icon-    folder.gif   Conference Reports for NATAP  
  22nd Conference on Retroviruses and
Opportunistic Infections
Seattle Washington Feb 23 - 26, 2015
Back grey_arrow_rt.gif
HIV Prevention at CROI 2015 - Jared Baeten MD, PhD, Connie Cellum, MD, MPH
  HIV Prevention at CROI 2015
Conference on Retroviruses and Opportunistic Infections
Seattle, WA, USA
February 23-26, 2015
Jared Baeten, MD PhD
Connie Celum, MD MPH
University of Washington
CROI 2015, held in unusually sunny Seattle, again brought together a broad group of researchers, clinicians, policymakers, advocates, funders and others to talk about the pathogenesis, prevention, treatment, and cure of HIV and associated infections. The past years have seen tremendous enthusiasm regarding new HIV prevention and treatment strategies, which might signal a radical change in the global impact of the epidemic. As in previous years, oral sessions were recorded and are available online (http://www.croiwebcasts.org/), along with full copies of abstracts (http://www.croiconference.org/scientific-program/electronic-materials). We have cited specific links here for certain oral sessions - these contain both slides and video, but alternative links with slides with audio only are available.
Pre-exposure prophylaxis (PrEP) for HIV prevention
A major focus of the last four meetings of CROI has been the evidence and the role for PrEP for HIV prevention. In PrEP, an HIV-uninfected person uses an antiretroviral medication ahead of an HIV exposure in order to prevent infection. Prior to CROI 2015, four clinical trials demonstrated that daily oral PrEP, based on the antiretroviral agent tenofovir disoproxil fumarate (TDF), alone or in combination with emtricitabine (FTC), is efficacious for the prevention of HIV acquisition; one additional trial demonstrated efficacy for pericoital use of an intravaginal gel containing tenofovir. HIV protection ranged from 39% to 75% compared with placebo in these trials, and two additional trials that enrolled young African women found no HIV protection - the differences across trials have been attributed largely to adherence. Indeed, across all PrEP studies, adherence to PrEP is strongly related to its HIV protection benefits, and strong evidence has found that those who took PrEP (e.g., based on detection of the PrEP medication in blood samples) reduced their risk of HIV by 90% or more. The US Food and Drug Administration (FDA) in 2012 approved daily oral FTC/TDF as the first PrEP agent with a formal label indication for prevention of sexual HIV acquisition. In addition to a flood of oral abstracts and posters at CROI 2015 related to PrEP, a number of ancillary sessions on PrEP were extremely valuable, including the first morning plenary, which addressed moving PrEP from discovery to implementation (http://www.croiwebcasts.org/console/player/25535?mediaType=slideVideo&). One symposium speaker suggested that CROI 2015 was PrEP's 'bar mitzvah' - where this new prevention strategy emerged fully into the discussion. The presentations, questions, and chatter in the hallways about PrEP have shifted to how to best implement this effective prevention intervention for maximal benefit and impact. New research related to PrEP was presented in both oral and poster sessions.
Two studies from men who have sex with men (MSM) populations in Europe emphasized the high risk of HIV acquisition and the incredible potential of PrEP for preventing HIV. Both studies had their use of a comparison group discontinued this past October by their independent data monitoring committees. The first was the PROUD study, a randomized trial of daily oral FTC/TDF PrEP among men in the United Kingdom (McCormack, abstract 22LB, http://www.croiwebcasts.org/console/player/25539?mediaType=slideVideo&). PROUD enrolled 545 MSM attending public sexual health clinics. A primary goal of PROUD was to evaluate PrEP in a real-world setting, and thus the use of public clinics, with minimal extra research funds, was key to its design. PROUD randomly assigned men willing to take PrEP to immediate initiation versus a 12-month deferred initiation; the use of a delayed PrEP arm was done because PrEP was not available within the UK National Health Service at the time of the study and there was a desire to obtain contemporaneous comparative information on the use and prevention impact of PrEP in the UK. In total 276 men were assigned immediate PrEP - of whom 2 were found to be HIV infected at enrollment and another 7 had no follow-up; 269 men were assigned deferred PrEP - 1 was HIV infected at enrollment and 12 had no follow-up. In total there were 22 HIV infections: 3 in the immediate arm and 19 in the deferred arm, at an incidence of 1.3 versus 8.9 per 100 person-years. This difference was 86% and was highly statistically significant (90% CI 58-96, p=0.00002). The three cases of HIV in the immediate PrEP arm all had suggestions that they were not truly PrEP-exposed at the time of HIV infection (for example, one had defaulted for more than a year). In fact, the initial 545 men enrolled in PROUD were supposed to have been a feasibility pilot only, but the very high background rate of HIV (8.9% per year in those not on PrEP) and the high degree of HIV protection seen in the immediate PrEP arm resulted in definitive evidence for HIV protection with just the pilot-sized project.
Indeed, the number of men needed to be treated with PrEP for one year to prevent one infection was just 13 - the lowest number seen in any PrEP study to date and a strong indication of public health bang for the buck. The rate of sexually transmitted infections (gonorrhea, Chlamydia, or syphilis) in the population was very high - approximately 50% during follow-up - further emphasizing the risk nature of the population; the rate of STIs was slightly higher in the immediate PrEP arm, but there was also a greater frequency of testing in that arm (perhaps as a result of engagement in care due to PrEP prescribing). Sexual behavior data are just preliminary for the study but indicate very high exposure but no difference in number of partners or condom use among those receiving versus not receiving PrEP. Notably, 31% of those in the deferred arm received post-exposure prophylaxis (PEP) at some point during follow-up - the fact that HIV incidence was still so high in the deferred arm in spite of almost one-third receiving PEP emphasizes the gap that PEP does not yet fill for this population. Thus, in summary, PROUD shows that a high-risk MSM population can seek access to PrEP in a public health setting with brief counseling and use it well, with high-level HIV protection.
The second European study to present results was the IPERGAY clinical trial (Molina, abstract 23LB, http://www.croiwebcasts.org/console/player/25540?mediaType=slideVideo&). This trial tested FTC/TDF PrEP with "on demand" dosing (i.e., around the time of sex) - a strategy that might be easier for some to use. The trial assigned 414 MSM from France and Canada to either PrEP or placebo; as in PROUD, use of a non-PrEP arm was justified because of non-availability of PrEP within the public health system in France. Also like PROUD, the degree of HIV protection was 86% (95% CI 40-99, p=0.002), with 2 infections among those assigned PrEP and 14 among those assigned placebo, and the 2 infections on PrEP occurring in men who had stopped receiving PrEP months before they acquired HIV. The HIV incidence in the placebo arm was 6.6% per year. The "on demand" dosing in IPERGAY is important to take into account: 2 tablets of FTC/TDF were to be taken 2-24 hours before sex was anticipated, another 1 tablet 24 hours later, and then an additional 1 tablet approximately 48 hours after the first dose, for a total of 4 pills around each sexual act. Men took an average of 16 pills per month in IPERGAY, or 4 per week, an amount that has been associated with 100% protection against HIV in the previously reported iPrEx study (IAS 2014). Thus, there is question whether men with less than weekly sexual activity (who would not take 16 pills per month then) would get as high of HIV protection as the average in IPERGAY. Nevertheless, these results emphasize that PrEP may be forgiving to different dosing schedules, at least for MSM (pharmacologic data suggest that more rigorous dosing may be needed for heterosexual exposures), and intermittent dosing could be an alternative for some men to daily dosing.
We evaluated PrEP delivered in an open-label context (Baeten, abstract 24, http://www.croiwebcasts.org/console/player/25541?mediaType=slideVideo&). The Partners Demonstration Project is an open-label interventional study providing integrated delivery of PrEP and antiretroviral treatment (ART), which also has tremendous prevention benefits, to heterosexual HIV serodiscordant couples in Kenya and Uganda. The study responds to calls for demonstration projects of PrEP to determine how best to deliver this new prevention strategy. A total of 1013 couples were enrolled in the project, beginning in November 2012. Couples are specifically recruited if they have characteristics indicating higher HIV risk - such as younger age, no children within the partnership (which may result in risk-taking behavior), limited condom use, and a higher viral load in the HIV infected partner. In the project, PrEP is offered to the HIV uninfected partner "as a bridge" to ART in the infected partner - i.e., during the period when the infected partner is not on ART (due to individual or systems delays in starting) and for the first six months after ART is started (to give time for ART to fully suppress viral replication). As reported in a separate abstract at CROI 2015 (Heffron, abstract 969), PrEP uptake (96%) and adherence (>80% by multiple measures, including detection of tenofovir in plasma) are very high in the population; ART uptake is also high, although an important proportion of couples delay ART for at least a few months. In the oral abstract, the rate of HIV transmission in the study was reported - to date, only 2 transmissions have been observed, at a rate of only 0.2 per 100 person-years. Using statistical modeling to compare this rate to an anticipated HIV incidence based on a validated risk scoring tool for HIV serodiscordant couples, a total of 39.7 HIV transmissions would have been expected so far, at an incidence of 5.2 per 100 person-years. This difference between 39.7 and 2 is a 96% reduction (95% CI 81-99, p<0.0001). Both of the HIV transmissions were in women who had apparently stopped taking PrEP-one had left her partner and the other remained in a partnership with a man who was not on ART. These results suggest that this public health approach to delivering PrEP - offering it to the uninfected member of a couple until his or her partner has an undetectable viral load - provides nearly complete risk reduction to this key population. The PrEP bridging to ART and viral suppression strategy tested in the Partners Demonstration Project was favorable in a cost-effectiveness analysis presented in a poster at CROI 2015 (Ying, abstract 1106).
In considering the results of these three studies, particularly PROUD and the Partners Demonstration Project, one should recognize that the degree of HIV protection found in these open-label studies is greater, substantially so and seemingly nearly complete, than was seen in the randomized trials of PrEP - in other words, effectiveness was greater than efficacy. That is rarely seen in health studies, but it emphasizes that individuals who want PrEP seemingly take it very well and as a result achieve high protection; in the clinical trials of PrEP, uncertainty about PrEP's efficacy (which had not yet been proven) and potential randomization to placebo may explain why some did not adhere well, and thus did not garner HIV protection. Several posters reported on PrEP use in open-label studies. In Boston, PrEP uptake is increasing among MSM, based on experience from the Fenway Community Center, in which a tripling of PrEP prescriptions was noted between 2013 and 2014 (Mayer, abstract 972). Across the US, the CDC surveyed physicians, nurse practitioners, or physician assistants who had completed training and who provided care to HIV-infected patients during June 2013-January 2014 to estimate the weighted prevalence of ever prescribing PrEP and to describe patients for whom PrEP was prescribed (Garg, abstract 974). Of the 64% of HIV care providers who responded to questions about PrEP prescription and who also reported providing care to HIV-infected and non-HIV-infected patients (n=935), 26% ever prescribed PrEP, of which 74% prescribed to men who have sex with men, 23% to men who have sex with women, 30% to women who have sex with men, 23% to uninfected partners in serodiscordant couples trying to conceive, and 1% to injection drug users. PrEP coverage and adherence with daily or event-driven dosing was studied among young women in a study in Cape Town (HPTN 067, Bekker, abstract 978LB.) After 6 weeks of directly observed dosing, participants were randomly assigned to one of three unblinded PrEP dosing regimens for 24 weeks of self-administered dosing: daily, twice weekly with a post-intercourse boost, or before and after intercourse. Pills were dispensed from a Wisepill device that recorded each opening, and tenofovir and emtricitabine metabolite levels in plasma and PBMC were analyzed at 10 and 30 weeks. Of 179 women randomized to the self-administered phase, adherence was greater with daily compared with twice weekly and event-driven dosing. When sex was reported in the prior week, both plasma TFV and PBMC TFV diphosphate were detected in more women in the daily dosing at weeks 10 and 30, compared with twice weekly with a post-intercourse boost, or before and after intercourse. Daily dosing resulted in better coverage of sex acts and adherence, and higher drug levels, may foster better habit formation and provide the most forgiveness for missed doses. In another oral abstract from the same session as PROUD and IPERGAY, a combination of survey data and modeling analyses suggested that PrEP use in San Francisco is currently at about 1/3 of what the goal should be (based on patient preferences and risks), and that reaching the goal would result in a substantial decrease in HIV that that highly-impacted city (Grant, abstract 25, http://www.croiwebcasts.org/console/player/25542?mediaType=slideVideo&). Disappointing results presented at CROI 2015 came from the findings of the FACTS 001 study (Rees, abstract 26LB, http://www.croiwebcasts.org/console/player/25543?mediaType=slideVideo&), a randomized trial of tenofovir vaginal gel, a microbicide for HIV prevention. Tenofovir gel was tested in two prior large studies: CAPRISA 004, which reported in 2010 that peri-coital use (one dose before and one dose after sex, somewhat like IPERGAY for oral PrEP) reduced HIV risk 39% compared to placebo, and VOICE, which reported at CROI in 2013 that daily use did not reduce HIV risk, likely because <30% of women used the gel. FACTS 001 was done to replicate the findings of CAPRISA 004, randomizing 2059 South African women to either active or placebo gel, for use in a peri-coital manner like CAPRISA 004. The population was young (70% were <25 years of age), 90% were single, and 62% were living with parents or siblings. A total of 123 HIV infections occurred, at a rate of 4.0 per 100 person-years: 61 among those receiving active tenofovir gel and 62 among those receiving placebo, a completely null effect. The study team estimated that women used gel for approximately 50-60% of sex acts overall, and there was a suggestion that those who used it well (because of detection of tenofovir in cervicovaginal samples collected in the clinic) had reduced HIV risk by 52% (95% CI 3-77, p=0.04). However, the fraction of women who were able to use the gel well enough to prevent HIV was clearly too small to result in an overall demonstration of HIV protection in this population. These findings serve to further emphasize how prevention options that are usable by women at risk continue to be needed.
The safety of PrEP was initially established in the clinical trials, and ongoing analyses presented at CROI 2015 reinforced PrEP safety. Both PROUD and IPERGAY found PrEP to be safe in their populations. In addition, several posters assessed aspects of safety. Analyses from the Partners PrEP Study found that PrEP-selected drug resistance detected in 9 participants using 454 ultrasensitive sequencing faded by 6 months following drug cessation (Weis, abstract 983), emphasizing that PrEP-selected HIV resistance, rare to begin with, appears to fade over time. A second poster assessed kidney PrEP safety, finding that the slight decline of 2 mL/min/173m2 in estimated glomerular filtration rate associated with PrEP use disappears within 4 weeks discontinuation of daily oral PrEP (Mugwanya, abstract 981). Finally, PrEP drug sharing was rarely reported (0.003% of visits) or detected based on tenofovir testing when measured in the heterosexual population who took part in the Partners PrEP Study (Thomson, abstract 988).
Long-acting PrEP agents may avert some of the adherence challenges faced by daily oral/vaginal PrEP - either as injectable agents or slow-releasing intravaginal rings. Data on safety and ex vivo challenges were presented for several new PrEP agents. Encouraging preliminary safety, tolerability and tissue challenge data were presented about the prodrug tenofovir disoproxil fumarate (TDF) ring (Keller, abstract 992LB), which is more potent than tenofovir (TFV) against HIV in vitro and has greater tissue permeability and cellular uptake. Safety was excellent and 11 of 14 had detectable levels after 14 days of the TDF vaginal ring use with tissue TFV-DP levels ranging from 52-550 fmol/mg. Tissue from women in TDF arm challenged ex vivo had a 90% reduction in HIV copy number compared to no change in the placebo arm. A rectal ex vivo challenge was conducted on rectal tissue obtained at baseline and 7 days after a single oral dose of maraviroc 300 mg or 600 mg, and showed no protection (Coll, abstract 964). A study of 26 Chinese rhesus macaques study of injectable cabotegravir (GSK1265744) followed by weekly intravaginal SIV mac251 challenges demonstrated significant protection with both the 30 and 50 mg/kg doses (Spreen, abstract 966LB) but no difference between the 10 mg/kg dose and placebo. A cabetogravir plasma concentration of 4 times the protein-adjusted IC90 level (i.e., 711 ng/mL) was associated with a 90% probability of in vivo protection after 7 SIV challenges.
HIV testing and linkages to care, and treatment as prevention
HIV testing is the starting point for all subsequent steps in the HIV treatment and prevention cascades, particularly initiation of antiretroviral therapy (ART) for those found to be HIV infected. ART is key to prevention strategies worldwide, as treatment results in near complete (if not complete) elimination of infectiousness.
The results of HPTN 065/TLC-PLUS were presented at CROI 2015 (El-Sadr, abstract 29, http://www.croiwebcasts.org/console/player/25546?mediaType=slideVideo&). Innovative strategies to facilitate linkage to care and viral suppression were evaluated in HPTN 065 through 38 testing/care sites in the Bronx NY and 38 sites in Washington DC between 2011-2012. HIV test sites were randomized to financial incentives ($25 gift card for HIV+ persons getting follow up lab tests done and $100 gift card for completion of provider encounter with development of care plan) or standard of care. HIV care sites were randomized to a $70 gift card for having HIV+ persons engaged in care and achieving VS (<400 c/ml) and a maximum of one gift card for every 3 months. The study outcomes were measured via the surveillance system and continuity of care was determined by having a CD4 or viral load in at least 4 of 5 quarters. For 1109 HIV+ diagnoses, 1061 coupons were dispensed, of which 79% were redeemed. The rates of linkage to care increased over time and were not increased with the financial incentives.
For the viral suppression intervention, there were 9641 patients eligible and 49,650 visits qualified for VS gift cards. The proportion with viral suppression also increased over time in both arms and overall was not increased with financial incentives, except in the subgroups of sites with a lower proportion (<65%) of virally suppressed patients at baseline, hospital-based sites, and smaller sites. These findings emphasize that financial incentives are feasible and, although they may have had benefits in some subgroups, were not a universally robust strategy to increase linkage to care and viral suppression.
HIV testing and positive status awareness rates were evaluated in a cross-sectional population-based survey of Individuals aged 15-59 years in Kwa-Zulu Natal, South Africa (Huerga, abstract 1014). Overall HIV prevalence was 25.2% with 30.9% in women and 15.9% in men. Although 81% reported ever having had an HIV test prior to the survey, the proportion was significantly higher among women (88% of women vs 70% of men) and a majority of HIV+ men were unaware of their diagnosis. These results were consistent with results presented from the Swaziland HIV Incidence survey (SHIMS, Ellman, abstract 1013). Of the 5,829 (32%) adults tested HIV-seropositive in SHIMS, 38% were unaware of their diagnosis. Undiagnosed HIV was more common among men than women (50% vs 32%). Effective HIV testing strategies to reach men are urgently needed.
This theme was also demonstrated in data on engagement within HIV care in South Africa were analyzed by triangulating national viral load and CD4 data with vital registration data and national HIV prevalence survey data (Takuva, abstract 154). An estimated 6.4 million people are living with HIV in South Africa based on 2012 data, of whom 3.3 million (51%) are estimated to be linked to care based on lab results, 2.2 million (34%) on ART, and 1.6 million (25%) virally suppressed. Of those on ART, 75% virally suppressed and among all PLWHA only 1 in 4 have viral suppression. Across all stages, males were less likely than females to be engaged in care, on ART and virally suppressed.
A hybrid mobile HIV testing approach of multi disease community health campaigns followed by home-based testing achieved population-wide coverage in the SEARCH trial in 32 communities in eastern and southwestern Uganda and western Kenya; 89% of stable adult residents and 80% of all adults were tested for HIV, 43% of whom reported no prior HIV testing. (Chamie, abstract 1101). Adult HIV prevalence was 9.4% (twice the national average) and persons were identified early in their course of infection with a median adult CD4+ count of 516 cells/μL.
At CROI 2014, one abstract that received considerable media attention was one describing HIV transmission risk in a prospective cohort of European HIV serodiscordant couples on ART (the PARTNERS study). There were no HIV transmissions within the partnerships during the study, translating into an upper 95% confidence interval of the risk estimate of 0.4% per year, and 1% per year for those practicing anal sex. A similar study, among MSM couples primarily from Australia also has demonstrated no HIV transmissions to date (Grulich, abstract 1019LB).
Other interesting talks that touched on testing, linkage, PrEP, and other themes included a symposium talk on implementation in the US (http://www.croiwebcasts.org/console/player/25776?mediaType=slideVideo&) and a Thursday morning plenary on HIV risk and prevention globally in sex workers (http://www.croiwebcasts.org/console/player/25787?mediaType=podiumVideo&).
Reproductive health and HIV
Hormonal contraception is used widely and plays an important role in preventing unintended pregnancies and reducing maternal morbidity and mortality. Some (but not all) prospective observational studies have found an increased risk for women to acquire HIV infection when they are using hormonal contraception, especially injectable depot medroxyprogesterone acetate (DMPA, otherwise known as branded Depo-Provera). This injectable contraceptive is a popular contraceptive method, with high use in southern and East Africa, where HIV is prevalent; it is also easy to use, fast to administer, and can be used discretely. Limited data also suggest that HIV infected women using DMPA might be more infectious to sexual partners. A poster discussion at CROI 2015 brought in new information that will further the conversation.
In an analysis of samples from the Women's Interagency HIV Study (WIHS), CD4, CCR5, and CXCR4 expression on peripheral blood mononuclear cells was assessed in women using different contraceptive methods, or no contraception (Tsibris, abstract 858, http://www.croiwebcasts.org/console/player/25615?mediaType=slideVideo&). Both the levonorgesterol IUD and DMPA were associated with increased CCR5 expression, suggesting one possible mechanism by which HIV susceptibility could be altered. Another study (Swaims, abstract 860, http://www.croiwebcasts.org/console/player/25616?mediaType=slideVideo&) also assessed peripheral blood mononuclear cells, in this case demonstrating increased CCR5 expression during the luteal phase of the menstrual cycle. In a third poster (Roxby, abstract 861, http://www.croiwebcasts.org/console/player/25617?mediaType=slideVideo&) 15 women initiating DMPA were followed, with vaginal fluid sampling for microbiologic studies and cytokines; after DMPA start, a sustained decline in inflammatory mediators IL-6, IL-8, and IL1-ra was seen as well as shifts in vaginal bacterial concentrations. Taken together, these studies raise hypotheses about how contraceptive methods and HIV risk could interact.
Medical male circumcision
A themed discussion discussed voluntary medical male circumcision (VMMC) and evolving knowledge and practice (http://www.croiwebcasts.org/console/player/25619?mediaType=slideVideo& & http://www.croiwebcasts.org/console/player/25625?mediaType=slideVideo&). Approximately 9 million VMMCs were conducted by the end of 2014, with numbers doubling between 2011-2013 including high numbers in Kenya, Tanzania and Uganda. With current trajectories of VMMC and current levels of flat funding, it is estimated that the target of 20 million men circumcised by 2016 will fall short by 4.5 million. Although the VMMC program was established to target men ages 15-49, 30-40% of clients served so far are ages 10-14. Overall, programs have not observed significant risk compensation, although resumption of sex before complete wound healing at 6 weeks has been observed in a significant minority.
Different service delivery models for VMMC in Tanzania include fixed clinics, campaigns, and mobile VMMC teams (Hellar, abstract 1086, http://www.croiwebcasts.org/console/player/25620?mediaType=slideVideo&), in which mobile VMMC services were more likely to see older men (>19 yrs) compared to fixed clinics and VMMC campaigns (38%, 22%, and 29%, respectively). PrePex was introduced into Rwanda VMMC programs in February 2014; 63% of 20,877 VMMC in Rwanda were performed with PrePex between February and August 2014 (Rugwizangoga, abstract 1087, http://www.croiwebcasts.org/console/player/25621?mediaType=slideVideo&). Acceptability was high, and uptake could have been even higher but was limited by stock-outs in certain sizes. The potential for self-selection, risk compensation, and characteristics of 1192 men who were circumcised was evaluated in comparison to a stratified random sample of 2384 men in the Rakai Community Cohort Survey (Kayaagi, abstract 1088, http://www.croiwebcasts.org/console/player/25624?mediaType=slideVideo&). Circumcised men were younger, more educated, less likely to be married, less likely to be currently sexually active, and more likely to report genital ulcers in the prior 12 months, suggesting self-selection for VMMC. Although there were indications of risk compensation among men who were circumcised with a modest (3%) increase in sexual activity among men ages 15-24 and a 10% increase in sex with a high risk woman, that did not attenuate the protective benefit of VMMC as circumcised men were 50% less likely to acquire HIV in this sample. The population level effect of VMMC was assessed in Rakai; VMMC coverage among non Muslim men was estimated to be 49% and a 10% increase in MMC among non-Muslim men was associated with a 12% decrease in community level HIV incidence (Kong, abstract 158, http://www.croiwebcasts.org/console/player/25818?mediaType=slideVideo&).
The Rakai program evaluated biologic effects of MC on penile HIV and HSV-2 shedding in HIV-infected men as surrogates for increased infectiousness after MC in HIV-infected men and the association between microbiome and local cytokines in uncircumcised men. Among 236 HIV-infected men, using penile lavage prior to MC and at weekly post-surgical visits for 12 wks, penile HIV shedding was detected in 13% of visits post MC visits, and was transiently increased at weeks 1 and 2 but was significantly decreased at 3 months (Manucci, abstract 30, http://www.croiwebcasts.org/console/player/25547?mediaType=slideVideo&). Penile HIV shedding was more likely in men with CD4<200, lower among men with a fully healed wound and on ART. Penile HSV-2 shedding was increased in the first 2 weeks following MC in 176 HIV/HSV-2 co-infected men in Rakai, as measured by penile lavage and HSV-2 PCR (Grabowski, abstract 1084); baseline HSV-2 shedding rates were 9.7%, which transiently increased to 12.9% at week 1 and 14.9% at week 2, and decreased to 6.9% by week 6. There were no differences in HSV-2 shedding by ART status or plasma HIV levels. Investigators recommend counseling about immediate post-MC risk of HSV-2 and HIV transmission and the need for abstinence until wound healing. The association between foreskin microbiota and local cytokines were evaluated in 138 uncircumcised men from Rakai, which demonstrated that anaerobic flora (Prevotella and Peptostreptococcus, which they have shown are decreased with MC) were associated with significant increases in penile IL-8 levels, while Corynebacterium (which is increased post-MC) was associated with lower levels of IL-8 (Liu, abstract 1085). Further research is needed to understand how the intact foreskin and removal of the foreskin of change the susceptibility to HIV in uncircumcised and circumcised men. Along this line, non-medical penile cutting (dorsal slit) procedures, which are part of cultural practices in Papau New Guinea, were associated with increased thickness of the inner foreskin stratum corneum, which may help explain lower risks of HIV associated with penile cutting in that setting (Shih, abstract 1089, http://www.croiwebcasts.org/console/player/25622?mediaType=slideVideo&).
Sexually transmitted infections
Two oral abstracts explored herpes simplex virus type 2 (HSV-2) in the context of HIV prevention. The first (Bender-Ignacio, abstract 27, http://www.croiwebcasts.org/console/player/25544?mediaType=slideVideo&) explored the effects of oral and vaginal gel tenofovir for reducing herpes simplex type 2 outbreaks and asymptomatic viral shedding in women without HIV infection but who had chronic HSV-2. Prior studies have already found that oral and vaginal tenofovir-based PrEP prevents acquisition of HSV-2 to some degree, likely because tenofovir is a weak anti-HSV-2 medication (in addition to being a strong anti-HIV one). The study found that oral tenofovir reduced HSV shedding and lesions by about 25-30% and vaginal gel reduced the quantity of virus shed, but not the rate of shedding. These results confirm that tenofovir has some in vivo activity against HSV-2, but the degree of anti-HSV-2 activity is too low to use this medication as a sole treatment for HSV-2, for which other medications that work very well (e.g., acyclovir and related compounds) are already available.
The second abstract (Grabowski, abstract 28, http://www.croiwebcasts.org/console/player/25545?mediaType=slideVideo&) assessed the effect of use of hormonal contraception on HSV-2 acquisition. As noted above, a potential effect of contraception on HIV acquisition and transmission is a topic of ongoing debate and investigation. Among 682 women in Rakai, Uganda who were seronegative for both HIV and HSV-2, 70 HSV-2 seroconversions occurred, at an overall incidence of 13.5 per 100 person-years. Compared to women not using contraception, women using the injectable contraceptive DMPA consistently during follow-up had a doubling in risk of HSV-2 (adjusted HR 2.26, 95% CI 1.09-4.69, p=0.03) and this elevated risk was preserved in sub-analyses restricted to women with a known HSV-2 infected partner. Pregnancy, oral contraceptive use, and inconsistent DMPA use were not associated with increased HSV-2 risk. The results of this study are suggestive - only one prior study has tried to assess whether contraception increases HSV-2 risk (and no increased risk was seen in that study) - but its small sample size (only 9 HSV-2 infections occurred in consistent DMPA users, only 4 of whom had a known HSV-2 infected partner) and the potential for confounding do not make these results definitive, a point raised repeatedly by the authors. This is a priority area for continued investigation.
Prevention of mother-to-child transmission of HIV
Prevention of mother-to-child transmission (PMTCT) of HIV is a tremendous public health achievement. Understanding how best to deliver PMTCT services is the current research priority. Tuesday morning featured a powerful oral abstract session on PMTCT-related topics.
Data from the PROMISE Study (Fowler, abstract 31LB, http://www.croiwebcasts.org/console/player/25551?mediaType=slideVideo&) PROMISE is a large, complex, multinational study with three levels of randomization that was designed to evaluate optimized PMTCT management in breastfeeding populations in the context of evolving availability of antiretroviral therapy. Pregnant women with CD4 counts ≥350 cells/μL are randomized in the antepartum period at triple ART (one of two different regimens lopinavir/ritonavir, with either zidovudine/lamivudine or tenofovir/emtricitabine) versus shorter-course zidovudine with intrapartum single-dose nevirapine with a post-delivery tail of tenfovir/emtricitabine (antepartum randomization). The other two levels of randomization are post-partum: for infants, to triple versus nevirapine-only prophylaxis during breastfeeding and for women, to continue triple ART versus stop treatment until clinically indicated. At CROI 2015, data from an antepartum randomization were reported, after an interim Data and Safety Monitoring Board review in November 2014 recommended changes to the study protocol. A total of 3529 women were enrolled in the antepartum component across 15 sites in 7 countries. Women were young and healthy: the median age was 26 years, 97% were African, the median CD4 count was 530 cells/μL, and 97% had WHO stage 1 disease. Mother-to-child HIV transmission through 14 days after birth was <2% for both randomization groups, but was significantly lower for women assigned triple ART than those assigned the single-dose nevirapine-based strategy (9 infections in 1710 infants [0.6%] vs. 25 infections in 1326 infants [1.8%], a difference of -1.28% in risk [95% CI -2.11 to -0.44]). Moderate grade abnormal laboratory results were more common in those receiving triple ART, although these were relatively uncommon (<10%). Adverse pregnancy outcomes of moderate (low birth weight <2500 g, preterm delivery <37 weeks) severity were more common for those receiving triple ART compared to the single-dose nevirapine-based regimen, and severe adverse pregnancy outcomes (birth weight <1500 g, delivery at <34 weeks, occurring in 9% of infants) were more common only in the triple ART arm using tenofovir/emtricitabine. There were significant differences in infant outcomes for infants in the triple ART versus single-dose nevirapine-based arm; however, within the triple ART arm, the rate of infant death was 0.6% for those exposed to zidovudine/lamivudine versus 4.4% for those exposed to tenofovir/emtricitabine (p=0.001). In summary, the antepartum results from PROMISE support the 2013 WHO recommendations for use of triple maternal ART during pregnancy to achieve the lowest risk of HIV transmission possible. Importantly, the two triple ART regimens tested both were associated with increased risk of moderate adverse pregnancy outcomes, including preterm birth and low birth weight, and there appeared to be a difference in more rare severe outcomes for the triple ART arm using tenofovir/emtricitabine compared to the zidovudine/lamivudine arm. The differences across arms will be a subject of ongoing follow-up in this trial population.
The cascade of care in pregnant and breastfeeding women, particularly its key endpoint, population viral load, was assessed in three population-based, multistage household surveys, from Malawi, South Africa, and Kenya (Maman, abstract 32, http://www.croiwebcasts.org/console/player/25552?mediaType=slideVideo&). Of 12,461 women identified in the sample, 11,550 were tested (92.7%). The prevalence of breastfeeding was 38% in Kenya, 34% in Malawi, and 13% in South Africa; HIV prevalence among pregnant and breastfeeding women was 22% in Kenya, 13% in Malawi, and 23% in South Africa. HIV testing at antenatal clinics was >85% in all three surveys - a strikingly positive finding. However, the cascade of care was troubling in all settings, with important gaps in all segments, from diagnosis to ART initiation to viral suppression. The gaps were greatest in the Kenyan survey, done in an area where Option A (i.e., not triple ART in the antepartum period but instead zidovudine and single-dose nevirapine in those with higher CD4 counts) was the practice at the time. Those with viral loads <1000 copies/mL were 27% of those in Kenya, 72% in Malawi, and 63% in South Africa; even in Malawi, though, 12% had a viral load >100,000 copies/mL. HIV incidence among women aged 15-29 who were breastfeeding was estimated using incidence assays at 3.8 per 100 person-years in Kenya, 0.9 in Malawi, and 3.2 in South Africa. These results emphasize important gaps in diagnosis, incident infection in pregnancy and breastfeeding, and limitations in implementing combination ART in pregnancy and breastfeeding.
An analysis from the Malawi BAN study (which previously demonstrated the efficacy of maternal and infant antiretrovirals in preventing HIV transmission) assessed delayed HIV detection in infants exposed to post-natal nevirapine prophylaxis (King, abstract 33, http://www.croiwebcasts.org/console/player/25553?mediaType=slideVideo&). During the study, infants were tested for HIV by the Gen-Prove Aptima HIV-RNA assy of the Roche Amplicor HIV-1 Qualitative DNA assay. A total of 28 infant HIV infections were detected after 28 weeks of age, of whom 13 had received post-partum infant nevirapine, a number that was higher than for infants who had not received prophylactic nevirapine. Of these, 9 (3 who had received nevirapine prophylaxis) were tested by an ultrasensitive droplet digital PCR assay at time points prior to their HIV diagnosis. Six of these 9 had HIV DNA detectable by the ultrasensitive assay - with a delay of 9 to more than 22 weeks. Only one infant had resistant virus, but the mother had the same mutation, suggesting transmission rather than selection. In summary, the ultrasensitive assay used in this study detected HIV transmission earlier than standard assays, and the delay to standard detection was greater in the context of post-natal nevirapine prophylaxis.
In the PMTCT poster sessions, interesting new data were presented. In a study from Kenya, the rate of viral decline after ART initiation was slower in women with acute versus chronic HIV infection (Drake, abstract 866), emphasizing yet another challenge of incident HIV infection in pregnancy, and in another study, a high rate of HIV superinfection was seen in HIV-infected pregnant women (Redd, abstract 869). In a multicenter study from France (Mandelbrot, abstract 867), no HIV transmissions were observed from 2588 HIV-infected women to their infants who had started ART prior to conception and continued it through pregnancy. In Botswana, scale-up for ART has resulted in a fall in MTCT rates to <1% (Powis, abstract 870). Further positive data about the potential benefits of ART in pregnancy was detailed in Option B+ reports from Rwanda (Bobrow, abstract 865), Mozambique (Sebastian, abstract 873), and Malawi (Herce, abstract 874), although retention in care for this generally healthy population is more challenging, as emphasized by a report from Haiti (Domercant, abstract 875).
Finally a Wednesday morning plenary overviewed PMTCT and management of pediatric HIV, with a wide vision (http://www.croiwebcasts.org/console/player/25663?mediaType=podiumVideo&).