HIV Prevention at CROI 2020
Conference on Retroviruses and Opportunistic Infections |
Boston, MA, USA & Virtual
March 8-11, 2020
Jared Baeten, MD PhD
Connie Celum, MD MPH
University of Washington
CROI 2020 was held virtually rather than in person because of the expanding global outbreak of COVID-19/SARS-CoV-2. Nearly forty years after the global HIV pandemic was first recognized, the SARS-CoV-2 pandemic raises many parallels - a new and deadly infection, quickly realized to have spread rapidly across the planet, having serious impact on the lives of many, healthcare workers bravely at the frontlines, and bringing fear and stigma with it as well. Many, many CROI attendees were simultaneously working on efforts to contain and mitigate the pandemic, and the work to fight COVID-19 increased exponentially immediately after CROI. We applaud the CROI leadership for their dedication to getting the science out there, their obvious passion for this work, and their singularity of purpose to make CROI 2020 the tremendous success that it was.
An excellent special COVID-19 session (http://www.croiwebcasts.org/console/player/44708?mediaType=slideVideo&) delivered up-to-date (for March 10th) information on the expanding pandemic. Its inclusion in the meeting was an incredibly valuable addition for an audience looking to understand the epidemiology, clinical manifestations, and pathophysiology of this new disease.
HIV prevention has long been a major topic at CROI, with game-changing new options in prevention frequently being reported for the first time there, including at CROI 2020. This year's CROI did not disappoint - there were important new prevention advances, many presentations on accelerating the prevention advances we have in our hands now, and then fantastic additional work addressing areas the relate to HIV prevention in the US and globally. 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/abstracts/search-abstracts/). 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.
HIV prevention in plenary sessions
CROI always has fantastic plenary sessions, and this year was no exception. Monday and Tuesday both included plenaries related to prevention work: an assessment of where we stand after completion of the multiple universal test and treat trials (De Cock, http://www.croiwebcasts.org/console/player/44560?mediaType=slideVideo&) and broad review of HIV prevention for persons who inject drugs (Strathdee, http://www.croiwebcasts.org/console/player/44662?mediaType=slideVideo&). The annual N'Galy-Mann Lecture, which celebrates collaborative global work, was an especially thoughtful assessment of what it will take at the front lines of HIV treatment and prevention in Africa to turn the course on the epidemic (Coutinho, http://www.croiwebcasts.org/console/player/44554?mediaType=slideVideo&).
HIV treatment and testing
HIV testing, linkage to care, initiation of ART, continued engagement in care, and viral suppression define the continuum that is key to maximizing the benefits of treatment as prevention. A number of sessions focused on HIV testing, particularly focusing on strategies to improve the steps in the testing-to-viral suppression continuum.
From Kenya, a study was presented evaluating assisted partner services (APS) among persons who inject drugs (PWID) (Guthrie, abstract 42, http://www.croiwebcasts.org/console/player/44596?mediaType=slideVideo&). The focus of the analysis was the "number needed to interview" to identify a new positive case, or one not on ART - a measure of yield for APS, which tends to range from 3-8 across different contexts in prior studies, a lower number is preferable. The process of APS has been shown repeatedly as an efficient, effective, acceptable, and safe method to identify previously-unidentified persons living with HIV who were not on ART. In the SHARP study, PWID living with HIV were enrolled and asked to identify all of their sexual partners for the past three years - making them an "index" for APS. A counselor, or the person themselves with counselor assistance, would then contact partners. If partners were found to have HIV themselves, they could then become a new index. The study enrolled 627 persons (325 men, 302 women), with a median age in the mid-30s, 24% men / 80% women engaged in transactional sex, ∼80% were on ART, and 27%/9% were infected with HCV. The median number of partners named was 4 for both men and women, and dual sexual and injection partners were rare. In total 2772 partners were named, 2526 were traced, and 724 (29%) were HIV positive but only 86 (12%) were unaware of their HIV status, reflecting Kenya's very strong testing program; however 166 (23%) were not on ART - a NNTI of 3.8 (5.7 for male and 2.8 for female indexes). The site in Nairobi (versus in coastal Kenya) and needle exchange centers (versus methadone programs) yielded a better NNTI. Those identified were highly likely to start ART. In conclusion, APS in PWID in Kenya was feasible and productive, with the best NNTI was greater among women, in Nairobi, and those from drop-in centers.
From Lesotho, a study evaluated multi-month ART dispensing (Tukei, abstract 43, http://www.croiwebcasts.org/console/player/44597?mediaType=slideVideo&), a form of community-based differentiated care delivery. 30 health facilities were randomized as clusters to: 1) 3-monthly ART supply at facilities (control), 2) 3-monthly ART supply through community-based delivery, 3) 6-monthly ART supply through community-based delivery as well. Eligible persons for this differentiated care system were on first-line ART, had a viral load of <1000 copies/mL, and were otherwise healthy. In total 5336 persons were enrolled (1898 in the control group, 1558 in the 3-monthly community group, and 1880 in the 6-monthly community group), out of just over 10,000 screened - the primary reason for non-enrollment was lack of a viral load in the prior 12 months. The primary outcome was retention in ART care: 96.1% overall, 97.1% in the control, 96.5% in those with 3-monthly supplies, and 94.7% in those with 6-monthy supplies, and no significant difference across the randomized arms and all were within the 3.25% non-inferiority limit defined for the trial. Viral suppression was high among those who had viral load results: 98.3% overall, and 98.6%, 98.1%, 98.3% in the three arms, respectively. Mortality was very low - <1%. In sum, retention in care and viral suppression were high in this cluster-randomized trial of community-based differentiated care, suggesting this is a feasible strategy to expand ART delivery options.
The SEARCH trial was a community-randomized study of universal testing and treatment for HIV in Kenya and Uganda; its results were reported in 2018 and did not show a difference between the randomized arms but did demonstrate a community-based strategy to provide multi-disease care, including for HIV, in efficient ways. An abstract from SEARCH assessed care engagement and viral suppression between the randomized arms (Hickey, abstract 45, http://www.croiwebcasts.org/console/player/44599?mediaType=slideVideo&). Among nearly 4400 persons (35% men, 8% youth, and nearly 3000 already on ART), streamlined care in the SEARCH model was associated with improved "time in care" (a measure of retention in care) and some improved viral suppression (for the subset that were ART-experienced and viremic: 67% vs. 47%, RR 1.41, 95% CI 1.04-1.92). Yet another example that there are good ways to make HIV care more simplified.
Similarly, from Philadelphia, a collaborative data-to-care intervention improved reengagement in care, retention in care, and viral suppression among persons who were out of care (Brady, abstract 46, http://www.croiwebcasts.org/console/player/44600?mediaType=slideVideo&). The study was a randomized trial: standard of care (which included appointment reminders by calls or letters and routine clinic outreach) vs. the intervention, in which a public health professional (a disease intervention specialist) assisted with reengagement. A total of 898 were enrolled, the majority were black, cis-gender male, and over age 40. Overall, patients assigned to the intervention had higher rates of reengagement in care within 90 days of randomization (62% vs. 43%, p<0.001), retention in care at a year (59% vs. 44%, p<0.001), and viral suppression within 1 year (64% vs. 56%, p=0.01). Thus, a public health-driven approach can improve reengagement in care, including with effects lasting a year, over standard clinic-based strategies.
From the four large UTT trials in Africa (mentioned above from the Monday plenary), this abstract analyzed the relationship between HIV viremia and suppression at the population level and HIV incidence (Peterson, abstract 46, http://www.croiwebcasts.org/console/player/44601?mediaType=slideVideo&). The four trials: Ya Tsie in Botswana, PopART in South Africa and Zambia, SEARCH in Kenya and Uganda, and TasP in South Africa. The analysis used community-level linear regression, taking into account the cluster-randomized nature of the trials. In total, there were 105 communities, with prevalence of HIV from 2-40%. Overall, HIV incidence increased by 0.07 (95% CI 0.05-0.1) per 100 person-years for each 1% increase in viremia and this relationship was true in each study. Non-suppression was also associated: each 10% non-suppression was associated with a 0.12 per 100 person-year incidence, and this association was true in each study. In sum, both viremia and non-suppression were associated with HIV incidence across the UTT trials. The findings support the thrust of the UTT trials: increasing population-level viral suppression decreases HIV incidence. A second abstract, from Australia, similarly showed a strong association between decreasing community viremia and decreasing HIV incidence (Callander, abstract 47, http://www.croiwebcasts.org/console/player/44602?mediaType=slideVideo&).
Finally, a randomized trial of community-based ART delivery was reported (Barnabas, abstract 48, http://www.croiwebcasts.org/console/player/44603?mediaType=slideVideo&). The trial took place in Sheema district in Uganda and KwaZulu-Natal in South Africa. The primary objection was to test the impact of community-based ART on viral suppression, with pre-defined subgroups looking at South Africa and particularly men in South Africa. Persons living with HIV and not on ART, who were clinically stable were randomized to: 1) a clinic group with clinic ART initiation and resupply, 2) a hybrid group with clinic ART initiation and mobile van resupply, and 3) home or mobile van ART initiation and mobile van resupply. The primary outcome was viral suppression (<20 copies/mL) at 12 months. Community ART was nurse-led. 8265 persons were tested for HIV, 2479 were living with HIV, and 1315 were randomized (most not randomized were on ART and virally suppressed already): 446 to clinic, 442 to hybrid, and 427 to community group; 305 were from Ugana and 1110 were from South Africa, and 51% overall were men and 37% were <30 years of age. For the primary endpoint, 63%, 68%, and 74% achieved viral suppression in the clinic, hybrid, and community groups, reflecting superiority of community over clinic (1.18, 1.07-1.29, p=0.00053) and noninferiority for hybrid versus clinic. For South Africa, the fractions were 60%, 67%, 72% with similar results. When limited to men, 51%, 65%, 72%, with both the community and hybrid superior to clinic. There were few adverse events, and no more in the community arms than the clinic. Notably, the overall viral suppression was 72% for men and 73% for women in community group, signaling parity. In sum, 93% of people were eligible for community-based ART, and community-based ART was superior overall and particularly among men and was safe.
In another study of differentiated care, a randomized trial from South Africa evaluated postpartum ART (Odayar, abstract 131LB, http://www.croiwebcasts.org/console/player/44814?mediaType=slideVideo&). The population was women living with HIV who had started ART in pregnancy or within 70 days post-partum and who had a viral load <400 copies/mL. The intervention was an adherence club, a group-based care model (20-30 people, with shorter visits, pre-packed medications, and annual lab testing and operating at a community venue), compared to continued follow-up in HIV clinics (standard of care). Follow-up occurred through 24 months. In total, 412 women were randomized: 206 to the adherence clubs and 206 to standard of care. Through 24 months, the proportion ever having a viral load >1000 copies/mL was 29% among those assigned to the adherence club versus 37% for those assigned to standard of care (HR 0.71, 95% CI 0.50-1.01), results were similar if the viral load cut-off was >50 copies/mL and were largely consistent across subgroups and in a per-protocol analysis. Safety outcomes - including death, TB diagnosis, and infant outcomes - were similar. Thus, post-partum differentiated care improved viral suppression and was safe.
In a themed discussion, presentations highlighted strategies and impact of partner notification in different settings. Namimbi (abstract 937, http://www.croiwebcasts.org/console/player/44633?mediaType=slideVideo&) presented about assisted partner services (APS) in Kampala, Uganda. Recent surveillance (the Uganda PHIA) indicates 6.9% HIV prevalence in Kampala, of whom only 77% HIV+ have been identified. APS were implemented from October 2107-September 2018 which trained HCW from 69 health facilities. HCW identified HIV+ clients >15 years of age who had a notifiable sex partner of unknown HV status; index clients gave written consent for APS. HIV testing was conducted at facility or in the community, and tested partners were linked to HIV care or prevention services. Of 53,162 HIV+ clients eligible for APS, 35,704 were interviewed for a 67% acceptance rate, and 48% of those notified were tested for HIV. On average, 1.3 sex partners were interviewed for each index partner and this strategy effectively reached men. 5690 (29%) of sex partners were HIV positive which is almost 5 times higher than the HIV prevalence in Uganda. Overall, 354 (<1%) reported experiencing post notification gender-based violence.
Dikjkstra (abstract 938, http://www.croiwebcasts.org/console/player/44634?mediaType=slideVideo&) reported on a combined intervention in coastal Kenya, in which TGW and MSM peer educators offered oral HIVST or an acute HIV symptom referral card. All had HIV RNA to detect acute HIV. If they tested HIV+, they were enrolled as index clients in study. Of 597 MSM and TGW, 445 (75%) accepted self-test kits and 188 (4%) accepted symptom referral. 454 HIV RNA tests were performed, of which 430 (95%) were HIV negative and 103 (25%) accepted PrEP. Ultimately, 7 (2%) were RNA positive and referred for ART and APS. A total of 28 HIV index clients were referred for APS, who referred 169 partners. Of partners who came for testing, 66 became through by provider referral and 17 by index client referral. 18 of the 28 index partners reported no social harm. Targeted peer mobilization and APS were feasible and safe. Peer educators understand their sexual networks and were able to reach partners.
Golden (abstract 939, http://www.croiwebcasts.org/console/player/44635?mediaType=slideVideo&) reported on APS scale-up in Botswana. A total of 52 clinics implemented APS between October 2018 and June 2019. In this model, counselors did not offer to directly notify partners, and key metrics of APS were monitored: reach, contact index, testing index, and case-finding index. Of 715 persons testing HIV+, 88% received APS (contact index of 0.85), of whom 66% partners tested (testing index 0.55) and a case finding index of 0.12. The HIV positivity rate was 22% in partners (similar to UK and lower than previous studies in Africa). The team concluded that they achieved high levels of reach, that case finding in Botswana was substantial lower than elsewhere in Africa maybe because the APS model didn't include direct notification of partners, suboptimal implementation, incomplete identification of partners at the time of testing, that a high proportion (90%) of persons with HIV have already been diagnosed, and/or selection bias in published studies.
Ebrahim (abstract 940, http://www.croiwebcasts.org/console/player/44636?mediaType=slideVideo&) optimizing HIV testing in 24 countries increases yield in HIV case finding, 2018-19. PEPFAR in 2019 recommended to reduce facility-based PITC through symptom and/or risk based screening, increasing HIV testing among exposed contacts of HIV+ index patients. Analyzed data from 2018-19 to assess the number of total tests (volume) and positivity rates which increased from 3.5 to 4.3%. The biggest change was in positivity rates through contact tracing from 7% to 19.3% whereas positivity rate through mobile testing and PITC testing remained stable. Decreased testing volume by 16% increased positive by 6% and increased yield by 12%. However, importantly, provider-initiated testing still accounted for 66% of all HIV+ tests. Symptom and/or risk based screening tools should be retained to increase yield of PITC tests.
Remera (abstract 941, http://www.croiwebcasts.org/console/player/44637?mediaType=slideVideo&) compared index testing approaches in Rwanda. Active case finding was implemented in all 23 health facilities in Kigali as pilot in 2018 using 3 methods: 1) client referral, 2) provider referral, 3) dual referral. Results found 2632 index cases (65% female) of whom 91% provided contact info, and 4116 partners were elicited for ratio of 1.7. Of these, 90% partners were invited, with 29013 partners identified of which 41% from client referral, 18% contract referral and 41% provider referral. HIV testing uptake was 85% and 9% were HIV infected, with 94% linkage to care.
Pre- and post-exposure prophylaxis (PrEP & PEP)
In PrEP, an HIV uninfected person uses an anti-HIV agent ahead of an HIV exposure in order to prevent infection. At each CROI since 2011, PrEP has been a major focal point, with major clinical trials, translational science, safety studies, and, most recently, implementation projects reporting findings. Large-scale efficacy clinical trials of PrEP - first as pills and then as an antiretroviral-containing vaginal ring - have demonstrated that PrEP is an effective and safe intervention for HIV prevention. Efficacy evaluations of injectable delivery of PrEP is ongoing, and additional delivery approaches (e.g., implants, less-frequent pills, etc.) are also being tested, all potentially heralding a future where multiple options exist for individuals to choose among for prevention. PEP has been part of HIV prevention for many years, for both occupational and sexual/injection exposures. Standard PEP is a 28-day course of combination antiretroviral medications; more simple options could be game-changing. A Tuesday morning oral abstract session addressed new options and opportunities for PrEP and PEP.
Three oral abstracts evaluated new potential PEP regimens in non-human primate (NHP) models. The first (Beckerman, abstract 87, http://www.croiwebcasts.org/console/player/44699?mediaType=slideVideo&) assessed a low-dose rectal SHIV challenge model, testing combination emtricitabine/tenofovir alafenamide (FTC/TAF, which is an approved PrEP medication for men when dosed daily), alone or with bictegravir (BIC), Standard doses of these medications were highly effective in this NHP model as PrEP (given 2 hours prior and 24 hours after exposure) but not particularly effective as PEP, protecting only 1/6 animals. When a higher dose of BIC was added (100 mg), however, protection was present - 82% (p=0.046) when given 12 and 36 hours after exposure, 80% (p=0.055) when given 24 and 48 hours after, and 67% (p=0.13) when given 48 and 72 hours after. In sum, FTC/TAF worked well as PrEP and FTC/TAF/BIC is promising for a simple PEP agent.
The second (Dobard, abstract 88, http://www.croiwebcasts.org/console/player/44700?mediaType=slideVideo&) evaluated a vaginal insert containing TAF and elvitegratir (20/16 mg)using a vaginal challenge model. TAF/EVG insert had previously been demonstrated to be effective as PrEP (CROI 2019); in the present abstract, protection was also 100% (6/6 animals protected) when provided 4 hours after an exposure. Pharmacokinetetic work showed that PBMCs had high concentrations of tenofovir diphosphate, a very interesting finding from topical delivery of a medication. The findings are very encouraging for on-demand TAF/EVG asa potential PrEP or PEP agent. Phase I human studies are ongoing for both vaginal and rectal use.
The third (Grubner, abstract 89, http://www.croiwebcasts.org/console/player/44701?mediaType=slideVideo&) tested islatravir (ISL), a novel nucleoside reverse transcriptase translocation inhibitor (NRTTI), in an intravenous NHP weekly challenge model. Dosing of ISL was oral, at a dose comparable to that being evaluated for HIV treatment by ISL. The experiment had a dose de-escalation approach, starting first with four weekly doses after a challenge, then three, two, and finally one, with stages separated by seven weeks (ISL is not detectable in this NHP species three weeks after last dose, making seven weeks a reasonable wash-out). The two, three, and four weekly PEP doses were all 100% protective, the single dose protected 4 of 6 animals. Pharmacologic modeling done by the team suggested that a single dose of ISL, if provided within 24 hours of an exposure, might provide high enough levels for protection in humans. A single-dose PEP strategy would of course be extremely exciting.
A vaginal film containing VRC01 and HSV8 monoclonal antibodies against HIV-1 and HSV-2 was tested in a safety study in women (Cu-Uvin, abstract 90, http://www.croiwebcasts.org/console/player/44702?mediaType=slideVideo&). The results were encouraging: safe, well-tolerated, easy dissolution of the film, no change in vaginal pH or Nugent score (a marker of bacterial vaginosis), no increase in inflammatory cytokines, and good concentrations of VRC01 and HSV8 through 24 hours after dosing, sufficient to neutralize virus in ex vivo evaluation.
On other fronts, real-time monitoring of adherence to daily oral FTC/TDF PrEP has been a topic of great attention. Urine testing has gotten some attention, given the ease in obtaining the relevant biologic sample, particularly if testing could be low-cost and easy. A lateral-flow assay (LFA) urine test - like a urine pregnancy test - was developed and tested (Spinelli, abstract 91, http://www.croiwebcasts.org/console/player/44703?mediaType=slideVideo&). Samples came from men and women from studies in Kenya, Uganda, and the US. The urine LFA was compared to urine tested by a standard ELISA. A total of 684 samples were tested. Sensitivity was 100% (505/505 samples that were positive for tenofovir in ELISA were positive by LFA), specificity was 98.3% (176/179 negative by ELISA negative by LFA), reflecting an overall 99.6% accuracy of the assay. This LFA may provide important assistance for adherence monitoring and counseling for individuals taking FTC/TDF (and arguably FTC/TAF) PrEP, as well as ART containing these antiretroviral agents. The LFA is being evaluated in field settings now for adherence promotion. A second abstract using this LFA was presented in a poster session (Stalter, abstract 2495), demonstrating a strong association between LFA detection and protection against HIV acquisition, using archived samples from a completed PrEP trial from Kenya and Uganda.
PrEP using daily oral FTC/TAF has received a lot of attention over the past year, following on the results of the DISCOVER trial presented at CROI 2019 and the subsequent FDA evaluation and approval for use in persons exposed rectally to HIV. An abstract (Ogbuagu, abstract 92, http://www.croiwebcasts.org/console/player/44704?mediaType=slideVideo&) presented the final, full 96 week follow-up for the DISCOVER trial. The study randomized 5399 men who have sex with men and transgender women to daily FTC/TAF or FTC/TDF PrEP. Retention on drug through 96 weeks was ∼81%. As reported last year at CROI, the profile of the population, in terms of sex without condoms and STIs, suggested high HIV risk. At week 48, presented at CROI 2019, there were 22 HIV infections recorded: 7 among those assigned FTC/TAF and 15 FTC/TDF. At week 96 there was only one additional infection (who was not taking PrEP at the time of infection), in the FTC/TAF group. The results at week 96 indicate an incidence rate ratio (IRR) of 0.54 (95% CI 0.23-1.26), demonstrating noninferiority of FTC/TAF compared to FTC/TDF as PrEP. Through week 96, FTC/TAF continued to be well-tolerated. Weight gain was observed in both arms, fasting glucose was comparable between both groups, FTC/TDF had greater decline in LDL and HDL. Thus, through the final 96 week follow-up in the randomized comparison of FTC/TAF to FTC/TAF in DISCOVER, non-inferiority remains demonstrated.
An abstract assessed acute HIV infection and HIV resistance risk, using data from national PrEP roll-out in Thailand (Colby, abstract 93, http://www.croiwebcasts.org/console/player/44705?mediaType=slideVideo&). The project assessed acute HIV infection in >2400 persons started on PrEP, and used pooled qualitative HIV RNA. A total of 7 (= 1 out of ∼350) had acute infection: 5 had a positive Aptima qualitative HIV test and 2 were negative. Resistance was detected in 3, all were M184V/I (FTC resistance), after PrEP use of 29-122 days, and none had TDF resistance. The assessment is that acute HIV infection among persons starting PrEP was uncommon and pooled HIV RNA testing was useful to identify cases of acute HIV. In addition, those who had less time on PrEP had less risk of antiviral resistance.
The interface of contraceptive use with HIV, both from a treatment and a prevention perspective, has been a prominent topic of discussion at several CROI meetings, and again was important at CROI 2020.
The potential for drug-drug interactions between hormonal contraceptives and antiretroviral medications has been a topic of global discussion for several years now, after epidemiologic and pharmacologic data found that concurrent use of efavirenz (EFV) and hormonal implants resulted in diminished implant effectiveness for pregnancy prevention. In the last couple of years, dolutegravir (DTG) has become global first-line therapy, but few data are available to know if DTG interacts with contraceptives. An abstract from Kenya presented data from an open-label pharmacokinetics study among women living with HIV and using a variety of contraceptives (Patel, abstract 129, http://www.croiwebcasts.org/console/player/44812?mediaType=slideVideo&). In Kenya, as in many countries in Africa with high burden of HIV in women, implant use is increasing. Women using hormonal implants - 23 living with HIV and using DTG and 25 HIV uninfected, as comparators - were followed for 24 weeks, with plasma etonorgestrel (ENG) concentrations measured every 2 weeks. Women using DTG had higher (not lower) ENG plasma levels throughout follow-up; safety and satisfaction were generally good. Thus, there was no evidence of a detrimental drug-drug interaction between ENG implants and DTG-containing ART. Higher ENG levels did not appear to be associated with safety concerns.
A poster discussion session on HIV susceptibility in the female genital tract occurred on Wednesday. An introductory talk (Mitchell, http://www.croiwebcasts.org/console/player/44870?mediaType=slideVideo&) led off and was great for framing the session. The first poster (Keller, abstract 1061, http://www.croiwebcasts.org/console/player/44872?mediaType=slideVideo&) did ex vivo work before and after treatment of bacterial vaginosis to assess enhancement of HIV infection, finding that treatment with metronidazole led to a reduction in HIV enhancing activity and correlated with a shift to a Lactobacillus-dominant microbiome. The second and third used samples and data from the ECHO Trial (Bunjun, abstract 1062, http://www.croiwebcasts.org/console/player/44873?mediaType=slideVideo& & Ramla, abstract 1063, http://www.croiwebcasts.org/console/player/44874?mediaType=slideVideo&), to assess cellular and inflammatory markers by randomized group. Cervical Th17 cells and Th17 cells expressing CD38 (indicating activated cells) were higher among women assigned injectable DMPA compared with the other two contraceptive methods, and cervical inflammatory cytokines were greater among women assigned the copper IUD. Of course, the ECHO Trial did not demonstrate substantial HIV acquisition differences across the three contraceptives tested, which is important to weigh when interpreting these results. Finally, a fourth abstract (Tobin, abstract 1064, http://www.croiwebcasts.org/console/player/44875?mediaType=slideVideo&) was a microbiologic analysis of a pharmacologic study of a combined estrogen/progestin vaginal ring among women living with HIV. Initiation of the ring led to a favorable shift in vaginal microbial communities for some women; removal of the ring was associated with a shift in vaginal microbial population to a more complex (i.e., unfavorable) one.
An excellent symposium on contraception and HIV was held on Tuesday, with four talks focused on drug-drug interactions between antiretrovirals and contraceptives (http://www.croiwebcasts.org/console/player/44771?mediaType=slideVideo&), challenges with rollout of the contraceptive implant in South Africa (http://www.croiwebcasts.org/console/player/44772?mediaType=slideVideo&), contraception and HIV risk (http://www.croiwebcasts.org/console/player/44773?mediaType=slideVideo&), and sexual and reproductive health services and rights globally (http://www.croiwebcasts.org/console/player/44774?mediaType=slideVideo&).
A special session was held to present the rationale for and results of the HVTN 702, a phase 2b study of the ALVAC and gp120 prime-boost regimen in South Africa (http://www.croiwebcasts.org/console/player/44777?mediaType=slideVideo& and http://www.croiwebcasts.org/console/player/44778?mediaType=slideVideo&). The rationale was that this regimen showed 31% efficacy in the RV144 Thai trial and was hoped to have higher efficacy if the ALVAC/gp 120 regimen was adapted to increase clade C coverage and increase antibody durability, which was done through the Pox Protein Public Private Partnership. A total of 5400 South African women and men from 14 sites were randomized in a 1:1 ratio to ALVAC (or placebo) at months 0 and 1 and ALVAC plus gp120 with the MF-59 adjuvant at months 3, 6, 9, 12 and 18 (or placebo). In the first stage of HVTN 702, the objective was to assess vaccine efficacy in the first 24 months. A total of 3786 women were enrolled, half of whom had more than one sex partner, 70% thought their partner had other partners, 94% reported inconsistent condom use., and 30% had a curable STI at baseline (primarily chlamydia). Of the men, 69% reported more than one partner, 16% transactional sex, 92% inconsistent condom use, and 18% had a curable STI (most commonly chlamydia). STI treatment was provided and PrEP was introduced during the trial, although only 2% of participants had detectable blood tenofovir levels suggesting very low uptake. The study had 6 DSMB meetings and the first interim efficacy analysis occurred in Jan 2020, which was triggered by sufficient 18 month follow-up data. Of the 267 incident HIV infections, there was no difference in HIV incidence (3.3/100 person-years) by study arm, a disappointing but very clearly non-efficacious result. These findings close the book on further prioritization of the RV144 vaccine regimen, after a decade and a half of research. The HVTN 702 team will investigate immunologic mechanisms for the lack of efficacy. Of note, HIV incidence was almost 4 times higher among women than men (4.2/100 p-yrs vs 1.2/100 p-yrs). The high HIV incidence among women is particularly alarming and highlights the need for biomedical HIV prevention interventions, including active provision of PrEP, for South African women in general and those participating in HIV prevention efficacy trials of products of unknown efficacy.
Epidemiology and HIV incidence
In the Tuesday themed discussion called "Coming into the Light: HIV in Transgender Women and Men," Kimani (abstract 878, http://www.croiwebcasts.org/console/player/44742?mediaType=slideVideo&) presented about PrEP uptake in 42 Kenyan MSM and 11 TGW. At month 12, 23% of participants were lost to follow-up. Most MSM reported insertive anal sex and TGW reported both insertive and receptive anal sex. PrEP adherence was assessed at month 6 by significantly more commonly used among MSM and TFSW. Any tenofovir diphosphate (TFV-DP) was detected in 63% of 8 TGW and 15% of 33 MSM with protective levels (TFV-DP levels associated with >4 doses/week on average) were detected in 43% of TWW and 0% of MSM. In-depth interviews indicated that PrEP reinforced female gender identity, felt at risk because of being trans which motivated their adherence in TGW, which suggested that in coastal Kenya, TGW had higher agency to take PrEP than MSM.
Hontelez (abstract 879, http://www.croiwebcasts.org/console/player/44743?mediaType=slideVideo&) presented on HIV prevalence and risk among 603 male and transgender sex workers (TMSW and TGSW) in Zimbabwe, which showed HIV prevalence of 28% in MSM, 38% in TFSW and 38% in TMSW. High rates of female clients were reported by all groups. Keruly (abstract 880) presented data on HIV, HCV, and HSV-2 status from the US Leading Innovation for Transgender Women's Health and Empowerment (LITE) study from 6 eastern and southern US cities. Of 562 individuals, seroprevalence of HIV was 29%, HSV-2 48% and HCV 5%, respectively; 44% were seronegative for all 3 infections with significant differences in prevalence by geography, race and ethnicity, family support and age.
Radix (abstract 881, http://www.croiwebcasts.org/console/player/44745?mediaType=slideVideo&) presented on 577 trans-men from the Callen Lorde clinic in New York, with 79% having received testosterone, 42% one or more gender-affirming surgeries, and only 42% had undergone HIV screening whom HIV prevalence was 2.9%.
In the Wednesday oral abstract session called "The good, the bad, and the ugly: HIV incidence and outcomes in selected populations, Solomon (abstract 147, http://www.croiwebcasts.org/console/player/44836?mediaType=slideVideo&) reported on an explosive epidemic of HIV and hepatitis C among 2512 PWID in New Delhi, India, who were recruited using a chain referral approach by naming persons they injected with in the past month. At baseline, 37% had HIV infection and only 7% were virally suppressed. HCV prevalence was 65%, heroin use was 27% and opioid use was 95%. Among the 1066 with at least one follow-up, HIV incidence was 26/100 person-years and HCV incidence was 34/100 person-yrs. Notably, 74% of incident HIV infections were associated with at least one viremic person in their egocentric network. These findings are alarming and highlight the importance of HIV and HCV testing, ART and harm reduction, and HCV treatment to curb what can be explosive micro-epidemics among PWID. Of course, high transmission of HIV among PWID has been known for decades, but the rapid spread seen here, the location of this epidemic (reflecting expansion of PWID epidemics to low- and middle-income countries), and the fact that we have very good strategies (ART, clean needles and works, HCV treatment) that can eliminate HIV and HCV transmission remind us strongly we still have far to go (see Strathdee plenary mentioned above for more on this topic).
Ussery (abstract 149, http://www.croiwebcasts.org/console/player/44838?mediaType=slideVideo&) analyzed HIV incidence in the Botswana Combination Prevention Project (YaTse) and characterized HIV incidence among repeat HIV testers. Of the almost 48,000 participants who tested HIV negative at baseline, 39% retested a median of 358 days later. A total of 195 become HIV infected, 79% of whom were women. The HIV incidence rate was 1.01 percent per year among women and 0.34 among men for a 3- fold higher HIV incidence among women. The highest HIV incidence was 1.87 among women ages 16-24, which was 7- fold higher than men ages 16-24. The authors hypothesize that potential sources of infection among young women is from men who are not virally suppressed either because they were missed by the YaTse testing and treatment intervention (e.g., because they did not know their status, were not on ART or not virally suppressed) or they were from outside the intervention communities (e.g., due to mobility). Two analyses presented support these possibilities. First, viral suppression was 72% among men vs 88% among women in the intervention communities. Second, phylogenetic mapping indicated highly diversified transmission networks with 72% lineages from 2 or more communities suggesting mobility and broad mixing between communities.
Grabowski (abstract 150, http://www.croiwebcasts.org/console/player/44839?mediaType=slideVideo&) presented data on the impact of combination HIV prevention on rapidly declining HIV incidence among men and women in the Rakai Community Cohort in Uganda, comparing the time periods before combination prevention scale up from 2006 to present to 1999-2004. Analyses were stratified by sex and age group (< or > 25 years). ART coverage increased steeply with the roll out of universal test and treat to 70% (82% among women and 73% among men) with viral suppression among 81% overall, exceeding the UNAIDS 90:90:90 targets (85% among women and 71% among men). Medical male circumcision increased from 15% to 65%. Although age at sexual debut increased over time, the proportion reporting multiple partnerships or consistent condom use did not change. HIV incidence was 1.2/100 person-years prior to and decreased to 0.6 by 2016 after combination HIV prevention scale up and further declined to 0.42 in Sept 2017. Notably, in age and gender-stratified analyses, there were no incident HIV infections among women ages 16-24 and incidence was 0.36 among women ages 20-24 in the last survey in 2017. HIV incidence was 0.43 among uncircumcised men and 0.31 among circumcised men. These encouraging reductions in HIV incidence in men and women, particularly young women, are associated with the roll-out of ART with viral suppression rates exceeding UNAIDS targets, VMMC scale up and possibly delayed sexual debut, and demonstrate the impact of and need for sustained investment to achieve population level impact. A number of excellent posters highlighted changing HIV incidence and analyzed factors associated with HIV incidence in different settings.
Several posters assessed important epidemiologic studies. Camlin (abstract 849) evaluated mobility metrics as a risk factor for HIV acquisition in rural Kenya and Uganda in the SEARCH trial, in which 704 incident HIV infections were observed. Several mobility metrics were significantly associated with a HIV acquisition, including mobility prior to enrollment, having lived outside the community for > 1 month in the prior year, and spending some nights away in the prior month for men only. Risher (abstract 851, poster) evaluated age-specific HIV incidence patterns among six population cohorts in Tanzania, Uganda, Malawi, South Africa and Zimbabwe, using a Bayesian model fit separately for sex in each site. From 2000-17, HIV incidence declined overall with the magnitude and timing varying by site. by age of infection was consistently higher for men compared to women. Notably, among adult women, 35-65% of new infections occurred in adolescent girls and young women ages 15-24. Lewis (abstract 856) analyzed factors associated with HIV seroconversion among young women ages 15-24 in KwaZulu-Natal, South Africa, using the prospective cohort data from the HIV Incidence Provincial Surveillance System. HIV incidence was 3.92/100 person-years. Among teenage girls ages 15-19, HIV acquisition was significantly associated with being an orphan (aHR=4.4), having a baseline STI (aHR 2.4), and negatively associated with having family support (aHR=0.5) and having a circumcised partner (aHR=0.6). For women aged 20-24 years, failure to complete high-school (aHR=1.8) and inconsistent condom use (aHR=2.7) were associated with HIV acquisition. These findings indicate that structural as well as behavioral and biologic factors are associated with increased risk of HIV acquisition among African AGYW, and that programs need to emphasize education, sexual health, male circumcision, condoms and frequency of HIV testing. Mwinnyaa (abstract 860) analyzed factors associated with HIV infection, HIV diagnosis, and lack of ART among men attending 3 emergency departments in the Eastern Cape of South Africa. 21% of 1458 men tested HIV+, of whom 41% were unaware of their HIV status. HIV prevalence increased with age and was associated with generalized weakness and signs of tuberculosis. ART use was <30% in men under age 35, none of men ages <25 were virally suppressed, and alcohol use was associated with lack of viral suppression. EDs in high prevalence settings in Africa are an important venue for identifying HIV+ men.