of eligible men starting PrEP. Targeted efforts to include transgender populations in the project were successful, and, for both MSM and TGW, prior knowledge of PrEP increased the likelihood of PrEP uptake. Two pre-conference satellite sessions - on achieving pregnancy safely with use of TasP and PrEP for prevention (SUSA02) and on new developments in injectable PrEP options (SUSA04) - covered additional new information for this field.
Thus, at IAS 2015, there was considerable discussion of PrEP. The final day closed with a plenary on translating PrEP science into program delivery - and described PrEP as "the missing piece in combination prevention" (Mesquita, abstract WEPL0101). Over the past four years, since the first PrEP efficacy trials demonstrated in late 2010 and through the first half of 2011 that PrEP was effective for preventing HIV, PrEP has emerged as a strategy ready for delivery and public health impact.
Treatment as prevention (TasP) is a potent antiretroviral based HIV prevention strategy. Like PrEP, the ultimate impact of TasP for HIV prevention depends on its coverage in a population (i.e., how many people who have HIV infection actually are receiving ART) and adherence (i.e., how well those receiving ART take it). Vancouver is a perfect venue to discuss TasP - the Canadian province of British Columbia (where Vancouver is located) has had been a leader in providing treatment with as great of coverage as possible; as a result, it has seen declines in the number new HIV infections. Building on the START results [http://www.natap.org/2015/IAS/IAS_21.htm], much of the discussion in the hallways in Vancouver was about treatment - earlier treatment for clinical care, earlier treatment for prevention, more widespread treatment for prevention, etc. How to translate the scientific successes of ART into public health impact is the big question - i.e., how to deliver ART, at scale, to enough people, with sustained use and adherence; for that big question, not all answers were forthcoming from IAS 2015, but the discussion was well underway.
In 2011, at this same IAS meeting the primary results of HPTN 052, a mulitcountry randomized trial of ART for prevention among 1763 HIV serodiscordant couples (97% heterosexual), which assigned couples in which the HIV infected partner had a CD4 counts between 350 and 550 (and who thus did not qualify for ART under treatment guidelines of the time) to immediate ART versus deferred ART (until the CD4 count declined). The primary findings demonstrated a 96% reduction in the risk of HIV transmission within the partnerships - 27 vs. 1 HIV transmission (although additional "unlinked" HIV transmissions from outside the partnerships occurred, in relatively equal numbers between the two randomization arms, which would be expected since ART only would protect against transmissions from the partner on ART, not from other sexual partners). After those primary results were reported, all couples were offered ART and followed for additional time; the final results of the trial were presented at IAS 2015 (Cohen, abstract MOAC0101LB [http://www.natap.org/2015/IAS/IAS_31.htm]). By end of study 98% of HIV infected partners initially assigned to the delayed arm were on ART. But, that proportion grew slowly - 1 year after 052 results were presented 16% were still delaying ART and 7% delayed at 2 years (in spite of the fact that all would have theoretically been ready for ART at the time of initial randomization). After April 2011, 32 new HIV infections occurred: 15 in the early ART arm and 17 in the delayed arm, of which 9 were linked within the partnership (2 early, 7 delayed) [http://www.natap.org/2015/IAS/IAS_46.htm]. Overall, for the entire study period of HPTN 052, early ART resulted in a 69% reduction in the risk of HIV infection overall (including both linked within the partnership and unlinked transmissions) and a 93% reduction in linked transmissions. There were a total of 8 infections after starting ART: 4 occurred soon after starting ART (during the period before viral suppression occurs) and 4 occurred after ART failure. Thus, the benefit of 052 is durable until the end of the study, and no linked HIV transmissions were observed when the index was stability suppressed on ART. However, delays in ART initiation (even in this highly motivated population), ART failures, and transmissions from outside partners emphasize that although when ART is present it is extraordinarily powerful there are important situations when ART is not present that need other prevention.
The question of adherence and viral suppression among asymptomatic HIV-infected persons who initiated ART at higher CD4 counts was assessed in the ANRS TasP trial in the Africa Centre, Kwa Zulu Natal (Iwuji, abstract MOAC0104 [https://www.youtube.com/watch?v=Uj6tpBvM8YI]). From March 2012-May 2014, 227 of whom had initiated ART and had 6 months HIV RNA data. Encouragingly, adherence ≥95% at 6 months was high (88% and 83% by pill country and visual analog scale self-assessment, respectively) with no evidence that this was associated with CD4 at initiation. Male sex was independently associated with < 95% adherence (2.58, 1.24-5.35, p= 0.01; ref. females). 83% (183/227) of those who started ART achieved HIV suppression by 6 months with no association with CD4 at initiation (1.13 per 100cells/mm3 higher, 0.96-1.33, p=0.40). Compared to those with ≥95% adherence by VAS, individuals with <95% adherence were somewhat less likely to suppress (0.44, 0.19-1.03, p=0.06). Thus, they found no evidence that among people newly entering HIV care, higher CD4 at ART initiation was associated with reduced adherence or poorer virological suppression, at least in the short-term.
A community-based randomized trial of linkage to HIV care and demand creation for voluntary medical male circumcision (VMMC) in rural communities in KwaZulu-Natal, South Africa and Sheema district, Uganda was reported (Barnabas, abstract MOAC0105LB [https://www.youtube.com/watch?v=ZRMsp04ONDg]). HIV testing was done at home or through mobile units. HIV-positive persons were randomly allocated to linkage to care strategies: clinic facilitation by lay-counselors at the initial clinic visit, lay-counselor follow-up visits at home, or standard clinic referral. HIV-negative uncircumcised men were randomized to VMMC demand creation strategies: lay counselor follow-up visits at home, SMS reminders, or standard VMMC promotion at the time of testing. Between June 2013 and February 2015, 15,332 persons received HIV testing and counseling.
Among 1,325 HIV-positive persons randomized to linkage strategies, overall clinic linkage was high (93%). Compared to standard linkage, lay counselor clinic facilitation increased linkage to care (RR=1.09, 95% CI: 1.05-1.13), and home follow-up visits increased ART initiation (RR=1.23, 95% CI: 1.02-1.47). In all arms, ART initiation was limited by bottlenecks in service delivery at the clinics, although 67% of those eligible initiated ART by 9 months. Overall, 82% of persons initiating ART achieved viral suppression without significant difference between study arms. Of 750 HIV-negative uncircumcised men randomized to VMMC promotion strategies, uptake of circumcision was 41% by month 3.
Compared to standard messages, VMMC uptake was significantly higher in the SMS promotion (RR=1.72, 95% CI: 1.36-2.17) and lay counselor follow-up arms (and RR=1.67, 95% CI: 1.29-2.14). This study indicates that community-based HIV testing and linkage to care and prevention can be effectively implemented and that simple strategies, such as SMS reminders or lay-counselor visits, increase linkage for ART initiation and male circumcision. However, additional strategies are needed to address clinic delays that are barriers to ART delivery.
ART is central to prevention of mother-to-child transmission of HIV (PMTCT) but numerous challenges impede perfect implementation of PMTCT services. There were a number of excellent PMTCT-focused sessions at IAS 2015, covering new science in reducing infant HIV infections.
From Malawi, Option B+ (i.e., initiation of life-long ART in women during pregnancy) showed high early dropout rates and poor adherence during the first 6 months of therapy (Haas, abstract MOAC0201). But, for those who remained on therapy, adherence improved after 6 months and retention was good once breastfeeding had ceased. Retention after 3 years was 70%. Malawi was one of the first countries in Africa to adopt Option B+ and it has received considerable attention for the program's successes and challenges. In many settings (not just Malawi) high rates of early discontinuation of ART after delivery have been observed and strategies aimed at stabilizing B+ drop out are priorities for new research.
In another study from Malawi (Rosenberg, abstract MOAC0202), a randomized trial to improve recruitment of male partners was reported. 200 newly-diagnosed HIV+ pregnant women newly initiated on Option B+ were randomized to one of two interventions to engage their partner for testing: 1) a written invitation to the partner including a potential appointment time and 2) the same invitation plus phone/community tracing of the male partner if he failed to present by day 7. The primary outcome was couples testing by day 30. Community tracing significantly increased the proportion of men tested by couples HIV testing and counseling: 74% vs. 52% for the written invitation alone. For the male partners (of these known HIV+ women), about 25% knew already they were infected, about 50% were HIV infected and learning this for the first time, and about 25% were HIV uninfected (thus a member of a serodiscordant partnership). Female loss to follow-up in the PMTCT clinic by 1 month was higher for the invitation only arm compared to the invitation plus active tracing arm (17% vs. 9%, p=0.09). Only 3 women reported social harms: 1 woman felt blamed, 2 couples separated, none reported violence. Thus, most women were willing to try partner testing, and invitation plus tracing resulted in higher uptake of couples testing, identified HIV+ men and serodiscordant couples, and kept more women engaged in care.
A late breaker abstract (Thirumurthy, abstract MOAC0302LB) evaluated a novel strategy to provide HIV testing to male partners of pregnant women - i.e., giving the woman an HIV self-testing kit to take home to provide to her partner.
This study, from Kenya, provided multiple self-test kits to women testing HIV negative in two clinics - one an antenatal and postpartum clinic (N=178, provided 3 tests to take home to partners) and one in a sex worker clinic (N=102, 5 kits provided). In all groups, women often used a test on themselves and frequently on a primary partner. Testing together with a primary partner was common, and social harms were rare (only 4 out of 280 women reporting any social harms). HIV+ persons were identified, more so for partners/clients of the sex workers but also for primary partners of women in ante- and postnatal clinic, and the majority linked to care. Thus, this novel strategy - providing self-test kits for home testing of partners (and clients) offers an opportunity to engage those who might not otherwise test.
In Zimbabwe (Cowan, abstract MOAC0203 [https://www.youtube.com/watch?v=YAz15ZW0wi4]), a multi-province evaluation of the impact of WHO Option A PMTCT (i.e., short-course antiretrovirals during labor and post-partum for women with higher CD4 counts) was presented. Option A was initiated in Zimbabwe on large scale in 2011, transitioning to Option B+ in late 2013. The project surveyed nearly 20,000 mother-infant pairs in two large waves of assessments in 2012 and 2014. HIV-free infant survival increased from 2012 to 2014 from 91% to 93% and HIV transmission decreased from 8.8% to 6.7%. In the subset of areas that had not truly started Option A by the 2012 survey, these differences were even greater - a 5+% improvement in both indices.
Testing and Linkage to Care
As HIV treatment guidelines are moving towards universal treatment regardless of CD4 and with implementation of effective biomedical prevention strategies such as medical male circumcision and PrEP, HIV testing becomes even more critical. Important priorities in HIV testing are community-based strategies, reducing stigma, and self-testing. The World Health Organization released its updated Consolidated Guidelines on HIV Testing Services (http://who.int/hiv/pub/guidelines/hiv-testing-services/en/) just prior to IAS 2015 and these were highlighted in a Monday evening satellite (MOSA04).
Presentations at IAS that discussed new findings in these areas were presented in an oral session called "HIV testing: the gateway to everything." In South Africa, data from the One Man Can intervention, a theory-based community mobilization intervention with efforts to engage men in HIV testing, address gender norms, and uptake of biomedical interventions, were presented (Lippman, abstract MOPDC0101). The intervention which was delivered in 11 intervention communities and compared to 11 control communities and included 2 day intensive workshops, outreach activities, door to door campaigns, street soccer, murals, and street theatre, and local mobilizers. The main finding was that for every standard deviation increase in the score for the 6 community mobilization domains, there was an increased 2.2-fold increased odds of reported HIV testing.
An intervention to reduce stigma and increase HIV testing was evaluated in a community randomized trial in Malawi (Derksen, abstract MOPDC0102). The intervention was an infographic to communicate that ART reduces HIV transmission by 96% and that the person who gets tested is a safer partner. They evaluated the impact through a survey 5 months later and found that 70% in intervention villages versus 18% in control communities believed that ART reduces transmission, preferred a partner on ART to an untested sexual partner, and more believed that a person on ART might find a new sexual partner. They also found that 3 months after the intervention, HIV testing rate was 60% higher in the intervention communities, based on data from 18 health facilities, among non-pregnant persons ages 15-49, and, importantly, the effect was similar for men and women. The study is currently evaluating the impact of the intervention on ART initiation rates.
From Seattle, an intervention focusing on self-testing among 230 MSM was reported (Katz, abstract MOPDC0103). HIV-negative MSM at high risk of infection (i.e., condomless anal intercourse, methamphetamine or popper use, a bacterial STI diagnosis, or 10 or more male anal sex partners during the past year) were randomized to use the OraQuick Advance Rapid HIV-1/2 Antibody Test on oral fluids or to seek testing as usual; follow-up was for 15 months. Self-testers got the OraQuick kit at baseline and could get up to one kit per month afterwards at no cost. All men were advised to test quarterly (the recommended standard for higher-risk MSM in Seattle), all men received testing reminders, and participants could test through any HIV testing source.
Self-testers reported an average of 5.3 tests during follow-up (most self-tests), compared with 3.6 in the control arm (p<0.0001). Proportions who got tested 4 times or more (as recommended) were 76% among self-testers and 54% among standard testers (p=0.001). Four self-testers and 2 standard-testers had a positive test.
Rates of condomless anal intercourse with a nonconcordant partner did not differ much between self-testers and standard-testers at 9 months (21% versus 22%) or 15 months (29% versus 24%).
At the final study visit 5.4% of self-testers versus 12.2% of standard-testers had a bacterial STI, for a risk difference of -6.8% (95% confidence interval [CI] -16% to 1.6%). Thus, self-testers had more frequent HIV testing and acquired fewer STIs, with no difference in self-reported sexual risk behaviors, indicating that self-testing may boost testing rates and cut the time people remain unaware of HIV infection, and that could lead to lower HIV transmission and improved individual care.
Behavioral economics
A highly anticipated session with two late breaker presentations was "HIV and Behavioural Economics: Where the Money Is?" (TUAC01). The high level finding was that neither conditional cash transfer studies conducted with high school students in rural South Africa (CAPRISA 007 and HPTN 068) reduced HIV incidence among young people, though HIV incidence in both of these studies, conducted among mostly school-attending youth, was low.
The first trial (Abdool-Karim, abstract TUAC0101LB [https://www.youtube.com/watch?v=qJu67zTJmqw]), CAPRISA 007, was a cluster randomized trial in KwaZulu-Natal among 3,217 consenting male (n=1,517) and female (n=1,700) grade 9 and 10 students. A locally-developed HIV prevention program, "My Life! My Future!" was implemented in all 14 schools. In addition, 7 schools (n=1,592 students) were randomly assigned to receive the intervention of cash incentives (maximum of $175 over 2 years) for fulfilling any combination of 4 conditions; annual HIV testing, performance in school tests, participation in "My Life! My Future!" and a written report on their community involvement project. HSV-2 and HIV serology was undertaken at baseline, 12 months and 24 months. HSV-2 prevalence at baseline was 9.0% in intervention schools and 7.3% in control schools. During follow-up, there were 319 new HSV-2 infections, with an incidence rate of 6.2 per 100 person-years in intervention schools compared to 8.7 per 100 person-years in control schools (IRR=0.70, 95% CI 0.57-0.86; p=0.007). Students who received larger cash incentives had lower HSV-2 incidence rates. The lower-than-anticipated overall HIV incidence rate of 1.6 per 100 person-years was similar in both groups of schools (IRR=1.26, 95%CI 0.66-2.39; p=0.4). A four-fold larger study would be required for 80% power to observe a 30% HIV incidence reduction. Thus, CAPRISA 007 was able to demonstrate an impact on HSV-2 incidence but not in HIV.
In the second study (Pettifor, abstract TUAC0106LB), HPTN 068, cash transfers conditioned on 80% school attendance were tested in a randomized trial among 2,488 young women in Agincourt, South Africa. The intervention did not result in a reduction in HIV incidence over three years of follow-up (hazard ratio intervention vs. control arm: 1.17, 95% CI 0.80-1.72, p=0.42). School attendance was high in both arms and was protective against HIV incidence; cash transfers were protective against several high risk behavioral outcomes. HSV-2 results are not available yet.
In a study from Kenya (Thirumurth, abstract TUAC0102 [https://www.youtube.com/watch?v=94te323Xzg4]), 903 uncircumcised Kenyan men aged 21-39 years were randomized in a 1:1:1 ratio to two intervention groups or a control group. One intervention group was offered compensation of US$12.50 conditional on voluntary medical male circumcision (VMMC) uptake. Compensation was provided in the form of food vouchers valid at shops in the study region. A second intervention group was offered the opportunity to participate in a lottery with high-value prizes upon undergoing circumcision. The primary outcome was VMMC uptake within 3 months. Men who were randomized to receive compensation of US$12.50 had the highest VMMC uptake (8.4%, 26/308), followed by those receiving lottery-based rewards (3.3%, 10/302) and those in the control group (1.3%, 4/299). Logistic regression analysis showed that compared to the control group, the US$12.50 group had significantly higher VMMC uptake (adjusted odds ratio (AOR) 7.1; 95% CI 2.4-20.8).
Participants in the lottery-based rewards group were not significantly more likely to become circumcised than participants in the control group (AOR 2.5; 95% CI 0.8-8.1). The effect of providing compensation of US$12.50 was largest among participants who were contemplating circumcision at the time of enrollment. Thus, cash transfers successfully increased uptake of VMMC in Kenya, but lottery-based approaches were significantly less effective than the cash transfers.
Key populations: epidemiology and prevention
Men who have sex with men (MSM)

An oral abstract session on Tuesday (TUAC03) focused on the global epidemic among MSM. From China (Xu, abstract TUAC0301), a multi-city study in 2012 and 2013 sought to characterize the HIV epidemic in MSM. More than 4000 men were recruited: HIV prevalence was 10% and 42% of those infected appeared to have recent infection. The results signal a still-growing epidemic of HIV among MSM in China.
Two studies from Seattle explored the MSM epidemic there. In the first (Chan, abstract TUAC0302 [https://www.youtube.com/watch?v=HJD2EyGJ6pY]), a life table analysis was conducted to illustrate the expanding, then shrinking, risk of HIV. Men born in the 1940s had a lifetime cumulative risk of 15%.
Those born 1960-1964 (thus who were in their 20s in the 1980s) had much higher risk: >40% were diagnosed with HIV by the age of 50. But, subsequent birth cohorts have had less risk of ever getting HIV: for example, those born in the late 1970s had a lifetime risk of just 7% and later birth cohorts are even lower. For black MSM, the trends were similar although for all cohorts the cumulative HIV risk is higher than for white MSM. In the second (Khosropour, abstract TUAC0305 [https://www.youtube.com/watch?v=XLnB6TIZwQA]), serial surveys of nearly 1000 HIV negative MSM attending an STD clinic assessed behavioral practices. HIV- men reported frequently asking about testing in sex partners (86% asked their last partner). Decisions not to use a condom were more common when the partner's last HIV test was <3 months prior. Similarly, when HIV- men had sex with known HIV+ partners, men seemed to make choices about condoms based on whether the HIV+ partner was using ART and/or had an undetectable viral laod. In sum, many MSM in Settle use HIV testing frequency/recency and ART use to guide sexual decision-making. Such behaviors are nuanced but may be HIV protective.
In another example of behaviors modified based on understanding transmission risk, a study from Australia, Thailand, and Brazil (Bavington, abstract TUAC0306 [https://www.youtube.com/watch?v=7OfSzsKpic0]) explored risk behavior in male HIV serodiscordant couples taking part in a prospective observational study. Couples attend visits at least twice yearly. HIV+ partners have a viral load and CD4 done, and info collected on ART use. The authors asked the question: does condomless sex increase when the HIV+ partner is perceived to have an undetectable viral load? 269 couples were enrolled, half in Australia. About 1/3 had outside partners, only a fraction of those had condomless sex with an outside partner (9% of the total). 80% overall were taking ART, and nearly all had a viral load <200 copies/mL. Most HIV- partners "knew" their partner's viral load result. 53% reported having condomless anal sex with their partner (more common in Australia). Overall, HIV- partners were 2.6 times more likely to report condomless anal sex with their study partner when the perceived viral load was undetectable (p<0.01), but this was driven by the Australian subsample. Thus, particularly in Australia where there has been strong community messaging about the transmission benefits of having an undetectable viral load, MSM couples make behavioral decisions based on viral load.
Injection drug users
The global HIV epidemic among persons who inject drugs continues to expand, in spite of high-quality evidence of interventions that can virtually eliminate HIV in this population - specifically harm reduction interventions including needle/syringe exchange, opiate substitution therapy, and other prevention interventions. A Tuesday plenary on the interface of public policy, injection drugs, and HIV, was told from the lens of the city of Vancouver, which has a substantial drug injecting population and at one time had very high HIV incidence in this population, which was reversed by aggressive prevention policies (Wood, abstract TUPL0103). In contrast, a Wednesday plenary on the epidemic in Russia, Eastern Europe, and Central Asia was a telling reminder of how policies can also exacerbate an epidemic (Kazatchkine, abstract WEPL0102 [https://www.youtube.com/watch?v=jTKXye52ctM]). The plenary discussed the still-rapidly growing epidemic among persons who inject drugs in that region, compounded by repressive drug policies, high levels of stigmatization, few prevention and opioid substitution therapy services, and low treatment coverage.
A recent outbreak of HIV in rural Indiana in the US has been a public health emergency and a reminder of how quickly HIV can spread in drug-using networks in the absence of prevention services. Two abstracts (Duwve, abstract MOAC0303LB, Galang, abstract MOAC0304LB presented by Brooks) discussed the Indiana outbreak.
In brief, in late 2014, 3 new infections were detected in a rural county that had only seen 5 new infections in the past decade. By January 2015, the public health department had determined that 2 of the 3 infections had shared needles, for the injection of the opioid oxymorphone. A public health emergency was declared by the governor of Indiana in March 2015 and by June a total of 170 new infections were diagnosed, all from a single strain. The demographics of the new infections are notable: about half male/half female, 100% white, high levels of poverty and unemployment, relatively young (median age 32, but range 18-57). Injection practices included multiple times per day injections, with multi-partner and multi-generational sharing of injection equipment. Multiple prevention strategies were implemented rapidly: access to immediate ART for those infected, expanded testing and early diagnosis, syringe exchange (which was available prior to the outbreak), PrEP, and opioid treatment services.
A cluster-randomized trial of a interventions to reduce HIV risk in Ukraine (TUAC0402 [https://www.youtube.com/watch?v=ehNRcdVkLXk]). Recruitment was of an index member and then at least two members of their injection network; these micronetworks were randomized to either standard of care (counseling and education) or a peer-led intervention based on social learning. For the 1200 HIV negative participants, 260 HIV seroconversions occurred (at an incredibly high overall incidence of 25% per year). This differed by intervention arm: 31% per year for the standard of care versus 18% per year for the intervention (at 41% decrease, p<0.001 in multivariate analysis). The peer-intervention arm had increased use of needle exchange programs.
Female sex workers
A Wednesday oral abstract session focused on epidemiology and prevention of HIV in female sex workers from a variety of settings (WEAC01 [https://www.youtube.com/watch?v=JWuu9EEdSfQ]). In Vancouver (Argento, abstract WEAC0101), HIV prevalence was 12% among female sex workers (very high given very, very low HIV prevalence in women otherwise in that setting) and social cohesion was associated with reduced rates of condom refusal by clients. In South Africa (Lane, abstract WEAC0103 [https://www.youtube.com/watch?v=g59ZZT8gJZQ]), a survey of sex workers in three large cities (Johannesburg, Cape Town, and Durban) showed high HIV prevalence (72%, 40%, 54%, respectively in the three cities) but the opportunity for intervention, as utilization of health services was high; thus, sex worker-focused prevention and treatment services could potentially be implemented for impact. In Malawi (Lancaster, abstract WEAC0104 [https://www.youtube.com/watch?v=c_7lW_3D-GE]), one example of such a service was done - providing mobile testing and point-of-care CD4 counts. HIV prevalence was 69%, 20% were newly diagnosed, 30% were ART eligible (by CD4 count) but not on ART.