4th HIV Research for Prevention conference (HIVR4P // Virtual), 27 & 28 January | 3 & 4 February 2021 (2025)

Evaluation of the kinetics of systemic distribution of IV injected monoclonal antibodies modified to alter host mediated Fc interaction in the rhesus macaque model

A.M. Carias¹, J. Schneider², M. McRaven¹, S. Xiao¹, K. Rogers³, M. Araínga³, R. Veazey⁴, F. Villinger³, T.J Hope¹

¹Northwestern University, Cell and Developmental Biology, Evanston, United States, ²Rush University, United States, ³New Iberia Research Center, United States, ⁴Tulane National Primate Research Center, United States

Background: Antibody‐mediated protection against HIV/SHIV transmission has been illustrated in non‐human primates (NHP) with IV injection of broadly neutralizing monoclonal antibodies (bNAbs). To gain insights into the kinetics and dynamics of BNAb distribution and localization, we utilized passively transferred fluorophore‐labeled VRC01, developing a platform in the living NHP model, without affecting antibody function. After injection, we followed antibody distribution over time, providing a unique perspective to observe how bNAbs reach different anatomical sites. Using this platform, with anti‐FcRn and mutated Fc‐function bNAbs, we are now unraveling the mechanisms of how FcR specificity affects antibody distribution.

Methods: Macaques were IV‐administered fluorescently tagged, anti‐FcRn, or a combination of VRC07‐523‐LS and/or Fc‐mutated VRC07‐523‐LS‐LALA (LALA) (mutated to disrupt FcR binding and complement) and/or VRC07‐523‐LS‐DEL (DEL) (mutated to enhance FcgRIII antibody binding and ADCC activity). Animals were necropsied between 6hours and 1week, tissues collected and imaged with deconvolution microscopy.

Results: By analyzing those animals that received fluorescently tagged anti‐FcRn, we are now able to discern the cellular locations of FcRn: on endothelial cells and select monocytes. Additionally, in columnar and brain tissues, we are able to visualize antibody distribution through the vascular system, with antibody‐FcRn interactions. Lastly, although studies are still ongoing, we visualize differences in VRC07‐523‐LS, VRC07‐523‐LS‐LALA and VRC07‐523‐LS‐DEL distribution amongst different tissues, across multiple time points.

Conclusions: These data build on our previous data looking at the distribution and localization of fluorescently labeled anti‐HIV bNAbs. However, and importantly, these studies provide novel insights into Fc receptor‐associated mechanisms of antibody delivery to different organs and tissues after IV injection into multiple animals. These experiments will provide critical insights into the mechanism(s) of distribution and localization of systemic antibodies by following the time course of distribution of passively transferred antibodies in a gain or loss of function experimental fashion.

Monoclonal antibodies with ultra‐long CDRH3 derived from vaccinated cows show exceptional neutralisation potency and breadth that is retained after Fc engineering

B. Heydarchi¹, D. Fong¹, A. Sethi², J. Edwards¹, N. Salazar Quiroz¹, T. Aktepe¹, C. Gonelli¹, S. Grimley¹, C. Mackenzie¹, B. Wales³, M. van Gils⁴, R. Sanders⁴, P. Gooley², I. Rouiller², D. Purcell¹

¹University of Melbourne, Doherty Institute, Melbourne, Australia, ²University of Melbourne, Bio21 Molecular Science and Biotechnology Institute, Melbourne, Australia, ³Dairy Production Sciences, Victorian Department of Economic Development, Jobs, Transport and Resources, Australia, ⁴Academic Medical Center, University of Amsterdam, Department of Medical Microbiology, Amsterdam, Netherlands

Background: A vaccine against HIV would ideally elicit broadly neutralising antibodies (BrNAbs). In HIV+ patients these immunoglobulins (Ig) are rare and often display long third heavy complementarity determining regions (CDRH3) with high levels of somatic hypermutation. The average length of bovine Ig‐CDRH3 is notably longer than other species and has potential for raising novel BrNAbs. We previously showed that cows vaccinated with HIV envelope (Env) produced BrNAbs that target the CD4 binding site (CD4bs). Here we used a stabilized trimer gp140 SOSIPv4.1 from the neutralisation resistant ADA subtype‐B strain to select a lineage of chimeric human monoclonal BrNAbs containing bovine V‐regions with ultra‐long CDRH3.

Methods: To produce monoclonal BrNAbs (mAbs), bovine memory B‐cells were isolated from cows vaccinated with AD8 gp140 (clade B), AD8 SOSIP gp140v4.1 or KNH1 SOSIP gp140 and BG505 SOSIP.664 gp140. The IgG+, SOSIP gp140 v4.1 AD8+ cells were FACS‐sorted and variable heavy (VH) and light (VL) genes were amplified and cloned into human constant region expression vectors. The mAbs were expressed and characterised for Env binding and neutralisation activity.

Results: Among many Env binding mAbs, fourteen showed heterologous neutralisation with CDRH3 length of 12 to 61 residues. Six mAbs with shorter‐length CDRH3 showed modest neutralisation against tier 1 clade B viruses. Another six mAbs with ultra‐long CDRH3 could neutralise tier 1 and tier 2 of clade B and clade C as well as tier 2 of clade A HIV virus. But three common‐lineage mAbs isolated from a KNH1/BG505 SOSIP vaccinated cow were highly potent (IC50 ~0.001µg/ml) in neutralising the global panel of twelve HIV pseuodoviruses in the TZM‐bl assay. Binding and neutralisation of Env mutant pseudoviruses confirmed these three potent bovine mAbs bound the CD4 binding site and required contact residues in C3, C4 and C5 of gp120. These potent mAbs strongly out‐competed VRC01 and other elite human BrNAbs for binding the CD4bs, yet were not polyreactive with human antigens. Antigen binding activity of these bovine V‐region chimeric mAbs was fully retained after re‐engineering cellular‐immune functionality into the human IgG Fc region.

Conclusions: BrNAbs with novel structured ultra‐long CDRH3 from Env vaccinated cows give new insights into conserved gp120‐CD4bs epitopes.

FcɣRIIa rs10800309 polymorphism is an HLA‐B57 and HLA‐B27 independent predictor of HIV control

C. Moog¹, R. Carapito², L. Mayr², A. Molitor², M. Verniquet², O. Tahar², P. Marialuisa³, D. Rey³, O. Lambotte⁴, S. Schmidt², S. Bahram²

¹U1109, Institute of Virology, France, ²U1109, France, ³NHC Strasbourg, France, ⁴U1018, France

Background: Fcɣ receptors (FcɣRs) are key immune regulatory receptors that connect antibody mediated immune responses to cellular effector functions. They are involved in the control of various immune functions including responses to infections. Genetic polymorphisms of FcɣRs coding genes (FCGR) have been associated with the regulation of HIV infection and progression.

Methods: In this study we analyzed the potential impact of five candidate FcɣR SNPs on viral control by genotyping 251 HIV controllers and 250 progressors.

Results: The rs10800309 AA genotype of the FcɣRIIa coding gene FCGR2A was found to be significantly associated with HIV control and this association was independent of HLA‐B57 and HLA‐B27 (OR, 2.84; 95% CI, 1.20 to 6.89; Pcor=0.033). We further confirmed the functional role of this polymorphism by showing an association of this same AA genotype with an increased in vitro FcɣRII expression on myeloid cells including dendritic cells (p=0.0032).

Conclusions: Together, these results suggest that the AA genotype of rs10800309 confers an improved immune response through FcɣRII upregulation in favor of a role of Fc mediated inhibition in HIV control. Moreover, this polymorphism may serve as an additional predictive marker of HIV control.

Serum IgA inhibits HIV‐specific broadly neutralising antibody Fc functions

S. Davis, A. Chung, S. Kent

The Peter Doherty Institute for Infection and Immunity, Department of Microbiology and Immunology, Australia

Background: The importance of antibody Fc functions were highlighted in the human HIV RV144 vaccine trial, however, serum IgA reduced vaccine efficacy and Fc functions. Elucidating how serum IgA modulates Fc responses is essential. Furthermore, macaque HIV‐broadly neutralizing antibodies (BnAbs) passive transfer studies suggest a vital role of Fc functions in protection. Here we endeavour to determine if serum IgA influences the Fc capacity of IgG from HIV individuals or BnAbs.

Methods: Pooled purified IgG from HIV individuals (HIVIG) along with a panel of BnAbs including PGT121 and VRC01, currently in human clinical trials, were assessed for their Fc functional capacity. The influence of IgA upon IgG was assessed by adding pooled HIV‐specific IgA (n=10), pooled HIV‐negative IgA (n=10), IgA1 or IgA2 and colostrum IgA.

Results: HIV‐specific IgA showed minor inhibition of phagocytosis (median=10.38%, IQR=8.09%, p>0.05). Intriguingly, significant inhibition was observed when HIV‐negative IgA was added (median=21.24%, IQR=14.28%, p<0.001). Similarly, significant inhibition was observed with IgA1 (median=23.11%, IQR=18.18%, p<0.001), IgA2 (median=19.88%, IQR=4.60%, p<0.001) and colostrum IgA (median=23.31%, IQR=5.76%, p<0.001) when added to HIVIG and BnAbs. Addition of FcαR block to these assays was capable of reconstituting Fc functions, suggesting that IgA inhibition is mediated through IgA‐FcαR binding.

Conclusions: HIV‐negative serum IgA, and to a lesser extent HIV‐positive IgA, reduced the functional capacity of HIVIG and BnAbs, suggesting IgA may inhibit through two mechanisms: epitope competition and IgA‐FcαR mediated inhibitory mechanisms. Understanding the mechanisms behind why IgA inhibits Fc responses could lead to improved future HIV vaccine design and educate passive transfer monoclonal antibody therapies.

Opening the HIV‐1 envelope for neutralization and antibody‐dependent cellular cytotoxicity

D.N. Nguyen¹, J. Richard², W.D. Tolbert¹, R. Gasser², S. Ding², D. Vézina², S.Y. Gong², G. Gendron-Lepage², H. Medjahed², S. Gottumukkala², J. Prévost², A. Finzi², M. Pazgier¹

¹Uniformed Services University of the Health Sciences, Medicine, Bethesda, United States, ²Centre de Recherche du Centre Hospitalier de l'Université de Montréal (CRCHUM), Montréal, Canada

Background: The apparent successes of cancer cure strategies that are based on therapeutic antibodies have caused a re‐evaluation of the potential for similar strategies to achieve a functional HIV cure. However, most of the currently explored antibody‐based therapies and eradication strategies are based on broadly neutralizing antibodies. By contrast, strategies for a functional cure using non‐neutralizing Abs (nnAbs) known to eliminate HIV‐infected cells through Fc receptor (FcR) effector functions including antibody‐dependent cell mediated cytotoxicity (ADCC) remain a significant yet largely unexploited avenue of research.

Highly conserved CD4 inducible (CD4i) epitopes within the constant region 1 and 2 (C1C2 or Cluster A) and the co‐receptor binding site (CoRBS) have been shown to be suitable targets for Abs capable of inducing ADCC against infected cells expressing Env in an “open” conformation. As strictly CD4 dependent, these potent ADCC targets become available for Ab recognition upon triggering of Env trimer with cell surface CD4 therefore are occluded and not accessible for Ab targeting on infected cells where expression of CD4 is downregulated.

Methods: Here we present a structure‐based design of new CD4‐antibody fusion molecules and evaluate their abilities to sensitize HIV‐1‐infected cells to ADCC‐mediated killing. We developed two classes of Ab fusion proteins in which domains 1 and 2 of soluble human CD4 are linked with either a C1C2 (class 1) or a CoRBS (class 2) specific monoclonal antibody through a flexible (G4S)6(G4T)2‐linker. We found optimal conjugation sites and linker lengths that allows each of these novel bispecific fusion molecules to recognize the native “closed” Env trimers and to initiate the structural rearrangements required for exposure of these conserved Env epitopes.

Results: Our in vitro functional testing shows that these fusion proteins efficiently target and eliminate HIV infected cells through ADCC activity. Surprisingly, attaching sCD4 to these otherwise non‐neutralizing parent cluster A and CoRBS specific Abs has rendered them potent neutralizing activity against tier 1 and 2 viruses. Competition experiments reveal that these biological activities rely on both Ab and CD4 moieties interaction with Env.

Conclusions: Altogether, our results raise the exciting possibility of using these fusion proteins in the ongoing efforts to achieve a functional cure.

Profiling the HIV epidemic with recency of infection instead of recency of diagnosis: 2years of experience in North Carolina, USA

S. Zhou¹, S. Sizemore², M. Moesers¹, S. Zimmerman³, E. Samoff³, V. Mobley³, S. Frost⁴, A. Cressman⁵, J. Eron², M. Clark¹, C. Jones², M. Cohen², J. Nelson¹, R. Swanstrom¹, A. Dennis⁶

¹University of North Carolina, Lineberger Comprehensive Cancer Center, Chapel Hill, United States, ²University of North Carolina, Chapel Hill, United States, ³North Carolina Department of Health and Human Services, United States, ⁴Microsoft Research, United States, ⁵University of North Carolina, Globl Health and Infect Disease, Chapel Hill, United States, ⁶University of North Carolina, Infectious Diseases, Chapel Hill, United States

Background: The identification of recent (incident) HIV infections among people who are initially diagnosed for the first time is critical to HIV prevention. We developed a Multiplexed Primer ID‐Next Gen Sequencing (MPID‐NGS) approach to identify recent infection by measuring the intra‐host viral diversity over multiple regions of the HIV genome. We examined the field implementations of MPID‐NGS to identify recent infection and drug resistance mutations (DRMs) among persons with new HIV diagnoses reported by the North Carolina State Laboratory of Public Health in 2018 and 2019.

Methods: The MPID‐NGS libraries were constructed for protease (PR), partial reverse transcriptase (RT), integrase (IN), and the V1 to V3 region of the env gene using remnant serological diagnostic tests. New diagnoses were restricted to sera collected within 30days of HIV diagnosis dates. The MiSeq platform was used for sequencing. The TCS‐DR pipeline was used for bioinformatics analysis and to identify DRMs. Recent infection was defined as infection <9months old, and the RT and V1/V3 regions were used to assess recency. We examined factors associated with recency using surveillance data.

Results: A total of 515 persons with new diagnoses from 2018 to 2019 were successfully sequenced. Overall, 202 (39%) had recent infection, 241 (47%) were chronically infected, and 72 (14%) had indeterminate recency at time of diagnosis. We detected a greater percentage of recent infections in 2019 than 2018 (44% vs. 35%, p=0.06). By comparing the characteristics of recent versus chronic infection at diagnosis, we found that young people were more likely to be diagnosed in the recent infection stage (p<0.01), and individuals diagnosed with recent infection were more likely to be in active transmission clusters than those with chronic infection (p<0.01). K103N was the most commonly seen DRM (approximately 10%) while other clinically significant DRMs were rarely seen.

Conclusions: We demonstrate an all‐in‐one platform to monitor HIV‐1 recency, DRMs, and phylogenetically linked transmission clusters in near real‐time. We believe this approach has the potential to be a useful tool as part of public health efforts to reduce new infections by monitoring the percentage of recency among new HIV diagnoses and providing opportunity to interrupt transmission within clusters.

Correlates of high HIV transmission areas in a generalized hyperendemic setting: findings from a national survey in Eswatini

N.M. Philip¹, G.S. Lovasi², R. Nkambule³, Q. Abdool Karim⁴, J. Justman¹, B. Mathema⁵

¹ICAP at Columbia University, Mailman School of Public Health, New York, United States, ²Drexel University, Dornsife School of Public Health, United States, ³Ministry of Health, Department of Health Services ‐ Public Health, Eswatini, ⁴Centre for AIDS Programme of Research in South Africa (CAPRISA), Durban, South Africa, ⁵Columbia University, Mailman School of Public Health, New York, United States

Background: Small areas with high uncontrolled HIV infection promote ongoing transmission and micro‐epidemics. Whether they result from geographic clustering of multiple HIV risk factors is unknown but key to designing place‐based, combination interventions for epidemic control.

Methods: A nationally representative, household‐based sample of adults, ages 18 to 49years, was enrolled from December 2010 to June 2011 from 575 enumeration areas in Eswatini. Consenting adults completed an interview and rapid HIV testing. Viral load was quantified using the COBAS AmpliPrep/Taqman HIV‐1 Test, v 2.0. Multi‐level latent class modeling identified statistically significant combinations of biomedical and behavioral risk factors and classified the combinations into small (enumeration) area risk profiles, categorized by HIV risk types. Linear regression assessed the correlation between area profiles and area prevalence of detectable viremia (≥20 copies/milliliter) among all adults regardless of HIV status.

Results: 18,172 surveyed adults were categorized into one of six HIV risk types, each showing a unique pattern of five risk factors that conveyed HIV transmission and/or acquisition risk propensity. The three most frequent composite prevalences of HIV risk types were categorized into area profiles: low‐moderate acquisition (low), moderate acquisition/transmission (moderate), and high acquisition/transmission (high). Detectable viremia prevalence increased from low [17.7%], moderate [25.4%], and high [35.1%] profiles and was 7.4% [p<.001] and 17.1% [p<.001] higher in moderate and high profile areas, respectively, when compared with low profile areas. High profile areas comprised the largest proportions of the highest transmission/acquisition risk types. The two highest risk types, high acquisition and very high transmission risk adults, were seronegative and undiagnosed seropositive, respectively, with the greatest likelihood of no prior HIV testing, multiple partnerships, and partners with unknown status.

Conclusions: Area HIV risk profiles can explain variation in area HIV viral measures. Co‐location of higher transmission and acquisition risk types in small areas may enable uncontrolled HIV viremia and geo‐located sources of transmission.

Progress toward HIV elimination goals: trends in and projections of treatment as prevention strategy in 38 African countries

P. Nguyen¹, S. Gilmour¹, P. Le¹, K. Onishi¹, K. Kato², H. Nguyen¹

¹St. Luke's International University, Graduate School of Public Health, Japan, ²Kanto Rosai Hospital, Department of Obstetrics and Gynaecology, Japan

Background: We aimed to estimate the trends and projections of indicators of Treatment‐as‐Prevention (TasP), the key global strategy for HIV elimination in Africa, and to calculate the probability of reaching key UNAIDS targets.

Methods: We included 1,456,224 sexually active adults age 15 to 49 in 38 African countries from 112 nationally representative population‐based surveys 2003 to 2018. Bayesian mixed‐effect models were applied to estimate the prevalence of annual HIV testing and condom use for every country and year to 2030, and the probability of reaching UNAIDS targets in 2020 and 2030.

Results: Most countries have upward trends in TasP outcomes, but seven saw downward trends in annual HIV testing and five saw decreases in condom use (Figure 1). The highest coverage of annual HIV testing in 2030 is predicted in Swaziland with 92.6% (95% CrI: 74.5%‐98.1%), Uganda with 90.5% (72.2%‐97.2%) and Lesotho with 90.5% (69.4%‐97.6%). Meanwhile, Swaziland, Lesotho, and Namibia will have the highest proportion of condom use in 2030 at 85.0% (57.8%‐96.1%), 75.6% (42.3%‐93.6%), and 75.5% (42.4%‐93.2%). The probabilities of reaching targets range from 0% to 28.5% for HIV testing and 0% to 12.1% for condom use and no country showed a high probability (>50%) of meeting either target (Table 1).

Conclusions: We observed limited progress on TasP in Africa and little prospect of reaching global targets for HIV/AIDS elimination. Varying outcomes in countries at the same development level suggest inadequate resource allocation and low effectiveness of HIV/AIDS programs. Although some funding agencies are considering withdrawal from supporting Africa, more fattention to funding and expanding testing and treatment are needed in this region.

HIV prevalence and incidence among FSWs participating in a HIV vaccine preparedness study in Dar es Salaam,Tanzania

D. Faini¹, F. Msafiri², P. Munseri³, M. Bakari⁴, A. Joachim², T. Nagu³, E. Tarimo⁵, E. Lyamuya², E. Sandström⁶, G. Biberfeld⁶, C. Nilsson⁶, C. Hanson⁶, S. Aboud⁶

¹Muhimbili University of Health and Allied Sciences, Epidemiology and Biostatistics, Tanzania, United Republic of, ²Muhimbili University of Health and Allied Sciences, Microbiology and Immunology, Tanzania, United Republic of, ³Muhimbili University of Health and Allied Sciences, Internal Medicine, Tanzania, United Republic of, ⁴Ministry of Health, Community Development, Gender, Elderly, and Children, Tanzania, United Republic of, ⁵Muhimbili University of Health and Allied Sciences, Tanzania, United Republic of, ⁶Karolinska Institute, Stockholm, Sweden

Background: PrEPVacc is a phase IIb multicenter HIV vaccine trial planned to be conducted at five sites in four sub‐Saharan African countries that aims to evaluate safety and efficacy of two combinations of HIV vaccine regimens; HIV DNA + gp120/alum and HIV DNA, MVA + gp140/MPLA. A PrEPVacc HIV‐negative registration cohort study was established to determine HIV prevalence and incidence among female sex workers (FSW) in Dar es Salaam, Tanzania.

Methods: Between October and December 2018, a total of 773 FSW aged 18–45years were screened for eligibility and 700 were enrolled. At screening and at 3‐monthly follow‐up visits, demographics and risky behavioural assessment and collection of blood samples were done. HIV testing was performed using two sequential rapid diagnostic tests; SD Bioline HIV1/2 and Uni‐Gold HIV‐1/2. HIV reactive samples were confirmed using Siemens Enzygnost HIV Integral 4 ELISA. Logistic regression was used to estimate odds Ratios for factors associated with HIV prevalence. Time‐to‐event analysis was performed using Poisson regression to estimate HIV incidence. Women were censored at 12months of follow‐up or earliest date of HIV seroconversion. Date of seroconversion was assumed to be midway between last negative and first positive HIV test results.

Results: HIV prevalence at screening was 8% (59/773), associated with older age (p<0.001), lower education levels (p<0.001) and being single (either never married or separated/divorced/widowed) (p<0.001). FSWs who reported being raped or having used drugs, were more likely be HIV‐infected than their counterparts (p=0.01 and p=0.002, respectively). Attendance at 12months was 80% (560/700) with women in the cohort contributing a total of 609 person‐years‐at risk (PYR). Twenty‐one FSWs seroconverted with the HIV incidence rate in the cohort of 3.45 per 100 PYRS (95% CI; 2.25 to 5.28/100 PYRS).

Conclusions: HIV prevalence and incidence were high among FSW in Dar es Salaam. These findings demonstrate feasibility of recruiting FSW for HIV vaccine prevention trials.

Abstract OA02.04‐Table 1.

Risk factors for HIV transmission in heterosexual men, men who have sex with men, and transgender women participating in the HVTN 702 “Uhambo” and HVTN 503/503‐S “Phambili” HIV vaccine trials

M. Malahleha¹, H. Janes², F. Laher³, L.‐G. Bekker⁴, B. S. Prigmore⁵, D. Grove⁵, J.J. Kee⁵, M. Allen⁶, M. Andrasik², M. Atujuna⁴, N. Singh⁷, D. Kalonji⁸, G. Meintjes⁹, P. Kotze¹⁰, N. Grunenberg², Y. Huang², Z. Moodie², J. A. Odhiambo¹¹, P. Smith⁴, G. Gray¹¹

¹Setshaba Research Centre, Clinical Research, Pretoria, South Africa, ²Fred Hutchinson Cancer Research Center, Vaccine and Infectious Disease Division, Seattle, United States, ³Perinatal HIV Research Unit (PHRU), Vaccine Research, Soweto, South Africa, ⁴The Desmond Tutu HIV Centre, University of Cape Town, Institute for Infectious Disease and Molecular Medicine, Cape Town, South Africa, ⁵Fred Hutchinson Cancer Research Center, Statistical Center for HIV/AIDS Research & Prevention, Seattle, United States, ⁶National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health (NIH), Division of Acquired Immunodeficiency Syndrome (DAIDS), Rockville, South Africa, ⁷South African Medical Research Council (SAMRC), HIV Prevention Research Unit (HPRU), Verulam Clinical Research Site, Durban, South Africa, ⁸South African Medical Research Council (SAMRC), HIV Prevention Research Unit (HPRU), Isipingo Clinical Research Site, Durban, South Africa, ⁹University of Cape Town, Faculty of Health Sciences, Clinical Platform of the Wellcome Centre for Infectious Diseases Research in Africa (CIDRI‐Africa), Cape Town, South Africa, ¹⁰Qhakaza Mbokodo Research Clinic, Clinical Research, Ladysmith, South Africa, ¹¹South African Medical Research Council, Parow Valley, South Africa

Background: South Africa has the highest HIV incidence globally. HIV risk has been extensively studied in South African cisgender women; however, less is known about risk drivers in cisgender men and transgender women (TGW). We characterised HIV incidence, and sexual behaviours and clinical characteristics associated with HIV acquisition, amongst cisgender men and TGW in two South African HIV vaccine efficacy trials.

Methods: We included data from heterosexual cisgender men (HCM), men who have sex with men (MSM) and TGW who participated in Uhambo (N=1611 randomised to vaccine/placebo, 2016 to 2020) and Phambili (N=219 randomised to placebo, 2007 to 2011; vaccinees excluded due to potential vaccine‐increased risk). Cox proportional hazards models were used to associate baseline variables—self‐reported last thirty‐days (Uhambo) or six‐months (Phambili) sexual behaviours and laboratory‐confirmed STIs—with HIV acquisition.

Results: Median age was 25 (IQR: 22 to 30). Most identified as heterosexual or reported no male partner (1636/1830 [89.40%]). More MSM and TGW versus HCM reported anal sex (90.21% vs. 5.01%), transactional sex (41.75% vs. 11.37%), ≥2 partners (82.99% vs. 65.28%), sex with alcohol/drugs (67.53% vs. 55.50%) or sex with an HIV‐positive partner (71.65% vs. 44.74%). Overall HIV incidence was 1.32% annually (95% CI: 0.99 to 1.74) and higher in MSM and TGW (8.84%, 95% CI: 5.40 to 13.65) or those with an STI (2.99%, 95% CI: 1.63 to 5.01). Based on multiplicity‐adjusted univariate analyses, anal sex (HR 6.34, 95% CI: 3.45 to 11.66), transactional sex (HR 3.42, 95% CI: 1.80 to 6.51), and MSM or homosexual identity (HR 15.62, 95% CI: 7.82 to 31.81) were significantly associated with HIV acquisition. In a multivariate model of published HIV risk factors, only MSM and homosexual identity (HR 12.90, 95% CI: 4.03 to 41.29; p<0.001) was significantly associated with HIV acquisition.

Conclusions: Identifying as MSM or homosexual is a strong predictor of HIV acquisition in South African men and TGW. While overall incidence in cisgender men and TGW is low relative to women, subpopulations of cisgender men and TGW experience exceptionally high incidence.

VRC01 antibody prevention of HIV

L. Corey, P.B. Gilbert, N.M. Mgodi, S. Edupuganti, M.S. Cohen

Fred Hutchinson Cancer Research Center, Seattle, United States

Background: We designed proof‐of‐concept trials to determine if a broadly neutralizing HIV antibody (bnAb) directed at the CD4 binding site of HIV‐1 (VRC01) was capable of preventing HIV acquisition.

Methods: We enrolled men and transgender persons who have sex with men (MSM/TG) in the Americas (HVTN 704/HPTN 085) and women in sub‐Saharan Africa (703/081) into two parallel trials. Enrollees received 10 intravenous infusions at eight‐week intervals of VRC01 10mg/kg, VRC01 30mg/kg, or placebo. Participants were assessed for HIV acquisition at four‐week intervals. In vitro neutralization sensitivity of VRC01 for each acquired isolate (IC80) was measured by the TZM‐bl assay.

Results: VRC01 reduced acquisition of HIV isolates with in vitro sensitivity to the antibody of IC80 <1µg/mL. Against this group of highly sensitive viruses, receipt of VRC01 achieved 75% protection over the 20‐month study period in both women at risk of HIV infection in sub‐Saharan Africa exposed to subtype C variants and MSM/TG persons in South America, Switzerland and the US exposed to subtype B variants. Susceptibility to the antibody was the important determinant of antiviral activity. The incidence of HIV among isolates sensitive to VRC01 (IC80 <1µg/mL) was 0.2/100 person‐years in VRC01 recipients versus 0.86 in control recipients (p<0.001) (estimated efficacy 75.4%, 95% CI: 45.5, 88.9). VRC01 did not prevent acquisition of isolates with IC80 >1.0µg/mL. The AMP trials were designed conjecturing that strains with an IC80 <10mg/mL would be effectively inhibited by VRC01, an in vitro cutoff met for 65% to 81% of strains in contemporaneous subtype B and C panels. Because only 30% of control arm‐acquired isolates had IC80 <1µg/mL, overall prevention efficacy was 26.6% (95% CI: −11.7, 51.8) for 704/085 and 8.8% (−45.1, 42.6) for 703/081 (p>0.10).

Conclusions: These studies provide proof‐of‐concept for bnAbs to prevent HIV acquisition. The breadth and levels of a bnAb required for effective prevention are predicted by the in vitro neutralization sensitivity of the viruses circulating in the community as measured in the TZM‐bl assay. These findings provide the guideposts for clinical development of combination bnAbs for the prevention of sexually acquired HIV.

Long acting injectable cabotegravir is safe and effective in preventing HIV infection in cisgender women: interim results from HPTN 084

S. Delany-Moretlwe¹, J. Hughes², P. Bock³, S. Gurrion⁴, P. Hunidzarira⁵, D. Kalonji⁶, N. Kayange⁷, J. Makhema⁸, P. Mandima⁵, C. Mathew¹, M. Mokgoro⁶, J. Mpendo⁹, P. Mukwekwerere⁵, N. Mgodi⁵, P. Nahirya Ntege¹⁰, G. Nair¹¹, C. Nakabiito¹², H. Nuwagaba-Biribonwoha¹³, R. Panchia¹⁴, N. Singh⁶

¹Wits RHI, University of the Witwatersrand, Johannesburg, South Africa, ²Statistical Centre for HIV/AIDS Research Prevention Fred Hutchinson Cancer Research Institute, Seattle, United States, ³Desmond Tutu TB Centre, University of Stellenbosch, Stellenbosch, South Africa, ⁴Kisumu CRS, KEMRI, Kisumu, Kenya, ⁵University of Zimbabwe College of Health Sciences Clinical Trials Research Centre, Harare, Zimbabwe, ⁶HIV Prevention Research Unit, South African Medical Research Unit, Durban, South Africa, ⁷Blantyre CRS, College of Medicine, Blantyre, Malawi, ⁸Botswana Harvard AIDS Institute Partnership (BHP), Gaborone, Botswana, ⁹UVRI‐IAVI, Entebbe, Uganda, ¹⁰Baylor College of Medicine Children's Foundation Uganda, Kampala, Uganda, ¹¹Desmond Tutu Health Foundation, University of Cape Town, Cape Town, South Africa, ¹²Makerere University ‐ Johns Hopkins University Research Collaboration, Kampala, Uganda, ¹³Eswatini Prevention Center, Columbia University Mailman School of Public Health, New York, United States, ¹⁴Perinatal HIV Research Unit, University of the Witwatersrand, Johannesburg, South Africa

Background: HPTN 084 is a Phase 3 randomized, double‐blind, double‐dummy superiority trial evaluating safety and efficacy of long‐acting injectable cabotegravir (CAB) compared to daily oral TDF/FTC for HIV prevention in cisgender women. The blinded trial was stopped at a planned interim DSMB review in November 2020.

Methods: HIV‐uninfected PrEP eligible cisgender women were randomized 1:1 to either active CAB plus TDF/FTC placebo or active TDF/FTC plus CAB placebo. Participants received 5weeks of daily oral product followed by intramuscular injections every eight weeks after an initial four‐week interval load, alongside daily oral pills. Participants who discontinued injections were offered open‐label daily TDF/FTC for 48weeks after last injection. The primary endpoints were incident HIV infection and ≥grade 2 clinical and laboratory events.

Results: Overall, 3224 participants were enrolled at sites in South Africa (n=7), Zimbabwe (5), Uganda (3), Malawi (2), Botswana (1), Eswatini (1) and Kenya (1). Median age was 25years (range 18, 45), 54% (n=1754/3210) had 2+ partners in past month, 32% (n=1019/3210) had partner with HIV or unknown status, 19% (n=601/3190) had chlamydia and 7% (n=211/3190) had gonorrhoea at enrolment. Participant visit completion at months 6, 12, 18, and 24 was 94%, 90%, 87% and 86%. Thirty eight incident infections were observed over 3808 person‐years (HIV incidence 1.0%, 95% confidence interval [CI] 0.71, 1.37); 4 in the CAB arm (incidence 0.21, 95% CI 0.06, 0.54) and 34 in the TDF/FTC arm (incidence 1.79%, 95% CI 1.24, 2.51) (hazard ratio 0.11 [95% CI 0.04, 0.32]). Injection coverage was 93% of person‐years. In a random subset of 375 TDF/FTC participants, 62% of plasma samples had detectable TDF/FTC; 46% had concentrations consistent with daily dosing. Adverse events were mild‐moderate and balanced by arm. Among CAB participants, injection site reactions were more common (32% vs. 9%), but most were mild. Nausea was more common in the TDF/FTC (9%) compared to the CAB participants (5%). Pregnancy incidence was 1.3 per 100 person‐years (95% CI 1.0, 1.7); no congenital anomalies were reported.

Conclusions: While both products demonstrated high prevention efficacy and were safe and well tolerated; CAB was superior to TDF/FTC in preventing HIV infection in cisgender women.

Safety and single‐dose pharmacokinetics of VRC07‐523LS administered via different routes and doses

S. Walsh¹, C. Gay², S. Karuna³, O. Hyrien³, T. Skalland³, K.H. Mayer⁴, M. Sobieszczyk⁵, P. Andrew⁶, C. Karg³, J. Baumblatt⁷, L. Polakowski⁷, W. Chege⁷, S. Hasan³, X. Han³, A. McDermott⁷

¹Brigham & Women's Hospital, Infectious Diseases, Boston, United States, ²University of North Carolina, Chapel Hill, United States, ³Fred Hutchinson Cancer Research Center, Seattle, United States, ⁴Fenway Institute, Boston, United States, ⁵Columbia University, New York, United States, ⁶FHI 360, Durham, United States, ⁷National Institute of Allergy and Infectious Diseases, Rockville, United States

Background: Broadly neutralizing antibodies (bnAbs) are a promising approach for HIV‐1 prevention. VRC07‐523LS targets the CD4‐binding site of Env and was engineered for increased breadth and half‐life. In the only bnAb HIV prevention efficacy studies, the AMP studies, another CD4‐binding site targeting bnAb, VRC01, was administered intravenously (IV). However, subcutaneous (SC) or intramuscular (IM) administration may be preferred. We present the first interim data comparing these routes of bnAb administration from the ongoing HVTN127/HPTN087 study.

Methods: 124 healthy, HIV‐uninfected participants were randomized to receive VRC07‐523LS via the IV (2.5mg/kg, 5mg/kg, 20mg/kg), SC (2.5mg/kg, 5mg/kg) or IM (2.5mg/kg, placebo) routes. Safety data were collected for 112weeks following the first administration. VRC07‐523LS serum concentrations were measured by ELISA at Day 0, 3, 6, 28, 56, 84, and 112. The log‐linear portion of the time‐concentration curve was used to estimate the elimination half‐life.

Results: Injections were well‐tolerated, with mild pain or tenderness reported commonly in the SC and IM groups and mild to moderate erythema or induration reported commonly in the SC groups. Infusions were generally well‐tolerated, with infusion reactions reported in 3 participants in the 20mg/kg IV group. VRC07‐523LS has an estimated median half‐life of ~40days. Median peak concentrations (with interquartile range) were 42.2 (35.1, 52.7) μg/mL, 80.3 (72.3, 106.1) μg/mL, and 353.6 (278.0, 461.97) μg/mL for the IV groups; 11.4 (8.4, 15.2) μg/mL and 24.5 (18.8, 27.0) μg/mL for the SC groups; and 17.8 (15.5, 19.1) μg/mL for the IM group. Geometric mean trough concentrations were 3.4 (2.5, 4.6) μg/mL, 6.5 (5.6, 7.5) μg/mL, and 27.2 (23.9, 31.0) μg/mL for the IV groups; 0.9 (0.6, 1.4) μg/mL and 3.1 (2.2, 4.3) μg/mL for the SC groups; and 2.6 (2.1, 3.3) μg/mL for the IM group. The peak VRC07‐523LS serum concentrations increased linearly with the administered dose. At a given dose, peak and trough concentrations were highest in the IV groups and lowest in the SC groups.

Conclusions: VRC07‐523LS appears to be safe and well‐tolerated across a range of doses and routes and is a promising bnAb for inclusion in HIV‐1 prevention regimens.

Safety and PK of potent anti‐HIV monoclonal AB VRC07‐523LS in HIV‐exposed infants

C. Cunningham¹, E. Capparelli², E. McFarland³, P. Muresan⁴, C. Perlowski⁵, E. Smith⁶, R. Hazra⁷, L. Purdue⁸, P. Harding³, A. McDermott⁹, J. Mascola⁹, B. Graham⁹

¹Duke University School of Medicine, Pediatrics, Durham, United States, ²University of California, San Diego, United States, ³University of Colorado Anschutz Medical Campus, United States, ⁴Harvard T.H. Chan School of Public Health, Boston, United States, ⁵FHI 360, Durham, United States, ⁶NIH, Bethesda, United States, ⁷National Institute of Child Health and Human Development, United States, ⁸DAIDS, NIAID, Bethesda, United States, ⁹Vaccine Research Center, NIAID, Bethesda, United States

Background: Despite the effectiveness of antiretroviral therapy, vertical HIV transmission continues. A potent, broadly neutralizing, monoclonal antibody (bNAb) administered to HIV‐exposed infants might reduce transmission. VRC07‐523LS is 5‐fold more potent and has a prolonged half‐life compared to VRC01. VRC07‐523LS may provide therapeutic levels over the duration of breastfeeding with infrequent doses.

Methods: This is an open‐label study of VRC07‐523LS administered to HIV‐exposed infants. Non‐breastfed infants receive 80mg subcutaneous (SC) within 72hrs of birth. Infants and mothers receive ART to prevent transmission. Infants have safety assessments and VRC07‐523LS levels at 24hrs, week 2, 4, 8, 12 and every 12weeks through week 96. The target week 12 plasma level is 10 mcg/mL: the level needed to neutralize>90% of tier II viruses in a multiclade panel. Plasma VRC07‐523LS levels are determined through week 12 and compared to previously reported levels of VRC01.

Results: The non‐breastfed cohort fully accrued (N=11) from US sites Jan‐Sep, 2019. All infants received 80mg VRC07‐523LS SC within 72hours of birth (mean 1.5days), resulting in an average dose of 28mg/kg (range 23 to 32mg/kg). Enrollees were 45% male, 73% Black, 18% Hispanic. One infant withdrew after 4weeks. VRC07‐523LS was well tolerated. Local reactions were rare and mild: 1 infant had injection site erythema of 0.5cm and 1 had tenderness. Five infants developed Grade 3/4 events within 28days of receipt of VRC07‐523LS (vomiting [N=2], neutropenia, hyperkalemia, and parainfluenza sepsis), none considered related to study treatment. Pharmacokinetic measures through week 12 show average levels of 68.7, 31.1, 16.3mcg/mL at weeks 4, 8, and 12, respectively. Mean VRC07‐523LS levels exceeded those previously reported for VRC01 20mg/kg SC at week 2, 4, and 8. Ongoing growth contributed to the fall in VRC07‐523LS concentration but levels remain over the target of 10 mcg/mL at week 12.

Conclusions: We identified no safety or tolerability findings when VRC07‐523LS is administered to neonates. Week 12 is an appropriate time for a second dose in infants with ongoing breastmilk exposure. VRC07‐523LS, with its enhanced potency and extended half‐life, could achieve target levels for the duration of breastfeeding with dosing every 3months.

Neutralization profiles of HIV‐1 subtype C breakthrough viruses from the Southern African VRC01 AMP trial (HVTN 703/HPTN 081)

N.N. Mkhize¹, R.E. Mapengo¹, V. Bekker¹, T. Modise¹, P. Kgagudi¹, B.E. Lambson¹, H. Kaldine¹, R.T. van Dorsten¹, N. Mgodi², S. Karuna³, S. Edupuganti⁴, L. Corey³, M.S. Cohen⁵, J. Hural³, J. McElrath³

¹National Institute for Communicable Diseases, HIV Virology, Johannesburg, South Africa, ²University of Zimbabwe, College of Health Sciences Clinical Trials Research Centre, Harare, Zimbabwe, ³Fred Hutchinson Cancer Research Center, Vaccine and Infectious Disease Division, Seattle, United States, ⁴Emory University, Department of Medicine, Atlanta, United States, ⁵University of North Carolina at Chapel Hill, Department of Medicine, Chapel Hill, United States

Background: HIV‐1 Env reference panels are used to guide the clinical development of broadly neutralizing antibodies (bNAbs) but need to be updated periodically as genetic drift may impact neutralization phenotypes. We assessed the neutralization sensitivity of breakthrough viruses from the southern African VRC01 AMP trial (HVTN 703/HPTN 081) as geographically relevant, current subtype C transmitted/founder viruses.

Methods: Envelope sequences of breakthrough infections from 30 women in the AMP trial (prior to unblinding) were used to produce transmitted/founder (T/F) pseudotyped viruses. These were tested against 14 bNAbs targeting the CD4bs (n=4), V3‐glycan (n=3), V2‐apex (n=3), gp120‐gp41 interface (n=2) and the MPER (n=2) in the TZM‐bl neutralization assay. Single bNAb neutralization data was used in the Loewe additive model (CombiNAber) to model the efficacy of bNAb combinations. Weakly neutralizing antibodies targeting normally occluded epitopes on HIV Env (V3, CD4i, V2 and gp41) were included to assess the tier phenotype of the viruses.

Results: All breakthrough viruses were of the tier 2 phenotype, typical of T/F viruses. At a concentration of 1µg/ml at IC50, 91% of viruses were neutralized by VRC07‐523LS (GMT IC50=0.16) representing the best coverage by a single bNAb. Combinations of two to four bNAbs increased the number of viruses neutralized with the 4 bNAb combination (CAP256‐VRC26.25/PGT121/VRC07‐523LS/35022) neutralizing 100% of viruses (GMT IC50=0.01). The best‐in‐class 3 bNAb combination (CAP256‐VRC26.25/PGT121/35022) provided 97% coverage (GMT IC50=0.01). Clinically advanced triple combinations such as CAP256‐VRC26.25.25/PGT121/VRC07‐523LS and/PGDM1400/PGT121/VRC07‐523LS both resulted in a coverage of 97% (GMT IC50=0.02) against this panel of 30 subtype C breakthrough viruses.

Conclusions: Our data confirm the need to use combination bNAbs in passive immunization trials to improve coverage of subtype C viruses. The exposure of a subset of these breakthrough viruses to VRC01 may have affected their phenotype, and this will be investigated when the AMP trial data is unblinded. Overall, these breakthrough viruses represent a unique resource for defining the sensitivity of contemporaneous circulating viral isolates to bNAbs.

Analysis of genetic diversity and VRC01 pressure on HIV‐1 breakthrough viruses from the AMP trial (HVTN 703/HPTN 081 and HVTN 704/085)

C. Williamson¹, D. Westfall², W. Deng², A. Pankow², D. Matten¹, B. Murrell³, T. York¹, A. Gwashu-Nyangiwe¹, M. Rolland^4,5, P. Edlefsen⁶, E. Giorgi⁷, C. Magaret⁶, D. Montefiori⁸, L. Morris⁹, M. S. Cohen¹⁰, L. Corey¹¹, J. Hural¹¹, J. McElrath¹¹, M. Juraska⁶, P.B. Gilbert⁶, J. I. Mullins², HVTN703/HPTN081 and HVTN704/HPTN083 protocol team, sequencing working group, community and participants

¹University of Cape Town, Institute for Infectious Diseases and Molecular Medicine, Cape Town, South Africa, ²University of Washington, Department of Microbiology, Seattle, United States, ³Karolinska Institutet, Department of Microbiology, Tumor and Cell Biology, Stockholm, Sweden, ⁴Henry M. Jackson Foundation for the Advancement of Military Medicine, Bethesda, United States, ⁵Walter Reed Army Institute of Research, U.S. Military HIV Research Program, Bethesda, United States, ⁶Fred Hutchinson Cancer Research Center, Statistical Center for HIV/AIDS Research and Prevention, Seattle, United States, ⁷Los Alamos National Laboratory, Los Alamos, United States, ⁸Duke University Medical Center, Duke Human Vaccine Institute, Durham, United States, ⁹National Institute for Communicable Diseases, National Health Laboratory Service, Johannesburg, South Africa, ¹⁰University of North Carolina at Chapel Hill, Department of Medicine, Chapel Hill, United States, ¹¹Fred Hutchinson Cancer Research Center, Vaccine and Infectious Disease Division, Seattle, United States

Background: The Antibody Mediated Prevention (AMP) trial evaluated if VRC01, a CD4 binding site broadly neutralizing antibody, could prevent HIV‐1 acquisition. To inform our understanding of how prevention efficacy depended on viral env gene charateristics, viral quasispecies and neutralization sensitivity of HIV‐1 breakthrough infections were analyzed.

Methods: The trial evaluated VRC01 in women from sub‐Saharan Africa (703/081); and men and transgender persons who have sex with men (704/085), from the Americas. Control samples from Thailand, Kenya and South Africa are being used for calibrating infection timing using sequence data. Rev‐env‐nef (REN) sequences were generated using Sanger and PacBio Single Molecule Real‐Time (SMRT) sequencing platforms. To improve PacBio accuracy and enable quantitation, each viral RNA molecule was labelled with a unique molecular identifier (SMRT‐UMI). Rev‐env genes from imputed founder viruses were synthesized for pseudovirus production and in vitro sensitivity to VRC01 determined using the TZM‐bl assay.

Results: Approximately 84000 unique REN sequences were generated from the first HIV‐1 positive visit from 218 infected individuals in both trials, and two to four weeks later from a subset of individuals, providing a median of 174 unique env sequences per participant per time point. In approximately 2/3 of individuals, viral diversity was low, consistent with infection with a single founder virus. Of the 1/3 of individuals with higher viral diversity, consistent with infection with multiple founders, a third had low frequency variants (<5% of sequences). We synthesized 1 to 4 Env‐pseudoviruses per person, and identified some individuals infected with variants with different neutralization phenotype, including infection with both sensitive (IC80<1µg/mL) and resistant (IC80>3µg/mL) viruses. In some individuals, the unique VRC01 resistance mutations were associated with distinct viral lineages, suggestive of infection with resistant viruses. However, we also found evidence of low frequency resistant mutations, that were likely derived from resistance acquired after infection.

Conclusions: SMRT‐UMI sequencing allowed a very detailed evaluation of viral populations following infection, including identification of minor founders not detected using conventional approaches. We found evidence of infection with viruses of mixed neutralization phenotypes, and in some cases, low frequency resistance mutations that were likely to be due VRC01 pressure. The study is not yet unblinded.

HIV‐1 bnAb M4008_N1 targets a novel site of vulnerability at the V3 crown

K.‐W. Chan¹, C. Luo¹, H. Lu², X. Wu², X.‐P. Kong¹

¹NYU School of Medicine, Department of Biochemistry and Molecular Pharmacology, New York City, United States, ²Columbia University Vagelos College of Physicians and Surgeons, Aaron Diamond AIDS Research Center, New York City, United States

Background: Most of the human monoclonal antibodies (mAbs) targeting the V3 crown of HIV‐1 gp120 can only neutralize tier 1 easy‐to‐neutralize viruses. We have isolated a broadly neutralizing mAb (bnAb) M4008_N1 encoded by IGHV1‐69 and IGKV1‐5 genes that neutralized about 40% of tier 2 circulating viruses. M4008_N1 binds to JR‐FL gp120 independent of glycans, and it effectively competes with known anti‐V3 crown mAbs, such as 447‐52D, 2219, and 2424, as well as bnAb PGT128 targeting the V3 base glycan region.

Methods: We determined a cryo‐EM structure of Fab M4008_N1 in complex with BG505 DS‐SOSIP at a 3.2Å resolution.

Results: Our structure revealed that M4008_N1 contacts a large envelope surface area surrounded by several glycans and centered at the V3 crown of gp120. M4008_N1 interacts with gp120 using both heavy and light chains, but primarily through its CDR H3 to form a beta sheet‐like interaction with the N‐terminal strand of V3 crown hairpin. It interacts as well the V1V2 stem and the C4 region. These contacts likely force V3 to remain sequestered underneath the V1V2 domain, compatible with the native closed conformation of the prefusion envelope trimer, and thus distinctive from other known anti‐V3 crown mAbs. In addition, its binding on SOSIP trimer spatially blocks the access to the CD4‐binding site on the neighboring gp120 protomer.

Conclusions: Our results revealed a new mode of bnAb approaching the N‐terminal strand of the sequestered V3 loop in the HIV‐1 native closed envelope trimer, thus a novel neutralization mechanism targeting a site of vulnerability at the HIV‐1V3 crown. This new mode of bnAb recognition is relevant for the design of immunogens targeting this site of vulnerability.

User assessment of a microarray patch for HIV PrEP and as a multipurpose prevention technology for HIV and pregnancy prevention: perspectives from Uganda and South Africa

M. Kilbourne-Brook¹, A. Ismail², S. Magni², T. Fellows², A. Ruhweza Katahoire³, F. Ayebare³, G. Siu³, F. Semitala⁴, P. Kyambadde⁵, D. Katuntu⁶, A. Rein‐Weston¹, C. Jarrahian¹

¹PATH, Medical Devices and Health Technologies, Seattle, United States, ²Genesis Analytics, South Africa, ³Makerere University, Child Health and Development Centre, Kampala, Uganda, ⁴Makerere University, College of Health Sciences, Department of Internal Medicine, Kampala, Uganda, ⁵Ministry of Health, MARPI, STD/AIDS Control Program, Kampala, Uganda, ⁶PATH, Uganda, Kampala, Uganda

Background: Innovative HIV and pregnancy prevention products that are easy to use and acceptable are needed to expand prevention options, especially for adolescent girls and young women (AGYW). A microarray patch (MAP)—a novel drug delivery system—is being developed to administer HIV PrEP and as a multipurpose prevention technology (MPT) to protect from HIV and unintended pregnancy. Feedback from user/stakeholder assessments early in product development can help refine product features to meet the needs of intended users.

Methods: Employing focus group discussions (FGD), key informant interviews, and mock‐use exercises with prototypes, we explored user/stakeholder perceptions of the MAP technology and needs/preferences regarding product features that could influence usability, acceptability, and programmatic integration. This study was conducted among 6 target audiences in 8 rural and urban sites in South Africa and Uganda. Overall, we conducted 14 FGDs with AGYW (18–24years), 2 with female sex workers (FSW), 8 with heterosexual men, and 2 with men who have sex with men. We also interviewed 20 HIV and family planning health care providers, 4 FSWs, and 6 policymaker/program managers. Seventy additional participants representing all user groups evaluated the usability of MAP prototypes. Findings were coded/analyzed using Atlas.Ti software.

Results: All groups expressed interest in the MAP technology, reporting potential advantages over methods such as pills and injectables. Most participants preferred a smaller MAP and long‐term protection (1–3+months) with some differences noted across groups. Self‐administration and discreet use were valued by all. Preferred application site and duration of application varied by MAP size. An MPT MAP was preferred over an HIV prevention‐only MAP by most AGYWs and FSWs. Participants wanted more confidence from the feedback indicator regarding correct MAP application and drug delivery. Participants wanted more information about how the MAP works and voiced concerns about potential drug‐related side effects and effectiveness.

Conclusions: This early‐stage user assessment of MAPs for HIV PrEP and MPT found high potential acceptability among users/stakeholders. MAP size and duration of protection are key attributes that will influence acceptability and uptake. Ongoing user assessments are essential to refine MAP prototypes to better meet users’ needs.

High protection against vaginal SHIV infection in macaques by a biodegradable implant releasing tenofovir alafenamide

I. Massud

Centers for Disease Control and Prevention, Laboratory Branch, Division of HIV/AIDS Prevention, Atlanta, United States

Background: To realize the promise of ending the global epidemic, long‐lasting HIV protection is needed through sustained release of antiretroviral drugs for months to years. Leveraging promising exploratory pharmacokinetics (PK) showing high TFV‐diphosphate (TFV‐DP) levels in PBMCs and dose escalation in rhesus macaques, we investigated the PK and efficacy of a polycaprolactone implant releasing tenofovir alafenamide (TAF) in pigtail macaques for preventing vaginal SHIV infection.

Methods: Implants were administered subcutaneously in the arm using a contraceptive trocar. The PK profile of implants releasing 0.35mg/day of TAF was evaluated in 3 pigtailed macaques. Efficacy was evaluated in 6 macaques that received two TAF implants (one per arm; total release‐rate=0.7mg/day). Macaques were exposed vaginally to SHIV162p3 twice‐weekly for 6weeks. Infection outcome was compared to 8 untreated controls. SHIV RNA was monitored in plasma by RT‐PCR. TFV‐DP in PBMCs and vaginal lymphocytes was measured by LC‐MS/MS. Skin biopsies near implantation were assessed by H&E staining.

Results: Median TFV‐DP levels in PBMCs over 84days with a single TAF implant were 954 [range 398 to 2080] fmols/106 cells. TFV‐DP was consistently detected in vaginal lymphocytes 24h post implant removal (50.7 [range 29.3 to 67.9] fmol/106 cells). Median TFV‐DP levels in PBMCs during the challenge phase were 1.6‐fold higher (1,519 [range 1068 to 1897] fmols/106 cells), which was consistent with the use of 2 TAF implants per animal. All macaques receiving TAF implants were SHIV RNA negative following 12 virus exposures and remained negative for 16weeks after implant removal. In contrast, 7/8 controls were infected after median of 4 SHIV exposures (Wilcoxon p‐value=0.0037). Despite slight to no skin reactions, H&E revealed marked deep dermal necrosis and inflammation by 7weeks post‐implantation. Recovered implants maintained a high degree of drug purity (>96%) and showed a strong correlation with in vitro release‐rates (median per animal=0.77mg/d).

Conclusions: TAF implants releasing 0.7mg/d resulted in high TFV‐DP levels in PBMCs that provided complete protection against vaginal SHIV infection. Findings define benchmarks needed for full protection against vaginal infection with single agent TAF. Improved TAF implants that maintain high efficacy while reducing local toxicity have the potential to provide long‐lasting protection against vaginal HIV infection.

Distribution of long‐acting (LA) cabotegravir (CAB) in plasma, mucosal tissues, and associated fluids after a single ultrasound‐guided intramuscular (IM) injection in healthy adult participants

E. Weld¹, J. Sadik Shaik², S. Edick³, E. Fuchs⁴, S. Riddler³, M. Marzinke¹, R. D'Amico⁵, K. Bakshi², Y. Lou⁶, C.W. Hendrix⁴, K. Han², S.L. Ford², D. Margolis⁵, W. Spreen⁵, P. Patel⁵

¹John Hopkins University School of Medicine, Department of Medicine ‐ Clinical Pharmacology, Baltimore, United States, ²GlaxoSmithKline, Research Triangle Park, United States, ³University of Pittsburgh, Pittsburgh, United States, ⁴John Hopkins University School of Medicine, Baltimore, United States, ⁵ViiV Healthcare, Research Triangle Park, United States, ⁶Precision Biosciences, Durham, United States

Background: CAB LA dosed at 2‐month intervals demonstrated virologic efficacy in maintaining HIV‐1 suppression as part of a dual regimen with rilpivirine LA and is undergoing Phase 3 evaluation as a single agent for HIV‐1 pre‐exposure prophylaxis (PrEP). CAB pharmacokinetics (PK) in plasma and mucosal tissues associated with sexual HIV‐1 transmission were evaluated following single CAB LA IM injection.

Methods: Participants received 4weeks of daily oral CAB 30mg, followed by 14 to 42day washout and a single ultrasound‐guided gluteal IM injection of CAB LA 600mg (3mL). PK samples were collected after oral dosing and through 12weeks post‐injection in plasma, rectal tissue (RT), cervical tissue (CT), vaginal tissue (VT), cervicovaginal fluid (CF), and rectal fluid (RF). CAB concentrations were determined by HPLC‐MS/MS. PK parameters were estimated using noncompartmental analysis. Pearson correlations of time‐matched tissue and plasma concentration pairs were determined.

Results: Nineteen participants (10F, 9M) enrolled with mean age 33years, weight 79.4kg, and BMI 27.2kg/m². Two participants withdrew before injections, one did not undergo tissue sampling, and one was hospitalized for unrelated serotonin syndrome. Median CAB concentration‐time profiles are presented by matrix (Figure 1). Median CAB tissue:plasma ratio was 9% to 10% in RT (r=0.96), 14% to 18% in CT (r=0.94), 14% to 16% in VT (r=0.92), 8% to 13% in CF (r=0.65), 32% to 45% in RF (r=0.19). Geometric mean (CVb%) plasma half‐life was 19.1 (81.4%) days.

Conclusions: Plasma CAB PK was consistent with prior studies, and CAB concentrations in tissue and fluid were proportional to plasma over time. Correlations with plasma concentrations were stronger for tissue (RT, CT, VT) than for luminal fluid (CF and RF). Tissue concentrations were 1/6^th (CT, VT) to 1/10^th (RT) of plasma concentrations. With sufficient distribution into mucosal tissues associated with sexual HIV‐1 transmission, CAB tissue:plasma ratios may serve as important measurements in evaluating CAB LA as an effective PrEP agent.

Design and testing of a cabotegravir reservoir implant for HIV prevention

T.J. Hope¹, D. Karunakaran², S.M. Simpson², J.T. Su³, E. Bryndza-Tfaily⁴, R. Vezy⁵, G. Dobek⁶, J. Qiu², D. Watrous², S. Sung², J.E. Chacon²

¹Feinberg School of Medicine, Northwestern University, McCormick School of Engineering and Obstetrics and Gynecology, Evanston, United States, ²Northwestern University, McCormick School of Engineering, Evanston, United States, ³Elon University, College of Art and Science, Elon, United States, ⁴Feinberg School of Medicine, Northwestern University, Cell and Developmental Biology, Evanston, United States, ⁵Tulane University, Division of Comparative Pathology, New Orleans, United States, ⁶Tulane University, New Orleans, United States

Background: Long‐acting antiretroviral implants may be an essential means of protecting high‐risk individuals from HIV infection. Long acting formulations can overcome issue of compliance associated with current pill forms of PrEP. Here we describe the design and testing of a reservoir subcutaneous implant capable of releasing cabotegravir for a year.

Methods: We compressed cabotegravir along with solubilizing excipients into cylindrical pellets and inserted them into a heat‐sealed hydrophilic polyether urethane membrane. The membrane was selected so that it can conduct the drug into the subcutaneous space. The implants were manufactured with dimensions of 47mm lumen length, 3.6mm outer diameter, and 200µm wall thickness. Four cabotegravir pellets were loaded in the core, with a total drug loading of 274±3mg.

Results: After implantation, the implant swelled and released an average of 0.35mg/day of cabotegravir in rhesus macaques. Five implants achieved an average cabotegravir plasma concentration of 373ng/ml in macaques. Adjacent implants demonstrated a PK less than anticipated, revealing that implant placement is important because of potential implant interference. The nonhuman primates tolerated the implant without gross or microscopic signs of toxicity compared to placebo after 1month and 3months. Cabotegravir plasma levels in macaques dropped below detectable levels within two weeks after the removal of the implants. Extended PK studies in Rats revealed stable release of cabotegravir for 6months.

Conclusions: Long‐acting formulations of antiretroviral compounds have the potential to increase the efficacy of PrEP by eliminating issues of compliance that can be associated with daily pill taking in high risk populations. We describe a cabotegravir reservoir implant designed to last 1year before removal and replacement (as needed). The removal of the implant avoids the drug PK tail associated with long‐acting injectable formulations. The release is stable for at least 6months and shows no evidence of adverse host reactions to the implant after histological analysis. Although the function of this reservoir implant in humans and the plasma cabotegravir concentrations required for protection in humans remains to be conclusively determined, this reservoir implant has the potential to provide an additional option for individuals seeking efficient PrEP for HIV prevention.

Trial design, enrollment status, demographics, and pharmacokinetics (PK) data from a blinded interim analysis from a phase 2a trial of Islatravir once monthly (QM) for HIV pre‐exposure prophylaxis (PrEP)

S. Hillier¹, L.G. Bekker², S. Badal-Faesen³, C. W. Hendrix⁴, S. A. Riddler⁵, S. Rasmussen⁶, H. Schwartz⁷, G. Nair², J. H. Lombaard⁸, Y. Caraco⁹, A. Peer¹⁰, M. Patel¹¹, B. Evans¹¹, B. Homony¹¹, V. Teal¹¹, P. Hwang¹¹, M. Robertson¹¹, R. Plank¹¹

¹Magee‐Womens Research Institute & Foundation, Pittsburgh, United States, ²Desmond Tutu HIV Centre, Cape Town, South Africa, ³University of the Witwatersrand, Johannesburg, South Africa, ⁴John Hopkins Hospital, Baltimore, United States, ⁵University of Pittsburgh, Pittsburgh, United States, ⁶Celerion, Lincoln, United States, ⁷Research Centers of America, Hollywood, United States, ⁸Josha Research, Bloemfontein, South Africa, ⁹Hadassah Medical Center, Jerusalem, Israel, ¹⁰Rambam Health Care Campus, Haifa, Israel, ¹¹Merck & Co., Inc., Kenilworth, United States

Background: Islatravir (ISL, MK‐8591) is the first nucleoside reverse transcriptase translocation inhibitor (NRTTI) in development for the treatment and prevention of HIV‐1. ISL is being evaluated as a once monthly tablet as PrEP. Trial design, enrollment demographics and status, and PK data from an ongoing phase 2a study are presented.

Methods: This randomized, double‐blind, placebo controlled, parallel‐group, multi‐center study is assessing the safety, tolerability and PK of oral ISL in adults aged 18 to 65 who have low‐risk for HIV‐1 acquisition. Participants were randomized in a 2:2:1 ratio into one of 3 groups receiving six doses of ISL 60mg, ISL 120mg, or placebo administered orally once monthly. ISL plasma levels were measured in all participants and ISL PK in peripheral blood mononuclear cells (PBMCs) and mucosal tissue (rectal, cervical and/or vaginal) was measured in a subset. Simulations using a previously developed population PK model were used to assess the interim observed plasma and PBMC PK data. Safety was assessed by adverse event (AE) reporting.

Results: As of 17 September 2020, 171 of a planned 250 (68.4%) participants have been randomized and dosed (68.4% Female, median age 33years, 70.8% white); 126 participants received all 6 planned doses, 38 completed the trial (through final follow‐up visit) and 8 discontinued the trial early. Blinded safety monitoring suggests that study drugs were well tolerated. Interim PK analysis shows ISL‐triphosphate trough concentrations following either 60mg or 120mg monthly doses were all above the pre‐specified PK threshold for HIV‐1 prophylaxis of 0.05 pmol/10⁶ PBMCs. Preliminary PK analysis of biopsied tissues suggest rapid, sustained and adequate distribution of ISL to sampled tissues. ISL PK exhibited linear dose proportionality (Figure 1).

Conclusions: This interim analysis suggests that monthly doses of ISL 60mg and 120mg achieved the pre‐specified efficacious PrEP PK threshold.

Role of HLA‐E antigen presentation on NK control of HIV infection 

L. Romero-Martin¹, C. Duran-Castells¹, M. Olivella², M. Rosas-Umbert¹, M. Ruiz-Riol¹, A. Olvera¹, C. Brander¹

¹Institut de Recerca de la Sida (IrsiCaixa), Hospital Universitari German Trias I Pujol, T Cell Immunity and Vaccinology, Barcelona, Spain, ²University of Pompeu Fabra, Spain

Background: MHC‐E restricted T‐cell responses have been observed in SIV‐vaccine strategies using CMV‐based vectors but their contribution to virus control is unclear. HLA‐E was originally identified as a ligand of some NKG2 family receptors expressed by NK cells. Two highly frequent, functional, HLA‐E alleles (*01:01 and *01:03) have been defined and T‐cell responses to bacterial and viral pathogens restricted by these molecules have been described.

Methods: In order to evaluate the interaction of HLA‐E‐presented viral epitopes with NK‐ and T‐cell receptors, we modeled the structural interaction of HLA‐E presenting canonical (VL9) and non‐canonical (HIV‐derived RL9 and PM9) epitopes with NKG2A/2C or TCR. We determined the ability of 14 peptides (including VL9, 1 CMV‐, 1 EBV‐ and 11 HIV‐derived) to stabilize HLA‐E *01:01 and *01:03 and how this modified NK degranulation and cytotoxic activity. In vitro inhibition of viral replication by NK cells was assessed in autologous HIV infected CD4+ T‐cells from HIV‐seronegative individuals (N=12). HLA‐E expression on these target cells was assessed by RT‐PCR and flow cytometry. The relationship of HLA‐E expression with in‐vivo HIV control was tested by measuring HLA‐E expression in HIV‐controllers (N=31) and non‐controllers (N=16) by RT‐PCR.

Results: Our structural models evidenced that TCR have less affinity for HLA‐E than NKG2 receptors. Interestingly, HIV peptides RL9 or PM9 presented by HLA‐E*01:01 or HLA‐E*01:03 showed a predicted increase in affinity to NKG2A and NKG2C, respectively. In vitro experiments indicated that HLA‐E*01:01 was generally less stable on the cell surface, but none of HIV‐derived peptides stabilized HLA‐E. HIV KG9‐HLA‐E*01*03 and RL9/TP9/VI9/YG9‐HLA‐E*01*01 decreased NK cytokine secretion. RL9/MD9‐HLA‐E*01:01 resulted in increased lysis of peptide pulsed target cells. NK‐mediated inhibition of viral replication correlated positively with HLA‐E expression. Conversely, in HIV‐infected individuals, HLA‐E expression on total PBMC was significantly higher in non‐controller individuals.

Conclusions: The elevated expression of HLA‐E in uncontrolled HIV infection and its potential differential interaction with NKG2 molecules depending on peptide binding indicates a pivotal role in NK during HIV infection. As such, HLA‐E presenting HIV‐derived epitopes may sensitize target cells for NK lysis in early infection whereas prolonged, elevated expression of HLA‐E may lead to NK reduced viral control.

Acute SIV Induces BHLH gene variants prior to adaptive natural killer cell formation

D. Ram, K. Kroll, M. Aid, R.K. Reeves

Beth Israel Lahey Health, Center for Virology and Vaccine Research, Boston, United States

Background: Adaptive natural killer (NK) cells have been demonstrated against a wide range of viral infections including CMV, influenza, SIV, and HIV, but are generally not detectable until late in infection due to their low magnitude. In this study we analyzed purified NK cells in acute SIV infection to evaluate potential molecular mechanisms of early adaptive NK cell formation.

Methods: Peripheral NK cells from a longitudinal acute SIV infection study in rhesus macaques (n=6) were sorted and RNA‐Seq performed using an Illumina platform. RNA‐Seq data was aligned using STAR aligner and Tophat suite was used for quantification of reads. To determine alternatively spliced transcripts the data was aligned with salmon and alternative transcripts were quantified and annotated using the IsoformSwitchAnalyzeR package. Custom probesets targeting various exon regions were designed to identify the desired isoform using RNA‐probe flow cytometry to complement up to 30‐parameter traditional flow cytometry.

Results: Rhesus NK cells were identified and sorted using a standard gating strategy: CD3‐CD14‐CD20‐NKG2A/C+. Using traditional RNA‐Seq data analysis we identified several groups of genes associated with normal NK cell antiviral responses, including IFN, cell cycle and mTOR‐associated genes. Interestingly, multiple BHLH transcription factor family members were significantly induced, including BHLHB42 (HES4) and BHLHE40 (DEC1) which clustered with multiple antigen processing and downstream memory cell signaling modules paralleling CD8 T cell memory formation. Using the IsoformSwitchAnalyzeR package revealed an even more restricted gene‐set of alternatively spliced genes including several TNFR and zinc finger transcription factor family members. RNA‐flow was then used to validate some of the identified genes, revealing several alternatively spliced transcript variants. For example, acute infection induced total ablation of the fully functional HES4 isoform while simultaneously favoring expression of a HES4 isoform lacking one of its DNA binding domains.

Conclusions: Our work provides the first insights into how retroviral infection indirectly influences alternative splicing in NK cells in any species. Further, we have identified several genes, including BHLH transcription factors that may play a role in the formation of adaptive NK cell responses. This may provide several opportunities for better directing NK cell responses during HIV infection.

Compartment specific changes of innate lymphoid cells within the intestinal mucosa of HIV‐1 infected individuals

O.E. Asowata¹, R. Fardoos², A. Singh¹, Y. Zungu¹, A. Ngoepe¹, F. Nene¹, A. Ntuli¹, F. Karim¹, A. Shalek³, F. Anderson⁴, A. Leslie¹, H.N. Kløverpris¹

¹Africa Health Research Institute, K‐RITH Tower Building, Durban, South Africa, ²University of Copenhagen, Department of Immunology and Microbiology, Denmark, ³Massachusetts Institute of Technology, Institute for Medical Engineering & Science and Department of Chemistry, Cambridge, Massachusetts, United States, ⁴University of KwaZulu‐Natal, Discipline General Surgery, Inkosi Albert Luthuli Central Hospital, South Africa

Background: HIV infection occurs predominantly in the mucosal surfaces of the gastrointestinal tract and is associated with compromised intestinal barrier integrity and dysbiosis that is not reversed by current antiretroviral therapy (ART). Innate lymphoid cells (ILCs) orchestrate mucosal barrier defences and are involved in the regulation of tissue homeostasis. However, the impact of HIV‐1 infection on ILCs in the human intestinal mucosa and gut draining lymphoid tissue is unknown.

Methods: Here, we present a large cohort of patients from a gastrointestinal clinic in KwaZulu‐Natal, South Africa recruited within extremely high HIV endemic areas. Human gut draining lymph nodes and intestinal biopsies were collected during surgical procedures. Phenotypic characterization of ILCs was done using flow cytometry and immunohistochemistry. Moreover, we employed single‐cell transcriptomics of pre‐sorted ILCs to examine the gene expression profile and ex vivo response to HIV infection.

Results: Total ILC levels in the gut of HIV uninfected individuals were comparable. However, we found ILC subtype‐specific and regional changes between the small and large intestine. Intraepithelial ILC1 (p<0.0001) and NK cells (p<0.0001) were significantly enriched in the duodenum compared to the colon, whereas ILC3s were significantly expanded in the colon compared to the duodenum (p<0.0001) in HIV uninfected individuals. In HIV infected participants, we observed expansion of duodenal ILC3s (p=0.03) that was not found within the colon compared to HIV uninfected participants. The gut compartment‐specific difference in ILC3 levels was independent of CD4 T‐cell depletion observed in both compartments (p<0.0001). We found modest changes in ILC3 levels within gut lymph nodes in HIV infected and uninfected individuals. Single‐cell RNAseq of pre‐sorted ILCs from lymph nodes reveals tissue‐specific transcriptional profiles between NK, ILC1 and ILC3 subsets with pending analysis of their response to HIV infection ex vivo.

Conclusions: Compartment‐specific differences in ILC subsets suggest distinct roles for these cells throughout the small and large intestine. Strikingly, ILC subset change in response to HIV infection within the small intestine with enriched ILC3s in HIV infected duodenal biopsies, irrespective of CD4 T cell depletion, suggests that ILC3s may play important roles in intestinal epithelial homeostasis that should be explored further for therapeutic potentials during chronic HIV‐1 infection.

Disruption of immune cell homeostasis within rectal mucosal tissue of HIV+ young men who have sex with men

S.A. Smith¹, P.K. Amancha¹, P.M. Murray¹, I.R. Pollack¹, C.G. Ackerley¹, R.R. Amara², C.F. Kelley¹

¹The Hope Clinic of the Emory Vaccine Center, Infectious Disease, United States, ²Yerkes National Primate Research Center, Atlanta, United States

Background: Ongoing inflammation within rectal mucosal tissue (RM) is a hallmark of HIV infection. While HIV infection is known to cause profound alterations in RM T cell populations in older adults, the effects of HIV infection on the immune cell ecosystem within RM tissue of young adults is not well understood.

Methods: We analyzed immune cell subsets in RM biopsies from 71 HIV‐ and 28 HIV+ young men who have sex with men (YMSM) aged 18 to 24years who were taking antiretroviral therapy with suppressed viremia (VL<400 copies/mL), and relatively preserved peripheral CD4+ T cells counts (median 633, range 232 to 1706), by multicolor flow cytometry. Median proportions of innate and adaptive immune cell subsets were compared between groups with Mann‐Whitney tests (threshold of discovery: p<0.01), and associated with peripheral CD4 counts via Spearman correlation coefficients.

Results: Despite generally healthy peripheral CD4 counts, we observed a significant decrease in CD4+ T cells within the RM of HIV+ YMSM (p=0.0007). This included reduced memory CD4+, CCR5+ CD4+, and IFNγ+ or TNFα+CD4+ (all p<0.009), but not Treg (p=0.34) or IL‐17A+ (p=0.60). Conversely, we observed an increase in Ki67+ CD4+ in the RM of HIV+ YMSM (as a percentage of CD4+ T cells, p=0.0002), as well as an increase in CD8+ T cells, including memory CD8+, non‐tissue resident CD69‐CD103‐CD8+, Ki67+ CD8+, and IFNγ+CD8+ (all p<0.002). We also observed increases in mucosal‐associated invariant T cells (MAIT) (p=0.010), accompanied by a decrease in CD16‐CD56+ NK cells (p=0.002). No differences were observed in RM γδ T cells, macrophages, B cells, neutrophils, CD1c+ myeloid dendritic cells, or plasmacytoid dendritic cells. Peripheral CD4 counts did not correlate with immune cell population proportions within the RM.

Conclusions: In addition to perturbations in the CD4+ and CD8+ T cell compartments, HIV infection among YMSM is associated with alterations in MAIT and NK cells populations within the RM, even in the context of effective ART and immune reconstitution. It will be critical to further understand the disruption of the RM immune environment in young HIV+ individuals, who are potentially facing a lifetime of chronic gut inflammation. To that end, additional transcriptomic and microbiome analyses are ongoing.

Pharmacokinetics, safety, and vaginal bleeding associated with continuous versus cyclic 90‐day use of dapivirine and levonorgestrel vaginal rings for multipurpose prevention of HIV and pregnancy

S. Achilles¹, C.W. Kelly², D.L. Blithe³, J. Long³, B.A. Richardson², B. Devlin⁴, C.W. Hendrix⁵, S.M. Poloyac⁶, M.A. Marzinke⁵, D. Singh⁷, J.M. Piper⁸, J. Steytler⁴, B.A. Chen⁹

¹University of Pittsburgh, Department of Obstetrics, Gynecology and Reproductive Sciences/School of Medicine, Pittsburgh, United States, ²Statistical Center for HIV/AIDS Research & Prevention/Fred Hutchinson Cancer Research Center, Seattle, United States, ³NIH/Eunice Kennedy Shriver National Institute of Child Health and Human Development, Division of Intramural Population Heath Research, Contraceptive Development Program, Bethesda, United States, ⁴International Partnership for Microbicides, Silver Spring, United States, ⁵Johns Hopkins University, Department of Medicine, Division of Clinical Pharmacology, Baltimore, United States, ⁶University of Pittsburgh, Department of Pharmaceutical Sciences, School of Pharmacy, Pittsburgh, United States, ⁷Magee‐Womens Research Institute, Pittsburgh, United States, ⁸NIH/National Institute of Allergy and Infectious Diseases, Division of AIDS, Bethesda, United States, ⁹University of Pittsburgh, Department of Obstetrics, Gynecology and Reproductive Sciences, School of Medicine, Pittsburgh, United States

Background: We compared pharmacokinetics, safety, and vaginal bleeding associated with continuous versus cyclic use of a 90‐day multipurpose prevention technology (MPT) vaginal ring for sustained delivery of dapivirine (DPV) for HIV pre‐exposure prophylaxis and levonorgestrel (LNG) for contraception.

Methods: In this Phase‐1 trial, healthy HIV‐uninfected women were randomized (1:1) to continuous or cyclic (28‐days in/2‐days out) use pattern of an MPT ring containing 200mg DPV/320mg LNG over 90‐days. We quantified plasma DPV and LNG via liquid chromatography‐mass spectrometry. We assessed vaginal bleeding and adverse events (AE) by daily text messaging.

Results: Twenty‐five evaluable participants had median age 36.0years (range 21 to 43) and BMI 27 (range 20 to 39), and 20 (80%) self‐identified as white. There were 84 AEs: 59 Grade‐1, 24 Grade‐2, and one Grade‐4 (anemia related to cyclic product use). The number of women with=Grade‐2 genitourinary AEs did not differ by continuous (3/12, 25%) or cyclic (5/13, 38%) use. With continuous use, median Cmax for DPV and LNG were 750pg/mL (IQR 475 to 1401) and 1675pg/mL (IQR 645 to 2575), respectively (Figure 1). Two days after ring removal (cyclic group), plasma DPV remained=concentration associated with previously demonstrated efficacy. Number of days with no bleeding, spotting/light, moderate, and heavy bleeding did not differ by use pattern (Table 1).

Conclusions: Serum DPV and LNG concentrations previously associated with efficacy for HIV and pregnancy prevention respectively are achieved or exceeded with cyclic and continuous use of this MPT ring without toxicity. Vaginal bleeding profiles did not differ.

Randomized, placebo‐controlled trial of safety, pharmacokinetics, and pharmacodynamics of 90‐day intravaginal rings (IVRs) releasing tenofovir (TFV) with and without levonorgestrel (LNG) among women in Western Kenya

N. Mugo¹, V. Mudhune², R. Heffron³, K. Thomas³, E. McLellan-Lemal⁴, S. Peacock¹, S. O'Connor⁴, B. Njoroge⁵, B. Nyagol², E. Ouma², R. Ridzon⁴, N. Isoherranen⁶, D. Erikson⁷, R. Haaland⁴, S. Morrison³

¹University of Washington, Kenya Medical Research Institute, Centre For Clinical Research, Kenya, ²Kenya Medical Research Institute, Centre for Global Health Research, Nairobi, Kenya, ³University of Washington, Global Health, International Clinical Research Center, Seattle, United States, ⁴Centers for Disease Control and Prevention, Office of Infectious Diseases, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Atlanta, United States, ⁵Kenya Medical Research Institute, Centre For Clinical Research, Nairobi, Kenya, ⁶University of Washington, Department of Pharmaceutics, Seattle, United States, ⁷Oregon Health & Science University, Endocrine Technologies Core, Oregon National Primate Research Center, Portland, United States

Background: Globally, young women face parallel epidemics of HIV infection and unintended pregnancy. Protection from both requires safe and effective prevention tools.

Methods: Healthy women ages 18 to 34years, not pregnant, HIV‐seronegative, HBsAg‐negative, not using hormonal contraception, of reproductive potential, and at low risk for HIV were randomized 2:2:1 to continuous use of a TFV/LNG, TFV or placebo IVR. We assessed genital and systemic safety, TFV concentrations in both plasma and cervicovaginal fluid (CVF) and LNG levels in serum using mass spectrometry. We further evaluated TFV pharmacodynamics (PD) through anti‐HIV and anti‐HSV activity in CVF ex vivo, and LNG PD using cervical mucus quality markers and serum progesterone measurement for ovulation inhibition.

Results: Among 312 women screened, 27 were randomized to use IVRs: TFV/LNG (n=11); TFV alone (n=11); and placebo (n=5). Most screen fails were due to vaginal infections. Median (IQR) days of ring use was 68 (36 to 90). Women reported more intermittent bleeding events with TFV/LNG IVR use than in the other two arms, but otherwise adverse events (AE) were distributed similarly among arms. There were two non‐product related AEs graded >2. No visible genital lesions were observed. Steady state geometric mean concentration (ssGMC) vaginal TFV amount was comparable in the TFV/LNG arm (44.0μg/swab (95% CI (31.2, 62.0)) and the TFV alone arm (30.3μg/swab (95% CI (18.1, 50.7)), p=0.25. Plasma TFV ssGMC was <10ng/mL for both TFV rings. In vitro, CVF anti‐HIV activity showed increased HIV inhibition over baseline following IVR use, from a median of −7% to 84% in TFV/LNG, 15% to 89% in TFV alone, and –27% (increased ex vivo infection) to –20% in placebo participants. LNG ssGMC was 240pg/mL (95% CI 170, 340) with rapid decline after removal to 90pg/mL (95% CI 60, 120) within 24hours. Following 15days of IVR use, progesterone levels in luteal phase indicated anovulation (<3ng/mL) more frequently using TFV/LNG (63.6%) compared to TFV (18.2%) and placebo (0.0%).

Conclusions: TFV/LNG and TFV alone IVRs were shown to be safe when used by African women. Pharmacokinetic characteristics and markers of protection against HIV‐1 and pregnancy suggest the potential for clinical efficacy of these IVRs.

Randomized, placebo controlled phase I trial of safety, pharmacokinetics, pharmacodynamics and acceptability of a multipurpose prevention vaginal ring containing tenofovir and levonorgestrel

A. Thurman¹, V. Brache², A.L. Ouattara¹, N. Chandra¹, T. Jacot¹, S. Jackson¹, M. Clark¹, M.M. Peet¹, H. Hanif¹, T. McCormick¹, S. Ju¹, J.L. Schwartz¹, M. Marzinke³, D.W. Erikson⁴, U. Parikh⁵

¹CONRAD, Eastern Virginia Medical School, Norfolk, United States, ²PROFAMILIA, Biomedical Research Department, Santo Domingo, Dominican Republic, ³Johns Hopkins University, School of Medicine, General Chemistry, Baltimore, United States, ⁴Oregon National Primate Research Center, Endocrine Technologies Core, Beaverton, United States, ⁵University of Pittsburgh, Infectious Diseases, Magee Womens Research Institute, Pittsburgh, United States

Background: Multipurpose prevention technologies that provide protection against HIV acquisition and unintended pregnancy, are safe and acceptable, and support flexible use, are critically needed. We report results from a Phase I study evaluating the safety, pharmacokinetics (PK), pharmacodynamics (PD), and acceptability of CONRAD's tenofovir/levonorgestrel (TFV/LNG) intravaginal ring (IVR) following 3months of continuous or interrupted use in women.

Methods: CONRAD A15‐138 was an outpatient, randomized, partially blinded, placebo‐controlled, parallel study conducted in Norfolk, VA and the Dominican Republic. Participants were healthy, 18 to 50years old, had a body mass index <30kg/m² and reported no use of exogenous hormones and regular menses. Participants were randomized to 1 of 4 study arms: TFV/LNG or placebo IVR worn continuously for ~90days or cyclically for 3 cycles of 28days of use with 3days removal then re‐insertion.

Results: We screened 68 women; 47 were randomized onto study and 40 completed all visits. IVRs were safe with no significant changes in cervicovaginal epithelium, immune cell populations or soluble immune and inflammatory markers from baseline. Most TFV/LNG IVR users reported no change in their menstrual cycle or fewer/lighter bleeding days. Median vaginal fluid TFV concentrations were 546 to 3077ng/mg throughout 90days of use. Median TFV‐DP tissue concentrations exceeded 1000 fmol/mg within 72hours of IVR insertion. At 1 and 3months of use, vaginal fluid of women using TFV/LNG IVRs had significantly greater inhibitory activity against HIV growth in vitro compared to baseline and to placebo (p<0.01). TFV/LNG IVR users had mean serum LNG concentrations exceeding 200pg/mL within 2hours of IVR insertion. All placebo users ovulated each month, while only 39% to 71% of TFV/LNG users ovulated during months 1, 2 or 3, consistent with previously tested, effective contraceptive LNG IVRs. TFV/LNG IVR users had significantly lower cervical mucus (CM) Insler scores and a higher proportion of poor or abnormal in vitro CM sperm penetration (p<0.05).

Conclusions: The TFV/LNG IVR was safe, acceptable and delivered high TFV concentrations locally correlating with local PD. The microdose of LNG caused changes in CM, sperm penetration and ovulation compatible with contraceptive efficacy, while inducing acceptable changes in menstrual bleeding patterns.

Heterosexual couples’ preferences for dual‐purpose prevention products for HIV and pregnancy prevention: the CUPID Study (MTN‐045) in Uganda and Zimbabwe

A. Minnis¹, J. Etima², M.K. Shapley-Quinn³, P. Musara⁴, D. Kemigisha², E. Browne³, M. Stoner³, N. Mgodi⁴, C. Nakabiito², M. Hartmann³, N. Macagna⁵, J. Piper⁶, A. van der Straten¹

¹RTI International, Women's Global Health Imperative, Research Triangle Park, United States, ²Makerere University‐Johns Hopkins University (MU‐JHU) Research Collaboration, Kampala, Uganda, ³RTI International, Research Triangle Park, United States, ⁴University of Zimbabwe College of Health Sciences, Harare, Zimbabwe, ⁵FHI 360, Durham, United States, ⁶National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, United States

Background: Most research with end‐users of future HIV prevention products has focused on women despite male partners’ important role in product use decision‐making. The CUPID Study assesses preferences for future dual‐purpose prevention (DPP) products for pregnancy and HIV prevention and examines relationship‐based influences pertinent to the development and use of DPP products.

Methods: CUPID is a multi‐methods cross‐sectional study targeting 400 heterosexual couples in Uganda and Zimbabwe (female partners aged 18 to 40 and HIV negative by self‐report) imitated in January 2020. Couples complete interviewer‐administered surveys individually, followed by a joint interview that includes an interactive Ideal Product Activity (IPA), self‐directed by the couple and observed by a trained researcher. The IPA asks couples to select the most preferred characteristics of a DPP product (e.g., delivery form, effect on menses, effects on return to fertility, timing of use and duration of effectiveness). We examined overall attitudes toward DPP products, preferred characteristics elicited in the IPA, and relationship decision‐making characteristics associated with these preferences using Poisson regression models with robust standard errors.

Results: Among 117 couples (mean age, females: 27, males: 32) enrolled through March 2020, nearly all (92%) indicated a preference for a dual vs. single purpose product. Primary advantages cited include ease of using a “2 in 1” product and, among females, ability to present the product as only a contraceptive; disadvantages included concerns regarding side effects in a combined product and the need to switch methods when pregnancy is desired. The majority (61%) of couples selected oral tablets as their ideal formulation, while 39% preferred a vaginally delivered product (ring, insert or film). Though longer‐duration products were favored (two to three months), one‐third indicated their ideal product would be monthly and 10% preferred on‐demand. Women who reported they usually made their health care decisions individually (vs. jointly with their partner) were less likely to indicate their ideal product as vaginally delivered (IRR: 0.48, 95% CI: 0.26, 0.90, p=0.02).

Conclusions: Couples, both individually and jointly, indicated high interest in DPP products that combine HIV and pregnancy prevention. Ideal product characteristics, particularly delivery form, signaled a role for partner engagement and disclosure to support use of vaginally delivered products.

A phase I study to assess safety, pharmacokinetics and pharmacodynamics of a topical multipurpose prevention insert containing tenofovir alafenamide fumarate and elvitegravir dosed vaginally

A. Thurman¹, A.L. Ouattara¹, N. Yousefieh¹, T. Jacot¹, H. Hanif¹, P. Anderson², L. Bushman², X. Fang¹, M.M. Peet¹, N. Vann¹, T. McCormick¹, V. Agrahari¹, O. Singh¹, M. Clark¹, G.F. Doncel¹

¹CONRAD, Eastern Virginia Medical School, Norfolk, United States, ²University of Colorado, Skaggs School of Pharmacy and Pharmaceutical Sciences, Aurora, United States

Background: A discreet, on‐demand, vaginal or rectal, pre‐ or post‐exposure prophylaxis product is unique and fills an important gap in human immunodeficiency virus (HIV) and herpes simplex virus type 2 (HSV‐2) multipurpose prevention. We describe the safety, acceptability and multi‐compartmental pharmacokinetics (PK) and pharmacodynamics (PD) after healthy women used a single vaginal insert containing tenofovir alafenamide fumarate (TAF) and elvitegravir (EVG).

Methods: This was a Phase I, open‐label, study. Women (n=16) self‐administered one TAF (20mg)/EVG (16mg) insert vaginally in the clinic and were randomized (1:1) into one of two sample collection time groups (4 and 48 vs. 24 and 72hours). All women were sampled 7days post dosing. We assessed safety by treatment emergent adverse events. We measured EVG, TAF and tenofovir (TFV) concentrations in plasma, vaginal fluid and tissue and TFV‐diphosphate (TFV‐DP) in vaginal tissue. PD was modeled ex vivo by quantifying the change in inhibitory activity of vaginal fluid and vaginal tissue against HIV and HSV‐2 after treatment, compared to baseline.

Results: The TAF/EVG insert was safe, well tolerated and acceptable following a single vaginal dose. We measured high cervicovaginal fluid TFV, with mean and median concentrations exceeding 200000ng/mL for up to 24hours post dosing (median 340854 IQR (232508, 474736ng/mL) and exceeding 1000ng/mL for up to 7days post dosing. All participants had vaginal tissue EVG concentrations of>1000ng/g at 4 and 24hours post dosing. TFV‐DP tissue concentrations exceeded 1000 fmol/mg by 24 and 72hours post dosing in 75% of participants. Vaginal fluid inhibition of HIV and HSV‐2 ex vivo was significantly (p<0.05) increased from baseline and was similarly high at 4 and 24hours post dosing. Consistent with high tissue TFV‐DP concentrations, p24 antigen production from ectocervical tissues infected ex vivo with HIV was significantly (p<0.05) decreased from baseline at 4hours post dosing.

Conclusions: A single vaginal dose of TAF/EVG topical insert was safe, acceptable and effective in preventing HIV‐1 and HSV‐2 infection ex vivo in this first‐in‐woman clinical study. PK data support an extended window of high mucosal exposure.

SMS reminders did not improve PrEP adherence in a randomized controlled trial among young Kenyan women

J. Haberer¹, N. Mugo², E. Bukusi², M. Pyra³, K. Oware², C. Kiptinness², K. Thomas⁴, L. Garrison⁵, N. Musinguzi⁶, S. Morrison⁴, P. Anderson⁷, K. Ngure⁸, J. Baeten⁴

¹Massachusetts General Hospital, Global Health, Boston, United States, ²KEMRI, Nairobi, Kenya, ³University of Chicago, United States, ⁴University of Washington, Seattle, United States, ⁵Massachusetts General Hospital, United States, ⁶Mbarara University of Science and Technology, Global Health Collaborative, Uganda, ⁷University of Colorado, United States, ⁸Jomo Kenyatta University, Kenya

Background: Pre‐exposure prophylaxis (PrEP) is highly effective in preventing HIV acquisition. However, adherence among young women has been challenging. SMS reminders have been shown to improve adherence to antiretroviral therapy in some contexts. We present results from a randomized controlled trial of SMS reminders to support PrEP adherence among young women in Kenya.

Methods: The MPYA (Monitoring PrEP among Young Adult women) study involved 18 to 24‐year‐old women with high HIV risk. PrEP adherence was measured over 2years with a real‐time electronic monitor (Wisepill) and pharmacy refill; tenofovir‐diphosphate (TFV‐DP) levels were determined from dried blood spots for 15% of participants. Participants were randomized 1:1 to SMS reminders versus no reminders. Reminders were initially sent daily, although participants could opt for “as needed” reminders (if they missed doses). The effect of SMS reminders on adherence was assessed by Poisson models adjusted for study site.

Results: Of 348 women in the study, 173 were assigned daily SMS reminders; 9% opted for “as needed” reminders. Participants were a median of 21years old and two‐thirds reported condomless sex in the month prior to enrollment. As shown in the figure, adherence decreased over time by both electronic monitoring and pharmacy refill. Correlation between electronically monitored adherence and TFV‐DP levels was high (R²=0.73). Among participants picking up PrEP, electronically monitored adherence was 26.9% over 24months and did not differ by arm (27.0% with SMS; 26.7% without SMS; adjusted incidence rate ratio 1.07 [95% CI 0.86, 1.33], p=0.55). No differences were seen between arms when assessing adherence for all participants regardless of PrEP refills, by pharmacy refill data, with only 12months of follow‐up, or by site.

Conclusions: SMS reminders did not overcome adherence challenges among young Kenyan women taking PrEP over two years. Given the overall low adherence in the trial, additional interventions are needed to support PrEP use in this population.

“I just decided to stop:” understanding PrEP discontinuation among individuals initiating PrEP in HIV care centers in Kenya and its implications for a public health approach to prevention

A. Dollah¹, F. Ongolly², K. Ngure³, J. Odoyo¹, E. Irungu², K. Mugwanya⁴, J. Morton⁴, E. Bukusi¹, N. Mugo², J. Beaten⁴, G. O'Malley⁴

¹KEMRI/RCTP, Research, Nairobi, Kenya, ²KEMRI/PHRD, Nairobi, Kenya, ³Jomo Kenyatta University of Agriculture and Technology, Research, Kenya, ⁴University of Washington, Gloabal Health, Seattle, United States

Background: Although PrEP discontinuation rates in clinical trials and demonstration projects have been well characterized, little is known about it in routine public health settings. Understanding discontinuation decisions in non‐study settings is important for calibrating expectations of PrEP persistence in national programs and strengthening public health approaches to HIV prevention.

Methods: In‐depth interviews were conducted with 47 individuals who initiated PrEP at 25 different HIV comprehensive care centers (CCCs) in Central and Western Kenya, whose clinic record indicated had not returned for scheduled refills. We explored decisions around initiation, discontinuation, and restarting PrEP. An inductive, thematic, content‐analytic approach was used to analyze the data.

Results: Clients initiated PrEP because they had one or more sexual partners who were either HIV positive or of unknown status. Many discontinued PrEP when their perceived risk decreased (i.e. because those relationships ended, because they were living apart from a primary partner, or when a known HIV positive partner became virally suppressed). These participants expressed willingness to re‐start PrEP if their partnership situation changed. Others reported discontinuation due to side effects (dizziness, nausea, weight gain) or found daily pill‐taking too burdensome, and preferred condoms for prevention purposes. Some participants (mostly women) discontinued PrEP due to their partner's insistence and wished for additional PrEP education and counselling to foster partner support. Though relatively few participants identified facility level factors as primary reasons for discontinuing PrEP, many described stigma‐related discomfort with accessing PrEP at CCCs, inconvenient clinic location and/or operating hours, long wait times, and short refill dates as barriers. Most individuals reported multiple reasons for deciding to discontinue PrEP and did not inform health facilities of their decision to stop.

Conclusions: Clients make intentional decisions to discontinue PrEP as they weigh different prevention options and navigate fluid and sometimes challenging relationships. Many clients will decide to discontinue PrEP when perceiving themselves to be at reduced risk and PrEP counseling approaches must include provisions for addressing ‘seasonal risk.’ PrEP will not be the right prevention method for everyone. However, expanding PrEP access points and increasing sex‐positive messaging may facilitate PrEP being a better option for many.

Limited knowledge of, interest in, and eligibility for event‐driven PrEP among MSM in two samples in the United States

M. Newcomb, D. Ryan, C. Xavier Hall, K. Macapagal, B. Mustanski

Northwestern University, Institute for Sexual and Gender Minority Health and Wellbeing, Evanston, United States

Background: Uptake of daily oral pre‐exposure prophylaxis (PrEP) among men who have sex with men (MSM) in the United States is increasing rapidly. Non‐daily event‐driven PrEP (ED‐PrEP, or “2‐1‐1”) also substantially reduces HIV risk and is recommended by the World Health Organization (WHO) for MSM with more infrequent sexual behavior. In the absence of national guidelines for ED‐PrEP use in the United States, it is unclear how many MSM in the U.S. are aware of, interested in, and eligible for this dosing strategy.

Methods: We analyzed data from 2 independent samples of HIV‐negative MSM in the U.S. First, an analytic sample of 424 MSM were drawn from RADAR, a longitudinal cohort study of >1200 young MSM in the Chicago area (current data collected October 2019 to March 2020). Second, 267 MSM were drawn from a larger U.S. national cross‐sectional survey recruited via social media focused on biomedical HIV prevention (collected November 2019 to March 2020). We conducted descriptive statistics using SPSS to examine use, knowledge, and interest in ED‐PrEP. We calculated eligibility for ED‐PrEP based on WHO guidelines (i.e., 2 or fewer sexual encounters/week, ability to anticipate/delay sex within 2 to 24hours of first dose).

Results: Across both samples, 2.3% of MSM reported using ED‐PrEP. In the U.S. national sample, 38.9% had heard of ED‐PrEP, but only 18.5% knew all details of the 2‐1‐1 dosing method. Among current PrEP users, a substantial minority of MSM were eligible for ED‐PrEP (46.7% Chicago, 30.5% U.S. national), but most preferred daily dosing (preference for ED‐PrEP was 21.2% Chicago, 17.7% U.S. national). A marginally higher number of MSM who were eligible for ED‐PrEP expressed preference for this method, compared to those who were ineligible.

Conclusions: Although many MSM in the U.S. would be eligible for ED‐PrEP based on WHO guidelines, knowledge of this strategy remains limited and most MSM report preferring daily dosing. As knowledge of ED‐PrEP increases, interest will likely also grow among MSM. Thus, it is critical that national U.S. health organizations develop guidelines for ED‐PrEP use and that MSM are given accurate information about how to effectively use this dosing strategy.

I'm taking PrEP for myself and not for people: PrEP disclosures influence adherence journeys for adolescent girls and young women in South Africa

J. Daniels¹, D. Bresenham², L. de Vos², R. Mawarire², M. Atujuna³, S. Hosek⁴, C. Celum⁵, L.‐G. Bekker³, A. Medina-Marino²

¹Charles R. Drew University of Medicine and Science, Psychiatry and Human Behaviors, Willowbrook, United States, ²Foundation for Professional Development, South Africa, ³Desmond Tutu HIV Foundation, Cape Town, South Africa, ⁴Cook County Health, Chicago, United States, ⁵University of Washington, Seattle, United States

Background: Prevention effectiveness of oral PrEP is directly associated with levels of adherence. To improve the prevention‐effective‐use of PrEP, we explored behavioral, social and community factors associated with PrEP disclosure and use among adolescent girls and young women (AGYW).

Methods: Individual in‐depth interviews (IDI) were conducted with 40 AGYW (aged 16 to 25years) participating in the Community PrEP study in Eastern Cape Province, South Africa. IDI guides were developed using Social Practice Theory (SPT). IDI domains included: PrEP knowledge and use, study site experiences, PrEP actions plans and influences, and PrEP discussions in social spaces. AGYW with≥12months of study participation were interviewed, with equal representation of AGYW with high and low PrEP adherence measured via tenofovir‐diphosphate (TFV‐DP) blood concentrations. IDIs were conducted in isiXhosa, audio‐recorded, transcribed and translated for analysis using an iterative process guided by SPT to assess factors influencing PrEP disclosure and adherence.

Results: Most participants (78.4%) were enrolled in high school, lived with family (97.3%), had a current partner (51.4%), and were sexually active (64.9%) within the last year. All participants considered their high HIV risk and future goals as primary PrEP motivators. Participants described PrEP disclosure events to family members, friends and boyfriend/partners, with 91.7% of those with high TFV‐DP blood concentrations disclosing to all three categories, while only 57.9% of those with low TFV‐DP blood concentrations disclosed to all three. Receiving PrEP support from multiple individuals helped in navigating PrEP disinterest from others, lack of PrEP education within the community, and mistrust in relationships. After disclosure, gossip and PrEP myths were not influential for those with multiple disclosures and sources of support. All participants described taking PrEP as an independent journey, but with support from different people in their lives.

Conclusions: PrEP initiation was strongly influenced by one's future goals, while PrEP adherence was seen as an independent journey influenced by disclosures to key individuals in different social spaces. AGYW with high TFV‐DP blood concentrations described larger circles of assembled support. Prevention‐effective‐use of oral PrEP among AGYW may be dependent on building disclosure skills to develop a PrEP support system.

High initiation and persistence on pre‐exposure prophylaxis (PrEP) in HIV‐uninfected pregnant women in Cape Town, South Africa

D. Joseph Davey¹, R. Mvududu², N. Mashele², M. Lesosky², L.‐G. Bekker³, P. Gorbach¹, T. Coates⁴, L. Myer², PrEP‐PP study

¹University of California Los Angeles, Epidemiology, Los Angeles, United States, ²University of Cape Town School of Public Health and Family Medicine, Division of Epidemiology and Biostatistics, Cape Town, South Africa, ³Desmond Tutu Health Foundation, Cape Town, South Africa, ⁴UCLA, Medicine, Los Angeles, United States

Background: Oral PrEP is a safe and effective prevention strategy to reduce women's risk of HIV in pregnancy and postpartum. Successful PrEP outcomes require PrEP adherence at the time of potential HIV exposure, especially in pregnancy when tenofovir plasma concentrations are lower than postpartum.

Methods: The PrEP in pregnancy and postpartum (PrEP‐PP) study is an ongoing prospective cohort which enrolls consenting pregnant, HIV‐uninfected women (>15‐years) at first antenatal care visit, followed through 12‐months postpartum. Interviewers collect data on participant socio‐demographics, relationships, HIV risk factors including: partner's serostatus, intimate partner violence, substance use and depression. At baseline we provided point‐of‐care STI testing/treatment of chlamydia, gonorrhea and trichomonas (GeneXpert). We evaluate factors associated with: PrEP initiation, PrEP persistence (returning for PrEP repeat prescription) and PrEP adherence (reporting taking PrEP≥5 of last seven‐days) at three‐months after PrEP start using multivariable logistic regression controlling for baseline age and gestational age.

Results: Between Aug'19 and Oct'20 we enrolled 712 pregnant women (median gestation=21w; median age=26y). Following counseling, 91% of women initiated PrEP at their first antenatal visit (n=649). Women who were married had lowest odds of taking PrEP compared to unmarried women (aOR=0.11; 95% CI=0.02 to 0.73), though over one‐fourth didn't know her partner's serostatus (27%). Thirty‐five percent of women were diagnosed with a STI. Those with a STI had a non‐significant greater odds of initiating PrEP (aOR=1.60; 95% CI=0.89 to 2.89). Pregnant women who reported high internalized PrEP stigma had lower odds of initiating PrEP (aOR=0.06; 95% CI=0.03 to 0.12). At 1m, 72% women returned for a repeat prescription; at 3m, 60% returned. Among those returning at 3m, 87% reported adhering to PrEP in past seven‐days. Being postpartum was associated with lower odds of PrEP persistence and adherence (aOR=0.31; 95% CI=0.16 to 0.61) adjusting for follow‐up study, maternal and gestational age. Retention was lower in women who reported side‐effects (22% of women reported nausea, dizziness and vomiting).

Conclusions: PrEP uptake was high especially in high‐risk women. PrEP persistence and adherence were higher than other SA populations. Anticipated stigma, side effects and the postpartum period presented challenges to optimal PrEP use, pointing to potential targets for interventions to improve PrEP adherence in pregnant/postpartum women.

Antibody isotype switching as a mechanism to counter HIV neutralization escape

C. Scheepers¹, V. Bekker¹, C. Anthony², S. Richardson¹, B. Oosthuysen¹, T. Moyo¹, P. Kgagudi¹, D. Kitchin¹, M. Nonyane¹, B. Mabvakure¹, Z. Sheng³, B. Lambson¹, A. Ismail¹, N. Garrett⁴, S. Abdool Karim⁴

¹National Institute for Communicable Diseases, HIV and STIs, Sandringham, South Africa, ²Institute of Infectious Disease and Molecular Medicine, South Africa, ³Columbia University, New York, United States, ⁴Centre for the AIDS Programme of Research in South Africa (CAPRISA), Durban, South Africa

Background: Neutralizing antibodies (nAbs) to highly variable viral pathogens show remarkable diversification during infection, resulting in an “arms race” between virus and host. Somatic hypermutation (SHM) in immunoglobulin variable regions enables maturing antibodies to neutralize emerging viral escape variants. However, antibody diversification also occurs through class‐switch recombination (CSR) resulting in intra‐lineage isotype variation. The immunoglobulin constant region can influence epitope recognition and viral escape, but whether CSR also contributes to the maturation of natural antibody lineages during the course of HIV infection has not been explored.

Methods: Longitudinal deep sequencing of an HIV‐directed nAb lineage, CAP88‐CH06, identified several co‐circulating isotypes (IgG3, IgG1, IgA1, IgG2 and IgA2), some of which shared identical variable regions. CAP88‐CH06 lineage members were expressed as multiple naturally occurring isotypes and tested for neutralization and binding against longitudinal single‐genome derived autologous envelope clones and epitope mutants.

Results: We show that higher levels of SHM were associated with improved neutralization potency against the T/F virus, and an increased ability to neutralize viruses from later time points. However, we also show differences in neutralization and binding of autologous viruses and epitope mutants, based on the isotype of the antibody. In all instances, IgG3 and IgA1 isotypes were better able to neutralize longitudinal autologous viruses and epitope mutants than IgG1 versions of the same antibody, in some cases by >70 fold. We further show that CSR directly impacts nAb lineage maturation, with some switches such as IgG3 to IgG1 resulting in reduced neutralization of circulating viruses, creating “dead‐end” antibody sub‐lineages. However, these detrimental switches could be rescued by a further CSR event, from IgG1 to IgA1, which restored neutralization capacity of these antibodies, and enabled further lineage maturation.

Conclusions: Our data suggest as with SHM, CSR may have beneficial or deleterious outcomes for antibody maturation. In the CAP88‐CH06 lineage, a beneficial CSR event following a detrimental switch, a process we term “switch redemption”, enabled the continued maturation of that sub‐lineage. Thus, CSR represents an additional immunological mechanism to counter viral escape from HIV‐specific antibody responses. Furthermore, vaccine design strategies aimed at promoting class‐switch recombination could enhance antibody responses and vaccine efficacy.

Vaccination induces maturation of diverse unmutated VRC01‐class precursors to HIV‐1 broadly neutralizing antibodies in an Ig‐humanized mouse model

X. Chen¹, T. Zhou¹, S. Schmidt¹, H. Duan¹, C. Cheng¹, G.‐Y. Chuang¹, Y. Gu¹, M. Louder¹, B. Lin¹, C.‐H. Shen¹, Z. Sheng², M.G. Joyce³, N. Doria‐Rose¹, L. Shapiro², M. Tian⁴

¹NIAID, VRC, Bethesda, United States, ²Columbia University, New York, United States, ³Walter Reed Army Institute of Research, Silver Spring, United States, ⁴Boston Children's Hospital, Boston, United States

Background: The vaccine elicitation of broadly neutralizing antibodies (bnAbs) is a key HIV‐research goal. The VRC01‐class of bnAbs targets the CD4‐binding site on the HIV‐envelope trimer and requires extensive somatic hypermutation to neutralize effectively. Despite substantial progress in developing germline‐engaging immunogens to activate VRC01‐class bnAb precursors and using transgenic mouse models for eliciting VRC01‐class antibodies, vaccine‐induced VRC01‐class antibodies starting from unmutated precursors have exhibited limited neutralization breadth, particularly against viruses bearing glycan at residue N276 (glycan276), which is present on most circulating strains.

Methods: Here, using an immunoglobulin (Ig)‐humanized mouse model in which the VRC01‐class germline VH gene recombines with diverse mouse D and JH gene segments and pairs with an unmutated rearranged light chain to create a diverse repertoire of unmutated VRC01‐class precursors, we compared a “multi‐strain, heterologous boost” sequential immunization strategy, which presents the conserved CD4bs in different gp120 cores or trimers for better immune focusing on the CD4bs, with previously tested strategies based on germline‐binding 426c cores with different levels of glycan shielding of the CD4bs (Tian et al., 2016) or repeated eOD‐GT8 60mer priming followed by diverse Env boosts.

Results: With both the new strategy and the 426c‐core strategy, we elicited cross‐clade neutralizing serum titers against a sentinel panel of eight HIV‐1 strains, including five bearing glycan276. We further identified multiple lineages of VRC01‐class bnAbs from the immunized animals, including two neutralizing >50% of a 208‐strain panel, and carried out mutagenesis and crystal structure analyses, revealing key sites of SHM and mechanisms for surmounting glycan276 to achieve neutralization breadth.

Conclusions: Overall, our study provides proof‐of‐concept for the induction of VRC01‐class bnAbs of greater than 50%‐neutralization breadth by sequential immunization, succeeds in eliciting antibodies capable of recognizing glycan276‐bearing strains, and uses longitudinal analysis to pinpoint the development of specific SHM in response to specific immunogens.

Combinations of scFv of HIV bNAbs demonstrate high breadth and potency against a multiclade panel of viruses

R.T van Dorsten, P. Moore, L. Morris

NICD, Center for HIV and STIs, Johannesburg, South Africa

Background: Broadly neutralizing antibodies (bNAbs) are currently being assessed in clinical trials for their ability to prevent HIV infection. Single chain variable fragments (scFv) of bNAbs have advantages over full antibodies as their smaller size may permit improved diffusion into mucosal tissues and facilitate vector‐driven gene expression. We have previously shown that scFv of bNAbs individually retain significant breadth and potency. Here we tested combinations of five scFv derived from bNAbs CAP256‐VRC26.25 (V2‐apex), PGT121 (N332‐supersite), 3BNC117 (CD4bs), 8ANC195 (gp120‐gp41 interface) and 10E8v4 (MPER) for their efficacy against multiple HIV subtypes.

Methods: We combined two or three scFv in equimolar amounts in a neutralization assay and tested them against a multiclade panel of 15 viruses. Experimental IC50 and IC80 data were compared to neutralization titres predicted using the Loewe additive model. This model was extrapolated to a 45‐virus panel and optimal combinations were calculated.

Results: Like full‐sized antibodies, combinations of two and three scFv showed significantly improved potency and breadth compared to single scFv. Two scFv combinations generally followed the additive model for breadth and potency, except for CAP256.25+10E8v4 scFv, which showed higher experimental potency compared to predicted titres. Low‐levels of synergy were also observed in 6/10 triple combinations, most of which contained CAP256.25 and/or 10E8v4. No significant antagonism was observed for any of the double or triple combinations. Extrapolation to a 45‐virus panel revealed that at 1μg/mL (IC50 titres) there was 100% coverage for one dual combination (CAP256.25+10E8v4) and three of the triple combinations containing these two scFv. The geometric mean potency of the best triple combination (CAP256.25+ 10E8v4+3BNC117) was significantly more potent than any of the single scFv (0.047μg/mL at IC50 and 0.18μg/mL at IC80).

Conclusions: Combinations of scFv show significantly improved breadth and potency over individual scFv, as previously observed for full‐sized antibodies. The V2‐specific bNAb CAP256.25 and the MPER bNAb 10E8v4, were the most common scFv in the best double and triple combinations and were also more likely to show synergy. Given their size advantage, combinations of scFv show potential for passive immunization.

Active tuberculosis co‐infection enhances HIV‐1 specific humoral immunity

B. Adeoye¹, A.J. Olson², L. Nakiyingi³, Y.C. Manabe⁴, K.R. Jacobson⁵, J. Ellner⁶, M. Sagar⁷

¹Boston University School of Medicine, Microbiology, Boston, United States, ²Boston Medical Center, Infectious Disease, Boston, United States, ³Makerere University College of Health Sciences, Kampala, Uganda, ⁴Johns Hopkins University, Baltimore, United States, ⁵Boston University School of Medicine, Infectious Disease, Boston, United States, ⁶Rutgers Robert Wood Johnson Medical School, Medicine, Piscataway, United States, ⁷Boston Medical Center, Medicine, Boston, United States

Background: Mycobacterium tuberculosis has been shown to enhance antibody responses against diverse viruses, potentially via the impact of lipid antigens or trained immunity. We hypothesized that active tuberculosis (TB) enhances the development of HIV‐1 neutralizing antibodies in HIV‐1 co‐infection.

Methods: We compared anti‐HIV‐1 antibody responses among treatment‐naive plasma samples from 15 HIV‐1 participants with TB (HIV‐1/TB) and 16 HIV‐1 participants without TB. Ability to inhibit 12 different tier 1 and 2 HIV‐1 variants of diverse subtypes in the TZM‐bl neutralization assay was used to estimate a neutralization breadth and potency (BP) score. ELISA was used to compare antibody titers against other latent infections.

Results: HIV‐1/TB and HIV‐1 infected participants had similar baseline plasma virus levels (p=0.33) and CD4 counts (p=0.40). HIV‐1/TB individuals had a significantly higher BP score (0.57±0.05, range 0.32 to 0.97) than HIV‐1 group (0.41±0.05, range 0.24 to 0.57, p=0.02). The plasma activity of 6 HIV‐1/TB individuals with high baseline BP scores clustered with CD4 binding site and membrane‐proximal external region targeting broadly neutralizing antibodies (bnAbs). After completing TB treatment and/or starting HIV‐1 therapy for 6months, HIV‐1/TB (0.64±0.09, n=6, range 0.22 to 0.99) as compared to HIV‐1 participants (0.48±0.09, n=8, range 0.26 to 0.78) still had higher neutralizing capacity, but the difference was not statistically significant (p=0.11). Neutralization BP score did not correlate with the total plasma IgG, baseline viral load, CD4 count, IL‐6, sCD163, and MCP‐1 concentrations. HIV‐1/TB (0.02±0.01) as compared to HIV‐1 (0.03±0.01; p=0.68) group had similar level of HIV‐1 envelope diversity. The HIV‐1/TB and HIV‐1 only participants had similar anti‐tetanus (p=0.44) and anti‐HSV (p=0.25) antibody titers.

Conclusions: Our results suggest that active TB enhances the neutralizing capacity of anti‐HIV‐1 antibodies, possibly leading to the emergence of bnAbs that target conserved envelope domains. TB neither potentiates pre‐existing antibody responses against latent infections nor correlates with other factors previously shown to be important for bnAb emergence. Mechanisms that account for the enhanced HIV‐1 neutralization in HIV‐1 individuals with active TB could be leveraged in the generation of a more effective humoral response in HIV‐1 vaccination and treatment.

Development of a novel VRC01‐class germline targeting immunogen derived from anti‐idiotypic antibodies

E. Seydoux¹, Y.‐H. Wan¹, J. Feng¹, A. Wall¹, S. Aljedani¹, L.J. Homad¹, A.J. MacCamy¹, C. Weidle¹, M.D. Gray¹, L. Brumage¹, J.J. Taylor^1,2,3, M. Pancera¹, L. Stamatatos^1,2, A.T. McGuire^1,2

¹Fred Hutch, VIDD, Seattle, United States, ²University of Washington, Department of Global Health, Seattle, United States, ³University of Washington, Department of Immunology, Seattle, United States

Background: Numerous broadly neutralizing antibodies (bNAbs), which exhibit remarkable breadth and potency in their ability to neutralize HIV‐1, have been isolated from HIV‐1‐infected individuals. VRC01‐class antibodies are among the most broad and potent bNAbs known and they were isolated from at least 10 different donors. They bind the CD4‐binding site on Env and use a unique combination of VH1‐2*02 VH gene paired with light chains expressing rare 5 amino acid long CDRL3 domains. Precursor VRC01‐class bNAbs display no reactivity to recombinant Env, which lead to the development of germline‐targeting immunogens to engage and stimulate naïve VRC01‐class precursor B cells. However, these immunogens also present off‐target epitopes that could hinder the maturation of VRC01‐class bNAbs.

Methods: As an alternative to Env‐derived germline‐targeting immunogens, we developed and characterized a panel of monoclonal anti‐idiotypic antibodies (ai‐mAbs) that can target and potentially activate putative VRC01‐class precursors with high affinity.

Results: In B cell sorting experiments, none of the ai‐mAbs were able to reliably engage VRC01‐class precursor B cells. By integrating ai‐mAb binding, structural and B cell sorting analyses, we engineered a bispecific molecule (iv4/iv9) derived from two ai‐mAbs (iv4 and iv9), from which the iv9 arm is specific for VRC01‐class heavy chains and the iv4 arm for VRC01‐class light chains. Compared to the parental ai‐mAbs, iv4/iv9 could bind ex vivo B cell receptors comprised of a germline VRC01‐class heavy or light chain, but it preferentially crosslinked and activated B cells expressing both VRC01‐class heavy and light chains. Using a murine adoptive transfer model, we observed that when used as an immunogen, the bispecific ai‐mAb was more efficient at engaging and expanding putative VRC01‐class precursor B cells in vivo than either iv4 or iv9 Fabs.

Conclusions: Our results are relevant not only to the development of an HIV‐1 vaccine aimed at eliciting VRC01‐class antibodies, but to general effort to activate specific B cell lineages that produce protective antibodies, and further suggest that ai‐mAbs‐derived immunogens may have general utility as germline targeting immunogens against diverse B cell targets.

BCG.HTI2auxo.int priming vaccination enhances the HIV‐1 specific T cell immune responses elicited by MVA.HTI

N. Saubi¹, A. Kilpeläinen¹, Y. Eto¹, C.‐W. Chen¹, À. Olvera², T. Hanke³, C. Brander², J. Joseph⁴

¹Vall d'Hebron Institut de Recerca, Microbiology, Barcelona, Spain, ²Irsicaixa, AIDS Research Institute, Badalona, Spain, ³The Jenner Institute, Nuffield Department of Medicine, Oxford, United Kingdom, ⁴Hospital Universitari de la Vall d'Hebron, Microbiology, Barcelona, Spain

Background: As part of our contribution to the EAVI2020 consortium research project to get a preventive HIV vaccine, we've constructed recombinant Mycobacterium bovis BCG expressing the EAVI2020 T‐cell immunogens.

Methods: In this study, we constructed an integrative E. coli‐mycobacterial shuttle plasmid, p2auxo.HTI.int, expressing the HIVACAT T‐cell immunogen (HTI). The plasmid was transformed into a lysine auxotrophic BCG strain (BCGΔLys) to generate BCG.HTI^2auxo.int. The DNA sequence coding for the HTI immunogen and HTI protein expression were confirmed and vaccine stocks were genetically and phenotypically characterized. We demonstrated that the vaccine was stable in vitro for. BALB/c mice were immunized with BCG. HTI^2auxo.int prime and MVA.HTI boost was safe and induced HIV‐1‐specific T‐cell responses. T‐cell responses were assessed by IFN‐γ ELISpot upon stimulation with 17 peptide pools spanning the HTI proteome.

Results: The highest total magnitude of IFN‐γ spot forming cells (SFC)/10⁶ splenocytes was observed in BCG.HTI^2auxo.int primed mice compared to mice receiving MVA.HTI alone or mice primed with BCGwt, although the differences between the vaccination regimens only reached trends. In order to evaluate the differences in the breadth of the immune response between mice primed with BCG.HTI^2auxo.int and BCGwt, we examined the number of reactive peptide pools per mouse. Interestingly, both BCG.HTI^2auxo.int and BCGwt primed mice recognized an average of 4 peptide pools per mouse. However, the variation was higher in BCG.HTI^2auxo.int primed mice with one mouse recognizing 12 peptide pools and three mice recognizing few or no peptide pools. On the other hand when comparing the percentage of responding mice per peptide pool, all BCGwt primed mice respond to the more dominant peptides IKProt and 2BRT, whereas the recognition profile appeared to be more spread out for BCG.HTI^2auxo.int primed mice and mice only receiving MVA.HTI, with these groups having a few mice responding to less recognized pools such as 1Dp17, IEp24, 1JProt. Mice were monitored weekly for signs of malaise. No vaccine‐related deaths, no local adverse events, and no associated systemic reactions were observed.

Conclusions: This vaccine candidate may be a useful tool in the development of an effective vaccine platform for priming protective responses against HIV‐1/TB and other prevalent paediatric pathogens.

Anti‐Env antibody‐independent protection of repeated intrarectal low‐dose SIVmac239 challenges in rhesus macaques by vaccination inducing Gag/Vif‐specific CD8+ T but not CD4+ T cells

H. Ishii¹, K. Terahara², T. Nomura¹, T. Tokusumi³, T. Shu³, H. Sakawaki⁴, T. Miura⁴, T. Matano¹

¹National Institute of Infectious Diseases, AIDS Research Center, Tokio, Japan, ²National Institute of Infectious Diseases, Department of Immunology, Tokio, Japan, ³ID Pharma Co., Ltd., Japan, ⁴Institute for Frontier Life and Medical Sciences, Kyoto University, Japan

Background: Virus‐specific CD4⁺ T cells are important for induction of effective CD8⁺ T cells but can be preferential targets for HIV/SIV infection. Recent studies have indicated the detrimental effect of vaccine‐induced CD4⁺ T cells on HIV vaccine efficacy (J Virol 88:14,232, 2014; Sci Transl Med 11:eaav1800, 2019), suggesting supporting a rationale for vaccine design inducing HIV‐specific CD8⁺ T‐cell responses without HIV‐specific CD4⁺ T‐cell induction. Recently, we have developed an novel immunogen, CaV11, consisting of tandemly connected overlapping 11‐mer peptides spanning viral Gag capsid (CA) and Vif. Immunization with DNAs and Sendai virus (SeV) vectors expressing CaV11 efficiently effectively induced CD8⁺ T cells but not CD4⁺ T cells targeting Gag and Vif with SeV‐specific CD4⁺ T‐cell help in rhesus macaques (J Virol 94:e01876‐19, 2020). In the present study, we investigated efficacy of the CaV11‐expressing vaccine against repeated intrarectal low‐dose SIVmac239 challenges in rhesus macaques.

Methods: Twelve rhesus macaques received four times of intramuscular vaccination with CaV11‐expressing DNAs at weeks 0, 1, 3 and 4 and four times of intranasal and intramuscular vaccination with CaV11‐expressing SeV vectors (SeV‐CaV11) at weeks 6, 7, 12 and 18. These twelve vaccinated and seven unvaccinated macaques were intrarectally challenged with low‐dose (200 TCID50) SIVmac239 biweekly starting from 6weeks after the last vaccination.

Results: All the vaccinated animals efficiently induced Gag/Vif‐specific CD8⁺ T‐cell responses with inefficient Gag/Vif‐specific CD4⁺ T‐cell induction after the SeV‐CaV11 vaccination as observed in our previous study. After eight times of SIV challenge, six of the seven unvaccinated macaques were infected, whereas eight of the twelve vaccinated were protected from SIV infection. Kaplan‐Meier analysis indicated a significant difference in the number of challenges for HIV acquisition between the unvaccinated and the vaccinated (p=0.0341 by Log‐rank test). Furthermore, even the SIV‐infected vaccinees showed significantly lower acute viral loads compared to the unvaccinated macaques.

Conclusions: The present study for the first time indicates that canonical CD8⁺ T cells induced by Env‐independent vaccination can protect the establishment of HIV infection after intrarectal virus exposure. Our results suggest that selective induction of virus‐specific CD8⁺ T cells by using this immunogen design is a promising HIV vaccine strategy.

Impact of HLA‐II associated HIV‐1 adaptations on vaccine‐induced CD4 T‐cell immune responses

J. Files, S. Sterrett, T. Fram, N. Erdmann, A. Bansal, P. Goepfert

University of Alabama at Birmingham, Medicine, United States

Background: Our group has previously used HLA‐II associated HIV viral polymorphisms to predict CD4+ T‐cell escape in HIV infected individuals. In both acute and chronic HIV infection, adapted epitopes (AE) encompassing these polymorphisms elicited weaker CD4 immune responses than the corresponding non‐adapted epitopes (NAE). A significant proportion (30% to 40%) of these HLA‐II associated viral polymorphisms are encoded in HIV‐1 vaccine immunogens. However, the impact of these HLA‐II associated adaptations on vaccine immunogenicity has not been previously studied.

Methods: Vaccine recipient and placebo PBMC samples were obtained from three different HIV‐1 vaccine trials. These included MRKAd5 or HVTN 502 (n=24), DNA/rAd5 or HVTN 505 (n=24), and DNA/MVA or HVTN 106 (n=40). Vaccine‐matched epitopes were then grouped together into adapted epitope and non‐adapted epitope pools. Immunogenicity was then tested utilizing a CD8‐depleted IFNg ELISpot assay (n=88) and a more sensitive flow‐based activation induced marker (AIM) assay (n=58) to identify antigen‐specific CD4+ T cells.

Results: In the 88 samples, we identified 10 positive IFNg ELISpot responses to non‐adapted epitopes, while we only identified 1 positive response to adapted epitopes (p=0.013, Fisher's Exact). Out of the 58 samples tested by the AIM assay, we identified a trend towards more non‐adapted epitope responses, with 28 responses to non‐adapted epitopes and 18 responses to adapted epitopes (p=0.087, Fisher's Exact). Overall, there was an increased magnitude of AIM marker expression in response to non‐adapted epitopes (p<0.0001, Wilcoxon). This shows that CD4+ T cells from vaccine recipients produce more IFNg and are more activated in response to vaccine‐matched non‐adapted epitopes. Finally, we also report a positive correlation between non‐adapted epitope‐specific CD4 response magnitude and Env‐specific IgG antibody generation in DNA/rAd5 (p=0.014, Spearman's), suggesting that these CD4 responses may play a role in eliciting HIV‐specific antibodies. Ongoing studies are being performed to determine the phenotype and function of these CD4+ T‐cell responses at the single‐cell level.

Conclusions: Our data demonstrate that non‐adapted epitopes encoded within human HIV‐1 vaccine immunogens are immunogenic. This suggests that modified vaccine inserts should be considered to enhance CD4 responses and that such optimization strategies may have implications for future HIV‐1 vaccine design.

Applying insights from HIV to guide the development of a T‐cell vaccine for SARS‐CoV‐2

A. Nathan¹, G. Gaiha¹, E. Rossin², C. Kaseke¹, R. Tano-Menka¹, B. Walker¹

¹Ragon Institute of MGH, MIT, and Harvard, Ragon Institute of MGH, MIT, and Harvard, Boston, United States, ²Massachusetts Eye and Ear, United States

Background: There is urgent need for a safe and effective vaccine for SARS‐CoV‐2. While most current vaccine candidates aim to induce broadly neutralizing antibodies, recent studies have highlighted the critical importance of CD8+ T‐cells in reducing COVID‐19 severity. We apply structure‐based network analysis to the SARS‐CoV‐2 proteome and a novel HLA class I‐peptide stability assay to define mutationally constrained CD8+ T‐cell epitopes across 18 globally relevant HLA alleles in order to the guide the rational design of an effective SARS‐CoV‐2 T‐cell vaccine.

Methods: Structurally constrained, mutation‐resistant regions within 15 SARS‐CoV‐2 proteins were identified using a structure‐based network analysis algorithm initially developed and validated on HIV. Epitope network scores were calculated on all 16,649 possible 8 to 11‐mers. Only 312 of the 2235 highly networked epitopes were strong HLA class I binders for 18 HLA alleles as predicted by NetMHCPan4.1. To experimentally confirm binding, we employed a novel HLA‐peptide stability assay using CRISPR/Cas9‐edited TAP‐deficient monoallelic HLA class I‐expressing cell lines. Experimentally validated binders were incubated with PBMCs from 23 healthy donors and 25 mild, convalescent COVID‐19 patients prior to assessment of CD137 expression in the CD8+ T‐cell pool as an indicator of epitope‐specific reactivity.

Results: Highly networked regions in SARS‐CoV‐2 demonstrated low sequence entropy across the SARS‐CoV‐2 proteome, SARS‐CoV‐1, MERS, and bat betacoronaviruses. HLA‐peptide stability assays identified 104 of these highly networked epitopes that have >50% relative stability to an immunodominant HIV epitope and provide broad HLA population coverage. Incubation of these peptides with PBMCs from convalescent patients showed significantly higher CD137+ CD8+ T ‐cell reactivity to the structural protein epitopes than did healthy donors, with 72% of patients responding.

Conclusions: Application of structure‐based network analysis and HLA class I‐peptide stability assays identified a set of highly conserved and stabilizing epitopes across 18 globally relevant HLA alleles, setting the stage for rational T‐cell vaccine design for SARS‐CoV2, with broad betacoronavirus coverage.

Spontaneous in vivo control of HIV replication is underpinned by the cross‐clade antiviral potency of HIV‐specific CD8 T cells

J. Makinde¹, N. Fernandez¹, P. Hayes¹, C. Ochsenbauer², J. Dalel¹, J. Hare¹, S. Joseph¹, D. King¹, I.P.C Investigators¹, E. Sanders³, M. Price^4,5, E. Hunter⁶, J. Gilmour¹

¹IAVI Human Immunology Laboratory, Imperial College London, London, United Kingdom, ²University of Alabama, Birmingham, Alabama, United States, ³Kenya Medical Research Institute‐Wellcome Trust Research Programme, Kilifi, Kenya, ⁴IAVI, New York, United States, ⁵University of California at San Francisco, Department of Epidemiology and Biostatistics, San Francisco, California, United States, ⁶Emory Vaccine Center, Atlanta, Georgia, United States

Background: CD8 T cells are critical in the resolution of acute HIV viraemia and long term spontaneous persistent virus control. Obtaining a better understanding of the specificity and potency of the CD8 T cells response during the resolution of acute infection and how if differs between individuals who show exceptional long term in vivo control of viraemia compared to individuals who fail to control, may direct the rational design of T cell based immunogens and better tools to assess clinical vaccine candidates.

Methods: Three groups of individuals were selected from a multisite early infection HIV cohort of 613 volunteers drawn from nine clinical research centres in five African countries. Ten HIV+ volunteers, who rapidly and persistently controlled HIV in vivo were selected, and the breath and anti‐viral potency of their CD8 T cell responses assessed against a broad panel of replication competent transmitted founder HIVs (TFVs). These samples were matched and compared with 2 further groups of 10 volunteers who either failed to control their virus, or had intermediated control. Polyclonally expanded CD8 T cells from PBMC were assessed at three timepoints (2 acute and 1 chronic) post‐infection for CD8 T cell‐mediated viral inhibition of a cross‐clade panel of TFV isolates tagged with Renilla reniformis luciferase. The cross‐clade panel of 10 TFVs selected represents approximately 53.2% of conserved predicted CD8 T cell epitopes within the cohort from which the subjects were drawn.

Results: At all three timepoints, polyclonally expanded CD8 T cells from the controllers showed significantly higher levels of inhibition of viral replication compared to the non‐controllers. Furthermore 50% of controllers at all three timepoints were capable of inhibiting all 10 TFVs in the cross‐clade panel, compared to just 30% of intermediates and 20% of the non‐controllers.

Conclusions: Our results suggest that intrinsic qualities of the CD8 T cell responses that allow for the recognition of more HIV‐1 variants might underpin the spontaneous control of infection in the absence of treatment. Understanding these qualities is the focus of ongoing investigations in the pursuit of the rational design of better immunogens aimed at prevention and cure.

Planning for success: generating an early roadmap with global stakeholders for increasing access and uptake of new biomedical prevention products in resource‐limited settings: The Biomedical Prevention Implementation Collaborative (BioPIC)

A.R. Rinehart¹, C.D. Amole², M. Warren³, M. Czarnogorski⁴, M. Gross⁵, L. Van Damme⁶, I. Mukui⁷, B. Grinsztejn⁸, C. Celum⁹, A. Gomez¹⁰, N. Madidi¹¹, S.Y. Jenkins¹², C. Sozi¹³, J. Reid¹², R. Baggaley¹⁴

¹ViiV Healthcare, Research Triangle Park, United States, ²Clinton Health Access Initiative, Global HIV Access, Washington, United States, ³AVAC, New York, United States, ⁴ViiV Healthcare, Innovation and Implementation Science, Research Triangle Park, United States, ⁵Bill and Melinda Gates Foundation, HIV and Integrated Delivery, Washington, United States, ⁶Bill and Melinda Gates Foundation, Global Health, HIV, Washington, United States, ⁷DNDi, HIV Access and Medical Affairs, Kenya, ⁸Instituto Nacional de Infectologia Evandro Chagas, FIOCRUZ, Rio de Janeiro, Brazil, ⁹University of Washington, Department of Global Health, Seattle, United States, ¹⁰Independent Consultant, United Kingdom, ¹¹Population Services International, Centurion, Zimbabwe, ¹²Clinton Health Access Initiative, Washington, United States, ¹³United Nations, Resident Coordinator, Ethiopia, ¹⁴World Health Organization, Department of HIV and Global Hepatitis Programme, Geneva, Switzerland

Background: Decreasing the time from Phase 3 efficacy results to product introduction and public health impact requires advanced planning in parallel with trials. A global collaborative approach, with insights from prior product introductions from stakeholders engaged early to develop a comprehensive and coordinated strategy, should help to ensure that activities are well‐designed, well‐timed, and appropriately funded. BioPIC tested this model in planning for the investigational long‐acting injectable cabotegravir (CAB LA) for HIV PrEP and developed a product‐adaptable framework to support the introduction of next‐generation biomedical HIV prevention products.

Methods: BioPIC convened a diverse group of 104 global health and HIV prevention experts across more than 80+ organizations and 20 countries (Jul 2018 to present). An Implementation Strategy Committee (ISC) was formed to convene an Advisory Board (AB) and four technical Working Groups (WG), facilitate discussion, and synthesize the output of the WG. Participants included key global stakeholders representing policymakers, Ministries of Health, normative agencies, funders, procurers, implementers, researchers, and advocates. The WG used CAB LA as a case study to identify critical components of product introduction (pre‐ and post‐licensure) across four areas: Guidelines & Policy; Resource Needs & Procurement; Service Delivery; and Additional Research Planning.

Results: From October 2018 the four WG met to identify and prioritize needs/concerns based on previous experience with PrEP introduction. A summary of critical needs was sorted into activities that could be initiated during trials for planning and priority setting, after regulatory approval, and during product scale‐up. The AB endorsed this framework and requested dissemination of the strategy and plans to track and monitor execution against the framework. The CAB LA strategy further informed BioPIC's product‐adaptable framework, which is a foundational roadmap for next‐generation products.

Conclusions: Over 100 global HIV prevention experts contributed to a comprehensive introduction framework for a novel HIV prevention product. This CAB LA strategy informed BioPIC's product‐adaptable framework for pipeline products. Validity and success of the BioPIC model with continued refinement will ultimately be judged by reduced time to public health impact of CAB LA and any next‐generation biomedical prevention product.

Pathways to global access for novel HIV prevention technologies

S. Malhotra¹, N. Sender¹, M. Keane¹, M. Price², A. Kurzman¹

¹IAVI, Global Access, New York, United States, ²IAVI, Epidemiology, New York, United States

Background: Early planning to identify and address potential barriers to access are critical in ensuring that once forthcoming technologies are available, they are widely accessible and deliver impact. This research aims were to identify key considerations and potential strategies for supporting broad availability and uptake of forthcoming long‐acting HIV prevention products by. The goal of this research was to inform efforts to accelerate future access to new HIV biomedical prevention innovations in LICs and LMICs.

Methods: In‐depth interviews were conducted between Q2 Q4 2019 with 45 representatives from 19 leading global health organizations and programs involved in supporting access to global health prevention products. Interviews were used to gather lessons learned from past experiences that could be transferable to new HIV prevention products and identify key challenges and opportunities for global access. Data from qualitative interviews was supplemented by an in‐depth review of the literature on HIV prevention technologies.

Results: Several enabling strategies to accelerate access to promising HIV prevention technologies were identified along different stages of the research, development, and product introduction continuum,

−Establishing platforms to support early information‐sharing and coordination across global health stakeholders

−Improved harmonization of regulatory procedures. Bridge funding and de‐linking pooled procurement from donor funding to ensure countries do not lose the benefits of large volume purchasing as they transition to financial independence.

−Mobilizing innovative collaborations and novel financing mechanisms to catalyze R&D, drive co‐investment, and support affordability.

−Ensuring a robust package of evidence that addresses cost‐effectiveness, programmatic suitability, and epidemiological impact to support timely and widespread adoption.

−Understanding potential acceptability and implementation barriers early in development to ensure they are factored in product design and roll‐out strategies.

Conclusions: While the availability of forthcoming LA‐HIV prevention products is eagerly anticipated, research findings suggest that these products could confront challenges resulting in access delays and inefficiencies, unless addressed. Beginning early in development, coordinated efforts are needed to support early engagement, drive broad availability, ensure affordability, facilitate adoption and support uptake. Concerted and coordinated action will be needed to deliver upon key recommendations articulated in this research.

Advocates’ perspectives on an efficacy trial of F/TAF as PrEP for women

S. Hannah¹, M. Warren², N. Luthuli³, D. Ouya⁴, Y. Raphael⁵, N. Yola⁵, B. Kanyemba⁵, M. Chatani², L. Mworeko⁶

¹AVAC, Research Engagement, New York, United States, ²AVAC, New York, United States, ³AVAC, Durban, South Africa, ⁴AVAC, Nairobi, Kenya, ⁵APHA, Durban, South Africa, ⁶ICW‐EA, Uganda

Background: The October 2019 FDA approval of F/TAF (Descovy) as daily oral PrEP excluded individuals at risk of HIV infection from receptive vaginal sex, and required Gilead, the product's developer, conduct an efficacy trial in cisgender women. Gilead plans to conduct the Descovy for PrEP (D4P) trial in women in sub‐Saharan Africa, using background HIV incidence from previous trials, including ECHO, to estimate efficacy.

Given scrutiny around the clinical development of F/TAF, AVAC, Advocacy for Prevention of HIV/AIDS (APHA) and International Community of Women living with HIV Eastern Africa (ICW‐EA) convened independent civil society consultations to engage with Gilead and inform the design of the D4P protocol.

Methods: APHA and ICW‐EA, in collaboration with AVAC, led engagement with civil society stakeholders in South Africa and Uganda, respectively, through the following activities:

•Series of six consultations with approximately 100 advocates, to build research literacy and discuss advocacy concerns;

•Online survey to gather perspectives and consolidate positions and priorities;

•Direct consultation with Gilead.

All activities were conducted virtually in light of COVID‐19.

Results: The quantitative survey revealed diverse perspectives, especially those related to inclusion of pregnant and breastfeeding women, providing important nuance in discussions with Gilead.

Feedback centered around:

•Safety profile of F/TAF among adolescents and pregnant and breastfeeding women and possibilities of including these population in the trial;

•Size, color and packaging of the pill;

•Standard of prevention, reproductive health package, adherence counselling and social‐behavioural support during the trial;

•Post‐trial access to PrEP;

•Efficacy calculations based on ECHO incidence, given the unknown impacts of COVID‐19 on HIV rates, and that Uganda was not an ECHO country;

•Sustained and diverse stakeholder and community engagement throughout the trial life‐cycle, ensuring local relevance and acceptance;

•Framing F/TAF as an additional, rather than a “better”, PrEP option to TDF/FTC.

Conclusions: Advocates’ input informed the D4P protocol and Gilead's ongoing negotiations with regulatory agencies. Engagement should be sustained to ensure stakeholder support and ownership of the trial, results, and product introduction.

The process provides a model for robust engagement of advocates with product developers around complex, next‐generation prevention trials.

Amplifying youth voices in HIV prevention: lessons learned from a community‐based adolescent health project in Durham, NC

A.C. Maragh‐Bass¹, J.T. Mitchell², N.L. Bhushan³, M. Stoner⁴, L. Riggins⁵, K. LeMasters⁵, M. Walker⁵, S.L. Debnam⁵, A.F. Lightfoot⁵, C. Golin⁵, A.E. Pettifor⁵

¹FHI 360, Division of Behavioral, Epidemiological, and Clinical Sciences, Durham, United States, ²Duke University Medical Center, Durham, United States, ³University of North Carolina, Institute for Global Health and Infectious Diseases, Chapel Hill, United States, ⁴RTI International, Women's Global Health Imperative, Berkeley, United States, ⁵University of North Carolina, Chapel Hill, United States

Background: African American (AA) youth have the highest national rates of HIV. We established a 20‐person working group to develop an agenda for multilevel HIV prevention research in the southeastern United States where HIV incidence is particularly elevated.

Methods: Adolescent Health Working Group (AHWG) stakeholders included local housing authority members and residents, city council, health department, faith‐based groups, and youth from Durham, North Carolina. Meetings solicited perspectives on critical issues related to HIV risk and pre‐exposure prophylaxis (PrEP) access and awareness. The research team processed meeting notes, debriefed challenges, recorded ‘a running list of topics,’ and identified informational needs.

Results: Six meetings were held. Youth (ages 17 to 24years old) felt that the AHWG was “dominated by adult voices,” thus we recruited additional youth, established a suggestions box for anonymous feedback, and youth team members co‐facilitated subsequent meetings. Adults expressed parental concern for PrEP safety and children consenting to PrEP without parental knowledge. Therefore, we provided educational materials about state minor consent laws and HIV/PrEP. In contrast, youth were interested in challenges including daily PrEP adherence. The AHWG recommended that HIV/PrEP needs are part of a larger issue of ‘sexual health education because it doesn't happen in schools.’ Some youth mentioned that the arts may be a form of communication they would appreciate and that ‘they would listen to someone their age who had HIV, not old people.’

Conclusions: Youth‐specific PrEP messages should be youth‐led, avoid terms such as “high risk” or stigmatizing terminology, and encompass larger conversations about stigma, sex, and specific needs of Black youth and LGBT‐identified youth. Our work highlights that:

a)youth investment in research requires using trusted media and information sources;

b)the importance of our additional work to create a protected space led by youth voices in addition to a combined youth/adult group;

c)PrEP awareness must be increased via larger conversations and tailoring messages to youth and their trusted adults.

Future community partnerships should consider a greater number of AHWG meetings to further facilitate genuine relationships and youth leadership building as a part of youth‐focused HIV prevention research.

Associations of climate shocks, HIV, and HIV‐related behaviors amongst women in Mozambique based on the 2015 Demographic Health Survey and Climate Hazards Group InfraRed Precipitation with Station Data

T. Carpino¹, A. Low²

¹Columbia University, ICAP at Columbia University, New York, United States, ²Columbia University, New York, United States

Background: The past three decades, Mozambique had climate shocks including drought, cyclones, and flooding. Low et al (2019) found that severe drought (SD) is associated with higher HIV prevalence in some communities of young women in Africa. Potential theories attribute these associations to alterations of sexual behavior i.e. transactional sex due to a loss of income. This study analyzes the associations of climate shocks, HIV and preventive behaviors in Mozambican women.

Methods: We used the 2015 Mozambique Demographic Health Survey to select a nationally representative sample of women who completed an interview and HIV testing. Deviations in rainfall for May 2014 June 2016 were measured using precipitation data from Climate Hazards Group InfraRed Precipitation with Station Data (CHIRPS), with severe drought (SD) set at<15% of the empirical proportion of rainfall from 1981 to 2016. The associations between climate and HIV‐related outcomes were analyzed using logistic regression, incorporating stratification by age, urbanicity, region, wealth index, and education.

Results: Flooding was strongly associated with HIV status in women under 20 (OR=2.77, 95% CI: 1.34 to 5.69). There was no significant association between flooding and condom use, although there was a significantly greater average partner number for women in areas of flooding. Drought was associated with lower odds of HIV infection in women aged 20 and older. Drought was associated with lower odds of condom use in women under 20 (OR=0.59, 95% CI: 0.38 to 0.90) and greater odds of condom use in women 20 and older (OR=1.79, 95% CI: 1.37 to 2.35). Significant associations were found between drought, partner alcohol consumption, and violence.

Conclusions: The findings of this study do not fully align with the existing literature. Sexual behaviors and prevention methods are differentially associated with the exposures based on age. Further review is needed into the drivers of these disparities and considerations of VLS, recent HIV infection, transactional sex, and an exploration of temporality or reverse causality concerns. We recognize that the environment can affect individual behaviors and health, and with the expected rise of climate disasters globally, we must further investigate these findings to determine how to best intervene.

The evolution of oral PrEP access: tracking trends in global oral PrEP use over time

K. Segal¹, L. Fitch², F. Riaz², J. Rodrigues², M. Warren²

¹AVAC, Product Introduction and Access, Washington, DC, United States, ²AVAC, New York, United States

Background: Since oral pre‐exposure prophylaxis (PrEP) was approved for HIV prevention in 2012, 78 countries have begun offering PrEP in some form. Measuring progress against global targets and sharing strategies across countries are critical to driving impact. Since 2014, AVAC has collected data from PrEP programs to track and generate insights on PrEP implementation.

Methods: The Global PrEP Tracker is a comprehensive database of PrEP projects and national programs. Data are collected through a quarterly survey to 253 programs, and include PrEP initiation numbers, geography, and other indicators. This analysis examined data from Quarter 3 (Q3) 2016 to Q2 2020 to highlight global and regional PrEP initiation trends.

Results: Global PrEP uptake has increased six‐fold in approximately four years, from 102,446 initiations in 2016 to 651,586 in 2020. Annual growth has slowed over time, from 104% from 2017 to 2018, to 55% from 2018 to 2019, to 18% from 2019 to 2020. At the regional level, Oceania has the highest rate of change, with total PrEP initiations increasing from 318 to 29,093, driven largely by Australia. Sub‐Saharan Africa has substantially expanded PrEP access, from 4154 initiations in 2016 to 290,981 by mid‐2020, comprising 44% of the global total. South Africa and Kenya have led this growth, counting 69,876 and 63,000 cumulative initiations, respectively. Brazil is leading PrEP uptake in Latin America and the Caribbean, accounting for two‐thirds of initiations, and Thailand comprises 51% of initiations in Asia. The United States has the most cumulative initiations at 203,837, about one‐third of the global total, but has comparatively modest growth rates. Shared traits in many of these settings are early adoption, national commitment to scale‐up, and programs tailored to populations at high risk offering rights‐based services and linkages to social support.

Conclusions: While PrEP initiations have grown exponentially in several countries, global uptake falls far short of UNAIDS’ target of 3 million users, indicating a need for sustained demand creation where PrEP programs exist, and scale‐up where PrEP is provided via demonstration projects with limited reach. Countries should replicate successful approaches, including service delivery models that meet end users’ needs, to achieve notable reductions in new HIV infections.

A multi‐country investigation of pre‐exposure prophylaxis preferences among young people at risk of HIV in sub‐Saharan Africa

J. Dietrich¹, N. Ahmed², S. Nash³, G. Tshabalala¹, T. Nematadzira⁴, S. Hornschuh¹, M. Atujuna², M. Mulaudzi¹, R. Muhumuza⁵, A. Ssemata⁵, A. Kakande⁵, L. Stranix-Chibanda⁶, N. Martinson¹, L.‐G. Bekker², H.A. Weiss³

¹Perinatal HIV Research Unit, Faculty of Health Sciences, University of the Witwatersrand, Wits Health Consortium, Johannesburg, South Africa, ²Desmond Tutu HIV Foundation, University of Cape Town, Desmond Tutu HIV Foundation, Cape Town, South Africa, ³MRC Tropical Epidemiology Group, Department of Infectious Disease Epidemiology, London School of Hygiene & Tropical Medicine, Infectious Disease Epidemiology, London, United Kingdom, ⁴University of Zimbabwe, College of Health Sciences, Clinical Trials Research Centre, Clinical Trials Research Centre, Harare, Zimbabwe, ⁵Medical Research Council/ Uganda Virus Research Institute and London School of Hygiene & Tropical Medicine Uganda Research Unit, London School of Hygiene & Tropical Medicine Uganda Research Unit, Entebbe, Uganda, ⁶University of Zimbabwe, College of Health Sciences, Department of Paediatrics, Harare, Zimbabwe

Background: Daily Pre‐Exposure Prophylaxis (PrEP) is highly effective when adhered to, but its personal and public health benefits it limited by cost, user‐acceptability and uptake. The Combined HIV Adolescent PrEP and Prevention Study (CHAPS) is a three‐country study to investigate the acceptability of different PrEP regimens amongst young people in sub‐Saharan Africa. We describe preferences for daily and on‐demand PrEP in Cape Town and Johannesburg (South Africa), Entebbe (Uganda) and Harare (Zimbabwe).

Methods: Trained interviewers conducted cross‐sectional, structured online surveys capturing socio‐demographics, HIV risk behaviours, preferences for daily and on‐demand PrEP (three dosing options; (1) two pills before and after sex, (2) two pills a few hours before sex, (3) two pills a few hours after sex). Data were analysed by site and gender using STATA.

Results: Of 1339 participants aged 13 to 24years enrolled, 49.7% were female, with most (70.7%) aged 18 to 24years. The majority (78.9%) had sex, 83.3% a current partner, and 38.3% knew to have sex two hours in advance. Overall, 22.2% reported first sex aged 14years or younger and 50.8% weekly sexual activity. About half (51.4%) reported using a condom at last sex (55.6% females; 47.6% males). Twenty‐five percent had ever heard of PrEP before the study (11.8% in Entebbe) and willingness to use PrEP was 95.8%. Overall, on‐demand PrEP was the preferred option (59.5%) except for Cape Town, where 68.2% preferred daily PrEP. Males (65.2% vs 53.8%) preferred on‐demand while females (46.3% vs 34.8%) preferred daily use. Despite an overall preference (59.5%) across sites for on‐demand PrEP versus daily PrEP, more participants (36.8%) reported that taking daily PrEP would be the easiest compared to the on‐demand regimen options (26.2% (1), 27.5% (2) and 9.5% (3)).

Conclusions: Awareness of PrEP was low, despite high willingness to use PrEP. PrEP demand creation needs to be reviewed, optimised and tailored to socio‐demographic differences and designed in conjunction with young people. Additionally, although overall more participants in our sample preferred on‐demand options, their sexual behaviour patterns may not support its use (e.g. sex events are unplanned for many). Strategies in sexual health education and promotion particularly around assessing and understanding sexual risk behaviour needs further exploration.

Surveillance data from public and private primary care facilities uncover implementation successes and gaps during pre‐exposure prophylaxis (PrEP) scale‐up: Results from the Jilinde project in Kenya

A. Musau¹, D. Were¹, J. Mutegi¹, P. Ongwen¹, B. Wakhutu¹, J. Reed²

¹Jhpiego, Jilinde Project, Nairobi, Kenya, ²Jhpiego, Baltimore, United States

Background: Government and private sector facilities constitute the majority of health facilities in Kenya. Primary care units in these facilities have supported a vibrant HIV program and yield enormous potential to accelerate PrEP scale‐up reaching diverse populations. PrEP is nascent in Kenya and lessons garnered through various implementation platforms are germane. This study employs a cascade approach to elucidate successes and gaps in PrEP delivery through primary care units in Kenya.

Methods: The Jilinde project supports PrEP scale‐up in ten Kenyan counties through 50 public and 12 private facilities. Routinely integrated PrEP is offered by trained providers who collect and aggregate service data using nationally approved tools, which is routinely analyzed to inform programmatic changes. We analyzed these data using a prevention cascade approach to unearth implementation gaps comparing trends in proportions of clients receiving services across the PrEP continuum: behavioral risk screening; clinical eligibility assessment; receipt of first PrEP prescription; and, receipt of refill prescriptions.

Results: Between May 2017 and October 2019, 334,068 individuals tested HIV‐negative, of which (93,429) 28% received behavioral risk screening. Among these, 16,673 (18%) were clinically assessed and deemed eligible. From the pool of eligible individuals, 11,087 (66%) initiated, consisting of 55% HIV‐negative individuals in serodiscordant relationships, 25% high‐risk general population individuals, 14% female sex workers, and 2% men who have sex with men. After a month, 3610 (34%) clients returned for a PrEP refill. Rates of attrition did not vary by population group or public vs. private facilities. The proportion of clients receiving month‐one refills in private facilities progressively increased from 25% in 2017 to 36% in 2018, to 44% in 2019 unlike in public facilities. Overall, receiving a month‐three refill, upon receiving a month‐one refill, increased from 26% in 2017 to 52% in 2018 and 69% in 2019.

Conclusions: Overall, 72%, 34%, and 62% of HIV‐negative clients in primary care settings did not receive risk screening, first prescription, or one‐month refill respectively, culminating in 322,981 individuals who might have benefited from PrEP that didn't receive it. Missed opportunities underscore the need for continuous surveillance to identify and respond to implementation challenges and opportunities to enhance PrEP's impact.

Integrating STI screening into PrEP services for adolescent girls and young women (AGYW) in two primary health care (PHC) facilities in Johannesburg: lessons from Prevention Options for Women Evaluation Research (POWER)

D. Travill¹, M. Ndlovu², L. Kidoguchi³, T. Tlou², L. Lunika², J. Morton³, R. Johnson³, J. Baeten³, C. Celum³, S. Delany-Moretlwe²

¹Wits RHI, Research, Johannesburg, South Africa, ²Wits RHI, Johannesburg, South Africa, ³University of Washington, Seattle, United States

Background: High rates of STIs have been demonstrated in AGYW initiating PrEP. Most infections are asymptomatic and missed by standard‐of‐care STI syndromic management. We assessed the prevalence and factors associated with return for STI treatment following urine STI screening during the POWER study in Johannesburg, South Africa.

Methods: POWER aimed to demonstrate models of PrEP delivery to AGYW. We enrolled a cohort of HIV negative AGYW aged 18 to 25 at a dedicated adolescent PHC clinic or a conventional PHC clinic. Urine from participants was tested for chlamydia (CT) and gonorrhoea (GC) by GeneXpert at enrolment. Participants with positive results were contacted by phone or WhatsApp (at least 3 attempts) and offered treatment. We assessed associations between baseline characteristics and treatment return using logistic regression.

Results: Of 776 screened, 763 (98%) were enrolled and 737 (96%) initiated PrEP. Median age was 21, 11% reported >1 partner in past 3months, 18% reported consistent condom use, and 3% reported genital symptoms at enrollment. 622/763 (82%) were screened for STI, and 211 (34%) had any curable STI, 27% CT, 3% GC, and 3% both CT/GC at enrollment. Of those with any STI, 136 (65%) were successfully contacted and treated.

Participants were more likely to return for treatment if they had reported consistent condom usage (OR 4.9, 95% CI 1.5 to 16.1) compared to those that never used condoms, or reported using any form of contraception at enrolment compared to those that did not (OR 2.04; 95% 1.33 to 3.66). Despite apparent lower STI prevalence at the dedicated adolescent clinic (29% vs 37%), participants were more than three times likely to return for STI treatment as compared to the conventional PHC (OR 3.08, 95% 1.66 to 5.73).

Conclusions: Prevalence of asymptomatic curable STIs is high in this PrEP initiating population. A third of participants did not return for treatment highlighting the need for point of care diagnostics and novel STI control interventions especially for those at highest risk for STIs and HIV, to interrupt transmission and prevent complications. More research is needed to determine reasons for low uptake of STI treatment. Quality, integrated, adolescent‐responsive health services are an essential component of AGYW STI/HIV prevention.

Transient changes in daily sexual behavior and pre‐exposure prophylaxis use after the implementation of COVID‐19 restrictions among men who have sex with men

V. W. Jongen¹, H. M. L. Zimmermann¹, A. Boyd^1,2, E. Hoornenborg¹, M.A.M. van den Elshout¹, U. Davidovich^1,3, Y.T.H.P. van Duijnhoven¹, H.J.C. de Vries^1,4, M. Prins^1,5, M.F. Schim van der Loeff^1,5, L. Coyer¹, o.b.o.t. Amsterdam PrEP Project team in the HIV Transmission Elimination Amsterdam Initiative⁶

¹Public Health Service of Amsterdam, Department of Infectious Diseases, Amsterdam, Netherlands, ²HIV Monitoring Foundation, Amsterdam, Netherlands, ³University of Amsterdam, Department of Social Psychology, Amsterdam, Netherlands, ⁴Amsterdam UMC, University of Amsterdam, Department of Dermatology, Amsterdam Infection and Immunity (AII), Amsterdam, Netherlands, ⁵Amsterdam UMC, University of Amsterdam, Department of Infectious Diseases, Amsterdam Infection and Immunity (AII), Amsterdam, Netherlands, ⁶H‐TEAM, Amsterdam, Netherlands

Background: We assessed how the Dutch restrictions of 15 March 2020 affected sexual behavior, pre‐exposure prophylaxis (PrEP)‐ and condom‐use among PrEP‐users from the Amsterdam PrEP demonstration project.

Methods: We collected daily data on

(1)PrEP‐use,

(2)anal sex acts and

(3)condom‐use, per partner type (steady [SP], known casual [KCP], unknown casual [UCP]), from a mobile application used between 1 December 2019 and 30 June 2020. We compared the period before versus after 15 March 2020 with respect to average proportion of days per week at which each endpoint was reported (multilevel logistic regression) and average proportion of anal sex acts covered by PrEP and/or condom (bivariate probit regression). We evaluated whether the difference in sex acts could also be observed before versus after the same date in the previous year.

Results: We included data from 136 MSM. After 15 March 2020, the proportion of days with anal sex increased with SP (OR=1.25; 95% CI=1.09 to 1.44) and decreased with KCP (OR=0.73; 95% CI=0.64 to 0.82) and UCP (OR=0.54; 95% CI=0.48 to 0.61), while these changes were not (SP/KCP) or less apparent (UCP) during the previous year (Figure 1). Shifts in partner types were most profound immediately after 15 March 2020, and returned to pre‐restriction levels mid‐May 2020. The proportion of days with PrEP‐use decreased from 74% before to 58% after 15 March 2020 (p<0.001). After 15 March 2020, PrEP coverage of sex acts decreased with UCP (β=−0.36; 95% CI=−0.72 to 0.00), but not with SP and KCP; condom use during sex acts decreased with KCP (β=−0.36; 95% CI=−0.67 to 0.04) and UCP (β=−0.24; 95% CI=−0.46 to 0.03), but not with SP; and use of PrEP and condoms during sex acts decreased with KCP (p=0.026) and UCP (p=0.011), but not with SP.

Conclusions: MSM decreased sex acts with casual partners and increased sex acts with SP, but changes were transient. Decreases in sex with casual partners paralleled decreases in PrEP‐use. However, condom‐use during sex with casual partners decreased, indicating the importance of continued sexual health services, including STI screening and PrEP care, during COVID‐19 restrictions.

Fine‐tuning of in vitro transcribed mRNA for optimising a vaccine platform against HIV‐1

S. Linares Fernandez, J. Moreno, B. Verrier, J.‐Y. Exposito

CNRS and University Claude Bernard Lyon 1, Laboratoire de Biologie Tissulaire et Ingénierie Thérapeutique, Lyon, France

Background: The main hypothesis of this study is that changes in mRNA sequences, coupled to production and purification shortcuts can drastically impact the mRNA translation.

Methods: Endpoint: Measurement of mRNA expression (eGFP) in DC2.4 and HeLa by FACS.

Reference mRNA Sequence:5′UTR of Human β‐Globin, eGFP, two copies in tandem of 3′UTR human of β‐Globin and 148 Poly‐A. Poly‐A: Three poly‐A sizes were tested (76A, 112A and 148A) and evaluated overtime (4, 24, 48, 72 and 144hours). 5′UTR: 6 different synthetic 5′UTR were tested and evaluated overtime (4, 24, 48 and 96hours).

Results: Poly‐A: Our data suggest the 148 poly‐A tail was better than the poly‐A 76 and the poly‐A 112.

5′UTRs: The synthetic 5′UTR‐(4) had an increase of 37% compared to the human β‐Globin mRNA Production: Vaccinia capping showed an increase >75%, compared to ARCA. mRNA purification: The depletion of dsRNA increased the translation (Arca 15%, Vaccinia 35%). Double purification had the greatest impact on (approximately +80% in ARCA and Vaccinia). Global Results.

Our optimisation results showed an increase of up to 280% with non‐optimized mRNAs.

Conclusions: We demonstrated that variations in untranslated sequence, production and purification methods had an impact in mRNA translation efficiency.

Comparison of co‐immunization of DNA and protein in the same anatomical sites and in contralateral sites to identify mechanisms of protective immune response

B. Felber¹, Z. Lu¹, X. Hu¹, A. Valentin¹, M. Rosati¹, J.A. Weiner², X. Shen³, G.D. Tomaras³, C. LaBranche³, D.C. Montefiori³, W.B. Williams³, K.O. Saunders³, S.G. Reed⁴, D.J. Venzon⁵, G.M. Shaw⁶

¹National Cancer Institute at Frederick, Center for Cancer Research, Frederick, United States, ²Thayer School of Engineering, Dartmouth College, Hanover, United States, ³Duke University Medical Center, Durham, United States, ⁴IDRI, Seattle, United States, ⁵National Cancer Institute, CCR, Rockville, United States, ⁶University of Pennsylvania, Philadelphia, United States

Background: We hypothesized that simultaneous recognition of immunogens delivered as DNA and protein by the draining lymph node may impact the quality of the vaccine‐induced immune responses. To address this hypothesis, we compared the protective efficacy of an HIV vaccine comprised of DNA (env and gag) and Env proteins by co‐administration of both components in the same muscle or by delivering the two vaccine components (DNA and Protein) separated in contralateral sites.

Methods: Female rhesus macaques (20 animals/group) were immunized with a 6‐valent vaccine including DNA plasmids expressing membrane‐anchored gp145 Env sequentially isolated from a HIV‐1 infected individual (CH505). The DNA was delivered by IM injection followed by in vivo electroporation. The vaccine also included a gp120 Env protein component adjuvanted in GLA‐SE matching the sequences encoded by the plasmid DNA. The DNA and protein vaccine components were administered in the same anatomical sites (‘Co‐administration group’) or in contralateral sites (‘Separate Administration group’). After vaccination, the macaques were challenged by weekly intravaginal exposures with low dose T/F tier‐2 SHIV CH505 stock. Detailed flow and proteomics/transcriptomics analysis was performed.

Results: Only the co‐administration vaccine group was protected against SHIV CH505 acquisition, with a 67% risk reduction per exposure after 15 weekly intravaginal challenges. Macaques in the co‐administration group developed higher Env‐specific humoral and cellular immune responses. Non‐neutralizing anti‐Env antibodies, ADCC and anti‐Env antibodies with strong binding affinity to Fc‐gamma Receptor IIIa were associated with decreased transmission risk. Analysis of vaccine‐induced and innate immune responses and their association with mechanisms contributing to protection will be discussed. Follow‐up boosting of the protected animals showed that the immune response can be further enhanced and the antibody specificity can be broadened.

Conclusions: Co‐immunization of DNA+Protein in the same muscle induces rapid and effective immunity able to protect from tier‐2 SHIV challenge. These data suggest that simultaneous recognition, processing and presentation of DNA+ Env protein in the same draining lymph node plays a critical role in the development of protective immunity. This vaccine regimen, using simultaneously DNA and protein administration, could also be beneficial in protecting against other pathogens.

Tetravalent immunogen assembled from conserved regions of HIV‐1 and delivered as mRNA demonstrates potent preclinical T cell immunogenicity and breadth

N.A. Moyo¹, E. Wee¹, B. Korber², K. Bahl³, S. Falcone³, S. Himansu³, A.L. Wong⁴, A.K. Dey⁴, M. Feinberg⁴, T. Hanke¹

¹University of Oxford, The Jenner Vaccine Research Institute, Oxford, United Kingdom, ²Los Alamos National Laboratory, Theoretical Biology and Biophysics, United States, ³Moderna Inc, Cambridge, United States, ⁴International AIDS Vaccine Initiative, United States

Background: A vaccine will likely be one of the key tools for ending the HIV‐1/AIDS epidemic by preventing HIV‐1 spread within uninfected populations and achieving a cure for people living with HIV‐1. The currently prevailing view of the vaccine field is to introduce protective antibodies, nevertheless, an effective vaccine may need to harness protective T cells. We postulated that focusing a T‐cell response on the most vulnerable regions of the HIV‐1 proteome while maximizing a perfect match between the vaccine and circulating viruses will control HIV‐1 replication. Here, we exploit the mRNA lipid nanoparticle (LNP) platform by designing tetravalent immunogens to assess induction of T‐cell responses in a pre‐clinical model.

Methods: Two Gag and four Pol protein regions of the HIV‐1 proteome were selected for their high conservation among the HIV‐1 Group M isolates and inclusion of beneficial regions. These were computed into two complementing mosaics, which together achieved over 80% match of potential 9‐mer T‐cell epitopes over most of the six HIVconsvX regions. Here, using a novel epigraph algorithm, two versions of the same six regions were sequentially designed to cover additional common variants of potential 9‐mer epitopes and make further improvements over the previous vaccine in matching the global HIV‐1 isolates. These two mosaics and 2 epigraphs were converted into mRNA and the fully synthetic tetravalent mRNAs, collectively called HIVconsvM, were encapsulated into LNPs.

Results: In a mouse model, intramuscular injection of the LNP‐encapsulated mRNA induced high frequencies of T cells, which recognized seven out of seven examined H‐2^d class I‐restricted epitopes in most animals. This combination of mRNA and tetravalent immunogens induced potent, broad and polyfunctional T‐cell responses. There was an overall trend of a direct correlation between the magnitude of vaccine responses and valency of the vaccine cocktail both for individual epitopes and the overall ‘global’ response.

Conclusions: Our data show a great potency of this vaccine platform for induction of T‐cell responses and support evaluation of the HIVconsvM mRNA LNPs in humans, the species in which HIV‐1 evolved and for which further vaccine optimizations and response analyses will be the most relevant.

A VSV‐based HIV‐1 vaccine provides protection in macaques against low dose cross‐clade SHIVenv_SF162_P3 challenge

E. Arts¹, A. Berger², J. Pedersen², M.‐A. Lafrance², J. Knapp¹, H. Azizi², Y. Li¹, F. Scholte², J. Mann¹, A. Kamen³, K. Fowke⁴, X. Yao⁴, C.‐Y. Kang¹, E. Cohen⁵, G. Kobinger²

¹Western University, London, Canada, ²Universite Laval, Quebec, Canada, ³McGill University, Montreal, Canada, ⁴University of Manitoba, Winnipeg, Canada, ⁵IRCM, Montreal, Canada

Background: We developed a series of 30+ VSV‐based constructs containing HIV‐1 Env chimeras with or without SIV gag in absence of VSV‐G and with or without the Ebola glycoprotein. After pre‐clinical analyses of production/expression and small animal testing, a subset of 4 were used to immunize macaques that were later challenge with low dose SHIV.

Methods: VSVdelG‐based vectors were produced to express codon optimized HIV‐1 ecto Env of NL4‐3 or subtype A, strain 74 Env with N425K Env with different transmembrane and intracellular tails (TMIC) of the VSV G, Ebola GP, mutated HIV‐1 and SIVmac239 Env. VSVdelG replication was driven either by the functional Env chimera or by cis addition of Ebola or Marburg GP. Expression and Env function for entry was tested and VSV vectors were produced for mouse immunizations to measure humoral (binding and neutralizing Abs) and CTL responses. The best candidates were then tested in rabbits and macaque challenge studies.

Results: Details of the three year preclinical screening and subsequent correlates of protection for the best vaccine candidates will be provided in the presentation. The best candidate, VSVdelG_A74‐gp140/SIV‐TMIC_Ebola‐GP was used as the prime followed by a boost with VSVdelG_SIV‐Gag + A74‐gp140/SIV‐TMIC_Ebola‐GP and then a boost with VSVdelG_A74‐gp140/SIV‐TMIC_Ebola_GP (Grp1). Eight of ten of the control macaques (Grp2) and 10 of 10 in an alternative vaccine regimen (Grp2) were infected with SHIV_SF162‐p3 by the final low dose 7 challenges. Four of ten animals of the best performing vaccinated group were infected by challenge 7th.

Conclusions: Building on the VSV‐Ebola vaccine technology, we engineered/tested an improved VSV‐based HIV vaccine and observed one of the best protections against low dose heterologous challenges in macaques to date. Findings also suggest that use of VSV‐HIV_Env construct with Gag may be important for protection, especially in boosting CTL response, but its inclusion in the final boost (groups 2 and 4) may promote an activated CD4+ T cell population capable of increasing SHIV infection. These findings suggest an important balance of cell‐based versus humoral immunity in prime/boost vaccine strategies for optimal protection. The current study is one of the first reports of meaningful cross‐clade heterologous protection.

Improved in vitro expression and in vivo immunogenicity of a candidate MVA‐vectored HIV‐1 vaccine compared to SAAVI MVA‐C

N. Douglass¹, M. Van Diepen², R. Chapman², S. Galant², E. Margolin², P. Ximba², T. Hermanus³, P. Moore³, A.‐L. Williamson²

¹University of Cape Town, Pathology, South Africa, ²University of Cape Town, South Africa, ³University of Witwatersrand, Johannesburg, South Africa

Background: SAAVI MVA‐C was previously developed in South Africa and tested in Phase 1 clinical trials. Following more recent advances in antigen design, modifications were made to the HIV‐1 genes expressed by modified vaccinia Ankara (MVA) and a head‐to‐head comparison was performed using the SAAVI MVA‐C and newly constructed MVAGD5 vaccines.

Methods: MVAGD5 was constructed with an HIV‐1 subtype C mosaic gag inserted in the A11R‐A12L locus and a modified form of the Du151 env gene inserted between I8R and G1L of MVA, both under vaccinia virus mH5 promoter control. The env gene was modified as follows: replacement of the native leader sequence with that of the human tissue plasminogen activator, replacement of the furin cleavage site with a flexible linker (2x GGGGS), inclusion of an I559P mutation and truncation of gp160 to gp150.

The two vaccines were compared for HIV‐1 gene expression in HEK293T cells, by western blot analysis and immunofluorescence.

Immunogenicity was tested in rabbits inoculated intramuscularly with 108pfu recombinant MVA at weeks 0 and 4; and 40µg soluble trimeric Du151 protein at week 12. Rabbits were bled at weeks 0, 4, 8, 12, 14 and 16 and serum was tested for binding antibodies by ELISA and neutralising antibodies using the TZM‐bl pseudovirus assay.

Results: Compared to SAAVI MVA‐C the modified MVA vaccine, MVAGD5, showed increased expression of both Gag and Env in HEK293T cells. These two proteins were detected in the clarified medium from MVAGD5‐infected cells, but not SAAVI MVA‐C‐infected cells. Co‐localisation of Gag/Grttn and Env was observed by fluorescence microscopy in infected HEK293T cells for both MVAGD5/SAAVI MVA‐C vaccines. MVAGD5 elicited binding antibodies after the first inoculation, whereas SAAVI MVA‐C only elicited binding antibodies after the protein boost. MVAGD5 elicited neutralising antibodies to Tier 1A and Tier 1B pseudovirions whereas SAAVI MVA‐C generated no significant neutralising response. Neither vaccine elicited autologous Tier 2 neutralising antibodies.

Conclusions: Modifications made to both gag and env genes in the construction of SHIP MVAGD5, compared to the historic SAAVI MVA‐C vaccine, resulted in superior in vitro expression of both HIV‐1 genes and improved immunogenicity in rabbits.

Implementing a rapid response to the COVID‐19 global pandemic in MTN‐034/REACH: an HIV prevention trial among adolescent girls and young women in Africa

T. McClure¹, J. Davis², M. Garcia², K.R. Eddy³, T. Palanee‐Phillips³, B. Bam⁴, G. Nair⁴, S. Kassim⁴, C. Agwau Akello⁵, P. Ndadziyira⁶, K. Ngure⁷, C. Celum⁸, L. Soto-Torres⁹, A. van der Straten¹⁰

¹FHI 360, Science Facilitation, Durham, United States, ²FHI 360, Durham, United States, ³Wits Reproductive Health and HIV Institute, University of the Witwatersrand, Johannesburg, South Africa, ⁴Emavundleni Research Centre, Desmond Tutu HIV Foundation, University of Cape Town, Cape Town, South Africa, ⁵Makerere University‐John Hopkins University Research Collaboration, Kampala, Uganda, ⁶University of Zimbabwe College of Health Sciences Clinical Trials Research Centre, Harare, Zimbabwe, ⁷Jomo Kenyatta University of Agriculture and Technology, College of Health Sciences, Kenya, ⁸University of Washington, Department of Medicine, Seattle, United States, ⁹National Institute of Allergy and Infectious Diseases, Department of AIDS, United States, ¹⁰Women's Global Health Imperative, RTI, Research Triangle Park, United States

Background: The MTN‐034/REACH trial, evaluating safety and preferences of oral pre‐exposure prophylaxis (PrEP) and the dapivirine vaginal ring among adolescent girls and young women (AGYW) in sub‐Saharan Africa, was in full implementation when the COVID‐19 pandemic erupted. Associated socio‐economic upheaval and mobility disruptions impacted the 247 enrolled AGYW, and staff, potentially jeopardizing study quality and development of HIV prevention methods. In response, accrual was paused and rapid modifications to follow‐up procedures were implemented to distribute study product, monitor safety, handle biological samples and testing, and support product adherence.

Methods: As sites prepared for impending lockdowns, guidance was issued regarding safety and confidentiality precautions, mitigating social harms (e.g. counseling/referrals, reimbursements for food security) and COVID‐19 education for participants and staff. A contingency plan for study conduct, tiered by operational status, was developed and implemented according to site capacity. All modifications, including telephonic visits in lieu of clinic visits, providing extra study product, and prioritizing in‐person procedures, were monitored in real‐time for operational and analysis considerations. Community engagement included study updates to local stakeholders and delivery of COVID‐19 key messages for community sensitization.

Results: Implementation modifications during country lockdowns resulted in nearly uninterrupted study product and contraceptive access for participants. Clinic access remained critical for participants without options for secure telephonic visits or with onsite needs, therefore sites operated in reduced capacity, with staff rotations and transport for staff and participants. Counselors expanded adherence support with remote reminders and check‐ins, using alternative documentation and confidentiality practices. Completion rates for expected study visits and procedures were high. Sites maintained contact with most participants despite lockdowns (Table 1).

Conclusions: Well‐orchestrated modifications to study implementation ensured participant and staff safety, and maintained study integrity and community engagement during COVID‐related disruptions. The MTN‐034/REACH strategy serves as a template for optimal study operations during times of upheaval, especially when supporting AGYW.

Social, economic, mental health and medical care impacts of COVID‐19 in a US cohort of sex and gender diverse adolescents and young adults at‐risk for HIV

P. Serrano¹, J. Diaz¹, A. Muñoz¹, S. Hosek², A. French²

¹Ruth M. Rothstein CORE Center, Research, Chicago, United States, ²John H. Stroger, Jr., Hospital of Cook County, Infectious Disease, Chicago, United States

Background: The United States’ national response to the COVID‐19 pandemic disrupted schools, work places, and healthcare. As the shutdown continued, the American Public Health Association warned of developing mental health crises across the US. In this study we aim to understand the scope of the pandemic on the emotional and financial well‐being as well as access to routine HIV/STI and pre‐exposure prophylaxis (PrEP) among sexual and gender minority youth and young adults.

Methods: Participants in the Keeping it LITE virtual cohort study were surveyed in May of 2020, and were administered an online survey asking them to reflect on their COVID‐19 related experiences since March of 2020. The survey collected demographic data and questions on the impacts of COVID‐19. Those enrolled in the parent study were ages 13 to 34, cisgender men or transgender/gender nonconforming individuals, who have behavioral risk factors that make them vulnerable to HIV. Descriptive statistics were collected and reported via Qualtrics.

Results: 2116 participants, or over half of the cohort responded (61.5%). Mean age was 26.21 (SD=4.84), a majority were male (81%), White (55%), and gay (69%). Many participants reported decreased wages (33.33%) or losing their jobs (27.41%). A majority of participants reported negative impacts on mental health (80.61%), including increased anxiety (58.3%) and depression (42.93%). A significant minority of PrEP users (42.3%) reported changing or stopping PrEP during the pandemic, due to disrupted PrEP follow‐up care (43.8%), while 20% reported difficulty accessing HIV/STI testing during the pandemic.

Conclusions: Most of the study participants experienced disruptions to their daily lives in the areas of employment and healthcare. During this relatively short time period, these vulnerable youth reported an alarming rise in mental health problems, financial distress, and difficulties in accessing HIV preventive and STI care. Longitudinal data are needed to quantify the full impact of the pandemic on the mental health and prevention behaviors of sexual and gender minority youth and young adults.

Acceptability of telemedicine and HIV self‐test among PrEP users during the COVID‐19 pandemic in Brazil

B. Hoagland¹, T.S. Torres¹, D.R.B. Bezerra¹, A. Garner², K. Geraldo¹, L. Freitas¹, C. Jalil¹, E. Carvalheira¹, A. Nabor¹, E. Bastos¹, D. Waite¹, J.R. Grangeiro¹, T. Santos¹, L. Monteiro¹, J. Freitas¹

¹Instituto Nacional de Infectologia Evandro Chagas INI/Fiocruz, Lapclin‐AIDS, Rio de Janeiro, Brazil, ²Hornet INC, United States

Background: In March 2020 telemedicine and HIV self‐testing (HIVST) were adopted by some Brazilian Public Health services to minimize disruptions in pre‐exposure prophylaxis (PrEP) access and delivery during social distancing and community containment recommendations related to the COVID‐19 pandemic.

Methods: During April‐May/2020, Brazilian cisgender men who have sex with men (MSM) and transgender/non‐binary people were recruited to complete a web‐based survey advertised on Hornet, Facebook and WhatsApp. Data from MSM and transgender women (TGW) aged≥18years receiving PrEP through a Public Health Service or the ImPrEP demonstration study were analyzed. The impact of social distance in their lives, acceptability of telemedicine and HIVST procedures were assessed.

Results: Of 3536 who completed the questionnaire, 2677 MSM/TGW self‐reported HIV negative/unknown status, 705 were PrEP users and 654 reported receiving PrEP from a Public Health Service (50.9%; 333/654) or the ImPrEP demonstration study (49.1%; 321/654) were included in this analysis. The median age was 35.1years (IQR:25 to 35), most were cisgender men (641;98.0%). For 66.5% (n=435) social distance had a high impact in their lives, mainly in the economics (252;38.5%) and affective/sexual (217;33.2%) aspects. Despite recommendations for social distancing, 44.5% (n=291) reported sex with casual partners; 54.3% (158/291) had condomless receptive anal sex with them. During this period, 72.6% (n=475) maintained daily PrEP use due to fear of getting HIV (284;59.8%), sex with casual partners (84;17.7%), HIV‐positive partner (64;13.5%) and belief that PrEP protected against COVID‐19 (43;9.1%). PrEP teleconsultation was experienced by 20.9% (n=137) and 89.1% (n=122) felt satisfied with this procedure. For those who did not experience a teleconsultation, 69.6% (n=360) reported being very comfortable to try it. Awareness about HIVST was high (488; 74,6%) but only 26.6% (n=174) had ever used it. Among those who never used, willingness to use was 78.3% (n=376). For 87.2% (n=570) home delivery of PrEP refills would be acceptable.

Conclusions: Telemedicine procedures for PrEP delivery including HIVST showed to be highly acceptable among PrEP users as well as home delivery PrEP refills. These results point out that such technologies could be an option to be implemented by Public Health Services in Brazil to avoid PrEP access shortage during the COVID‐19 pandemic.

The COVID‐19 pandemic and transition to telehealth: appointment attendance and patient perspectives at an adult HIV clinic

I. Auchus, K. Jaradeh, A. Tang, B. Boslett, J. Marzan

University of California, 360 Wellness Center, San Francisco, United States

Background: In the United States, up to 35% of HIV primary care patients struggle to routinely attend follow‐up appointments. The COVID‐19 pandemic forced many institutions to transition to a telehealth‐focused model of care to maintain patient and provider safety. However, it was unknown how telehealth would impact patient attendance and perceptions about their care, particularly in populations at high risk of appointment non‐adherence. To understand the impact of telehealth on retention in care for a vulnerable HIV‐infected population, we assessed patient perspectives of telehealth and its effect on appointment attendance.

Methods: We studied patients at the UCSF 360 Wellness Center, an HIV Primary Care Clinic at the University of California, San Francisco, during the 2019 to 2020 fiscal year. To assess the effect of the city‐wide shelter‐in‐place orders and subsequent transition to telehealth, we compared appointment attendance before and after the order. Patient perceptions were collected via anonymous survey through a secure online portal. The survey included multiple‐choice questions and free‐text responses. Patients who did not have portal access were reached by telephone. Responses were analyzed and coded for themes.

Results: Before March 16, telehealth comprised 5.2% of all clinic visits; after, telehealth comprised 86.9% of visits. No‐show rates across all in‐person visits before March 16th averaged 16.5% during the prior 6months. Telehealth visits after March 16th had a no‐show rate of 13.7%. Of the 134 survey responses, 19.6% of patients reported preference for in‐person visits, 9.8% reported preference for telehealth visits, and 70.7% had no preference. Perceived strengths of telehealth included:

1)improved convenience;

2)safety in avoiding other patients; and

3)increased privacy.

Several disadvantages included:

1)technical barriers;

2)lack of familiarity with video visit platform; and

3)lack of human connection as compared to in‐person visits.

Conclusions: The majority of our respondents (80.5%) said they are equally or more likely to attend telehealth visits as compared to in‐person visits, and our data demonstrated improvement in appointment attendance. We plan to continue telehealth visits as a permanent and prevalent visit option to support HIV care for our patients. To further improve the telehealth experience, patients would benefit from more robust technical support.

Utilizing a robust HIV prevention community engagement framework to engage communities in COVID‐19 prevention efforts

M. Andrasik, F. Rentas, L. Oseso, S. Wallace, G. Broder

Fred Hutchinson Cancer Research Center, Vaccine and Infectious Disease Division, Seattle, United States

Background: The HIV Vaccine Trials Network (HVTN) has built a solid foundation of community engagement (CE) rooted in Good Participatory Practice, driven by the belief that engaging communities at each phase of research from pre‐study implementation through results dissemination fosters trust and mutual understanding. CE efforts utilize Community‐Based Participatory Research approaches and are informed by models of health behavior change. With the onset of the SARS‐CoV‐2 pandemic, the HVTN joined a large group of partners to form the COVID‐19 Prevention Network (CoVPN). The CE infrastructure existing within the HVTN was tapped to lead the CoVPN's vaccine study efforts.

Methods: The CoVPN developed a seven‐part strategic plan which has facilitated robust CE. The plan includes: (1) development, production and dissemination of materials for participants/potential participants; (2) development, production and dissemination of materials for clinical research sites; (3) utilization of Priority Population Scientific Expert Panels for protocol review and guidance; (4) convening a Community Working Group; (5) stakeholder engagement activities to build and enhance trust; (6) developing a Marketing and Community Influencers plan; (7) developing a CoVPN website with a centralized registry; and (8) convening CE advisory committees for sponsor organizations. The plan identifies priority populations most impacted by COVID‐19: Native/Indigenous, African American/Black, Latina/o/x/Hispanic, Older Adults (aged 65+)/Veteran populations, and essential workforce. All activities are tailored for each priority population and the overall focus is on building and maintaining relationships. The plan supported early identification of recruitment and retention barriers and built on existing infrastructure, allowing for expeditious identification of strategies to mitigate CE challenges.

Results: Long standing relationships with partners helped disseminate materials, open additional channels for dissemination, and identify and nurture new relationships to build trust. Support at the federal level facilitated new partnerships among the research network, Dept. of Defense, Veterans groups, and other NIH teams. Implementation of the CoVPN CE strategies resulted in a surge of interest in the PreventCOVID.org website and increases in racial and ethnic minority enrollment in the CoVPN registry.

Conclusions: The strong CE infrastructure developed by the HVTN allowed for a seamless pivot to engaging communities and addressing challenges in COVID‐19 vaccine trials during the SARS‐CoV‐2 pandemic.

Improved killing of HIV‐infected cells by a combination of three neutralizing and non‐neutralizing antibodies

M. Tuyishime¹, C. Garrido², S. Jha¹, M. Moeser², D. Mielke¹, C. LaBranche¹, D. Montefiori¹, B.F. Haynes³, S. Joseph⁴, D. Margolis⁵, G. Ferrari⁶

¹Duke University, Surgery, Durham, United States, ²UNC HIV Cure Center, United States, ³Duke Human Vaccine Institute, Duke University, Medicine, Immunology, Durham, United States, ⁴UNC HIV Cure Center, Lineberger Comprehensive Cancer Center, UNC, Microbiology and Immunology, United States, ⁵UNC HIV Cure Center, UNC, Medicine, Microbiology and Immunology, and Epidemiology, United States, ⁶Duke University, Surgery 9 Department of Molecular, Durham, United States

Background: Pre‐clinical and clinical antibody‐dependent cellular cytotoxic (ADCC) responses correlating with protection from HIV‐1 infection provide the rationale to leverage them for treatment purposes. We evaluated ADCC mediated by 62 monoclonal antibodies (mAbs) combinations to identify those effective against latent reservoir viruses (LRVs) isolated from ART‐suppressed HIV+ donors using Latency Reversing Agents (LRA)‐exposed resting latently infected CD4+ T cells.

Methods: HIV‐1 envelope‐specific CD4‐binding site CH557, C1/C2 A32, V1/V2 PG9, V3 PGT121, gp120‐gp41 interface PGT151 and MPER DH511.2K3 mAbs were tested at a concentration≤1µg/mL for their ability to mediate ADCC individually and in combinations. Target cells were either primary CD4+ T cells infected in vitro with a panel of 10 clade B LRVs or ART‐suppressed HIV+ donor cells exposed to LRAs. We used an Infected Cell Elimination assay and a novel ex vivo modified viral outgrowth assay (Latent Clearance Assay) in presence of autologous primary NK cells, respectively.

Results: We observed that within 2hours of incubation at≤1µg/mL, 3 mAbs combinations mediated significantly higher killing (p<0.005) compared to single or paired mAbs. Addition of more than 3 mAbs did not further increase specific killing. A32+ DH511.2K3+ PGT121 was the best combination with a mean of 33.1% ADCC activity. Three LRVs were resistance to neutralization by PGT121, one of which (P800) was also resistant to PG9 and DH511.2K3. The resistance was caused by the escape mutations within HIV‐1 Env of these LRVs. These results were confirmed by the loss of the binding of PGT121 and/or DH511.2K3 to cells infected with these LRVs. After 24hours of incubation, we observed remarkably increased ADCC activity against 4 LRVs tested with the A32+ DH511.2K3+ PGT121 combination, and demonstrated elimination of cells infected with 3‐bNAbs‐resistant P800 LRV. Further, the A32+ DH511.2K3+ PGT121 combination efficiently cleared ex vivo latently infected LRA‐treated resting CD4+ T cells from P800 donor by autologous NK cells during a 25‐day culture.

Conclusions: MAbs combinations targeted the diverse HIV‐1 Env epitopes on the surface of infected cells, upon reactivation of the latent infection, more efficiently than individual mAbs. The results strongly indicate the importance of mAb combinations to achieve the broadest activity against HIV‐1 infected cells in future cure strategies.

Unprimed CD8+ lymphocytes promote the maintenance of HIV latency in CD4+ T cells

L. Franchitti, Z. Zhang, M. Statzu, H. Wang, M. Paiardini, G. Silvestri, D. Kulpa

Emory University, Department of Pathology and Experimental Medicine, Atlanta, United States

Background: Therapeutic strategies to reactivate the latent reservoir to facilitate clearance of HIV‐infected cells have had limited success in vivo. Recent studies have demonstrated that the presence of CD8+ T cells may inhibit virus replication during treated HIV/SIV infection; however, the mechanisms responsible for this antiviral effect remain poorly understood.

Methods: We used our primary cell based in vitro model of HIV latency to study the CD8+ T cell mediated suppression of HIV expression. To examine the impact of CD8+ T cells on the activity of latency reversing agents (LRA), CD4+ T cells generated in our in vitro latency model were treated with IL‐15, the IL‐15 superagonist N‐803, or TCR stimulated in the presence or absence of activated autologous CD8+ T lymphocytes (1:1 or 1:5 target: effector ratios). To identify the phenotype of CD8+ T lymphocytes with pro‐latency activity, we sorted memory and naïve subsets based upon CD45RA, CD27 and CCR7 surface expression, followed by co‐culture with HIV‐infected CD4+ T cells. All assays were monitored for intracellular Gag expression on CD4+ T cells, and the frequency of integrated HIV DNA was determined by qPCR.

Results: Treatment with LRA in the presence of activated CD8+ T lymphocytes resulted in a significant suppression of HIV latency reversal. TCR activation induced a high frequency of Gag+ cells in the monoculture (~27%, n=8), that was reduced 3‐fold at 1:1 (p=0.0002) and 17‐fold at 1:5 ratio (p<0.0001). Treatment with IL‐15 or N‐803 induced a lower frequency of Gag expression in the monocultures (~4% and 3% respectively), but also showed a significant decrease in latency reversing activity in the presence of CD8+ T cells at a 1:1 (p=0.0041; 0.005) or 1:5 (p<0.0001) target: effector ratio. CD45RA‐CD27+ CCR7+ CD8+ T cells displayed the highest suppression activity in our assays.

Conclusions: Our studies demonstrated a CD8+ lymphocyte mediated suppression of HIV expression in CD4+ T cells capable of maintaining latency in the presence of activation signaling. Understanding the mechanisms by which CD8+ lymphocytes suppress virus transcription and ultimately promote HIV latency in ART‐treated HIV‐infected individuals may provide critical insight to the design of HIV eradication approaches.

Analysis of the early responses to HIV‐1 in matched treatment naïve individuals reveals early soluble proteins that are associated with in vivo virus control

J. Makinde¹, E.W. Nduati², A. Freni Sterrantino³, C. Streatfield¹, C. Kibirige¹, J. Dalel¹, S.L. Black¹, P. Hayes¹, G. Macharia¹, J. Hare¹, D. King¹, S. Joseph¹, E. McGowan¹, B. Abel¹, E. Hunter⁴

¹Imperial College London, IAVI Human Immunology Laboratory, Department of Infectious Diseases, London, United Kingdom, ²Kenya Medical Research Institute‐Wellcome Trust Research Programme, Kilifi, Kenya, ³Imperial College London, MRC‐PHE Centre for Environment and Health, Department of Epidemiology and Biostatistics, London, United Kingdom, ⁴Emory University, Emory Vaccine Centre, Yerkes National Primate Research Centre, Atlanta, United States

Background: In attempting to define the correlates associated with natural protection against HIV, extreme heterogeneity in the datasets generated from systems methodologies can be further complicated by the inherent variability encountered at the population, individual, cellular and molecular levels. Furthermore, such studies have been limited by the paucity of well‐characterised samples and linked epidemiological data, including the duration of infection and clinical outcomes.

Methods: We identified newly HIV‐infected individuals from a multisite HIV incidence cohort drawn from nine clinical research centres in five African countries. Selected volunteers were further characterized with regards to early and persistent in vivo control of HIV‐1 replication, in the absence of antiretroviral treatment. We utilised a propensity score matching approach to control for the influence of five factors (age, risk group, virus subtype, gender, and country) known to influence disease progression on causal observations. The concentrations of fifty‐two soluble plasma proteins were assessed. This approach is unique because it focuses on the specific period of dynamic immunological control of viral replication for all volunteers in the absence of treatment.

Results: Among 603 volunteers, we identified three distinct groups of individuals (Low, Intermediate and High viral load volunteers) matched on all five factors and for whom samples were available at two timepoints within the dynamic phase of immunological control of in vivo replication following peak viraemia (i.e., within 0 to 365days post‐infection). We were able to confirm some of the factors related to, or already known to influence disease progression such as the possession of B*57 HLA Class I allele, and the infecting virus subtype. Our results also indicate possible roles for IL‐17C and MIP‐1α in the early and sustained control of infection.

Conclusions: Our results highlight the need to consider factors that could potentially introduce heterogeneity in datasets and mask valid differences when designing studies to define immune correlates.

Targeting sphingosine‐1‐phosphate signaling prevents cell‐to‐cell transmission of HIV‐1

R. Resop¹, R. Fromentin², D. Newman¹, H. Rigsby², L. Dubrovsky¹, M. Bukrinsky¹, N. Chomont², A. Bosque¹

¹George Washington University, Microbiology, Immunology and Tropical Medicine, Washington, D.C., United States, ²CRCHUM and Université de Montréal, Microbiology, Infectious Diseases and Immunology, Mongtréal, Canada

Background: Sphingosine‐1‐phosphate (S1P) is a modulator of a myriad of cellular processes and therapeutic targeting of S1P signaling is under clinical investigation. We sought to determine the role of S1P in cell‐to‐cell transmission of HIV.

Methods: Using a primary cell model, we examined whether targeting S1P receptors altered HIV transmission in memory CD4+ T cells. We activated and expanded naïve human CD4+ T cells and infected them with HIV, in the presence of the S1PR functional antagonist FTY720 before or after infection. In addition, we examined transmission of HIV from infected macrophages to autologous CD4+ T cells in a coculture system in the presence of FTY720. We quantified productive infection by p24‐Gag flow cytometry, measured frequency of cells harboring total and integrated HIV DNA by qPCR, and examined reactivation of latent virus following T cell receptor stimulation. Mechanistically, we investigated internalization of virions, cell cycle signaling, and innate HIV‐1 restriction factor activity following FTY720 treatment.

Results: FTY720 treatment reduced HIV transmission between CD4+ T cells as well as between macrophages and CD4+ T cells in co‐culture assays. Mechanistically, treatment with FTY720 targeted early stages of the HIV‐1 life cycle, as evidenced by 1) Reduced internalization of virions by viral entry assay; 2) Relative increase in the active form of SAMHD1, an innate restriction factor; 3) Reduction in total and integrated HIV‐1 DNA; and 4) Reduced frequency of productively infected cells. Moreover, FTY720 reduced susceptibility to infection by decreasing the percentage of transcriptionally active, cycling cells, and downregulated Cyclin D3. These alterations were consistent with reduction in cell‐to‐cell transmission, which directly corresponded to reduced latent virus in CD4+ T cells.

Conclusions: Our results demonstrate that targeting S1P signaling with FTY720 inhibits cell‐to‐cell transmission in multiple cell types. FTY720 targets early stages of the HIV‐1 life cycle, including entry of virions and events prior to reverse transcription. Targeting S1P appears to modulate SAMHD1 phosphorylation as well as modify the cell cycle to promote a less infection‐permissive state. Our research suggests that therapeutic targeting of this pathway early in infection may aid in the development of strategies to prevent establishment of initial infection and the latent reservoir.

Impact of HIV kick‐and‐kill therapy on host epigenetic and transcriptional programs in PBMC, and viral rebound after cART interruption

B. Oriol-Tordera^1,2, A. Esteve-Codina³, M. Berdasco^4,5, E. Gonçalves⁶, M. Esteller^4,5,7, T. Hanke^8,9, J. Moltó¹⁰, B. Clotet^1,10,11, M.L. Calle¹¹, B. Combadiere⁶, A. Sanchez‐Pla^12,13, B. Mothe^1,10,11, M. Ruiz-Riol¹, C. Brander^1,11,14

¹IrsiCaixa AIDS Research Institute, Institute for Health Science Research Germans Trias i Pujol (IGTP), Badalona, Spain, ²Autonomous University of Barcelona, Cerdanyola, Spain, ³Centro Nacional de Análisis Genómico (CNAG), Parc Científic de Barcelona, Barcelona, Spain, ⁴Cancer Epigenetics and Biology Program (PEBC), Bellvitge Biomedical Research Institute, L'Hospitalet de Llobregat, Spain, ⁵Josep Carreras Leukaemia Research Institute (IJC), Badalona, Spain, ⁶Sorbonne Université, Centre d'Immunologie et des Maladies Infectieuses – Paris (Cimi‐Paris), INSERM U1135, Paris, France, ⁷Department of Physiological Sciences II, School of Medicine, University of Barcelona, Barcelona, Spain, ⁸The Jenner Institute, University of Oxford, Oxford, United Kingdom, ⁹Joint Research Center for Human Retrovirus Infection, Kumamoto University, Kumamoto, Japan, ¹⁰Fundació Lluita contra la Sida, Infectious Diseases Department, Hospital Universitari Germans Trias i Pujol, Badalona, Spain, ¹¹Universitat de Vic‐Central de Catalunya (UVic‐UCC), Vic, Spain, ¹²Statistics Department, Biology Faculty, University of Barcelona, Barcelona, Spain, ¹³Statistics and Bioinformatics Unit Vall d'Hebron Institut de Recerca (VHIR), Barcelona, Spain, ¹⁴Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Spain

Background: BCN02 was a pilot kick‐and‐kill clinical trial that combined therapeutic vaccination (MVA.HIVconsv) with the latency reversing agent romidepsin (RMD) followed by a monitored antiretroviral pause (MAP) in 15 early‐treated HIV+ individuals (NCT02616874). Out of 12 evaluable participants for an omics subanalysis, 8 participants showed early (pVL>2000 copies/mL<4weeks) and 4 a more delayed viral rebound (pVL>2000 copies/mL>4weeks) in MAP. Systems biology analyses identified epigenetic and molecular mechanisms associated with response to vaccination and RMD and viral rebound kinetics.

Methods: Genome‐wide gene expression (Ilumina HiSeq2500) and DNA Methylation (Infinium ® HumanMethylation450 BeadChip) were assessed in frozen PBMC‐pellets at baseline, 1week after vaccination and 1week after the 3rd RMD infusion (post‐RMD). After pre‐processing and normalization steps we applied principal component analyses (PCA) and differential expression/methylation analyses (limma‐R/Biocondcutor). GSEA was used for functional analysis, and sPLS‐DA (MixOmics‐R/Bioconductor), to identify pathways explicative of differential viral rebound kinetics.

Results: The largest impact on host transcriptional and DNA methylation (DNAm) profiles was observed after the combined effect of vaccination and RMD, with 733 differentially expressed genes and 5695 differentially methylated positions being detected between baseline and post‐RMD (adjusted p<0.1). Modulated pathways at gene expression and/or DNAm level, were mainly associated with processes including cell cycle, DNA repair or metabolism and HIV, innate and adaptive immunity (GSEA q‐value<0.15). Of note, PCA revealed that only DNAm levels after the combined intervention segregated participants by their early or late viral rebound kinetics in MAP. We summarized the therapy‐modulated pathways with eigenvectors of DNAm at post‐RMD, and identified HIV, innate immunity and T‐cell pathways among the most relevant to discriminate the two viral rebound profiles in MAP. Interestingly, 3 CpG positions in the HIV category, 37 in the innate immunity group and 10 in the T cell category were differentially methylated between individuals with different viral rebound profile (adjusted p<0.1).

Conclusions: Host transcription and epigenetic programs provide a deeper understanding of the molecular mechanisms induced during HIV cure therapies. While DNAm after a kick‐and‐kill strategy could be used to predict viral rebound kinetics after ART interruption, further study is warranted in future controlled studies.

Heterologous vaccination with DNA and two different poxvirus vectors, expressing HIV‐1 envelope on the surface of Gag virus‐like particles, elicit autologous Tier 2 neutralising antibodies

R. Chapman¹, M. van Diepen¹, N. Douglass¹, S. Galant¹, E. Margolin¹, P. Ximba¹, T. Hermanus², P. Moore², E. Rybicki³, A.‐L. Williamson¹

¹University of Cape Town, Pathology, South Africa, ²National Institute for Communicable Diseases of the National Health Laboratory Service, Johannesburg, South Africa, ³University of Cape Town, Moelcular & Cell Biology, South Africa

Background: The only HIV vaccine trial (RV144) to show any efficacy used a canary poxvirus vector and recombinant protein for immunisation; leading to the further exploration of poxviruses as HIV vaccine vectors. Our group has shown that heterologous prime boost immunisations of mice and macaques with a novel capripoxvirus vector, lumpy skin disease virus (LSDV) and modified vaccinia Ankara (MVA) induced high magnitude, broad cellular immune responses. Here, we present data on further modifications to the vaccines, which were tested in different regimens, for immunogenicity in rabbits.

Methods: DNA (D), MVA (M) and LSDV (L) vaccines expressing HIV‐1 CAP256SU gp150 and subtype C mosaic Gag (Gag^M) were constructed. The expression of Env and Gag by all three vaccine was characterised in vitro. Groups of five rabbits were inoculated with four different vaccine regimens: DDMMLL; DDMLML; DDLMLM and DDLLMM at 0, 4, 8, 12, 16 and 20weeks. The titres of binding and neutralising antibodies in the rabbit sera were determined.

Results: The expression and secretion of HIV‐1 CAP256SU gp150 and Gag^M by all three vaccine vectors was verified in vitro. In addition, the formation of Gag virus‐like particles containing gp150 was confirmed using negative stain electron microscopy and western blotting following density gradient centrifugation. The quaternary structures of gp150 on the cell surface was assessed by immunostaining with human monoclonal binding antibodies (MAbs) to HIV‐1 Env, and fluorescent microscopy or FACS analysis. Binding of MAbs, PG16, PGT145 and CAP256 VRC26_08, confirmed that all three vaccines expressed some native, trimeric Env. All four vaccination regimens elicited high titres of neutralising antibodies (NAbs) to Tier 1A pseudovirus MW965.26 two weeks after the final immunisation. However, 5/5 rabbits in the DDLMLM, 3/5 in the DDLLMM, 2/5 in the DDMLML and none in the DDMMLL group developed NAbs to Tier 1B pseudovirus 6644. Two of the rabbits from the DDLMLM group also developed low levels of autologous Tier 2 NAbs.

Conclusions: Low levels of autologous Tier 2 NAbs were induced in 2/5 rabbits inoculated with the DDLMLM vaccination regimen. Further studies are planned to determine if these vaccines elicit good cellular immune responses.

In vitro and in vivo analyses of HIV‐1 clade C Envelope trimers highlight optimal antigenic profiles of novel HIV‐1 Env‐based vaccine candidates

A. Hauser¹, D. Peterhoff¹, T. Hartinger¹, M. Schachtner¹, N. Friedrich², A. Trkola², C. Moog³, W. Weissenhorn⁴, Q. Sattentau⁵, R. Sanders⁶, R. Wagner¹

¹University of Regensburg, Institute of Medical Microbiology and Hygiene, Regensburg, Germany, ²University of Zurich, Zurich, Switzerland, ³Université de Strasbourg, France, ⁴Institut de Biologie Structurale, France, ⁵University of Oxford, United Kingdom, ⁶Academic Medical Center, Amsterdam, Netherlands

Abstract not available.

Novel trimer‐only (TO) producing HIV‐1 envelope glycoprotein constructs for inducing broadly neutralizing antibody responses by genetic vaccination

I. del Moral Sánchez¹, E.G. Wee², S.M. Koekkoek¹, E.E. Schermer¹, W.H. Lee³, S. Kumar⁴, R. Zwolsman¹, J.D. Allen⁵, A. Torrents de la Peña³, M.M. van Haaren¹, E. Reiss¹, M. van Gils¹, M. Crispin⁵, G. Ozorowski³, A.B. Ward³

¹Amsterdam UMC, Medical Microbiology, Amsterdam, Netherlands, ²The Jenner Institute, University of Oxford, Nuffield Department of Medicine, United Kingdom, ³The Scripps Research Institute, Department of Integrative Structural and Computational Biology, United States, ⁴All India Institute of Medical Sciences, New Delhi, India, ⁵University of Southampton, School of Biological Sciences, Highfield Campus, Southampton, United Kingdom

Background: Soluble immunogens that properly mimic the HIV‐1 envelope glycoprotein (Env), such as SOSIP and native flexibly linked (NFL) trimers, are used as constituents for a number of HIV‐1 vaccine approaches. However, most NFL and SOSIP constructs are not only expressed as trimers, but also produce undesired monomers and dimers that expose non‐neutralizing epitopes. Usually, chromatography procedures are used to isolate the desired trimers for vaccination. However, this also implies that conventional SOSIP and NFL constructs might be less suitable for genetic vaccination. Here, we describe a novel HIV‐1 Env construct that was designed to express as trimers only (TO).

Methods: First, we combined the TO design with BG505 env to generate a BG505 TO‐SOSIP construct as a proof‐of‐concept. Next, we used TO‐SOSIP as a design template to generate a number of different HIV‐1 Env immunogens: 1) TO‐SOSIP in which immunodominant strain‐specific glycan holes were hidden by glycan masking (GM) (TO‐SOSIP.GM). 2) TO‐SOSIP containing germline‐targeting (GT) mutations that enable binding by naïve precursors of CD4bs broadly neutralizing antibodies (bNAbs) (TO‐SOSIP.GT).

Results: Transient transfection followed by lectin affinity chromatography showed that the BG505 TO‐SOSIP construct expressed only as trimers, of which >85% displayed a native‐like conformation, while BG505 SOSIP and NFL‐SOSIP preparations contained significant amounts of dimers and monomers. Furthermore, TO‐SOSIP interacted efficiently with quaternary dependent bNAbs, such as PGT145 and PGT151, while showing weak binding to most non‐NAbs. Both glycan masked and germline targeting variants also expressed only trimers. TO‐SOSIP.GM showed negligible reactivity to glycan hole‐targeting antibodies, while several TO‐SOSIP.GT variants showed efficient binding to inferred germline versions of several CD4bs bNAbs. Remarkably, the TO design approach also allowed us to generate influenza hemagglutinin (HA) immunogens that express only as trimers without heterologous trimerization domains.

Conclusions: The TO design is a useful platform for generating Env and HA immunogens and is especially suitable for genetic vaccination purposes. In the context of DNA prime and protein boost vaccination strategies, the glycan masked and germline targeting TO‐SOSIP constructs could prove essential for the early activation of potentially cross‐neutralizing epitopes.

Structural basis of neutralizing antibodies targeting the CD4‐binding site and fusion peptide elicited by immunization with heterologous HIV‐1 Env in NHP

T. Zhou¹, A. Nazzari², X. Chen¹, J. Gorman¹, M. Sastry¹, A.S. Olia¹, K. Xu¹, B. Zhang¹, C.S. Cheung¹, C. Cheng¹, A.R. Corrigan¹, E.K. Sarfo¹, K. McKee¹, R. Casner³, G. Cerutti³

¹National Institute of Health, Vaccine Research Center, Bethesda, United States, ²National Institute of Health, Bethesda, United States, ³Columbia University, New York, United States

Background: Elicitation of broadly neutralizing antibodies remains a challenge to the development of an effective HIV‐1 vaccine. While most neutralizing epitopes identified are present on the prefusion closed form of HIV‐1 envelope trimer (Env), only limited neutralization breadth has been achieved with previous HIV‐1 Env immunizations.

Methods: Here we characterize the antibody response in non‐human primates (NHP) following sequential immunization with HIV‐1 Env variants, priming by Env with glycans removed around the CD4‐binding site (CD4BS), boosting with Envs with gradually restored glycans and eventually with natively glycosylated Envs. Sera were evaluated with a sentinel panel of viruses. PBMCs of a selected group of NHPs with broad sera neutralization activity were analyzed using probe‐based B cell sorting and monoclonal antibody cloning. Antibodies were selected by binding affinity to gp120 core and Env trimer. Large panel neutralization and structural studies were conducted to map major sites of antibody targeting.

Results: The sequential Env‐only immunization scheme elicited substantial sera neutralization activity in a subgroup of NHP and NHPs primed/boosted with deglycosylated 4501dG5 trimers elicited higher titers against the fusion peptide (FP). Isolated antibodies with gp120‐binding activity had limited breadth, only neutralizing HIV‐1 strains lacking glycan on Asn276. Structural studies revealed that these antibodies target the CD4BS, however, in an orientation that fails to accommodate glycan 276. In contrast, one of the Env‐interactive antibodies, VRC7324, showed 34% breadth on a 208‐virus panel and a median IC50 of 0.98µg/ml. VRC7324 bound to diverse FP and competed with FP‐directed antibodies. Cryo‐electron microscopy was carried out to determine the structure of VRC7324‐Env complex and elucidate how a single antibody can recognize different FPs to achieve broad neutralization.

Conclusions: Immunization with glycan‐removed Env increases immune responses. However, optimal glycan shield around the CD4BS may be essential to train developing antibodies to accommodate glycans on Env. In contrast, Env‐only immunizations can elicit broadly neutralizing FP‐targeting antibodies and we identified that the 4501dG5 Env can be a promising priming immunogen for eliciting the FP‐targeting antibodies.

Cross‐reactivity between HIV‐1 bnAbs and parasite glycans

I. Huettner¹, S. Krumm^1,2, E. Landais³, K. Brzezicka⁴, S. Serna⁴, A. van Diepen⁵, J. Angulo⁶, F. Allan⁷, A. Emery⁷, N. Reichardt⁴, R. Hokke⁵, P. Poignard^3,8,9, K. Doores¹

¹King's College London, School of Immunology and Microbial Sciences, London, United Kingdom, ²BioNTech, Mainz, Germany, ³International AIDS Vaccine Initiative, Neutralizing Antibody Center, La Jolla, United States, ⁴CICbiomaGUNE, Biofunctional Nanomaterials Unit, San Sebastian, Spain, ⁵Leiden University Medical Center, Parasite Glycobiology Group, Leiden, Netherlands, ⁶University of East Anglia, School of Pharmacy, Norwich, United Kingdom, ⁷Natural History Museum, Life Sciences Department, London, United Kingdom, ⁸The Scripps Research Institute, Department of Immunology and Microbiology, La Jolla, United States, ⁹Université Grenoble Alpes, Institut de Biologie Structurale, Grenoble, France

Background: Many HIV‐1 broadly neutralizing antibodies (bnAbs) isolated from HIV‐1 infected donors directly engage, or are dependent upon, N‐linked glycans on the HIV‐1 envelope‐trimer for neutralization. It can take many years for bnAbs to develop following HIV‐1 infection making it a challenge to identify immunization strategies that will re‐elicit these bnAbs through vaccination. Since other glycosylated pathogens, such as bacteria and parasites, can raise a strong anti‐glycan antibody response, the question arises whether the development of glycan‐dependent HIV‐1 bnAbs could be guided by Abs targeting glycans on other pathogens. We therefore set out to study cross‐reactivity of glycan‐binding bnAbs with parasite glycans and prevalence of seroreactivity to parasite Schistosoma mansoni in the IAVI Protocol C cohort.

Methods: To measure cross‐reactivity of known glycan‐binding bnAbs with non‐self parasite‐derived glycans, we used a synthetic parasite glycan microarray and a ‘shotgun’ parasite glycan microarray. The mode of glycan binding was assessed using STD NMR. We further confirmed cross‐reactivity with S. mansoni using confocal microscopy and immunohistochemical analysis. To understand the seroprevalence of this parasitic infection in HIV‐1 infected individuals in S. mansoni endemic regions we screened plasma from the well‐studied IAVI Protocol C cohort for seroreactivity to S. mansoni antigens by ELISA.

Results: We observed cross‐reactivity of glycan‐specific bnAbs to self and non‐self S. mansoni glycans and were able to further confirm cross‐binding of bnAbs to S. mansoni glycoproteins by confocal microscopy and immunohistochemical analysis. STD NMR analysis demonstrated that bnAb PGT121 contacts both self and non‐self components of a representative parasite glycan. We detected an overall 41% seroprevalence against S. mansoni in a subset of IAVI protocol C donors, with a higher occurrence in donors with medium and top plasma neutralization (88% vs. 63%). We further observed cross‐reactivity of N332/V3 specific PCDN76 bnAb lineage, including the unmutated common ancestor, with S. mansoni.

Conclusions: These findings suggest antibody responses against parasite glycoproteins are readily found in HIV‐1 positive donors. The cross‐reactivity with parasite glycoproteins, particularly of the unmutated common ancestor of a bnAb suggest a non‐self‐glycan response could direct the development of glycan reactive HIV‐1 bnAbs. These observations will inform and aid HIV‐1 immunogen design.

Trends in incidence of bacterial sexually transmitted infections among gay and bisexual men using PrEP in Australia

M. Traeger¹, J. Asselin¹, C. El‐Hayek¹, J. Dittmer¹, E. Wright², A. Carter³, T. Vickers³, P. Patel³, C. Fairley⁴, B. Donovan³, R. Guy³, M. Hellard¹, M. Stoové¹

¹Burnet Institute, Disease Elimination, Melbourne, Australia, ²Alfred Health and Monash University, Department of Infectious Diseases, Melbourne, Australia, ³UNSW Sydney, Kirby Institute, Sydney, Australia, ⁴Melbourne Sexual Health Centre, Australia

Background: Multiple PrEP studies suggest incidence of bacterial STIs among gay and bisexual men (GBM) increases following PrEP initiation, possibly due to changes in behaviour, sexual networks or testing. Regular, asymptomatic STI testing of PrEP users (Australian guidelines recommend three‐monthly) may also help suppress incidence longer‐term. Using national sentinel surveillance data, we report trends in STI incidence among GBM PrEP‐users.

Methods: Retrospective patient data were extracted from 35 clinics, with individuals’ records linked between clinics. HIV‐negative GBM with evidence of TDF/FTC prescription (excluding PEP) contributed person‐time from their first subsequent STI test and were censored at their last test or six months after their last PrEP script. Censored participants could re‐enter at PrEP re‐initiation. Infection date was imputed as a random date between diagnosis and previous negative test. Using Poisson regression, we estimated half‐yearly incidence and assessed trends (July 2016‐December 2019) for syphilis, chlamydia and gonorrhoea, overall and stratified by anatomical site and age (</>35years). To account for selection bias associated with differences in characteristics of GBM initiating PrEP over time, a closed‐cohort analysis was conducted among GBM continuously using PrEP across the study period.

Results: Among 17,250 PrEP‐users contributing 24,823 person‐years (median time between tests=84days) chlamydia (51.3/100py to 39.1/100py; p<0.001) and gonorrhoea (43.1/100py to 32.8/100py; p<0.001) incidence declined, whereas syphilis incidence increased (6.4/100py to 9.9/100py; p>0.001). Among 3498 continuous‐PrEP‐users (10,538 person‐years), declines were observed for overall chlamydia (54.0/100py to 42.3/100py; p<0.001) and gonorrhoea (43.8/100py to 34.9/100py; p<0.001) incidence; pharyngeal chlamydia (p=0.116) and urogenital gonorrhoea (p=0.973) were stable, and syphilis incidence increased (6.8/100py to 11.8/100py; p<0.001). For both cohorts, gonorrhoea and chlamydia incidence were greater among GBM<35years, with declines most prominent among those≥35years. Syphilis rates and trends did not differ by age.

Conclusions: Representing the largest cohort analysis of STI incidence among PrEP‐users reported internationally, these data suggest that initial increases in incidence of some STIs following PrEP initiation may be attenuated over time. Frequent testing or changes in the risk profile of PrEP initiates over time may explain some reductions. Risk‐mitigation strategies may be needed for younger GBM and to curtail syphilis incidence.

For MSM attending sexual health clinics in England, rates of new HIV infections decreased and rates of new rectal GC diagnosis increased between 2010 and 2018

D. Donnell¹, K. Zewdie², N. Ratna³, V. Miller⁴, J. Saunders³, V. Delpech³, N. Gill⁵, H. Mohammed³

¹Fred Hutchinson Cancer Research Center, VIDD, Seattle, United States, ²University of Washington, Department of Epidemiology, Seattle, United States, ³National Infection Service ‐ Public Health England, Blood Safety, Hepatitis, Sexually Transmitted Infections (STI) and HIV Science, United Kingdom, ⁴University of California, Forum for Collaborative Research, School of Public Health, Berkeley, United States, ⁵National Infection Service ‐ Public Health England, STI and HIV Department, United Kingdom

Background: Incidence of rectal gonorrhea (GC) has been hypothesized as a correlate of HIV exposure risk in HIV prevention trial cohorts, because both infections occur after condomless anal sex. High rectal GC rates might be a useful measure for selecting sites for HIV prevention or vaccine trials, and high rectal GC incidence in trial participants of new biomedical prevention drugs may provide supportive evidence for ongoing HIV risk.

Methods: Rectal gonorrhea and HIV are routinely tested in sexual health clinics (SHCs) throughout England and pseudonymised, patient‐level data are reported to Public Health England. Through evaluation of repeated visits to SHCs by individual clients, we assessed HIV incidence and new rectal GC diagnoses in HIV‐negative MSM from 2010 to 2018, predating widespread roll‐out of PrEP. We used meta‐analysis regression to assess the population‐level association between (log)HIV and rectal GC incidence over the 9years, both overall and adjusted for geographic region.

Results: Between 2010 to 2018, 625,273 MSM attended SHCs in England, 322,857 person years were included in the assessment of HIV incidence, 205,946 in assessment of rectal GC. HIV incidence assessed each calendar year among MSM attending SHCs fell from 1.45/100PY in 2010 to 0.46/100PY in 2018. Rectal GC rates increased from 20/100PY to 34/100PY over the same period. Figures 1a and b show a negative correlation between rectal GC and HIV both overall (slope=0.139, p=<0.001), and when adjusted for different regions (slope=0.100, p<0.001).

Conclusions: Among MSM attending SHCs in England, rectal GC incidence substantially increased while HIV incidence substantially decreased from 2010 to 2018. HIV incidence likely decreased through expanded HIV testing, prompt ART initiation and increased viral suppression in HIV‐positive (U=U), interventions that did not decrease rectal GC. Interpretation of rectal gonorrhea as a correlate of HIV exposure risk is complex and context‐dependent, given effective HIV prevention interventions in both HIV‐positive and uninfected MSM.

Comparing applicator vs. “as lubricant” delivery of rectal dapivirine gel (MTN‐033)

K. Ho¹, C. Dominguez-Islas², D. Szydlo³, N. Macagna⁴, R. Brand¹, J. Piper⁵, J. Bauermeister⁶, M. Marzinke⁷, S. Riddler¹, S. Edick¹, S.L. Hillier⁸, J. Steytler⁹, J. Nuttall⁹, C.W. Hendrix¹⁰

¹University of Pittsburgh, Medicine, Pittsburgh, United States, ²Fred Hutchinson Cancer Research Center, Vaccine Research, Seattle, United States, ³Fred Hutchinson Cancer Research Center, Seattle, United States, ⁴FHI 360, Durham, United States, ⁵National Institutes of Health, Bethesda, United States, ⁶University of Pennsylvania, School of Nursing, Philadelphia, United States, ⁷Johns Hopkins University School of Medicine, Pathology, Baltimore, United States, ⁸University of Pittsburgh, Obstetrics and Gynecology, Pittsburgh, United States, ⁹International Partnership for Microbicides, Silver Spring, United States, ¹⁰Johns Hopkins University School of Medicine, Medicine, Baltimore, United States

Background: Rectal microbicides are promising strategies for local, non‐systemic delivery of HIV pre‐exposure prophylaxis, especially in men who have sex with men. Tenofovir gel administered intrarectally with an applicator had lower acceptability as compared to a daily oral PrEP regimen and traditional lubricant use in prior phase II studies. Data to enhance the behaviorally congruent delivery of rectal microbicides as lubricants are scant.

Methods: MTN‐033, a single site, open label, sequence randomized, crossover study, enrolled HIV negative men to receive 0.05% dapivirine (DPV) gel by intrarectal dosing using an applicator (2.5g) and self‐administered on an artificial phallus as lubricant (up to 10g). Plasma, rectal fluid, and rectal biopsies were collected over 24hr post dose. Participants completed behavioral acceptability surveys after each dosing regimen. We compared delivery methods using Wilcoxon tests for PK parameters.

Results: 16 participants used DPV gel by applicator and 15/16 participants used gel as lubricant (mean 1.8g, SD 0.8). No related adverse events were reported. Participants all felt the gel was easy to use. Peak DPV concentration in plasma (pg/mL) was higher using an applicator, median 319 (interquartile range [IQR] 247, 603), than use as lubricant, 85 (IQR 54, 218; p<0.01). Plasma DPV 24hour area under the curve (ng‐hr/mL) after applicator (median 3.5; IQR 3.1, 6.2) was higher than after use as lubricant (median 1.3, IQR 0.7 to 2.4; p<0.01) Mean DPV concentrations in rectal fluid 1hr post dose were similar with applicator and as lubricant, 20.9 (SD 33.6) and 25.6 (SD 23.8) pg/mg respectively. DPV was detected in only 1/31 rectal biopsies collected at 1 and 4hr post dose.

Conclusions: Evidence of local delivery and systemic absorption of DPV when dosed as an anal lubricant indicates potentially feasible development of lubricant delivered rectal microbicides. However, roughly 3‐fold lower DPV exposure in plasma and lack of detectable DPV in tissue biopsies indicate formulation changes are necessary to achieve protective tissue concentrations informed by vaginal microbicide experience. Our data suggests that DPV gel was safe and easy to use. Further development of a lubricant microbicide strategy is warranted.

Phase 1 safety and pharmacokinetic study of candidate rectal microbicide PC‐1005 rectal gel (MIV‐150/zinc acetate/carrageenan) in HIV‐1 seronegative adults (MTN‐037)

K. Ho¹, C. Hoesley², P. Anderson³, C. Kelly⁴, Y. Jiao⁴, S. Edick⁵, N. Reddy², R. Brand⁵, R.K. Ayudhya⁶, J. Piper⁷, R. Scheckter⁸, B.A. Friedland⁹, G.W. Creasy⁹, I. McGowan¹⁰, C.W. Hendrix¹¹

¹University of Pittsburgh, Medicine, Pittsburgh, United States, ²University of Alabama at Birmingham, United States, ³University of Colorado, United States, ⁴Fred Hutchinson Cancer Research Center, Seattle, United States, ⁵University of Pittsburgh, Pittsburgh, United States, ⁶Magee Womens Research Institute, United States, ⁷National Institute of Allergy and Infectious Disease/DAIDS, United States, ⁸FHI 360, Durham, United States, ⁹Population Council, New York, United States, ¹⁰Orion Biotechnology, Spain, ¹¹Johns Hopkins University, Department of Medicine, Baltimore, United States

Background: On demand topical PrEP may meet the needs of those who prefer episodic, non‐systemic PrEP or struggle with or reject daily oral PrEP. PC‐1005 gel, which contains MIV‐150, is active against HIV‐1, HPV, and HSV‐2 ‐ an attractive multipurpose prevention technology candidate. We evaluated rectal safety and pharmacokinetics (PK) of PC‐1005 gel applied rectally.

Methods: HIV‐uninfected persons >18years of age each received a series of 3 rectal gel doses at increasing volumes of 4, 16, and 32mL with a washout period between doses. Following each dose, plasma, rectal tissue and fluid, and vaginal fluid were collected up to 7 times over 48hours.

Results: We enrolled 13 participants, 7 men and 6 women (birth sex), median 34years of age, including 6 white, 5 black, and 2 mixed race/ethnicity. 12 participants completed all 3 doses.

Thirteen adverse events were reported, all Grade 1 or 2, of which 5 were judged related to study drug without relationship to dose volume. Median (interquartile range) plasma MIV‐150 concentration peaked 1 or 2hours after dosing at 0.074ng/mL (0.052, 0.086), 0.184 (0.162, 0.211), and 0.171 (0.098, 0.316), after 4, 16, and 32mL doses, respectively; median concentrations were below assay quantitation limits (BLQ) 24hours after dosing. Median half‐life was 1.4 to 1.9hours across dose volumes. Rectal Tissue MIV‐150 peaked 0.5 to 1hours after dosing at 1.4ng/g (ng/mL) (0.8, 1.9), 46.0 (30.7, 831.0), and 79.7 (11.9, 116.0), respectively, after each dose volume. Median tissue concentrations were BLQ (0.7ng/g) at 1.5 to 3, 24, and 3.5 to 5hours after each escalating dose volume, respectively. Median rectal fluid concentrations were >14ng/mL (ng/g) through 5hours in all arms; vaginal fluid samples were all BLQ (0.2ng/mL). Across dose volumes, 83%‐92% of participants rated the gel as comfortable or very comfortable.

Conclusions: PC‐1005 gel was well‐tolerated with low systemic exposure. MIV‐150 rectal tissue concentrations were transient and below the 100ng/g target; no MIV‐150 was detectable in vaginal fluids. Based on concentration data, a longer‐acting reformulation delivering more MIV‐150 to rectal tissues is likely needed to support further development of PC‐1005 as an on demand HIV rectal microbicide.

A randomized, double blind, placebo‐controlled, phase 1 safety and pharmacokinetic study of dapivirine gel (0.05%) administered rectally to HIV‐1 seronegative adults (MTN‐026)

C. Hoesley¹, K. Ho², E. Dunne³, C. Dominguez Islas⁴, M. Marzinke⁵, R.P. Kunjara Na Ayudhya⁶, J. Piper⁷, J. Bauermeister⁸, S. Johnson⁹, H. Gundacker⁴, B. Devlin¹⁰, J. Nuttall¹⁰, C.W. Hendrix⁵, I. McGowan⁶, R. Cranston²

¹University of Alabama at Birmingham, Medicine, United States, ²University of Pittsburgh, Pittsburgh, United States, ³Silom Community Clinic CRS Thailand MOPH‐U.S. CDC Collaboration, Thailand, ⁴SCHARP/Fred Hutchinson Cancer Research Center, Seattle, United States, ⁵Johns Hopkins University, Medicine (Clinical Pharmacology), Baltimore, United States, ⁶Microbicide Trials Network, United States, ⁷National Institute of Allergy and Infectious Diseases, Division of AIDS, United States, ⁸University of Pennsylvania, Philadelphia, United States, ⁹FHI 360, Durham, United States, ¹⁰International Partnership for Microbicides (IPM), Silver Spring, United States

Background: HIV prevention needs persist, in part, because of daily oral PrEP challenges, hence, interest in on demand topical options. The dapivirine vaginal ring demonstrated 30% risk reduction in HIV‐1 infections, but has not been evaluated for rectal use. We evaluated rectal safety, pharmacokinetics (PK) and pharmacodynamics (PD) of a dapivirine gel formulation.

Methods: We enrolled HIV‐uninfected men and women, 18 to 45years of age, at Thailand and United States sites. We randomized eligible participants 2:1 to receive blinded dapivirine (0.05%) or placebo gel via rectal applicator. A single dose phase was followed by observed daily dosing for seven days. Plasma, and fluid and tissue from both rectum and cervix were collected at baseline and 5 times over 72hours after each dose for PK, ex vivo HIV‐1 biopsy challenge (PD), tissue histology, and flow cytometry.

Results: We randomized 28 participants; two terminated early. Nine participants were female and 19 male (sex at birth); 12 were white, 11 Asian, 4 black, 1 other race/ethnicity. Mean years of age were 28.5 and 34.2 in the dapivirine and placebo arms, respectively.

Thirty adverse events occurred (Grade 1 or 2, except one unrelated Grade 3) without study arm differences or any drug related discontinuation. Dapivirine rectal tissue concentrations (median [interquartile range]) 0.5 to 1 and 2, hours after a single dose were 256ng/gm (below limit of quantitation [BLQ], 666) and BLQ (BLQ, 600), respectively, then BLQ (BLQ, BLQ) from 24 to 72hours; multiple dose concentrations were similar. Peak dapivirine plasma concentrations were 0.33ng/mL (0.15, 0.48) after one dose and 0.40 (0.33, 0.49) after seven doses. Cervical tissue and fluid concentrations were all BLQ. Flow cytometry indicated no changes or study arm differences. Ex vivo biopsy challenge showed ~10‐fold reduction in viral replication, but only through 2hours after dosing.

Conclusions: Dapivirine gel was well‐tolerated with dapivirine plasma concentrations similar to those of the dapivirine vaginal ring observed in clinical trials. However, the rectal ex vivo antiviral effect and tissue concentrations were only transient and well below vaginal tissue results with the dapivirine vaginal ring. A long‐acting reformulation or higher dose is likely needed to provide protection from anal sex.

Eliciting participant feedback in focus group discussions to strengthen the implementation of a HIV prevention study

I. Hawley¹, S. Roberts¹, M.K. Shapley-Quinn¹, A. Young¹, M. Garcia², K. Ngure³, L. Soto-Torres⁴, D. Kemigisha⁵, M. Chitukuta⁶, F. Mathebula⁷, N. Mangxilana⁸

¹RTI International, Women's Global Health Imperative, Research Triangle Park, United States, ²Family Planning International, United States, ³Jomo Kenyatta University of Agriculture and Technology, Kenya, ⁴NIAID, DAIDS, Bethesda, United States, ⁵Makerere University ‐ Johns Hopkins University (MU‐JHU) CRS, Kampala, Uganda, ⁶Spilhaus Clinical Research Site, Zimbabwe, ⁷Wits Reproductive Health and HIV Institute CRS, Johannesburg, South Africa, ⁸Emavundaleni CRS, South Africa

Background: Adolescent girls and young women (AGYW) in sub‐Saharan Africa are disproportionately affected by HIV, yet there is limited guidance on their engagement in biomedical HIV prevention research. Engaging AGYW through focus group discussions (FGDs) on study implementation may help inform and improve study procedures, address concerns and challenges, and ensure their needs are met in a timely manner.

Methods: Implementation‐focused FGDs were conducted in MTN‐034/REACH, an ongoing trial evaluating the safety of and adherence to the monthly dapivirine vaginal ring and daily oral PrEP among AGYW (16 to 21) in three African countries (South Africa, Uganda, Zimbabwe). Trained social scientists facilitated 8 FGDs using a semi‐structured guide with topics related to motivation to join the study, interactions with staff, challenges to product adherence, disclosure of product use, and community perceptions of the products and trial. Each FGD was conducted with participants who have completed two to six months of the 18‐month study. Summary reports of each FGD were disseminated to behavioral researchers, clinical researchers, and staff within and across sites.

Results: These procedures offer an innovative bottom‐up approach to improve study implementation in real‐time using participant feedback. Creating positive environments for receiving feedback and systematically distributing feedback across the study team facilitated the creation of actionable plans to address participants’ concerns and challenges. Feedback communication channels included.

(1)FGD facilitators presenting findings to clinic staff at internal meetings,

(2)site investigators consulting staff mentioned in FGDs, and

(3)study management sharing findings with adherence counselors.

Reviewing the objectives of these FGDs in advance improved staff's receptivity to feedback and criticisms. These FGDs resulted in actions such as organizing laboratory tours for community members to dispel rumors related to the trial, limiting the number of staff present during clinical procedures to increase privacy, and improving the efficiency of clinic flow during participants’ visits.

Conclusions: For clinic staff and behavioral scientists involved in the design and implementation of HIV prevention trials for AGYW, early‐stage implementation‐focused discussion groups can help identify challenges participants face, open communication channels among research staff, and address the needs of young participants.

A synergistic model of engagement: the ECHO Global Community Advisory Group and HC‐HIV Advocacy Working Group

J.W. Mwangi¹, Y. Raphael², L. Mworeko³, M. Happy⁴, E. Bass⁵, T. Njoki⁶, A. Forbes⁷

¹AFROCAB Treatment Access Partnerships, Programs, Nairobi, Kenya, ²APHA, ED, Durban, South Africa, ³ICWEA, Programs, Uganda, ⁴Advocacy for Quality Health Uganda (AQH‐Uganda), Entebbe, Uganda, ⁵AVAC, New York, United States, ⁶Athena, United States, ⁷Independent Consultant, United States

Background: In 2019, the Evidence for Contraceptive Options in HIV Outcomes (ECHO) study found no substantial difference in risk of HIV infection among women using one of three safe and effective contraceptive methods (DMPA‐IM, LNG implant, copper IUD). The trial used a range of advisory mechanisms that aligned to global standards in stakeholder engagement, including an ECHO‐specific Global Community Advisory Group (GCAG).

Methods: The HC‐HIV Advocacy Working Group includes highly visible women leaders from diverse geographies who advocate on the global, national and local stage; their advocacy paved the way for the formation of the ECHO GCAG. The GCAG included a subset of advocates from the Working Group that advised and communicated regularly with the ECHO Consortium around key trial issues‐particularly community engagement.

Results: The strength of the GCAG/HC‐HIV Advocacy Working Group model was the high degree of synergy between the two groups. The GCAG workshopped issues with the Group, and together ensured the voices of women were well‐represented in the study. Following the trial, the Group provided a space for GCAG members to continue advocacy efforts ‐ including communicating the implications of the trial results to their communities, pushing for integrated sexual and reproductive health services in national Task Forces, and monitoring WHO and Ministry commitments. This demonstrates how the model can advance advocacy from research to policy and implementation.

Conclusions: The model is now driving discussions around engagement for future HIV prevention clinical trials. A study looking at a new daily oral PrEP using F/TAF, planned in communities similar to the ECHO study, highlighted the model as part of their stakeholder consultations. Advocates engaged in the study include members of the GCAG, and study sponsors are utilizing existing advocacy networks to inform the engagement strategy. Similar to the ECHO study, the F/TAF trial has global implications for women. It remains to be seen if the study will replicate or adapt facets of the model; however, it is clear that there is much that can be learned from the relationship between the Working Group and GCAG and how it can support engagement and implementation.

The value of outside engagement‐civil society led engagement in PopART

C. Chilende¹, C. Kasanda¹, J. Salzwedel², N. Jain², F. Mwanza³

¹Treatment Advocacy & Literacy Campaign, Lusaka, Zambia, ²AVAC, New York, United States, ³Treatment Advocacy & Literacy Campaign, Luanda, United States

Background: Between 2013 and 2018 the PopART study, a randomized study evaluating community level impact of a package of HIV prevention interventions on HIV incidence, was conducted in South Africa and Zambia. Community Advisory Boards (CABs), a common mechanism of community engagement, were established early in the research process and served as a main form of communication. In Zambia, CAB engagement included quarterly meetings, protocol review, and participation in national HIV events. At community level, however, the CAB had limited spaces to provide feedback to community health structures, civil society organizations (CSOs) and the general population. To help address this gap, seven CSOs formed the Community Partnership Platform (CPP). This “outside‐in” or CSO led model of engagement built broader support for the study intervention.

Methods: The CPP partnered with the PopART trial team to build capacity of stakeholders on the Good Participatory Practice guidelines (GPP). The CPP used a mixed method approach to identify the impact of the trial sites’ stakeholder engagement programs. Score cards were created to anonymously evaluate perceived stakeholder engagement quality by the community in line with GPP. Focus groups were conducted with a range of stakeholders to provide a robust picture of community involvement throughout the research process. Findings were reported back to the trial in real‐time.

Results: Through community trainings, the CPP identified treatment literacy needs and supported development of effective training tools and evaluation measures. Through networking, the CPP held meetings with at‐risk populations to create demand, increase awareness, and ensure meaningful engagement with the study team. At the national level, the CPP represented community groups and CSOs in the National Steering Committee of the PopART study. This helped amplify the community voice during Pop ART study forums.

Conclusions: The use of CSOs as a community engagement mechanism is critical to optimizing the research process. The CPP initiative provided a parallel opportunity for engaging broader stakeholders in research and ensures the wider viewpoints of a community are included. This approach has the potential to improve transparency and accountability at a range of stakeholder levels from trial staff to policy makers.

Examining systemic racism in “empowerment‐based” HIV prevention research: reflections of a US‐France research partnership

T. Albritton¹, K.B. Coulibaly², I. Zoumenou³, M. Paige¹, C. Davis⁴, S. Carillon⁵, D. Matthews¹, P. Saintonge Austin⁶, L. Miles¹, V. Frye¹, A. Gosselin⁷

¹CUNY School of Medicine, Community Health and Social Medicine, New York, United States, ²Université de Paris, Centre Population et Development, France, ³Afrique Avenir ‐ Médiatrice en Santé, Afrique Avenir, Paris, France, ⁴Compassioniton, New York, United States, ⁵Université Paris Descartes, CEPED, UMR IRD‐Equipe SAGESUD, ERL INSERM U 1244, Paris, France, ⁶Children's Aid NYC, New York, United States, ⁷Institut Pierre Louis d'Epidémiologie et de Santé Publique (IPLESP/INSERM_S1136), Social Epidemiology Department (ERES) Faculté de Médecine de St Antoine, Paris, France

Background: HIV prevention research projects often apply “empowerment” approaches to achieve scientific goals and redress power imbalances between the study team and marginalized communities, particularly communities of color. Simultaneously, there have been calls for both scientific leadership and study methods that recognize the role of systemic racism within study team and the power differentials between study leadership and the communities.

Methods: ETOILE, Experiences of Tensions in Organizations and Interventions Leveraged for Empowerment and Prevention, is a French‐American cooperative project to explore and characterize how researchers in France and the US have addressed tensions related to race/racism‐related power imbalances within two empowerment‐based HIV prevention projects and how these tensions influence study implementation. Each project involved has predominately white leadership and predominately Black (US) or Black African/Black French (France) staff and community representatives. ETOILE applies reflexive, comparative, qualitative process evaluation methods to identify how tensions related to race/racism‐related power imbalances manifest in various aspects of the study. Originally planned as a face‐to‐face intervisitation between French and US teams, we have conducted with three semi‐structured, web‐based group interviews, where study scientific leadership, staff, and community representatives, discussed within‐project power relations, representation in study leadership and the role of race/racism, privilege and voice, related to community location, education and status. Interviews were conducted via Zoom and Skype and transcribed. Analysis is ongoing.

Results: Preliminary analyses have identified several overarching themes, including racialized organization of labor within each team; resistance to sharing power; need for strategies and opportunities for Black scholars in HIV prevention research; the role of country‐specific history and contexts; and need for white researchers to systematically undo marginalization of black researchers.

Conclusions: Failing to reflexively consider the role of within study team‐based, race/racism‐related power imbalances in empowerment‐based HIV prevention research may threaten the integrity and impact of the research and potentially reinforce the social conditions the research may be attempting to mitigate. Our hope is that by identifying how these tensions emerge in our own research and engaging in a structured analysis of the similarities and differences across contexts, we will identify best practices that reduce race/racism‐related power imbalances and advance both in US and French HIV prevention research.

Redefining ownership: the mismatch between female sex workers’ research priorities and the field in Kenya

G.V. Owino¹, R. Kasiba², J. Oluoch³, S. Kabura⁴

¹IAVI, Global Affairs, Nairobi, Kenya, ²Sex Worker Research Committee, Kenya, ³Community‐Led Research Advisory Committee, Kenya, ⁴Coast Research Advisory Committee, Kenya

Background: Participation of Female Sex Workers (FSWs) in research remains challenged due to significant stigma, discrimination and in some cases violence from the society and law enforcement. This demeans their priorities in health and research programs that eventually affect ownership and implementation of research. This warrants a different approach of engagement that emphasizes involvement and inclusion of FSWs’ priorities and ownership. We aimed at identifying the challenges faced by FSW in research, develop their own research agenda and a mechanism to take it forward.

Methods: A total of 65 participants (58 FSWs and 7 researchers) convened, from three participatory forums in three regions: Kisumu, Nairobi and Coastal Kenya. The forums were held over 2days per region to identify challenges and gaps in FSW engagement, develop research questions and mechanisms to take this forward.

Results: Gaps were identified in research literacy, engagement and ownership of past and ongoing research, some FSW had knowledge of some past research studies, were involved in some stages, most did not know about past or ongoing research, nor involved in the research life cycle. A disconnect between FSWs’ and researchers’ priorities was identified. FSW research priorities were heavily around structural issues, and to understand complex processes in biomedical research initiatives. FSW top structural priorities included: forms, causes and effects of sexual violence, mental health, existing mechanisms to engage FSW in research, decriminalization of sex work, impact of stigma and discrimination on accessing services, involvement in health policies and decision making, and impact of police and local authorities’ sensitization in dealing with violations. Common biomedical questions were: factors that facilitate FSW to participation, NPTs, PrEP, HIV vaccines and cure research, comorbidities. We developed a research agenda and advisory committees to help address these gaps.

Conclusions: Incorporation of FSWs’ and their research priorities in all stages of research is key for building trust, ownership and successful implementation. Therefore, researchers need to invest in research literacy for biomedical research activities for a better understanding. Utilize FSW advisory mechanisms as a model to empower FSW to be involved in defining and informing research especially structural priorities and results shared with all stakeholders.

Early Vaccine‐Induced V1V2 Antibody Responses in Four Pox‐Protein Public Private Partnership (P5) HIV Vaccine Trials

F. Laher¹, C. DiazGranados², C. Innes³, E. Andersen-Nissen⁴, G.E. Gray⁵, H. Lu⁶, J. Hutter⁷, K. Mngadi³, K. Marshall⁶, L. Polakowski⁷, L.‐G. Bekker⁸, M. Koutsoukos⁹, M. Allen⁷, M. Hosseinipour¹⁰, N. Grunenburg⁶

¹University of the Witwatersrand, Perinatal HIV Research Unit, Johannesburg, South Africa, ²Sanofi Pasteur, United States, ³The Aurum Institute, South Africa, ⁴Cape Town HVTN Immunology Lab, Cape Town, South Africa, ⁵University of the Witwatersrand, Johannesburg, South Africa, ⁶HIV Vaccine Trials Network, United States, ⁷National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, United States, ⁸Desmond Tutu HIV Centre, Cape Town, South Africa, ⁹GSK, Belgium, ¹⁰UNC Project Lilongwe, Malawi

Background: RV144 demonstrated modest HIV preventive vaccine efficacy in Thailand. Amongst multiple identified correlates, IgG V1V2 antibodies were associated with 43% HIV risk reduction. The P5 aimed to further this by investigating subtype C vaccine regimens with different vectors (canarypox, DNA plasmid), adjuvants (alum, MF59, AS01B, or none), gp120 protein doses (40mcg, 200mcg, 600mcg), dosing strategies (prime‐boost, co‐administration) and administration methods (needle‐syringe, Biojector). We compare IgG V1V2 responses across P5 trials with available data.

Methods: In the HVTN100, HVTN107, HVTN108, and HVTN111 trials, serum specimens were obtained two weeks after the fourth vaccination. We measured the frequency and, amongst positive responders, the magnitude (MFI*: background‐subtracted mean fluorescence intensity minus blank responses) of HIV‐1‐specific IgG (1:50 dilution) responses against V1V2 antigens for subtype C, gp70‐TV1.GSKvacV1V2/293F (matched to subtype C protein vaccine), or subtype CRF01_AE, AE.A244 V1V2 Tags/293F (matched to subtype B/E protein vaccine), depending on protein administered. Responses were measured on Bio‐Rad using a standardized custom Luminex HIV‐1 binding antibody multiplex assay.

Results: Amongst these subtype C regimens, the lowest response rate was induced by a regimen of two subtype C ALVAC primes followed by three boosts of subtype C ALVAC +200mcg unadjuvanted gp120 (4/17=23.5%, CI 9.6%‐47.3%); median magnitude 727.1MFI* (IQR 430.1 to 1152.1). The highest response rates were in co‐administration groups of adjuvanted protein plus DNA or ALVAC at all timepoints). Rates, median magnitudes were 93.8%, 12124.8MFI* (DNA +40mcg gp120+AS01B at 3 timepoints); 91.4%, 21756.9MFI* (DNA+200 mcg gp120+AS01B at 3 timepoints); 89.1%, 7147.25MFI* (DNA+ 200mcg gp120+MF59 by needle‐syringe at 3 timepoints); 88.5%, 7745MFI* (DNA by Biojector+200mcg gp120+MF59 by needle‐syringe at 3 timepoints); 86.7%, 6205.1MFI* (subtype C ALVAC+ 200mcg gp120+MF59 at 4 timepoints). For subtype C regimens, pooled response rates were higher for: (i) co‐administration regimens versus regimens omitting protein in the first two doses (152/169= 89.9% versus 233/374=62.3%, p<0.00,001) and (ii) groups with DNA vaccines versus those with ALVAC (202/253=79.8% versus 187/299=62.5%, p<0.00,001).

Conclusions: Strategies which optimized IgG V1V2 responses in P5 subtype C vaccine trials included adjuvanted protein, and co‐administration of adjuvanted protein + DNA or ALVAC vaccination. Responses to DNA administration by needle and syringe or Biojector were equivalent.

Similar antibody responses to subtype C ALVAC‐HIV/gp120 vaccine regimens using different adjuvants (MF59 vs. alum) but enhanced response rates and magnitude when ALVAC‐HIV and gp120 were co‐administered

P. Goepfert¹, K. Mngadi², Z. Moodie³, N. Grunenberg³, J.J. Kee³, K. Seaton⁴, G. Tomaras⁴, E. Andersen-Nissen⁵, O. Dintwe⁵, F. Laher⁶, L. Stranix-Chibanda⁷, N. Naicker⁸, L. Polakowski⁹, M. Koutsoukos¹⁰

¹University of Alabama at Birmingham, Medicine, Birmingham, United States, ²The Aurum Institute, Clinical Research, South Africa, ³Fred Hutch, Vaccine and Infectious Diseases, United States, ⁴Duke University, Surgery, Durham, United States, ⁵Cape Town HVTN Immunology Laboratory, Cape Town, South Africa, ⁶PHRU Vaccines Research, Soweto, South Africa, ⁷University of Zimbabwe College of Health Sciences, Harare, Zimbabwe, ⁸Centre for the AIDS Programme of Research in South Africa, Durban, South Africa, ⁹National Institute of Health, NIAID, Bethesda, United States, ¹⁰GlaxoSmithKline, Belgium

Background: The partially efficacious RV144 HIV‐1 vaccine trial utilized a gp120 protein boost adjuvanted with alum, whereas the non‐efficacious HVTN 702 preventive HIV vaccine utilized a gp120 protein boost adjuvanted with MF59. We compared Env‐specific IgG and IgA responses in HVTN 107, a phase 1/2a partially double‐blinded clinical trial, utilizing a Clade C ALVAC‐HIV prime and 100mcg gp120 protein boost adjuvanted with alum or MF59 vs. a protein‐only boost.

Methods: We randomized 132 participants into 4 groups and analyzed Env‐specific IgG and IgA binding antibodies at Months 6.5 and 12 (NCT03284710).

Results: Robust IgG (96.3% to 100%) and IgA (92% to 93%) responses were seen for all trial arms at the month 6.5 timepoint to gp120 Env proteins contained in the vaccine regimen (ZM96, TV1.C and 1086.C). Compared to all other trial arms, T3 (MF59‐adjuvanted co‐administration) exhibited significantly greater response rates and magnitudes to vaccine‐matched Env proteins at the month 12 timepoint. Comparable IgG response rates and magnitudes to vaccine‐matched Env proteins were observed between the alum and MF59 prime‐boost trial arms. Compared to other groups, T3 exhibited significantly higher IgG response rates to V1V2 panel antigens gp70_B.CaseA_V1V2 (83% vs. 18%‐48%, p <=0.010) and TV1_V1V2 (87% vs. T2: 56% p=0.017; vs. T4: 24% p<0.001) at the 6.5month timepoint. Median response magnitude among positive responders was higher in T3 vs. T1 for TV1_V1V2 (p=0.006). At month 12, response rates to the V1V2 panel antigens were low to moderate across treatment groups, ranging from 0% to 25.0%.

Conclusions: Env‐specific antibody response rates between Alum and MF59‐adjuvanted vaccine regimens were similar, including V1V2 IgG response rates that correlated with decreased HIV risk in RV144. On the other hand, co‐administration of ALVAC‐HIV with gp120 in MF59 induced the highest antibody response rates and magnitudes. Additional studies of vaccine‐elicited antibody specificity and function will inform adjuvant‐specific induction of potentially protective humoral immune responses.

Alum‐adjuvanted protein administered in combination with ALVAC‐HIV downregulates early serum cytokine responses to the vaccine whereas MF59 enhances the early cytokine burst

E. Andersen-Nissen¹, V. Voillet², A. Naidoo², J. Kee¹, A. Fiore-Gartland¹, O. Dintwe², N. Grunenberg¹, L. Polakowski³, L. Fleurs⁴, I. Jani⁵, M Sebe⁶, F. Laher⁷, L. Stranix-Chibanda⁸, N. Naicker⁹, K. Mngadi¹⁰

¹Fred Hutchinson Cancer Research Center, Vaccine and Infectious Diseases Division, Seattle, United States, ²Hutchinson Centre Research Institute of South Africa, South Africa, ³NIAID, Division of AIDS, Bethesda, United States, ⁴Desmond Tutu HIV Foundation, Cape Town, South Africa, ⁵CISPOC, Maputo, Mozambique, ⁶Aurum Tembisa Clinical Research Site, Tembisa, South Africa, ⁷University of Witwatersrand, Perinatal HIV Research Unit, Johannesburg, South Africa, ⁸University of Zimbabwe College of Health Sciences, Harare, Zimbabwe, ⁹Centre for the AIDS Programme of Research in South Africa, Durban, South Africa, ¹⁰The Aurum Institute, Clinical Research Department, South Africa

Background: Vaccine adjuvants are detected by the innate immune system and influence adaptive immune responses. The modestly efficacious RV144 Thai trial used alum to adjuvant the gp120 protein boost, whereas the non‐efficacious HVTN 702 South African trial used MF59. Volunteers enrolled in HVTN 107 (NCT03284710) were administered these adjuvants with a subtype C pox‐protein vaccine, affording a unique opportunity to contrast systemic innate immune responses and potentially inform the results of HVTN 702.

Methods: Eighteen volunteers were enrolled per group. T1, T2 and T4 received 2 doses of ALVAC‐HIV followed by 3 doses of ALVAC‐HIV plus subtype C gp120 administered with MF59 (T1), alum (T2), or un‐adjuvanted (T4). T3 received 4 doses of ALVAC‐HIV with MF59‐adjuvanted gp120. Innate immune responses were measured pre‐vaccination and at days 1, 3 and 7 after the first (T3) or third (T1,T2,T4) vaccinations. Alterations in circulating leukocytes were assessed by CBC. Thirty‐one serum immunomodulatory mediators were quantitated.

Abstract OA18.02‐Table 1.

Results: Neutrophil concentrations increased on day 1 in all groups (FDR‐p's < 0.02) and monocyte concentrations increased in T2,T3 and T4 (FDR‐p's < 0.12); lymphocyte concentrations significantly decreased at day 1 only in T1 (FDR‐p=0.03). Serum cytokines were also significantly altered post‐vaccination in all groups (FDR‐p's < 0.2). On Day 1, nine cytokines were induced in T1, including pro‐inflammatory IFN‐g, CXCL10, TNF‐a, IL‐6, as well as CCL4 and IL‐2, with many factors remaining elevated at Day 3 and returning to baseline on Day 7. In contrast, only IL‐2 and IL‐6 were induced in T2 on Day 1 and twelve factors were repressed on Day 7 relative to pre‐vaccination, including IFN‐g and monocyte‐chemotactic factors CCL8, CCL13, and CCL22. Responses in the MF59 groups differed when co‐administered at the first vaccination (T3) relative to the third vaccination (T1) with 12 cytokines repressed on Day 1 in T3.

Conclusions: Alum‐adjuvanted protein administered in a prime/boost regimen including ALVAC‐HIV reduced early systemic serum cytokine and chemokine responses to the vaccine, in stark contrast to MF59, where induction of a diverse immunomodulatory cytokine profile was observed in Southern African volunteers. Work on elucidating the differential effect of these two adjuvants on the types of innate immune cell responses in humans continues.

Heterologous prime boost immunisations with improved DNA, MVA and protein HIV‐1 subtype C vaccines elicit Tier 2 neutralising antibodies in a Chinese rhesus monkey model

R. Chapman¹, C. Adams¹, A. Keyser¹, M. van Diepen¹, N. Douglass¹, L. Morris², P. Moore², A.‐L. Williamson¹, G. Chege³

¹University of Cape Town, Pathology, Cape Town, South Africa, ²National Institute for Communicable Diseases, Centre for HIV and STIs, Johannesburg, South Africa, ³South African Medical Research Council, Primate Unit and Delft Animal Centre, Cape Town, South Africa

Background: We previously reported establishment of a simian‐human immunodeficiency virus (SHIV) challenge model using Chinese‐origin rhesus macaques (ChRM) for testing the efficacy of HIV vaccines in South Africa. In the current study, we sought to establish proof of concept using a vaccine regimen that had elicited autologous Tier 2 neutralising antibodies (NAbs) in rabbits.

Methods: The Env glycoprotein consensus sequence of the ChRM‐adapted SHIV was determined and utilised in the vaccines designed in this study. DNA and MVA vaccines expressing SIV Gag and HIV Env antigens were constructed and in vitro expression confirmed. A soluble gp140 protein was expressed from a stable HEK293 cell line and purified using lectin affinity chromatography and gel filtration. Six ChRM were inoculated with two DNA, followed by two MVA and finally two protein vaccinations on weeks 0, 4, 8, 12, 20 and 28. Vaccine‐induced T cell immunity was measured by IFN‐γ ELISPOT using peptide pools derived from SIV Gag and subtype C consensus Env while the NAbs were evaluated against Tier 1A (MW965.26), Tier 1B (6644.v2.c33) and Tier 2 (ZM109.B4) pseudovirions. The macaques were then challenged weekly from week 32 until they became infected. An unvaccinated control group was also challenged weekly until they became infected.

Results: The expression and secretion of HIV‐1 Env and SIV Gag by all three vaccine vectors was verified in vitro. Following vaccination, all the animals developed IFN‐γ ELISPOT responses after the DNA vaccinations (median: 255 sfu/million) which were boosted by the MVA inoculations (median: 1031 sfu/million). After protein boost, all animals had NAbs to MW965.26 (median titre: 426.5) and ZM109.B4 (median titre: 29.5) pseudovirions and 3 of 6 to 6644.v2.c33 pseudovirions. The animals in the vaccine group became infected following challenge at a similar rate to the controls, however, median peak viraemia in the vaccine group (1.8x10³ copies/ml) was lower than the controls (1.6x10⁴/ml). Viral replication kinetics were similar in all animals with rapid decline to undetectable levels by 12weeks post infection.

Conclusions: These data provide proof of concept regarding the utility of our ChRM virus challenge model and support further testing of our novel vaccines using this model.

Multivalent antigen presentation increases the antibody binding breadth and neutralizing potency upon the immunization with a self‐assembling HIV env vaccine

J.‐L. Chen¹, C. Fries², M. Dennis³, J. Eudailey³, A. Moody³, S. Permar³, J. Collier², G. Fouda³

¹Duke University, Molecular Genetics and Microbiology, Durham, United States, ²Duke University, Biomedical Engineering, Durham, United States, ³Duke University, Duke Human Vaccine Institute, Durham, United States

Background: Although previous studies have demonstrated that antigen valency is important to diversify the B cell repertoire, the ability of multivalent HIV vaccines to increase the breadth of vaccine‐elicited antibodies and to promote neutralization remain unclear. Herein, we utilized a novel nanomaterial platform (Q11) to evaluate the influence of antigen valency on the HIV vaccine‐elicited antibody responses.

Methods: Self‐assembly of the fiber‐like nanoscale structure of Q11 was triggered by increasing the solvent ionic strength, followed by conjugation of HIV Envelope gycoprotein120 of clade C strains. We constructed vaccines with distinct valency by changing the concentration of antigen in the conjugation. Mice and rabbits were immunized with 15μg of gp120 or Q11‐conjugated gp120 vaccine (gp120‐Q11) along with a TLR7/8 and 9 agonist adjuvant STR8SC. A bead‐based multiplex assay was used to measure antibody binding to heterologous Envs of subtypes B, C, and CRF_AE, and the TZM‐bl cell assay was used to measure neutralization.

Results: Mice immunized with gp120‐Q11 demonstrated higher antibody titers against the autologous Env (p =0.027) after three immunizations; and higher binding to the 4 heterologous Env after each immunization than mice immunized with gp120 (Figure 1). The increased magnitude and breadth were only observed in mice immunized gp120‐Q11 with 3 to 4 antigens on each fiber but not with gp120‐Q11 with lower valency, suggesting that multivalent antigen presentation on Q11 contributed to the enhanced response. Immunization of rabbits indicated that Q11‐gp120 also induced higher neutralization titer against the autologous tier 1 virus than gp120 (p =0.0285).

Conclusions: We demonstrated that increasing the antigen valency by Q11 conjugation enhanced the humoral response in two distinct animal models, showing the potential of such nanomaterial in vaccine delivery. As a more clinically relevant model is required to further assess this platform, we plan to conjugate SOSIP trimer to Q11 and assess its immunogenicity in non‐human primates.

Impact of LACTIN‐V (Lactobacillus crispatus CTV‐05) on genital immunology following standard bacterial vaginosis treatment: Results from a randomized placebo‐controlled trial

E. Armstrong¹, A. Hemmerling², S. Miller³, K. Burke⁴, S. Newmann², S. Morris⁵, H. Reno⁶, S. Huibner⁷, C.R. Cohen², R. Kaul^1,7,8

¹University of Toronto, Institute of Medical Science, Toronto, Canada, ²University of California San Francisco, Department of Obstetrics, Gynecology, & Reproductive Sciences, San Francisco, United States, ³University of California, San Francisco, Department of Laboratory Medicine, United States, ⁴Ruth M. Rothstein CORE Centre and Stroger Hospital of Cook County Health, Chicago, United States, ⁵University of California San Diego, Department of Family Medicine and Public Health, San Diego, United States, ⁶Washington University, Division of Infectious Diseases, St. Louis, United States, ⁷University of Toronto, Department of Medicine, Toronto, Canada, ⁸University Health Network, Toronto General Hospital, Immunodeficiency Clinic, Toronto, Canada

Background: Bacterial vaginosis (BV) is associated with genital inflammation and enhanced HIV risk. In contrast, a vaginal microbiome predominated by Lactobacillus crispatus is associated with lower proinflammatory cytokine levels and reduced HIV risk. We investigated the genital immune impact of a L. crispatus‐based live biotherapeutic (LACTIN‐V), which was recently demonstrated to reduce BV recurrence following standard BV treatment (NCT02766023).

Methods: In this pilot immune study, genital cytokine levels were assayed for 48 participants who reported near‐perfect adherence to assigned treatment during the phase 2b randomized double‐blind placebo‐controlled trial of LACTIN‐V to prevent BV recurrence. All participants received topical metronidazole for 5‐days and were then randomized 2:1 to LACTIN‐V or matched placebo for 11weeks. Vaginal swabs were collected before and after metronidazole therapy, and then at 4‐, 8‐, 12‐, and 24‐weeks. Cytokines were assayed using the MSD multiplex platform. The primary comparison was the impact of LACTIN‐V use on vaginal levels of IL‐1a, the prototypic inflammatory cytokine elevated during BV.

Results: Vaginal IL‐1a levels fell promptly in all participants immediately following metronidazole treatment (3.319 vs. 2.832; p=0.0055). While this reduction was sustained for at least 24weeks among participants subsequently receiving LACTIN‐V (n=32), vaginal IL‐1a levels rebounded to baseline levels by 24weeks in the placebo arm (n=16; t=−2.638, p=0.012). At 24weeks, L. crispatus CTV‐05 was detected by PCR in 16 participants (50%) in the LACTIN‐V arm and 0 participants (0%) in the placebo arm (p=0.001).

Conclusions: Treatment with LACTIN‐V following standard antibiotic treatment of BV not only reduced BV recurrence and increased L. crispatus CTV‐05 colonization, but also resulted in more sustained reductions in vaginal IL‐1alevels. Given the link between genital proinflammatory cytokines and female HIV acquisition, LACTIN‐V may represent a novel strategy to reduce HIV risk among women.

Construction of a comprehensive bacterial genome catalogue for the female genital tract identifies hundreds of new species

M. Hayward¹, S. Bloom¹, M. France², N. Mafunda¹, J. Xu¹, J. Rice¹, M. Ghebermichael³, J. Marrazzo⁴, C. Mitchell⁵, J. Ravel², C. Huttenhower⁶, D. Kwon¹

¹Ragon Institute of MGH, MIT and Harvard, Kwon Lab, Boston, United States, ²Institute for Genome Science, University of Maryland School of Medicine, United States, ³Ragon Institute of MGH, MIT and Harvard, Boston, United States, ⁴University of Alabama at Birmingham, Division of Infectious Diseases, United States, ⁵Massachusetts General Hospital, Obstetrics and Gynecology, United States, ⁶Harvard TH Chan School of Public Health, Boston, United States

Background: Specific cervicovaginal bacteria are associated with elevated inflammation in the female genital tract (FGT) and a number of poor reproductive outcomes, including HIV acquisition. These associations have been identified using bacterial 16S rRNA gene sequencing, which has limited resolution and often fails to achieve species‐level taxonomic assignment. Genome sequencing of bacteria affords strain‐level resolution, revealing that many species contain enough genes to make multiple independent genomes, reflecting a tremendous amount of diversity in potential phenotypes (virulence factors, antibiotic resistance, metabolic function, etc.). Few studies have used these methods to characterize the FGT microbiome, leaving much of the microbial diversity unexplored.

Methods: We isolated 1400 bacteria using selective media and performed whole genome sequencing, applying cutting‐edge methods for combining these data with assembled microbial genomes from over 1200 newly shotgun sequenced metagenomes from multiple geographic locations, to produce a comprehensive genome catalogue. Imposing species‐level phylogenetic clustering on the resulting genomes allowed us to identify both known and previously unknown species. Mapping of shotgun metagenomic reads to these genomes allowed us to determine prevalence and amount of genetic diversity accounted for by these new species across multiple demographic factors (race, ethnicity, geography, pregnancy and menopause).

Results: We identified 605 species in the female genital tract, 130 of which have not been previously described. Of these, 46 were only found in samples from African American and African women. Many of these new species were highly prevalent and co‐occurred with known and previously unknown species in community assemblages. Furthermore, we significantly increased observable genetic diversity for known species associated with poor reproductive outcomes. One notable example is Prevotella bivia, a bacterium associated with increased HIV acquisition risk, was found to be different in both metabolic and antibiotic resistance gene content in South African isolates when compared to those from North America.

Conclusions: We have massively expanded the observable genetic diversity in the FGT microbiome and identified both new species and previously unrecognized sub‐species structures. This catalogue will contribute to developing novel approaches to modulate the vaginal microbiome and reduce poor reproductive outcomes, including HIV acquisition risk.

Transcriptional analysis of cervical cell populations reveals inflammatory signatures and differential epithelial keratinization associated with Depo‐Provera use

E.H. Byrne¹, M. Farcasanu², S. Bloom³, N. Xulu⁴, J. Xu², S. Handley⁵, J. Bramante², C. Mitchell⁶, A.‐C. Villani⁷, D.S. Kwon²

¹Ragon Institute, Harvard Medical School, United States, ²Ragon Institute, United States, ³Massachusetts General Hospital, Infectious Disease, United States, ⁴Ragon Institute, South Africa, ⁵Washington University School of Medicine, Pathology and Immunology, United States, ⁶Massachusetts General Hospital, Obstetrics and Gynecology, United States, ⁷Massachusetts General Hospital, Center for Immunology and Inflammatory Diseases, United States

Background: Injectable progestin‐only contraceptives (IPCs), including Depo‐Provera (DMPA), have been associated with HIV acquisition risk in some epidemiological studies and in animal models. This risk has not been consistently observed and the risk of IPCs must be weighed against the known benefits of effective contraception. Regardless, understanding how IPCs shape the mucosal environment of the female genital tract can help elucidate biological mechanisms underlying their potential impact on HIV acquisition risk and may lead to new HIV prevention strategies.

Methods: Cervical cytobrushes were collected from HIV‐uninfected women aged 18 to 23years through the FRESH (Females Rising through Education, Support, and Health) study in Umlazi, South Africa, and cervical epithelial and CD4+ T cells were FACS‐sorted. RNAseq was performed on the sorted cells and data was analyzed using STAR, HTSeq, DESeq2, and GSEA.

Results: Cervical epithelial cells were compared between women who reported using DMPA (n=6) and women in the follicular phase of the menstrual cycle by reported last menstrual period (n=6). tSNE clustering showed clear segregation between groups, and 369 genes were significantly differentially expressed. Of these genes, two of the most highly expressed in DMPA users were CCL3 and CCL3L3, mediators of T cell homing. GSEA analysis demonstrated that the epithelial gene sets most significantly over‐expressed in DMPA‐using women were keratinization, cellular response to IFNg, and keratinocyte differentiation (FDR q<0.05). Parallel analysis of cervical CD4+ T cells identified 5 significantly differentially expressed genes (fold change >90) and a significantly enriched gene set in DMPA users, “positive regulation of T cell mediated immunity” (FDR q‐val 0.087).

Conclusions: In DMPA users, cervical epithelial cells express higher levels of the T cell homing cytokines CCL3 and CCL3L3, while cervical CD4+ T cells exhibit inflammatory gene set enrichment. At the same time, keratinization pathways may be altered in DMPA users, as suggested in previous literature. Together, these results suggest possible mechanisms by which progestins may contribute to genital inflammation and/or differential barrier function in the genital mucosa. These findings may point towards potential biological approaches to reduce HIV acquisition risk in women.

Silk fibroin as a mucosal delivery vehicle for Griffithsin and antiviral compounds: effective protection of macaques and high acceptance among user groups

P. LiWang¹, K. Crakes², C. Herrera³, J. Morgan¹, A. Lopez⁴, M. Palupi Rasajan⁴, S. Dandekar², N. Burke⁴

¹University of California Merced, Molecular Cell Biology, Merced, United States, ²University of California Davis, Department of Medical Microbiology & Immunology, Davis, United States, ³Imperial College, Department of Medicine, London, United Kingdom, ⁴University of California Merced, Public Health, Merced, United States

Background: While there are methods to prevent HIV infection, there is still a need for a prevention device that is effective, user‐controlled, easily transported without refrigeration, inexpensive and discreet. Importantly, such a device should be attractive to the user to avoid the problem of low adherence, particularly among at‐risk youth. We have developed the silk fibroin (SF) platform to encapsulate and release a single or combination of HIV inhibitors, including the lectin Griffithsin (Grft), that does not need refrigeration. These small silk devices can be inserted vaginally and rectally, and be formulated either for quick‐dissolve or sustained release.

Methods: We have used silk‐Grft inserts in female rhesus macaques (n=11), placing them vaginally and rectally. These compartments were swabbed to test for amount of Grft released and changes in local microbiome. Tissue biopsies were obtained and tested for protection against SHIV infection.

To determine the user acceptability of silk inserts, we held four focus groups in California during summer 2020. These groups were composed of sexually active women and men, both straight and gay, who were shown a variety of silk fibroin inserts. Participants discussed HIV prevention knowledge and practices in the context of varying relationship statuses, impacts of COVID‐19 on their sexual activity, and their perceptions of the silk inserts.

Results: The Silk‐Grft inserts readily dissolved in the vaginal and rectal compartments in macaques, where Grft levels were substantially above 100‐fold IC50, indicating effective release of Grft inhibitor. Tissue biopsies from both compartments were fully protected against ex vivo SHIV infection, with neither inflammation nor change in microbiome.

In focus groups, young, sexually active adults uniformly expressed positive sentiments and high interest when presented with silk inserts in an array of shapes and colors. These opinions included repeated preference for silk inserts over oral PrEP, due to the utility of silk inserts as spontaneous protection that did not require daily dosing.

Conclusions: Together, these recently obtained results show that silk fibroin is an excellent candidate for formulating HIV inhibitors, being both highly effective in preventing infection and also attractive to young, sexually active adults, which may lead to higher adherence and protection.

Pharmacokinetic and pharmacodynamic study of tenofovir and tenofovir alafenamide for PrEP in foreskin tissue

C. Herrera¹, L. Else², S.D. Penchala³, A.‐D.A. Pillay⁴, T.B. Seiphetlo⁴, L. Lebina⁵, C. Callebaut⁶, N. Martinson⁵, J. Fox⁷, S. Khoo²

¹Imperial College London, Department of Infectious Disease, London, United Kingdom, ²The University of Liverpool, Department of Molecular and Clinical Pharmacology, Liverpool, United Kingdom, ³The University of Liverpool, Liverpool, United Kingdom, ⁴University of Cape Town, Division of Immunology, Cape Town, South Africa, ⁵University of the Witwatersrand, Perinatal HIV Research Unit, Johannesburg, South Africa, ⁶Gilead Sciences, Foster City, United States, ⁷Guys and St. Thomas’ NHS Foundation Trust and King's College London, London, United Kingdom

Background: Pre‐exposure prophylaxis (PrEP) studies have focussed predominantly on the efficacy in female reproductive and colorectal tracts, with limited research on the efficacy of PrEP candidates in the male genital tract. We assessed the ex vivo pharmacological profile of Tenofovir (TFV) and Tenofovir alafenamide (TAF) in foreskin tissue to inform the design of the CHAPS oral PrEP trial (NCT03986970).

Methods: Foreskin specimens were obtained with signed informed consent from HIV‐negative males who voluntary requested medical circumcision. Inner mucosal and outer skin were cut in explants and exposed to serial dilutions of TFV or TAF for one hours prior to addition of HIV‐1BaL at a high (HVT) or a low viral titre (LVT) for two hours. Infection was assessed at different time points during 15days of culture by measurement of p24 in culture supernatants. TFV, TAF and TFV‐diphosphate (TFV‐DP) concentrations were measured in tissue, culture supernatants and dosing and washing solutions using LC‐MS methods.

Results: Dose‐response curves were obtained for both drugs against the two viral titres tested with greater inhibitory potency observed against LVT. Inhibitory equivalency mimicking oral dosing was defined between 1mg/mL of TFV and 15μg/mL of TAF against HVT for the dosing post‐ex vivo challenge included in the design of CHAPS trial. Concentrations of TFV‐DP in foreskin explants were at least 5 times higher after ex vivo dosing with TAF vs. TFV. Statistically significant negative linear correlations were observed between explant TFV‐DP levels and p24 concentrations following HVT (r²=0.6867, p =0.0001 for TFV and r²=0.6696, p =0.0002 for TAF).

Conclusions: Pre‐clinical evaluation of TAF reveals greater potency than TFV against penile HIV transmission. Ex vivo dose‐challenge studies in human foreskin explants can be used as surrogate for in vivo studies to compare doses and preventive agents to be included in clinical trials.

Alternative T cell effector functions are linked to humoral responses to HIV infection

L. Romero-Martin¹, M.L. Rodriguez de la Concepción², J. Carrillo², J. Blanco², B. Mothe³, M. Ruiz-Riol², C. Brander², A. Olvera²

¹Institut de Recerca de la Sida (IrsiCaixa), Hospital Universitari German Trias I Pujol, T Cell Immunity and Vaccinology, Barcelona, Spain, ²Institut de Recerca de la Sida (IrsiCaixa), Hospital Universitari German Trias I Pujol, Barcelona, Spain, ³Fundació Lluita contra la Sida, Infectous Diseases Department, Hospital Universitari Germans Trias i Pujol, Spain

Background: T cell populations showing a particular cytokine polarization may have a key role in inducing and maintaining effective cellular and humoral immunity to HIV. In addition to the commonly assessed Th1 and CTL responses, other T cell profiles involved in the immune response to HIV infection are not well characterized and their contribution to virus control remains unknown.

Methods: Cryopreserved PBMCs from HIV‐1 controllers (n=15, median plasma HIV‐1 RNA, pVL: 210 copies/mL, median CD4+ T‐cell count: 786 cells/mm³) and non‐controllers (n=15, median pVL: 52,435 copies/mL, median CD4+ T‐cell count: 406 cells/mm³) were stimulated with 17 peptide pools covering the entire HIV proteome. Effector function profiles of virus‐specific T‐cells were evaluated by “boosted flow cytometry”; a novel methodology based on labeling cytokines that characterize specific T cell profiles with the same fluorochrome (Th1/Tc1: INFg, IL‐2, TNFa; Th2/Tc2: IL‐4, IL‐5, IL‐13; Th17/Tc17: IL‐17A, IL‐22; Treg: IL‐10, TGFb and Tfh/Tfc: IL‐4, IL‐21). We also evaluated the humoral response to HIV (IgM, IgA and IgG) to identify T‐cell profile associated with humoral response. Finally, cellular and humoral responses correlated with respective clinical parameters (pVL, CD4 count).

Results: The total magnitudes of HIV‐specific CD4+ Tfh and CD8+ responses with a Tc2 or Tfc profiles were significantly higher in HIV controllers. Total Tfh/Tfc (CD4+ and CD8+) responses correlated negatively with pVL. Different polarization profiles were observed for responses targeting different HIV proteins, and overall a broader range of alternative effector functions were seen in cells from HIV controllers (Rev/Vpr/Env‐specific Tc2, Pol/Env‐specific Tfc and Pol‐specific Tfh responses). HIV controllers also showed higher IgG responses to HIV‐gp120 while their virus‐specific IgM antibodies titers were lower than in HIV non‐controllers. IgM titers correlated positively with pVL and negatively with HIV Tfc‐specific responses. The frequency of Tfc cells correlated with the gp120‐IgG/IgM ratio, supporting a possibly pivotal role of those in HIV control and antibody isotype switching.

Conclusions: We have identified a subpopulation of CD8+ T‐cells with Tfc‐like cytokine profile targeting HIV Pol and Env that correlated with pVL and antibody class switching. Further characterization of those Tfc‐like cells is needed to confirm their role in natural HIV control.

HIV‐resistant dual CD28/4‐1BB costimulated CAR T cells mitigate HIV pathogenesis in humanized mice

D. Claiborne¹, C. Maldini², K. Okawa³, T. Chen¹, D. Dopkin², X. Shan², K. Power¹, R. Trifonova¹, K. Krupp¹, M. Phelps¹, V. Vrbanac¹, S. Tanno¹, T. Bateson¹, G. Leslie², J. Hoxie²

¹Ragon Institute of MGH, MIT and Harvard, Massachusetts General Hospital, Boston, United States, ²University of Pennsylvania, Philadelphia, United States, ³Ragon Institute of MGH, MIT and Harvard, Boston, United States

Background: A potent and sustained T cell response will likely be a requisite component of an effective HIV cure. Chimeric antigen receptor (CAR) T cells, whereby autologous T cells are engineered with specific antigen‐targeting and functional attributes, have been effectively employed against a number of chemotherapy‐refractive cancers. However, a highly effective CAR T cell against HIV has yet to be developed, which could represent a powerful approach to target the HIV reservoir.

Methods: Bone marrow, liver, thymus (BLT) humanized mice were used to iteratively test CAR T cell efficacy against both acute and disseminated HIV infection. A CD4 ectodomain‐expressing CAR was used to target HIV Env‐expressing cells. We compared CAR T cells expressing the 4‐1BB and CD28 co‐stimulatory domains either individually, in a tandem “3^rd generation” construct, or as a novel Dual CAR T cell simultaneously expressing independent 4‐1BB and CD28 co‐stimulated CARs. A CXCR4‐C34 fusion inhibitor was co‐expressed to protect CAR T cells from HIV infection.

Results: HIV‐specific CAR T cells expressing 4‐1BB or CD28 costimulatory domains exhibited either enhanced proliferation and survival or enhanced effector function respectively, and the 3^rd generation tandem construct took on the phenotype of the CD28 membrane proximal domain; all constructs had negligible impact on HIV pathogenesis in vivo. In contrast, Dual CAR T cells simultaneously expressing independent 4‐1BB and CD28 costimulated CARs exhibited the proliferation and survival of 4‐1BB and the effector function of CD28 and prevented HIV‐induced CD4+ T cell loss despite persistent viremia. Importantly, protection of the Dual CAR T cells from HIV infection through expression of the C34‐CXCR4 fusion inhibitor significantly increased their efficacy resulting in reductions in acute phase viremia, accelerated viral suppression in combination with ART, and reduced viral burden in lymphoid tissues.

Conclusions: These data describe a novel CD4‐based Dual CAR approach, harnessing the favorable attributes of both 4‐1BB and CD28 costimulation. When protected from HIV infection, these Dual CAR T cells were capable of mitigating HIV pathogenesis in the context of a robust and fully disseminated HIV infection in vivo, while reducing tissue viral burden, a goal of HIV cure strategies.

Immune perturbation is more profound in newborn than in infant macaques during acute SHIV infection

N. Haigwood¹, M. Shapiro², T. Cheever¹, S. Pandey¹, E. Mahyari³, K. Onwuzu³, J. Reed⁴, H. Sidener⁵, J. Smedley¹, L. Colgin⁵, A. Lewis⁵, A. Johnson⁵, B. Bimber³, J. Sacha¹, A. Hessell¹

¹Oregon Health & Science University, Pathobiology & Immunology, ONPRC, Beaverton, United States, ²Oregon Health & Science University, Molecular Microbiology & Immunology, Beaverton, United States, ³Oregon Health & Science University, Genetics, ONPRC, Beaverton, United States, ⁴Oregon Health & Science University, VGTI, Beaverton, United States, ⁵Oregon Health & Science University, Comparative Medicine, ONPRC, Portland, United States

Background: Progress in preventing vertical HIV‐1 transmission has stalled, with 160,000 infants infected annually. Newborns and infants exhibit higher viral loads and more rapid disease progression than adults and older children, with the lowest survival rates in those infected as newborns. While immunologic immaturity likely promotes pathogenesis and poor viral control, little is known about immune damage in infants. Neonatal immunity is distinct from that of adults in ways that are presumed to hinder newborns’ ability to control viral infection.

Methods: Here we examined virologic and immunologic outcomes in rhesus macaques exposed to SHIVSF162P3 at one to two weeks or four months of age. To answer this question, we compared virologic outcomes, adaptive immune responses, frequencies and phenotypes of key leukocyte subsets, and transcriptome profiles during pathogenic SHIV infection in Newborn (one to two weeks old) and Infant (15 to 16weeks old) rhesus macaques that were lacking the two major MHC‐I alleles for post‐acute viral control.

Results: Although differences in plasma viremia and seeded reservoirs during acute infection were minimal, we observed age‐dependent alterations in leukocyte subsets and gene expression. Compared with infants, newborns had stronger skewing of monocytes and CD8+ T cells toward differentiated subsets and little evidence of type I interferon responses by transcriptomic analysis. Infants had evidence of a robust antiviral program mediated by type I IFN responses. In contrast, newborns had a transcriptional signature dominated by the downregulation of genes involved in mitotic, cell cycle, and activation processes; the near‐absence of a viral infection response; and upregulation of CXCL10, a biomarker of greater disease severity.

Conclusions: Taken together, our findings suggest a role for defective innate defenses and elevated immune activation in age‐dependent differences in pathogenesis. We conclude that SHIV, like HIV‐1, wreaks greater havoc in newborns than in infants infected at older ages, reinforcing the validity of this model for understanding mechanisms of pathogenesis in vertically acquired HIV infection. Ultimately, therapeutic intervention as early as possible after birth is more likely to have a durable effect on disease progression by limiting damage to the immune system.

Tissue‐resident memory CD4+ T cells in ectocervical versus endocervical specimens

A. Ssemaganda¹, K. Nyambura², F. Mulhall³, F. Nuhu¹, N. Jahan¹, K. Downing³, B. Sandberg³, H. Elands³, R. Groff³, A. Altman³, C. Robinson³, V. Poliquin³, J. Arsenio², L.R. McKinnon¹

¹University of Manitoba, Department of Medical Microbiology and Infectious Diseases, Winnipeg, Canada, ²University of Manitoba, Departments of Internal Medicine and Immunology, Winnipeg, Canada, ³Health Sciences Winnipeg, Winnipeg, Canada

Background: HIV susceptibility is likely determined in women at the female genital tract (FGT), with initial targets of the virus during transmission being CD4+ T cells. HIV target cells can be divided into subsets based on several criteria including their memory status. Little attention has been paid to a relatively new subset of tissue‐resident memory cells (TRM) in the FGT, including their relative frequency and ability to regulate the genital tract milieu and HIV infection risk.

Methods: We enrolled reproductive‐aged women from gynecological clinics in Winnipeg, Canada. Specimen collection included vaginal swabs, SoftCup^TM samples, endocervical cytobrushes, and ectocervical biopsies. Immune cells collected from the latter two specimen types were analyzed by flow cytometry and sorted for single‐cell RNA sequencing using the 10x genomics chromium system. Sociodemographic, behavioral and clinical data were collected using a standard questionnaire. This abstract includes pilot results from an ongoing analysis.

Results: We analyzed samples from 5 women by flow cytometry included a subset that was sorted (CD45+) for single‐cell analyses. TRM (defined as live CD3+ CD4+ CD69+ CD103+ cells) were ~4 times more frequent [Mean (SD)] in ectocervical biopsies [12.14 (7.78)] compared to endocervical cytobrushes [3.67 (3.41)], p=0.098. Phenotypically, Treg (CD25+ CD127low FoxP3+) were enriched by ~3fold in the CD103‐CD69+ T cell subset compared to CD103+ CD69+ TRM, while Th17 (CCR6+ CD161+) frequencies were similar between subsets. To date, all biopsies contained sufficient cells for capturing ~6000 CD45+ cells for single‐cell RNA sequencing.

Conclusions: CD4+ TRM were more commonly isolated in the ectocervical tissue compared to from endocervical cytobrush samples, possibly because the latter sample type is from columnar vs stratified squamous epithelium, and/or because it represents a superficial collection method. In depth characterization of the phenotype and function of CD4+ TRM from cervicovaginal tissue could have important implications for understanding microbiota interactions, genital inflammation, and HIV risk in women.

HIV‐1 infection and the control of viral replication are associated with greater expression of interleukin‐21 receptor on CD8+ T cells

J. Dalel, U.S. Kuong, P. Hayes, L. Black, S. Joseph, D. F. King, J. Makinde, J. Gilmour

Imperial College London, Human Immunology Laboratory, London, United Kingdom

Background: Interleukin‐21 plays an important role in the development of the immune response where it has been linked with the generation of virus‐specific memory CD8+ T‐cells, limiting exhaustion and promoting effector function during viral infection. However, little is known about the expression of interleukin‐21R (IL‐21R) during HIV‐1 infection or its role in HIV‐1 specific CD8+ T cell maintenance and subsequent viral control.

Methods: We compared IL‐21R expression on total and memory subsets of CD8+ T‐cells from HIV‐1 positive and HIV negative donors. We also measured IL‐21R on antigen specific CD8+ T cells in volunteers who were positive for HIV‐1 and had CMV responding T cells.

Results: IL‐21R expression was significantly higher on CD8+ T cells (p = 0.0050), and on naive (p =0.0003), central memory (p =0.0008) and effector memory (p =0.0105) CD8+ T cell subsets from HIV‐1+ve individuals relative to HIV‐1‐ve individuals. For those infected with HIV‐1, the levels of IL‐21R expression on HIV‐1‐specific CD8+ T cells correlated significantly with visit viral load (r= 0.8571, p = 0.0238, n=7). Lastly, CD8+ T cells from individuals with lower set point viral loads and demonstrated viral control had the lowest levels of IL‐21R expression.

Conclusions: Our data demonstrates significant associations between IL‐21R expression on peripheral CD8+ T cells and viral load and disease trajectory. This suggests that the IL‐21 receptor could be a novel marker of CD8+ T cell dysfunction during HIV‐1 infection.

Effect of navigation services to achieve viral suppression among men who have sex with men (MSM) and transwomen living with HIV who attend public HIV care services in Lima, Peru

G. Calvo¹, K.A. Konda¹, C.F. Cáceres¹, X. Salazar Lostanau¹, A. Silva-Santisteban¹, A. Maiorana², S.M. Kegeles²

¹Universidad Peruana Cayetano Heredia, Center for Interdisciplinary Research in Sexuality, AIDS and Society, Peru, ²University of California San Francisco, Center for AIDS Prevention Studies, United States

Background: In Peru, considerable gaps remain in achieving the goal of the HIV cascade, undetectable viral load. Services provided by navigators to facilitate linkage and retention in HIV‐care have proven successful; however, their utility has rarely been studied in Latin America, nor specifically among MSM/Transwomen (TW). We studied if navigation services delivered by peer navigators can facilitate MSM/TW HIV‐care engagement in Lima.

Methods: As part of the Orgullo+ intervention, four peer counselors living with HIV already working in the intervention area were trained as navigators to use strengths‐based counseling. MSM/TW living with HIV attending public health services in this area were assigned to a navigator. Navigators worked with the MSM/TW to help them link or reengage in HIV‐care, while the comparison area's cohort at other health facilities received standard services. We collected health care data from these cohorts of MSM/TW to estimate the effect of navigators on achieving undetectable viral load using multivariable Cox regression.

Abstract OA21.01‐Table 1.

Results: We enrolled, 364 MSM/TW, 192 were assigned a navigator. Individuals with a navigator included more TW, more reporting difficulties in linkage‐to‐care and in attending their appointments, all (p<0.001). At the endline, 79% of MSM/TW with a navigator had an undetectable viral load compared to 67% of MSM/TW without a navigator (p=0.008). In multivariable model, MSM/TW with a navigator achieved viral suppression 79% faster than MSM/TW without (p<0.001) and those reporting difficulties in attending their appointments achieved undetectable viral load 35% later than those without difficulties. No significant differences in achieving undetectable viral load were found by population group, age, linkage‐to‐care difficulties or time to linkage‐to‐care (all p>0.05).

Conclusions: Working with a navigator helped MSM/TW achieve an undetectable viral load in a shorter time compared to those who received standard services. Our study suggests that receiving navigation services can help to successfully engage and retain MSM/TW in HIV‐care.

Methadone program participation and current ART use among people who inject drugs in Kenya

L. Mbogo¹, A. Monroe‐Wise², B. Sambai¹, B. Guthrie², D. Bukusi³, W. Sinkele⁴, P. Macharia⁵, N. Ludwig⁶, J. Herbeck⁶, M. Dunbar⁶, E. Gitau⁴, C. Farquhar², H. Musyoki⁵, S. Masyuko⁵

¹University of Washington/KNH, HIV Testing and Counselling, Kenya, ²University of Washington, Global Health, Seattle, United States, ³Kenyatta National Hospital, HIV Testing and Counselling, Kenya, ⁴SAPTA, Kenya, ⁵NASCOP, Nairobi, Kenya, ⁶University of Washington, Seattle, United States

Background: In Kenya HIV prevalence among people who inject drugs (PWID) is estimated at 18% versus 5.6% in general population. To reduce opioid addiction and HIV risk, methadone was introduced in Nairobi and Kenya's coastal region in 2013. Studies in high‐income countries have shown that methadone is associated with antiretroviral therapy (ART) uptake, however, it is not known whether participation in a methadone program is associated with increased ART use among PWID in Kenya.

Methods: Participants were recruited from an ongoing study of Assisted Partner Services (APS) to identify, test and link to care the sexual and injecting partners of HIV‐positive PWID in Kenya. Recruitment for index participants was done from NSP programs and methadone clinics in Nairobi, Kilifi and Mombasa countie where they were receiving HIV care. In this study we interviewed all HIV‐positive index participants who had injected drugs in the last one year about linkage to care, ART uptake, methadone use and current injecting practices. We used logistic regression to evaluate associations between methadone program participation and current ART use, adjusting for demographic and behavioral characteristics.

Results: A total of 668 HIV‐positive index participants were enrolled. Of these, 138 (21%) were on methadone and 527 (79%) reported they were taking ART. The majority of participants were male (52%) and median age was 36years (IQR 31 to 42). Of 138 currently enrolled in methadone programs, 128 (93%) were on ART compared to 399 (75%) of the remaining 530 not taking methadone. After adjusting for age, sex, marital status, and living conditions, those on ART were over 4 times more likely to take methadone (adjusted Odds Ratio [aOR]=4.21, p<0.001). Of 346 men 89 (26%) were on methadone compared with 49 (15%) of 322 women (aOR=0.45, p<0.001). Women were also less likely to be on ART compared to men although this difference was not statistically significant (aOR=0.72, p=0.08).

Conclusions: Methadone use was strongly associated with reported ART use amongst PWID in Kenya. Women were significantly less likely than men to take methadone and were also less likely to be taking ART. These findings can guide policies and practices for targeted methadone support programs in Kenya.

Using an HIV risk assessment tool to increase frequency of HIV testing in men who have sex with men in Beijing, China: an app‐based randomized, controlled trial

Q. Luo¹, Z. Wu², G. Mi²

¹Binzhou Medical University, School of Nursing, Yantai, China, ²China Center for Disease Control and Prevention, Beijing, China

Background: Low risk perception hinders human immunodeficiency virus (HIV) testing among men who have sex with men (MSM). We aimed to promote HIV testing uptake and reduce sexual risk behaviors through a social networking application‐based intervention.

Methods: This was a randomized control trial conducted from October 2017‐ September 2018 among MSM in Beijing, China. Participants were randomly assigned into three groups: Group 1, Group 2, and Control received routine, comprehensive HIV education. In addition, Group 1 took an HIV risk assessment tool and received tailored feedback based on their risk assessment score, while Group 2 took the HIV risk assessment only. The number of HIV tests over twelve‐month study period was assessed using an intention to treat analysis, and the proportion of self‐reported unprotected anal intercourse (UAI) over the six‐month period after randomization was assessed by modified intention‐to‐treat analysis.

Results: 9280 MSM were recruited and randomly assigned into Group 1 (n=3028), Group 2 (n=3065), or the Control group (n=3187). Over the twelve‐month study period, Group 1 took 391 tests (mean 2.51 tests/person), Group 2 took 352 tests (mean 2.01 tests/person), and the Control group took 295 tests (mean 1.72 tests/person). Participants in Group 1 had an increased mean number of HIV tests within twelve months compared to the Control group (IRR: 1.32, [95% CI: 1.09–4.58, p=0.01]). The proportion of UAI was not statistically different among the three groups.

Conclusions: Implementing repeated HIV risk assessments coupled with tailored prevention messaging, particularly using an app‐based tool, could be incorporated into routine HIV prevention to promote HIV testing among MSM in China.

Peer‐distributed HIV self‐test kits to increase demand for HIV prevention and care services in rural KwaZulu‐Natal, South Africa: a three‐armed cluster‐randomised trial comparing social‐networks versus direct delivery

M. Shahmanesh¹, N. Mthiyane², C. Herbst², O. Adeagbo², M. Neuman³, P. Mee³, J. Dreyer², N. Chimbindi², T. Smit², N. Okesola², T. Zuma², G. Harling¹, N. McGrath⁴, J. Seeley³, F.M. Cowan⁵

¹University College London, Institute for Global Health, United Kingdom, ²Africa Health Research Institute, Durban, South Africa, ³London School of Hygiene and Tropical Medicine, London, United Kingdom, ⁴University of Southampton, Southampton, United Kingdom, ⁵Liverpool School of Tropical Medicine, Zimbabwe

Background: HIV elimination in South Africa calls for innovative approaches. We investigated HIV self‐testing (HIVST) for increasing demand for HIV care and prevention services, comparing two peer‐distribution approaches (direct distribution and an incentivized social‐network approach) against peer‐navigators providing information and referrals only.

Methods: Restricted randomisation (1:1:1) allocated 24 peer‐navigator pairs in rural Kwa‐Zulu Natal into: (1) incentivized‐peer‐networks (IPN): peer‐navigators recruited “seeds” to distribute 5 packs to 18 to 30year olds within their social networks. Packs contained 2 HIVST kits (OraQuick) and standard of care (SOC) materials. “Seeds” received 20 Rand (US$1.5) for each recipient who went on to distribute packs themselves; (2) peer‐navigator‐distribution (PND): peer‐navigators distributed HIVST packs and SOC materials directly; (3) SOC: peer‐navigators distributed barcoded clinic referral slips and sexual health information promoting HIV testing, pre‐exposure prophylaxis (PrEP) and antiretroviral therapy (ART). PrEP/ART linkage rates were defined as numbers screened for PrEP eligibility or starting ART within 90days of referral slip distribution per peer‐navigator outreach month (pnm). The primary outcome compared PrEP/ART linkage rates between arms for women aged 18 to 24years. Secondary outcomes included linkage rates for youth (18 to 30years). Intention to treat analysis was used. Investigators and statisticans were blinded to allocation.

Results: Between March and December 2019, 7563 (2055 IPN, 2069 PND, 2539 SOC) packs were distributed during 144 peer‐navigator outreach months, with 272 young adults linked to PrEP/ART (1.9/pnm). Linkage rates for women aged 18 to 24years were lower for IPN (n=26, 0.54/pnm) than PND (n=45, 0.80/pnm) and SOC n=49, 0.85/pnm), although not significantly so (rate ratios [RR] 0.68, 95% CI 0.28 to 1.66 and 0.64, 95% CI 0.38 to 2.36, respectively). Adding HIVST kits to peer‐navigator distribution (PND vs SOC) did not change linkage rates (RR 0.95, 95% CI 0.38 to 2.36). Youth (18 to 30years), results were similar but with stronger evidence of lower linkage rates (0.88/pnm) for IPN than PND (2.11/pnm, RR 0.42, 95% CI 0.18 to 0.98) and SOC (2.07/pnm, RR 0.42 95% CI 0.18 to 1.02).

Conclusions: Peer‐based HIVST distribution reached large numbers of young men and women, but did not increase demand for PrEP/ART, unless combined with peer‐navigator PrEP/ART promotion. Incentivized peer network models resulted in fewer linkages compared to direct peer‐navigator mobilization with or without HIVSTs.

Registration: NCT03751826.

Partners support, disclosure and side effects: facilitators of high maternal PrEP adherence in Cape Town, South Africa

D. Joseph Davey¹, L. Knight², K. Dovel³, N. Tsawe⁴, N. Mashele⁴, J. Markt-Maloney¹, Y. Gomba⁴, P. Gorbach¹, L.‐G. Bekker⁵, T. Coates³, L. Myer⁴, PrEP‐PP study

¹University of California Los Angeles, Epidemiology, Los Angeles, United States, ²University of Cape Town School of Public Health and Family Medicine, Cape Town, South Africa, ³UCLA, Medicine, Los Angeles, United States, ⁴University of Cape Town School of Public Health and Family Medicine, Division of Epidemiology and Biostatistics, Cape Town, South Africa, ⁵Desmond Tutu Health Foundation, Cape Town, South Africa

Background: HIV incidence is high during pregnancy and postpartum in many settings. PrEP is safe and effective but requires adherence during potential HIV exposure, yet the facilitators of high maternal adherence are not well understood in high HIV burden settings.

Methods: We conducted semi‐structured interviews with women who reported high adherence (PrEP use ≥25days in last 30‐days) within a PrEP service for pregnant and postpartum women located in a large primary care facility in a high‐HIV burden township. Topics for interviews included: individual and interpersonal risk, disclosure, anticipated PrEP stigma, safety, side‐effects, and facility‐level factors effecting adherence. A thematic approach guided an iterative process of coding (reviewed to ensure intercoder reliability) and analysis using NVivo 12.

Results: We interviewed 25 postpartum women with high PrEP adherence who were on PrEP for a median of nine‐months, median age 26‐years, and median baseline gestational age 24‐weeks. Themes identified as key drivers of optimal PrEP use were HIV risk perception – primarily due to partner's perceived risky sexual behaviors and unknown serosatus and a strong desire to have a baby free of HIV. Reported disclosure of PrEP use to family, partners and friends facilitated PrEP adherence. Women continued PrEP postpartum because they felt empowered by PrEP and did not want to “go backwards” and increase their HIV risk as before PrEP. Women who reported high adherence all discussed having community support and reminders to take PrEP on time. The primary barriers were anticipated or experienced stigma, which most overcame through education of partners/family about PrEP. Pregnant women experienced transient side effects, but found ways to continue, including taking PrEP at night. Women believed PrEP education and counselling were accessible when integrated into antenatal care which contributed to high PrEP adherence.

Conclusions: Facilitators of optimal PrEP use through pregnancy and postpartum included fear of HIV acquisition for self and infant, mostly due to partner sexual behaviors and unknown serostatus, along with PrEP disclosure, and encouragement from partners and family. PrEP programs for pregnant and postpartum women should integrate strategies to assist women with realistic appraisals of risk and teach skills for securing support for significant others.

R5 tropic HIV‐1 is preferentially translocated of across genital mucosa while X4 tropic HIV‐1 is selectively sequestered in genital epithelial cells and activates type I IFN signaling

A. Nazli¹, M.A. Zahoor², C. Kaushic¹

¹McMaster University, Department of Pathology and Molecular Medicine, Hamilton, Canada, ²University Health Network (UHN), 3Toronto Center for Liver Disease (TCLD), Toronto, Canada

Background: Women constitute more than 50% of the population currently living with human immunodeficiency virus (HIV‐1) worldwide. Majority of HIV‐1 transmission occurs in women through heterosexual intercourse. Although both CCR5‐tropic (R5) and CXCR4‐tropic (X4) HIV‐1 strains are present in semen, transmission occurs predominantly through R5 HIV‐1. The mechanism underlying this preferential selection of R5 HIV‐1 is incompletely understood. In the female genital tract, HIV‐1 has to first cross the epithelial barrier lining before it can infect target cells. We hypothesized that interactions between X4 and R5 strains of HIV‐1 and genital epithelial cells (GECs) may be responsible for preferential selection of R5 strains for mucosal transmission.

Methods: Viral translocation was studied by adding HIV‐1 strains on apical side of confluent polarized GEC monolayers and translocated virus was detected in basolateral supernatant by P24 ELISA and TZMbl indicator cell line. Presence of coreceptors and interferon stimulated gene expression (ISGs) were detected by qPCR. Type I IFN production was measured by ELISA.

Results: When X4 and R5 HIV‐1 were added to GEC monolayers, X4 HIV‐1 was selectively sequestered in endosomal compartment in GECs, while majority of R5 virus was translocated through the cells to the basolateral side. To determine if the uptake of HIV was differentially mediated through co‐receptors expression, CXCR4 and CCR5 expression was determined before and after HIV‐1 exposure. Both co‐receptors were expressed on GECs but while both HIV‐1 strains upregulated expression of CXCR4 co‐receptor, CCR5 co‐receptor expression was downregulated by both X4 and R5 HIV‐1. Our previous studies have demonstrated that Type I IFN is induced in GECs through a TLR‐2 dependent mechanism following HIV‐1 exposure. When we examined the involvement of IFN pathway in differential selection of HIV‐1, we found that X4 HIV‐1 induced significantly higher levels of TLR2, ISGs gene expression and IFN‐β production in GECs compared to R5‐tropic HIV‐1.

Conclusions: Altogether, we found that GECs show a differential response to X4 vs R5 HIV‐1 both in the uptake and transcytosis of the virus and innate immune responses, which could explain the preferential transmission of R5 over X4 HIV‐1 across female genital mucosa. Better understanding of transmission mechanisms could provide information about prevention strategies.

Dynamics of mucosal immune responses elicited by systemic prime/boost vaccination

A. Schuetz¹, Y. Phuangngern², M.A. Eller³, S. Akapirat², J. Dhitavat⁴, P. Pitisutthithum⁴, S. Nitayaphan⁵, S. Chariyalertsak⁶, N. Phanuphak⁷, C.A. DiazGranados⁸, J.H. Kim⁹, M.L. Robb³, N.L. Michael¹⁰, R.J. O'Connell¹⁰, S. Vasan³

¹MHRP Thailand, Retrovirology, Thailand, ²AFRIMS, Retrovirology, Bangkok, Thailand, ³MHRP, Retrovirology, United States, ⁴Mahidol University, Thailand, ⁵AFRIMS, Bangkok, Thailand, ⁶Research Institute for Health Sciences, Thailand, ⁷SEARCH, Thailand, ⁸Sanofi Pasteur, United States, ⁹International Vaccine Institute, Korea, Republic of, ¹⁰Walter Reed Army Institute of Research, Silver Spring, United States

Background: Mucosal surfaces play a critical role in HIV‐1 transmission and disease pathogenesis, hence new vaccine strategies should induce protective mucosal immune responses. Previous studies demonstrated that additional boosting of the 6months (mo) RV144 regimen with longer boosting intervals improved/maintained immune responses. Here we assessed cellular mucosal responses elicited after the RV144 ALVAC‐HIV/AIDSVAX B/E prime/boost intramuscular vaccine regimen followed by additional late boosts (RV306).

Methods: Sigmoid biopsies were collected two weeks post the 6mo priming regimen and the 15mo and 18mo late boosts with ALVAC‐HIV/AIDSVAX B/E. Dynamics of mucosal cell populations and vaccine‐specific cellular immune responses were assessed by flowcytometry.

Results: After the late boost at 15/18mo we observed an increase in the frequency of mucosal ß7^highCD4⁺ T cells (21.5%) compared to the 6mo boost (11%, p=0.01). This was accompanied by a decrease of ß7^highCD4⁺ T cells in the periphery after the 15/18mo boost to 3.85% compared to 4.65% post 6mo (p=0.04). After the 15/18mo boost, 100% of vaccine recipients developed mucosal TH023‐specific CD4⁺ T cell TNFa responses compared to just 30% after 6mo. Correspondingly, the magnitude of those responses after 15/18mo increased to 0.92% from 0.04% post 6mo (p=0.009). A similar trend was observed for mucosal TH023‐specific Th17 cells that increased to 0.11% after the 15/18mo boost while not being detectable post six months (p=0.04). This is in strong contrast to univariate peripheral CD4⁺ T cell TNFa responses that peaked post 6mo and were not augmented by late boosting, with no peripheral TH023‐specific Th17 responses detected at any timepoint. We also observed an increase in mucosal IgA‐producing plasmablasts upon 15/18mo boost to 17.2% compared to 8.6% after 6mo (p=0.04), however no vaccine‐specific IgA was detected in rectal secretions. The frequency of peripheral plasmablasts and vaccine‐specific plasma IgA did not increase after the late boosts.

Conclusions: Late boosts with ALVAC‐HIV/AIDSVAX B/E induced mucosal CD4⁺ T cell homing and improved vaccine‐specific mucosal CD4⁺ T cell responses including the induction of mucosal Th17 cells and increase in the frequency of IgA‐producing plasmablasts. Detailed characterization of mucosal immune responses in future vaccine studies is warranted given that systemic vaccination induces differing patterns of immune response between compartments.

Temporal and spatial characterization of SIV infection dynamics in rhesus macaque mucosal tissues

D. Maric, T.J. Hope, M. McRaven, G. Cianci, L. Corbin

Northwestern University, Cell and Developmental Biology, Chicago, United States

Background: The objective of this study is to characterize the early infection events using the wild‐type virus at the mucosal sytes. We hypothesized that as infection progresses the infected cells will be more heterogeneous due to the recruitment of different immune cell types to the site of infection. Also we hypothesized to see differences in infected cell profiles between stratified squamous tissues of anus and simple columnar epithelia of the rectum.

Methods: Twelve female rhesus macaques were challenged rectally with a mixture of a single round luciferase reporter virus and wild‐type SIVmac239. Animals were sacrificed either 48‐, 72‐ or 96h later. Luciferase signal was used to quickly scan large areas of the tissue and hone in on small regions that likely contain infected cells. Infected cells were identified microscopically by staining for SIV viral proteins and were validated by spectral imaging and nested PCR.

Results: Comprehensive phenotyping of nearly 2000 SIV infected cells revealed that the Th17 T cells infection rate does not vary much over the first 96h. However, from 48h to 96h, there is a pronounced decrease in immature dendritic cells infection rate and an increase in infection of other T cell subtypes, suggesting immune cell recruitment to the site of infection. Since anorectal tissues are composed of two structurally different tissues‐ squamous epithelia in the anus and columnar epithelia in the rectum, we compared the infected cell phenotypes in these tissues. Much like in FRT the infected cells in the anus are mostly T cells (75%) while those in the rectum are mostly non‐T cells (69%). We are currently in the process of studying if these two cell types are targeted preferentially or if targeting were solely a function of relative abundance.

Conclusions: We demonstrated that early infection events at the mucosal sites are very dynamic. While certain cell types are infected at the constant rates others increase or decrease as the infection progresses. In our future work we hope to paint the full picture of the HIV/SIV sexual transmission in time and in space by developing three dimensional imaging techniques to visualize the infected cell foci and immune cell responses.

Immune oscillatory nature through menstrual cycle regulation drives SHIV susceptibility from vaginal challenge

A. Kohlmeier¹, J. Brody², A. Sheth³, F. Hardnett⁴, S. Sharma⁴, I. Ofotokun⁵, I. Massud⁴, G. Garcia-Lerma⁴

¹CDC, DHAP, Atlanta, United States, ²Emory University, Department of Physics, Atlanta, United States, ³Emory University, Department of Medicine, Atlanta, United States, ⁴CDC, Atlanta, United States, ⁵Emory University, Atlanta, United States

Background: AIDS‐related illness is a leading cause of death globally in women aged 15 to 49. Increased susceptibility to HIV or SHIV infection has been proposed to occur during the luteal phase of the menstrual cycle when levels of the immunosuppressive sex‐hormone progesterone are fluctuating. We sought to better define the contribution of the menstrual cycle to HIV susceptibility through studying longitudinal immune properties.

Methods: Immune properties were defined throughout the menstrual cycle in 9 pigtail macaques with standard cycles (average length: 32.8days). Plasma progesterone and cytokines levels were measured by enzyme or bead‐based immunoassay. CD38, CCR5, CXCR3, PD‐1, FoxP3, IFNg and TNFa from CD4 T cells in PBMC were measured using fluorochrome‐conjugated antibody recognition. Power spectral analysis by Lomb‐Scargle algorithms was used to detect molecular oscillations in progesterone and CCR5 expression over reproductive cycles. Relationships between SHIV susceptibility and phase of the menstrual cycle were defined in macaques infected during repeated exposures to low (50TCID50) doses of SHIV162p3. PBMCs from 10 healthy women were used to define inflammatory behaviors during the cycle. Statistical modeling to evaluate pairwise comparisons of means were fit using generalized estimating equations.

Results: Menstrual cycle phase was associated with variations in circulating adaptive cellular characteristics, including CD4 T cell expression of CCR5, activation‐associated molecules PD‐1, CD38, and CXCR3, and functional responses defined by IFNg and TNFa. Comprehensive innate immune characterizations corroborated T cell variations and identified elevated type‐1 inflammatory properties predominantly occurring in the late luteal phase. Using power spectral analysis, we identified sinusoidal expression patterns of CCR5 that synchronized with reproductive cycles. SHIV163P3 infection in macaques was linked to challenges during periods of increasing inflammatory properties, while expression abundance or cycle phase alone was not strongly correlated. Analogous fluctuations in T cell inflammatory properties were found in healthy menstruating women.

Conclusions: We demonstrate that periodic shifts in the immune landscape under menstrual cycle regulation drives type‐1 inflammatory properties and dictates infection opportunities. We posit that novel prevention strategies that reduce local inflammation events in the FRT, such as occurs under menstrual cycle regulation, may help prevent HIV transmission events in women.

The impact of penile‐vaginal sex on penile immune correlates of HIV susceptibility

A. Mohammadi¹, S. Bagherichimeh¹, Y. Choi², A. Fazel¹, E. Tevlin³, S. Huibner², W. Tharao³, R. Kaul²

¹University of Toronto, Medicine, Canada, ²University of Toronto, Canada, ³Women's Health In Women's Hands Community Centre, Toronto, Canada

Background: The penis is the primary site of HIV acquisition in heterosexual men. The per‐act probability of HIV acquisition is enhanced by the presence of a foreskin, sexually transmitted infections (STIs) and/or microbiome dysbiosis, with elevated penile inflammatory cytokines acting as a common central pathway. However, the impact of sex itself on the penile immune milieu is unknown. We hypothesized that condomless penile‐vaginal sex would transiently increase penile cytokines, with increases being more profound and sustained in uncircumcised men.

Methods: To address this hypothesis we recruited 37 STI‐free heterosexual couples to the Sex, Couples and Science Study (SECS), an observational prospective study ran from July to November 2018 through the Women's Health in Women's Hands Clinic (Toronto, Canada). Approximately half of the male partners were circumcised. Baseline penile swabs, semen samples, cervical secretions and blood were collected after a period of abstinence, and 48hours later couples engaged in either condomless (n=30) or condom‐protected (n=8) penile‐vaginal sex. One couple participated twice, once in each group. Repeat penile swabs and blood were collected after one to two hours, seven to eight hours and fourty‐eight to seventy‐two hours. Soluble immune factors were quantified by multiplex immunoassay, and blood immune cell subsets were determined by flow cytometry; the study co‐primary immune endpoints were the penile levels of IL‐8 and MIG, cytokines previously linked to HIV acquisition. Statistical comparisons were performed using the Wilcoxon Signed Ranked test.

Results: One hour after sex there was a dramatic increase in coronal sulcus levels of IL‐8 and MIG, regardless of condom use, with a return to baseline by 72hours; similar patterns were observed for other chemoattractant chemokines. The extent and duration of these increases were similar in circumcised and uncircumcised men, and correlated closely with cytokine levels in the female partner's genital tract (after condomless sex) or in the participant's semen (after condom‐protected sex).

Conclusions: Relatively sustained increases in penile inflammatory cytokines after penile‐vaginal sex appear to reflect penile coating with cervico‐vaginal secretions and/or semen. Since external application of cytokines enhanced HIV acquisition in a foreskin explant model, these novel findings have important implications for the biology of penile HIV acquisition.

Screening for common mental disorders among anti‐retroviral therapy (ART) users: Implications of case‐finding for treatment of mental disorders, ART adherence and viral suppression

A. Kagee¹, J. Bantjes², W. Saal², L. Andersen³

¹Stellenbosch University, Psychology, South Africa, ²Stellenbosch University, South Africa, ³University of Cape Town, South Africa

Background: In South Africa (SA), the prevalence of common mental disorders (CMDs) is high among PLWH. CMDs interfere with adherence to HIV care, contributing to greater morbidity and mortality and greater likelihood of ongoing HIV transmission. Limited data exist on whether screening instruments are useful in detecting caseness for CMD's.

Methods: We administered the major depression, posttraumatic stress, and alcohol use disorder modules of the SCID, the Centre for Epidemiological Studies Depression Scale Revised (CESD‐R), the PTSD Checklist for DSM‐5 (PCL‐5), and the Alcohol Use Disorder Identification Test (AUDIT) to 688 patients receiving ART services at HIV clinics near Cape Town.

Results: 43.17% had any CMD as determined by the SCID; 24.8% met the criteria for major depressive disorder; 14.7% had posttraumatic stress disorder; and 22.1% had alcohol use disorder. In the receiver operating characteristic curve analyses, the area under the curve for each of the measures ranged from 0.89 to 0.95. Sensitivity, specificity, and positive and negative predictive values for the three screening instruments were as follows: CESD: 81.28, 82.33, 60.26 and 92.98; PCL‐5: 88.12, 88.18, 56.33, and 97.72; AUDIT: 88.03, 88.62, 89.93, and 86.51.

Conclusions: All three measures, the CESD, PCL‐5, and AUDIT were excellent in predicting non‐cases, i.e. true negatives for major depression, posttraumatic stress, and alcohol use disorder, respectively. Their utility in predicting caseness, i.e. true positives for these conditions was somewhat lower. Yet, the accurate detection of cases of CMD's may play an important role in ensuring that those in need of treatment may be accurately referred. Referral of large numbers of non‐cases will overload the health care system. Non‐referral of cases, i.e. persons meeting the diagnostic criteria for CMD's will mean non‐treatment of people who may benefit from it. CMD's, especially mood disturbance and harmful alcohol use, are implicated in poor ART adherence, sexual risk behaviour, and viral transmission. Effective screening, referral and treatment may help in efforts to prevent new cases of HIV.

Social‐media intervention: effects on HIV risk perception, intention for testing and risky sexual behaviours among tertiary institutions students in south‐western Nigeria

O. Olaleye, A. Ajuwon

University of Ibadan, Health Promotion and Education, Nigeria

Background: In Nigeria, HIV disproportionately affects young persons, underscoring the need for enhanced prevention intervention targeting this population. Many young Nigerians are ardent social‐media (SM) users but few studies are available on the role of this technology in HIV prevention. This study therefore, investigated the effects of SM‐delivered intervention on HIV risk perception (HRP), intention for testing (IfTe), and risky sexual behaviours (RSB) among students of selected tertiary institutions in South‐western Nigeria.

Methods: The quasi‐experimental study was conducted in two selected public polytechnics in Oyo and Lagos States which were randomly allocated into Experimental Group (EG) and Control Group (CG) respectively. A total of 101 (EG) and 99 (CG) participants were selected through a four‐stage simple random sampling technique. Baseline data collected using a pre‐tested self‐administered questionnaire included; types and frequency SM used, 12‐point HRP scales, IfTe among HIV yet to be tested participants, and 10‐point RSB scales. Scores≤6, 7 to 12, >12 were respectively categorised as low, average, and high HRP. Baseline findings were used to design and implement a four‐month intervention using WhatsApp; the most commonly used SM among the participants. A post‐intervention survey was conducted after a one‐month follow‐up using the instrument earlier used at baseline. Data were analysed using descriptive statistics, Chi‐square, and t‐test at α0.005.

Results: Respondents’ mean age was 21.4±2.7years, while males consisted of 61.4% and 70.7% in EG and CG, respectively. High HRP was 42.6% (EG) and 42.5% (CG) at the baseline, at the post‐intervention, high HRP significantly increased in EG (75.3%) compared with 43.6% in CG. At the baseline, 67.3% (EG) and 64.6% (CG) have never been tested for HIV while IfTe among them was 82.6% (EG) and 87.5% (CG). At the post‐intervention, significantly all yet to be tested participants (100%) in EG were willing to undergo HIV test while only 67.5% were willing in CG. Mean score on RSB significantly decreased from 4.2±2.6 to 2.4±1.4 in EG while CG increased from 3.7±2.3 to 4.2±2.8, comparing baseline with post‐intervention.

Conclusions: Social‐media intervention increased HRP and IfTe and reduced RSB among the participants. SM interventions are therefore recommended to reach young persons on HIV preventive related matters.

Race, minority stress and HIV prevention: the impact of intersectional stigma on PrEP engagement among black men who have sex with men

P. Burns¹, A. Omondi²

¹University of Mississippi Medical Center, Population Health Science, Jackson, United States, ²Jackson State University, Jackson, United States

Background: Sexual and racial minorities are disproportionately impacted by HIV. Globally, studies in HIV pre‐exposure prophylaxis continue to reveal persistently alarming low rates of uptake among key populations, particularly among Black gay and bisexual men. There is an urgent need to reduce stigma and discrimination and promote racial/ethnic equity in PrEP access.

Methods: Utilizing data from ACCELERATE! Initiative, an HIV prevention program targeting Black MSM, we examined 322 African‐American/Black who were surveyed at clinical and community‐settings in a medium‐sized city in the Southern region of the United States, the epicenter of the HIV epidemic. Between January 2016 and September 2017, participants were invited to take a computer‐assisted self‐interviewing (CASI) survey which assessed intersectional stigma and engagement in PrEP.

Results: Of the 322 participants, 170 were HIV‐negative and potentially eligible for PrEP; 42% (n=72) had taken PrEP in the past 12months; 46.5% (n=79) reported they had never taken PrEP. Next, we assessed the prevalence of minority stress among our sample of African‐American/Black MSM. Thirteen percent reported the following about their sexual orientation: Nearly 1 in 5 (19%) felt like an outcast; 18% felt like an unusual person, 13% resented being gay or bisexual; 11% were embarrassed and 9% were depressed. Respondents who were not taking PrEP were 10% more likely to feel like an outcast.

Conclusions: Findings point to high rates of minority stress ramong Black MSM who experience social marginalization. There is an urgent need for combination HIV prevention interventions to address both social and structural forms of stigma to increase uptake of PrEP among key populations.

Socio‐behavioral predictors of HIV serostatus among adults (≥15years) in Eswatini: evidence from the 2016 to 2017 Eswatini HIV Incidence Measurement Survey (SHIMS 2)

M.C. Shongwe¹, L.P. Dlamini², M.S. Simelane³, S.K. Masuku⁴, F.S. Shabalala⁴

¹University of Eswatini, Midwifery Science, Mbabane, Eswatini, ²Taipei Medical University, School of Nursing, College of Nursing, Taipei, Taiwan, Province of China, ³University of Eswatini, Statistics and Demography, Faculty of Social Sciences, Matsapha, Eswatini, ⁴University of Eswatini, Community Health Nursing Science, Faculty of Health Sciences, Mbabane, Eswatini

Background: In Eswatini, a country with the highest HIV prevalence in the world (at 27%), only 73% of those on antiretroviral therapy are virally suppressed; meaning, nearly a quarter can readily transmit the virus if having unprotected sex. For that reason, there is a need to understand the social and behavioral determinants of HIV seropositivity in the general population. The study sought to identify the socio‐behavioral factors (i.e. age at sexual debut, number of lifetime multiple sexual partners (LMSPs), marital status, alcohol intake, condom use at first sex and access to condoms) predicting HIV serostatus among the adult general population in Eswatini.

Methods: This study was a secondary analysis of data for adults (15 to 80years) who had a valid HIV test result in the 2016/17 Swaziland HIV Incidence Measurement Survey (SHIMS 2), a nationally representative household survey designed to assess the impact of HIV treatment programs. We performed gender‐stratified, multivariate binary logistic regression analyses among 7,121 participants (56.6% women).

Results: Adjusting for age, region, residence, educational level and wealth quintile in all models, women who initiated sex at <15years (adjusted odds ratio [AOR]=1.92, 95% confidence interval [CI]: 1.32, 2.79), at 15 to 17years (AOR=1.23, 95% CI: 1.06, 1.42), and who had≥3 LMSPs (AOR=1.41, 95% CI: 1.05, 1.90), had higher odds of being HIV seropositive. However, those who had one LMSP (AOR=0.66, 95% CI: 0.49, 0.89) and those with 2 LMSPs (AOR=0.71, 95% CI: 0.52, 0.99), had lower odds of being HIV seropositive. For men, those who were ever married/cohabited (AOR=1.90, 95% CI: 1.55, 2.33) and those who reported difficulties in getting condoms (AOR=1.61, 95% CI: 1.21, 2.15) had higher odds of being HIV seropositive. Neither alcohol consumption nor condom use at first sex was significantly associated with HIV serostatus.

Conclusions: Behavior change interventions should target married/cohabiting men and should emphasize the risk of lifetime multiple partners across the genders. Condom distribution campaigns should solicit and incorporate the views of men, including on how the barriers towards obtaining condoms can be overcome. Adolescent HIV prevention programs should encourage delaying sexual debut until the legal age of sexual consent (18years).

Dynamics and complexities of sexual and reproductive health for young people living with HIV in Gauteng, South Africa

B. Ndlazi, T. Masango

UNISA, Health Studies, Pretoria, South Africa

Background: Young people between 10 and 24‐years‐old continue to be vulnerable to HIV and its effects. Adolescents, particularly girls, living in settings with a generalized HIV epidemic are in most cases socially and economically disadvantaged. Despite the scourge of HIV, the infected young population's sexual and reproductive health (SRH) and rights are faced by challenges.

Methods: Convergent parallel mixed methods was conducted in five facilities; selected from 3 metro districts in Gauteng Province. Data were collected among YPLHIV age 18 to 24. Semi‐structured (106 respondents) and in‐depth (11 participants) interviews were conducted. STATA software stable release version 15.1 used for quantitave data analysis with the Statistical significance set at p<0.05. Thematic data analysis was conducted for the qualitative data with themes divided into catergories. The results and findings were then merged to identify conversion and diversion.

Results: A majority (66%) of respondents have never discussed sex related matters with parents. A cultural belief that portrayed SRH discussion with parents as disrespectful was found to be a common barrier to information. Majority (85.7%) indicated that they were currently dating. About 46.3% indicated condom usage on first sexual encounter. Intimate partner disclosure remains low at 45% accompanied by low condom usage. Fear of rejection was reported by 64.8% of the respondents as the major cause for non‐disclosure to intimate partners. One participant alluded to the findings by saying, “My partner was distance and kind of afraid of me after I disclosed my status to him”. Being male was associated with low (28.5%) serostatus disclosure compared to 53.6% of females.

Conclusions: Parental and family support is critical in addressing SRH needs and rights. The results from this study have proven that there's a compulsion for an improved and innovative way of implementing the SRH programme for adolescents and young people living with HIV. The adjustments are required so that the programme could possess the power to respond to the specific needs of young people living with HIV. The study highlights the need for programme integration at all level to eliminate the social and structural factors affecting the level of care provided to young people living with HIV.

Syndemic conditions and PrEP use are independently associated with rectal inflammation in sexual minority men

G. Tapia¹, T. Glynn², C. Miller³, A. McGaugh³, J. Manuzak³, C. Broedlow³, R. McIntosh², J. Bauermeister⁴, C. Grov⁵, R. Parisi⁶, D. Martinez⁶, N. Klatt³, A. Carrico⁷

¹Loyola University Chicago Stritch School of Medicine, United States, ²University of Miami, United States, ³University of Miami Miller School of Medicine, United States, ⁴University of Pennsylvania School of Nursing, Philadelphia, United States, ⁵The City University of New York School of Public Health, United States, ⁶AIDS Healthcare Foundation, Los Angeles, United States, ⁷University of Miami Miller School of Medicine, Public Health Sciences, United States

Background: Syndemic theory is a way to contextualize co‐occurring epidemics (e.g., substance use, mental health) that act synergistically to drive HIV incidence. Indeed, a positive dose‐response relationship has been shown between number of syndemic conditions with HIV risk behavior (condomless receptive anal intercourse; CRAI) and HIV seroconversion among sexual minority men, a group disproportionately burdened by HIV in many Western nations. However, little is known about the biological pathways whereby a syndemic could amplify biological vulnerability to HIV. Rectal mucosal inflammation is one possible mechanism for increased risk of HIV acquisition.

Abstract PE01.08‐Table 1.

Methods: This cross‐sectional study examined the association between number of syndemic conditions and PrEP use with biomarkers of rectal inflammation among 92 HIV‐negative, sexual minority men recruited at four AIDS Healthcare Foundation STI clinics in South Florida (USA), a region with disproportionately higher HIV incidence. All participants reported CRAI and no antibiotic use in the past three months. Rectal inflammatory cytokines were quantified using LEGENDplex. Participants completed screening measures for the following syndemic conditions: depression, PTSD, hazardous alcohol use, and any stimulant use. A count variable (0 to 4) was created by summing conditions participants screened positive for.

Results: After adjusting for age, race/ethnicity, and number of CRAI partners, greater number of syndemic conditions was associated with higher log10 levels of IFN‐γ, IL‐8, and IL‐23. PrEP use was independently associated with higher log10 levels of IFN‐γ, IL‐6, IL‐8, IL‐17a, and IL‐23.

Conclusions: These findings are among the first to demonstrate that sexual minority men with more syndemic conditions and PrEP users display greater rectal inflammation. An increase in local cytokines may mobilize the innate immune response, potentially amplifying biological vulnerability to HIV as well as other STIs.

Developing a chatbot for HIV risk assessment among young people living in Soweto, South Africa

G. Tshabalala, M. Mulaudzi, S. Hornschuh, E.K. Okyere-Dede, J. Dietrich

Wits Health Consortium, Perinatal HIV Research Unit, Johannesburg, South Africa

Background: Young people in South Africa have low uptake of HIV testing and treatment initiation. Lack of confidentiality in accessing HIV prevention services are due to long waiting queues, and sharing HIV testing facilities with other health care services. There is a need for remote HIV prevention methods that tap into platforms that young people can engage with in a non‐intrusive way. This study aimed to develop an internet‐enabled HIV Risk Assessment through an iterative approach with young people.

Methods: Two groups of participants stratified by gender (females and males, aged 18 to 24 years) participated in two focus group discussions each, conducted over 2days. Additionally,18 in‐depth interviews were conducted once‐off with young key populations (i.e. Gay, Bisexual, Lesbian, Transgender), aged 18 to 24years. To enable the development process, participants completed a paper‐based HIV risk assessment questionnaire. Thereafter, they provided feedback on the HIV risk assessment content, their opinion towards using an internet‐enabled HIV Risk Assessment, and recommendations to adapt the HIV risk assessment questionnaire into an internet‐enabled HIV Risk Assessment. Qualitative data were analyzed thematically using NVivo qualitative data analysis software.

Results: Discussions with participants reiterated their preference for an interactive and informational online HIV Risk Assessment, allowing to ask questions about their health and about nearest youth‐friendly clinics for HIV testing. Young people preferred discrete and private means to assess their own risk for HIV. The majority concurred that the design and logo of the internet‐enabled HIV Risk Assessment should not depict anything related to HIV. Participants wanted to receive notifications or reminders from the online HIV Risk Assessment to test for HIV every three months. HIV risk assessment questions (such as penetrative vaginal/anal sex; receive vaginal/anal/oral sex) identified as confusing and difficult were revised (vaginal/anal/oral sex). During the software development phase, participants’ opinions and recommendations from FGDs and IDIs were considered in developing a Chabot for HIV Risk Assessment.

Conclusions: A collaborative and user‐driven process is crucial when designing and developing an HIV prevention tool for targeted groups. Privacy and confidentiality are important features that may promote acceptability and willingness to use internet‐enabled HIV prevention methods.

Sharing objective measures of adherence to a vaginal microbicide promoted candor about actual use and bolstered motivation to prevent HIV during MTN‐025/HOPE

B. Kutner¹, R. Giguere¹, C. Lentz¹, C. Kajura-Manyindo², C. Dolezal¹, S. Butheliezi², M. Gwande³, S. Nampiira⁴, T. Ndlovu², P. Mvinjelwa⁵, W. Mwenda⁶, I. Balán¹

¹HIV Center for Clinical & Behavioral Studies at NY State Psychiatric Institute and Columbia University, Division of Gender, Sexuality and Health, New York, United States Minor Outlying Islands, ²South African Medical Research Council Clinical Trials Unit, South Africa, ³University of Zimbabwe College of Health Sciences Clinical Trials Research Center, Harare, Zimbabwe, ⁴Makerere University – Johns Hopkins University Research Collaboration Clinical Research Site, Kampala, Uganda, ⁵Emavundleni Clinical Research Site, South Africa, ⁶Queen Elizabeth Central Hospital, College of Medicine Clinical Research Site, Malawi

Background: Discrepancies between self‐reported and actual adherence to biomedical HIV interventions can compromise the integrity of clinical trials. One solution is to monitor adherence more objectively through biological assays. However, it is not well understood how to communicate these metrics in ways that foster honest reporting rather than defensiveness.

Methods: MTN‐025/HOPE was a Phase 3b open‐label trial of a vaginal ring to prevent HIV, conducted across four sub‐Saharan African countries. The trial involved testing each ring to measure its residual drug level (RDL), an objective marker of adherence. During successive adherence counseling sessions at Months 3, 6, 9, and 12, counselors shared with participants each ring's RDL, re‐conceptualized as a level of protection ranging from 0 “No Protection” to 3 “High Protection.” After the trial, from July 2018‐May 2019, we interviewed 22 counselors and conducted a matrix analysis to characterize their perspectives about RDL conversations.

Results: Counselors perceived that participants appreciated RDL feedback as an indication of their protection from HIV. Indeed, reactions varied depending on the RDL. Higher drug levels (RDL=2 or 3) stimulated elation and relief whereas lower levels (RDL=0 or 1) resulted in disappointment and, more rarely, in anger when participants self‐reported higher adherence. A nonjudgmental stance and support for autonomy to choose alternatives to the ring promoted disclosure of causes of lower adherence that otherwise might have remained forgotten (e.g., taking the ring out during menses) or concealed (e.g., preferring not to use the ring). Reframing RDL monitoring as “protection” rather than “adherence” also helped pivot from numerical results toward the trial's ultimate goal of HIV prevention. This emphasis on women's motivations to prevent HIV, rather than on ring use, encouraged consistent users to continue and infrequent users to switch to an alternative HIV prevention approach. Counselors noted that adherence conversations might have been more cursory if based solely on self‐report, without the anchoring metric of a woman's current protection against HIV.

Conclusions: Personalizing feedback from objective adherence ratings is complex and requires careful navigation to minimize defensiveness but can also be implemented in ways that motivate disclosure of non‐adherence and evoke commitment to preventing HIV acquisition.

Experiences participating in rectal microbicide clinical trials: insights from MTN‐026 & MTN‐033

R. Tingler¹, J. Bauermeister¹, S. Johnson², N. Macagna², J. Piper³, K. Ho⁴, C. Hoesley⁵, E. Dunne⁶, R. Cranston⁷

¹University of Pennsylvania, School of Nursing, Philadelphia, United States, ²FHI 360, Durham, United States, ³National Institute of Allergy and Infectious Diseases, Bethesda, United States, ⁴University of Pittsburgh, Department of Medicine, Division of Infectious Diseases, Pittsburgh, United States, ⁵University of Alabama at Birmingham, Birmingham, United States, ⁶Centers for Disease Control and Prevention, Atlanta, United States, ⁷University of Pittsburgh, Medicine, Division of Infectious Diseases, Pittsburgh, United States

Background: Successful implementation of clinical trials depends on participants’ engagement with and commitment to study protocols. Increasingly, implementation scientists have noted the importance of examining how internal (e.g., experiences with study procedures and staff) and external (motivations for participating) factors influence participants’ engagement with clinical trials. As part of two Phase I rectal microbicide clinical trials (MTN‐026 and MTN‐033), we ascertained how internal and external factors facilitated participants’ engagement throughout the clinical trial experience.

Methods: Forty‐four participants (age range: 19 to 47; 35 male, 9 female) were enrolled across three sites (Bangkok, Thailand and Birmingham, Alabama and Pittsburgh, Pennsylvania, USA) and completed semi‐structured, video‐facilitated in‐depth interviews (IDI). During these interviews, we explored participants’ study experiences, including motivations to participate in the trial and comfort with clinical (e.g., anoscopy) and behavioral (e.g., surveys and IDIs) procedures. Interviews were conducted by trained qualitative interviewers, transcribed, coded, and analyzed for key themes. Thematic analysis examined convergent and divergent themes emerging from the interviews.

Results: Majority of participants noted numerous positive motivations for participating in the trial (e.g., contributing to the development of HIV prevention products, healthcare access), and voiced how their interactions with clinical staff supported their trial participation (e.g., flexible scheduling). Participants highlighted their appreciation for staff being receptive to feedback, clearly explaining study procedures, and asking for consent before performing a procedure. While some participants mentioned feeling discomfort during clinical procedures (e.g., biopsies), they discussed the importance of the study and offered behavioral strategies to overcome discomfort during study procedures (e.g., watching Netflix on their phone, listening to music, or looking away from the monitors). No thematic differences regarding study procedures were observed between Thai and US participants.

Conclusions: Ensuring participants’ optimal engagement with study protocols is a pivotal aspect of any trial. Our findings support the need for multiple strategies staff can advise participants to engage in to overcome discomfort during study visits. Additionally, beyond acceptability of an investigational drug, our findings underscore the need to support planning and implementation efforts seeking to enhance internal (e.g. scheduling of and interactions during clinic visits) and external (e.g. altruistic motivators) factors linked to participants’ trial experience and engagement.

Association of social support and HIV‐related stigma on detectable viral load and condomless anal sex (CAS) among a diverse sample of HIV‐positive MSM enrolled in an mHealth study

K. Horvath¹, O. Shalhav², A. Talan², J. Rendina², S. Lammert³

¹San Diego State University, Psychology, San Deigo, United States, ²City University of New York, Pride Research Health Consortium, New York, United States, ³University of Minnesota, United States

Background: We assessed the impact of change in social support and HIV‐related stigma on detectable viral load and condomless anal sex (CAS) among a diverse community sample of HIV‐positive MSM residing in New York City and participating in an mHealth antiretroviral (ART) adherence intervention, called Thrive with Me (TWM).

Methods: Demographic, sexual behavior, viral load (VL), and psychosocial factors (including internalized, anticipated, and enacted HIV stigma subscales and emotional, tangible, affectionate, social interaction social support subscales) were collected at baseline, and during follow‐up (Month‐5, Month‐11, and Month‐17). Level of detection for VL was defined as <20 copies. Social support and stigma change was calculated by averaging the measures over follow‐up and subtracting from the baseline score. Social Support measures were dichotomized as highest decreased support (decrease in support >1.5*SD) vs consistent/increased support. Stigma measures were dichotomized as highest increased stigma (increase in stigma >1.5*SD) and consistent/decreased stigma. Detectable VL was defined as any report of a detectable VL during follow‐up and CAS was defined as 2+ follow‐up periods with self‐reported CAS. Risk Ratios and 95% confidence intervals were calculated using generalized linear models (GLM).

Results: 401 participants were recruited for TWM. Participants, on average, were 39years old (IQR=30–48). More than three‐quarters of participants were racial or ethnic minorities, including 57% African American and 27% Hispanic/Latino. At baseline, 154 (38.5%) of participants had a detectable VL and more than half (58.6%) reported CAS in the past 3months. Among participants with follow‐up, 246 (61.9%) had at least one detectable VL while 165 (41.2%) reported at least two follow‐up periods with CAS. A high decrease in emotional (RR=1.31; p=0.048), tangible (RR=1.38; p=0.003), affectionate (RR=1.43; p<0.001), and social interaction support (RR=1.42; p<0.001) were associated with an increased risk of a detectable VL. No associations were found between change in HIV‐related stigma and detectable VL. No associations were found between either social support or HIV‐related stigma and CAS.

Conclusions: Future interventions may benefit from activities that buffer against substantial changes in social support to protect against loss of viral control. HIV care practice may consider evaluating measures of social support frequently as a sudden decrease may predict poor virologic outcomes.

PrEP stigma affects PrEP uptake: attitudes towards pre‐exposure prophylaxis (PrEP) amongst HIV vaccine trial participants in Soweto, South Africa

F. Laher¹, T. Salami², S. Hornschuh¹, L.M. Makhale¹, M. Khunwane¹, M. Andrasik³, G.E. Gray¹, H.‐V. Tieu⁴, J.J. Dietrich¹

¹University of the Witwatersrand, Perinatal HIV Research Unit, Johannesburg, South Africa, ²University of Texas, United States, ³HIV Vaccine Trials Network, United States, ⁴New York Blood Center, Laboratory of Infectious Disease Prevention, United States

Background: South Africa has the most annual new HIV infections, 14% of the global total. Daily oral Truvada for HIV pre‐exposure prophylaxis (PrEP) was investigated in 3 clinical trials in South Africa, showing low to modest effectiveness (‐4% to 44%). In 2015, South Africa licensed Truvada for PrEP. In 2016, the national phased implementation began, prioritizing key populations, with low uptake and adherence. In 2017, the SAMRC‐HVTN established a programme to bolster PrEP availability at HIV vaccine efficacy trial sites in southern Africa. We explored attitudes to PrEP amongst trial participants.

Methods: We conducted a qualitative study with 38 participants enrolled in the HVTN 702 preventive HIV vaccine efficacy trial in Soweto, South Africa. Stratified by age, gender and sexual orientation, 5 focus group discussions were conducted during Sep‐Oct 2018. Discussions were audio‐recorded, transcribed, translated and thematically analyzed.

Results: Participants’ median age was 26 (IQR=23 to 30) years, 50%(n=19) identified as male, 47%(n=18) female, and 3%(n=1) transgender. Two major themes emerged.

(i) PrEP stigma. A barrier to uptake and adherence was the concern that others would see and misidentify the pills as HIV treatment, resulting in PrEP users being mistaken as HIV‐infected: “…there is a stigma that goes with tablets with my people when people see you drinking that medication. They automatically assume that it's ARV's [for treatment]” (MSM/TG_18 to 35years). PrEP side effects also carried stigma and the risk of inadvertent disclosure: “I am taking it secretly. So what I don't want is a situation where I become ill…and my partner will have those questions about what's making me ill” (F_25 to 35yearss). PrEP could hamper socializing: “You cannot take the pill because you are at a party” (M_25 to 35years). Participants stigmatized PrEP use as being connected to infidelity and sexual promiscuity.

(ii) Disadvantages of pills. Forgetfulness was common: “…even if I make alarms and put systems in place in order for me not to forget, I always do. Yeah, so like the tablets are not for me.” (MSM/TG_18 to 35years). Further disadvantages were difficulty swallowing large tablets and fear of side effects.

Conclusions: PrEP stigma poses uptake and adherence barriers. Strategies to reduce PrEP stigma in South Africa need investigation.

Community beliefs and practices during pregnancy and their potential effect on HIV prevention products use in Sub Saharan Africa

P. Mutero¹, J. Ryan², K. Reddy³, D. Kemigisha⁴, D. Gondwe⁵, T. Chitowa¹, P. Musara¹

¹University of Zimbabwe College of Health Sciences, Clinical Trials Research Centre, Harare, Zimbabwe, ²Women's Global Health Imperative (WGHI) RTI International, RTI International, Berkeley, United States, ³Wits Reproductive Health and HIV Institute, School of Clinical Medicine, University of the Witwatersrand, School Of Clinical Medicine, Johannesburg, South Africa, ⁴Makerere University – Johns Hopkins University Research Collaboration, Kampala, Uganda, ⁵Johns Hopkins Project‐College of Medicine, University of Malawi, Blantyre, Malawi

Background: Pregnant and breastfeeding (P/BF) women are at high risk of HIV acquisition due to biological and behavioural factors and need better prevention options. Uptake of new prevention products may be impacted by beliefs and practices during these periods. The MTN‐041/MAMMA study explored the hypothetical acceptability of a vaginal ring and oral pre‐exposure prophylaxis (PrEP) use during pregnancy and breastfeeding in Sub‐Saharan Africa. We explored pregnancy beliefs and practices and how they may impact future HIV prevention product use.

Methods: 23 Focus Group Discussions (FGDs) and 36 In‐depth interviews (IDIs) were conducted among 226 participants in Malawi (N=51), South Africa (N=47), Uganda (N=68) and Zimbabwe (N=60). Participants included P/BF women aged 18 to 40 (median age 26), men aged 18+ with P/BF partners (median age 30), grandmothers aged 18+ (median age 49) and key informants aged 18+ (median age 50). FGDs and IDIs were conducted in local languages, transcribed, coded using Dedoose software (v7.0.23) and analysed using a socio‐ecological framework. Data analysis was done by comparing and contrasting data across sites and study groups.

Results: Across sites, participants in all study groups described that pregnant women perform cultural practices to promote their health and the health of the unborn baby, ease birthing process and for spiritual guidance. Most participants admitted engaging in these practices. Additionally, women often register for antenatal care for detection of pregnancy abnormalities. Birth preparation practices for opening the birth canal include inserting fingers, herbs, drinking herbal mixtures and birth canal lubricants like okra. Pregnant women consult traditional healers and prophets because they are believed to deal with bad spells that cause prolonged labour, caesarean section and abnormal babies. Participants believed that birth canal opening practices could cause ring expulsion, while herbal interaction with HIV prevention drugs could harm both mother and baby. Seeking care from traditional healers and spiritual leaders could impact product use since some are against use of medications.

Conclusions: Cultural beliefs and practices during pregnancy may impact decision making around HIV prevention product use. As such, community roll out of prevention methods needs to be accompanied by education that takes these beliefs and practices into account to encourage uptake.

Choice period counselling for dapivirine vaginal ring and truvada in MTN‐034 study; experiences of counsellors at MU‐JHU CRS site Kampala, Uganda

F. Asiimwe Biira, R. Nakalega, H. Kalule Nabunya, E. Mulumba, B.G.M. Gati Mirembe, C.N. Nakabiito, E.K. Kyomukama, J. Etima, C. Agwau Akello

Makerere University‐John Hopkins University Research Collaboration, Kampala, Uganda

Background: Adolescents Girls and Young Women (AGYW) in Uganda are typically not empowered to make their own choices in many matters related to sexual and reproductive health. The MTN‐034/REACH study involves AGYW who were randomized to either the Dapivirine vaginal ring or oral Truvada. At the end of a one‐year crossover period after using both products, participants are offered an opportunity to select a product of their choice or neither product for a final 6‐month choice period. We will explore the practices and experiences of counselors to prepare for the choice period of this study where there is no random assignment.

Methods: Two training sessions were held for counsellors to prepare them for the choice period, during which counselling manuals and peer rating tools of counselling sessions were reviewed to identify best practices to be adopted to prepare participants for visits during the choice period. Additionally, counsellors held 4 meetings at the site with other study staff and they brainstormed on strategies of how to best deliver the information about “choice’. Notes were taken of the best practices and experiences were documented in the counselling reports.

Results: The team adopted pre‐choice counselling in individual or group sessions as the main strategy to prepare AGYW for the choice period. During these sessions, the counsellors use skills that include: active listening, empathy, being non‐judgmental as they answer questions and provide information in an unbiased manner about the ring and Truvada. Counsellors have learnt that it is hard for AGYW to make a choice and sometimes they request that the team makes a choice for them. The counseling team however, empowered participants to make informed choices and supported their decisions. AGYW should not be rushed to make decisions but rather be motivated as they tend to oscillate from option to option before making a final decision. Participants were also given the option of choosing neither product and informed about other available HIV prevention methods like condoms. AGYW appreciated the counselling and support provided.

Conclusions: Participants felt empowered to make the best HIV prevention choices and counsellors learnt that providing appropriate information while maintaining impartiality is key.

Does the ring work? Perceptions and understandings of the efficacy of a vaginal ring for HIV prevention amongst women participating in the MTN‐032/AHA study

J. Etima

Makerere University‐Johns Hopkins University Research Collaboration, Psychosocial Support, Kampala, Uganda

Background: MTN‐020/ASPIRE showed the dapivirine vaginal ring as a safe and partially effective (in part due to suboptimal adherence) HIV prevention option. ASPIRE was placebo‐controlled, and a potential influence on participants’ ring use was their perception of its efficacy. The MTN‐032/AHA study was an exploratory qualitative study with women who exited ASPIRE to understand reasons for nonuse and use of the ring during the trial.

Methods: Following ASPIRE, clinicians disseminated results of a 27% reduction in HIV acquisition for all women and no reduction among women under 25 through group meetings with participants. In AHA, single in‐depth interviews (n=98) and 12 focus group discussions (n=89) were conducted with women at 7 sites in Malawi, South Africa, Uganda and Zimbabwe. Eligibility included participation in the ASPIRE active arm, and ring use for >3months. Interview guides elicited an in‐depth assessment of how perceptions of ring efficacy impacted adherence and sexual risk behaviors during ASPIRE, and how perceptions and understanding of efficacy results impacted future ring use interest. Interviews were audio‐recorded, transcribed into English, coded and thematically analyzed.

Results: AHA women were on average 27years old, mostly unmarried (62%), and 48% had completed secondary school. Many didn't know their partner's HIV status (37%) or if their partner had other sexual partners (58%). Perceptions of ring efficacy while in ASPIRE varied; majority described efficacy as a binary assessment: the ring worked or not. Women who believed it worked described simply trusting it, or having confidence in the ring because they, or others in risky situations remained HIV‐uninfected. Participants recounted ASPIRE results through feelings of happiness and pride that the ring they used, and their trial participation, were successful. Many did not remember exact figures because of lack of comprehension or memory but recalled key details about differences in efficacy by age. The majority expressed interest in future ring use since it provided some level of protection.

Conclusions: Perceptions and understanding of efficacy influenced some women's adherence to and eventual interest in future ring use. Numeric trial efficacy results were not well comprehended/understood suggesting the need for cultural/linguistic specificity and end‐user involvement in message development.

Response to a novel technology‐assisted HIV self‐testing intervention in Kampala, Uganda by age and gender

K. Horvath¹, J.M. Bwanika², S. Lammert³, H. Banonya⁴, D. Musinguzi⁴, S. Magambo⁴, A. Kiragga⁴

¹San Diego State University, Psychology, San Deigo, United States, ²The Medical Concierge Group, Director of Research, Uganda, ³University of Minnesota, United States, ⁴The Medical Concierge Group, Uganda

Background: There were an estimated 22,000 new HIV infections among Ugandan youth (15 to 24years of age) in 2018, of which women accounted for 15,000 (68%). We assess variability in response to a novel technology‐assisted HIV self‐testing (HIVST) intervention among adults in Kampala, Uganda by age and gender groups.

Methods: In 2019, our team conducted a single‐arm trial of a technology‐assisted intervention to promote HIVST among at‐risk adults living in Kampala, Uganda over a 3‐month period. Participants received theoretically grounded messages and had access to an existing call‐in center to order free HIVST kits and have them delivered to a location of their choice. Intervention acceptability and outcomes were examined by the following age (18 to 24 vs. 25 to 49years) and gender groups: younger women (YW; n=33); younger men (YM; n=27); older women (OW; n=16), and older men (OM; n=23).

Results: At baseline, over half (53%) of YW were tested for HIV in the past 6months, compared to 56% OW, 61% OM, and 76% YM. Women were less likely to have heard of (16% of women; 42% of men) or ever used (1% of women; 22% of men) HIVST at baseline. A lower percentage of OW reported that HIVST was very easy to complete (58% vs. 77% YW, 80% OM, 81% YM), that the instructions were very clear (83% vs. 93% OM, 95% YM, 96% YW), and that they were very satisfied with HIVST (92% vs. 93% OM, 95% YM, 96% YW) compared to other groups. Across all groups, only 1 older man would not recommend HIVST. A lower percentage of OM (83%) ordered HIVST kits for themselves compared to other groups (85% YW, 83% OM, 89% YM). Once delivered, nearly all participants across all groups used the test and 1 younger man tested positive for HIV and was successfully linked to care.

Conclusions: Gender and age group differences in HIVST awareness and use at baseline (lowest among YW), intervention acceptability (lowest among OW), and ordering of HIVST kits (lowest among OM) were found. These findings may be important for future HIVST intervention trials and community roll out of HIVST programs in Uganda and other sub‐Saharan countries.

Counselors’ acceptability of adherence counseling session recording, fidelity monitoring, and feedback in a multi‐site HIV prevention study in four African countries

R. Giguere¹, C. Lentz¹, C. Kajura-Manyindo¹, B.A. Kutner¹, C. Dolezal¹, M. Buthelezi², I. Lukas³, S. Nampiira⁴, C. Rushwaya⁵, E. Sitima⁶, A. Katz⁷, A. van der Straten⁸, I.C. Balán¹

¹Columbia University and New York State Psychiatric Institute, HIV Center for Clinical and Behavioral Studies, New York, United States, ²South African Medical Research Council, South Africa, ³Wits Reproductive Health and HIV Institute, Johannesburg, South Africa, ⁴Makerere University ‐ Johns Hopkins University Research Collaboration, Kampala, Uganda, ⁵University of Zimbabwe College of Health Sciences, Harare, Zimbabwe, ⁶Johns Hopkins University ‐ Blantyre Clinical Trials Unit, Malawi, ⁷RTI International Women's Global Health Imperative, Research Triangle Park, United States, ⁸RTI International and University of California at San Francisco, Women's Health Global Imperative and Center for AIDS Prevention Studies, Research Triangle Park, United States

Background: To retain effectiveness, evidence‐based counseling interventions must be delivered with fidelity. We report on acceptability of the adherence counseling fidelity monitoring process used in MTN‐025/HOPE Study, the largest biomedical HIV prevention trial to integrate fidelity monitoring using audio recordings of counseling sessions.

Methods: The MTN‐025/HOPE Study, a Phase 3B open label extension trial across fourteen sub‐Saharan African sites involving 1456 women, explored safety and adherence to the dapivirine vaginal ring for HIV prevention between August 2016 and October 2018. Adherence counselors were trained to conduct Options Counseling, a Motivational Interviewing‐based intervention. A sample of their audio‐recorded sessions were monitored by a New York‐based multilingual team, to support intervention fidelity. The rating team evaluated sessions using a fidelity monitoring tool and was itself assessed monthly for interrater reliability (IRR). To understand acceptability and feasibility of fidelity monitoring, we surveyed 42 counselors and interviewed 22 counselors, and examined spontaneous mentions of session recordings by 10 study participants among 91 interviewed. Quantitative data were analyzed using descriptive statistics. Qualitative data were coded and thematically summarized.

Results: In total, 5366 Options sessions were audio‐recorded, of which 1238 (23%) were reviewed for fidelity. On a scale of 1 to 7, counselors indicated that session ratings were very helpful (mean rating of 6.64). Most counselors reported progressing from apprehension to confidence about the fidelity monitoring process after conducting 25 or fewer sessions, with 88% reporting feeling confident in their abilities and 90% likely to use skills learned through Options in the future. In interviews, study participants had mixed reactions to recorded sessions; some reported receiving better counseling when sessions were monitored, and others found them time‐consuming and burdensome. For raters, assessment of IRR was essential, as drift in ratings occurred over time. Furthermore, internal reviews of rating forms revealed the need for specific training to shape written feedback in a supportive, client‐centered manner.

Conclusions: Fidelity monitoring of counseling sessions in large multi‐site biomedical HIV prevention studies is acceptable and can feasibly guide and support counselors by providing structured feedback. Future international trials using behavioral interventions should include fidelity monitoring to ensure adherence to effective practices.

Demystifying myths and misconceptions about use of the Dapivirine Vaginal Ring and Truvada among adolescent girls and young women in the MTN‐034 study; observations from Kampala site

E. Mulumba¹, R. Nakalega², S. Nanyonga², H. Kalule Nabunya², F.A.B. Asiimwe Biira², E. Kyomukama², J. Etima Ongom², B. Gati Mirembe², C. Nakabiito², C. Agwau Akello²

¹Makerere University‐John Hopkins University Research Collaboration, Psycho‐social, Kampala, Uganda, ²Makerere University‐John Hopkins University Research Collaboration, Kampala, Uganda

Background: Communities in Africa have a major influence on the behaviors of adolescent girls and young women (AGYW) and their opinions are highly valued in the decision making processes of this young and vulnerable population at risk of HIV infection. Since myths and misconceptions have the potential to affect the utilization of the dapivirine vaginal ring (Ring) and Truvada while also affecting the research process, it is imperative that researchers understand the concerns and also find means to address them. We seek to explore the myths and misconceptions about the use of the Ring and the Truvada tablet among AGYW in the MTN 034 study at Kampala and also describe the strategies adopted to mitigate them.

Methods: The MTN‐034 is an ongoing, cross‐over study of daily oral PrEP and monthly dapivirine vaginal ring safety and preferences, From June 2019 to Feb 2020, 26 adherence support meetings were held with about 15 to 20 participants attending each of these bi‐weekly sessions. Additionally, one‐on‐one counselling sessions, community sensitization meetings, and parents/guardians meetings are held during study implementation. Summary notes are used to record the myths and misconceptions including challenges and proposed solutions that arise during these sessions.

Results: The team addressed these myths and misconceptions by providing accurate information about the study and study products to participants and community members. Participant and community engagement sessions plus scheduled counselling visits were a platform to dispel misconceptions.

Conclusions: Implementation of interventions requires sensitization of communities prior to study initiation to mitigate myths and misconceptions. Continuous education and engagement of communities, as well as participants, is key to dispelling myths and misconceptions.

Perceived severity of HIV infection in the biomedical era and its association with sexual risk behavior among HIV‐negative men who have sex with men

W.P. van Bilsen, H.M. Zimmermann, A. Boyd, U. Davidovich

Public Health Service of Amsterdam, Department of Infectious Diseases, Amsterdam, Netherlands

Background: Health‐related behaviors are closely linked to the perceived severity and potential consequences of a disease. With increasingly tolerable HIV‐treatment and the acceptance of the Undetectable=Untransmissable message, HIV may be seen more than ever as an easily manageable condition. This might negatively affect HIV‐prevention uptake. We investigated current perceptions on the general severity and potential consequences of an HIV‐infection and its association with sexual risk behavior among HIV‐negative MSM living in the Netherlands.

Methods: In‐depth interviews with recently diagnosed MSM were used to develop a questionnaire measuring the severity and anticipated consequences of HIV‐infection. The questionnaire was distributed online using gay dating sites/apps and social media between April‐July 2019. A structural equation model was constructed to explore which anticipated consequences contributed most to the general perceived severity of HIV‐infection and to assess the association between perceived severity and sexual risk behavior (i.e. condomless anal sex with casual partners without PrEP‐use).

Results: We analyzed 1072 HIV‐negative MSM completing the survey, of whom 28% reported sexual risk behavior in the preceding 6months. 77% perceived HIV as a severe illness. Anticipated negative consequences of HIV on sex/relationships were strongly related to the general perceived severity of HIV (β = 0.32, 95% CI 0.23 to 0.42, p<0.001) (Figure 1). Moreover, anticipated psychological (p<0.001), disclosure‐related (p<0.001) and health‐related negative consequences of HIV‐infection (p=0.04) were also related to general severity perceptions. Finally, a higher general perceived severity of HIV was correlated with lower sexual risk taking (β = ‐0.06; 95% CI=‐0.11,‐0.01; p=0.02).

Conclusions: One‐quarter of HIV‐negative MSM did not perceive HIV as a serious illness, which was associated with more prevalent sexual risk taking. Perceptions on the severity of an HIV‐infection mainly consisted of anticipated burdensome interpersonal aspects of living with HIV. These data imply that prevention strategies are challenged by the perceptions of low HIV‐infection severity among some HIV‐negative MSM.

Interpersonal and structural stigma towards HIV pre‐exposure prophylaxis among community‐based physicians delivering sexual health services in Taiwan

I.Y.‐H. Chu¹, C. Strong², C.‐W. Li³, S.W.‐W. Ku⁴, N.‐Y. Ko⁵, A. Bourne⁶, H. Burchett¹, F. Hickson¹

¹London School of Hygiene and Tropical Medicine, Department of Public Health, Environments and Society, Faculty of Public Health and Policy Faculty of Public Health and Policy, London, United Kingdom, ²National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Department of Public Health, Tainan, Taiwan, Province of China, ³National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Department of Internal Medicine, Tainan, Taiwan, Province of China, ⁴Taipei City Hospital Renai Branch, Division of Infectious Disease Department of Medicine, Taipei, Taiwan, Province of China, ⁵National Cheng Kung University Hospital, College of Medicine, National Cheng Kung University, Department of Nursing, Tainan, Taiwan, Province of China, ⁶Australian Research Centre in Sex, Health & Society, La Trobe University, Melbourne, Australia

Background: Engagement with HIV pre‐exposure prophylaxis (PrEP) by healthcare providers can be negatively impacted by both interpersonal and structural stigma (Hatzenbuehler and Pachankis, 2016). Since 2018, Taiwan has implemented government‐funded and self‐paid PrEP but few physicians provide PrEP services. We qualitatively explored PrEP‐related stigma among physicians to understand barriers to PrEP scale‐up.

Methods: From June to October 2019, sixteen face‐to‐face in‐depth interviews were conducted with physicians delivering sexual health services at community‐based clinics. We applied stratified sampling to ensure diversity of geography, specialities and experience in PrEP prescribing. Interviews were audio‐recorded, transcribed, and thematically analysed.

Results: Interpersonal stigma included participants’ own biases towards PrEP users (e.g. association with condomless sex, STIs and wasting public health resources) and expected biases of other physicians (e.g. the exclusiveness of PrEP for gay men, reluctance in taking blood samples from PrEP clients and worries about the spread of HIV drug resistance).

Structural stigma was related to PrEP being an HIV prevention measure. Negative social norms relating to HIV which now spill over to PrEP (“When it comes to HIV, even though PrEP is just a medication, many people become frightened”) and reinforced HIV stigma on PrEP service users (“The general public won't come to my clinic as they think many HIV patients are often here”). Sexual stigma disproportionately suppressed PrEP‐prescribing behaviour among criticism‐averse physicians. Some clinicians treated STI patients without taking sexual histories nor offering information on PrEP to those eligible. (“If you actively ask [about their sexual histories], patients will become annoyed and doubt whether you mean they are promiscuous.”). Interpersonal stigma and the health system were interrelated. The cloud‐based electronic health record (EHR) in Taiwan's healthcare system could amplify interpersonal PrEP stigma. Physicians worried that documenting PrEP on users’ EHR might reinforce other providers’ biases (“If I document [PrEP] clients as sex buyers in the [EHR] system, how can that be right!”).

Conclusions: Multifaceted PrEP stigmas perceived or operationalised by physicians impede the impact of PrEP in Taiwan. Future PrEP implementation programmes could promote medical professionalism without moral judgement, encourage conversations about HIV prevention at consultations, and foster trust in patient‐physician relationships.

Prevalence and correlates of intimate partner violence among high‐risk adolescents aged 14 to 19years in Kampala, Uganda

O. Kamacooko¹, Y. Mayanja², J.F. Lunkuse², J. Seeley², P. Kaleebu²

¹MRC/UVRI & LSHTM Uganda Research Unit, Statistics, Entebbe, Uganda, ²MRC/UVRI & LSHTM Uganda Research Unit, Entebbe, Uganda

Background: Intimate partner violence (IPV) is of public health significance as it impacts adversely on wellbeing. Adolescents involved in high risk behaviours are exposed to violence but often have limited access to prevention and treatment services. We studied the prevalence and factors associated with IPV among 14 to 19year old high‐risk adolescents in Kampala, Uganda.

Methods: We conducted a cross‐sectional study among male and female volunteers aged 14 to 19years. We defined IPV as a volunteer reporting prior experience of at least one form of violence (physical, emotional and sexual) from sexual partners. Indicators of spousal violence were combined to form two categories: never experienced the specific violence, and experienced at least one type of the specific violence (binary outcome). Data on socio‐demographics, sexual behaviour and substance use were collected. We analysed factors associated with IPV using logistic regression (Stata 15.0).

Results: We enrolled 365 participants, mean age 17.8 (±SD 1.02) years, of whom 64% were female. Of 184 (50%) who reported ever experiencing IPV, 105 (57%) had experienced IPV 3months prior to the interview. IPV was most prevalent among female participants, (79%). Prevalence of IPV by type was: emotional (35%), physical (28%) and sexual (24%). Females commonly reported sexual violence (89%) while males commonly reported emotional violence (22%). Emotional and physical violence were mainly perpetrated by regular non‐paying partners while sexual violence was mainly perpetrated by casual non‐paying partners. After adjusted analysis participants who were aged <18years (aOR = 0.57; 95% CI 0.34 to 0.96) and single (aOR = 0.46; 95% CI 0.23 to 0.93) were less likely to suffer IPV while earning from paid sex (aOR = 3.16; 95% CI 1.68 to 5.95) and alcohol dependency (aOR = 3.19; 95% CI 1.36 to 7.46) were highly associated with IPV.

Conclusions: Many adolescents experienced IPV, highlighting a need for support to reduce high‐risk behaviour and situations that expose them to IPV.

Factors associated with transactional sex in the African Cohort Study

N. Dear¹, A. Esber¹, M. Iroezindu², E. Bahemana³, H. Kibuuka⁴, J. Owuoth⁵, J. Maswai⁵, T. Crowell¹, J. Ake¹, C. Polyak¹

¹Walter Reed Army Institute of Research, U.S. Military HIV Research Program, Silver Spring, United States, ²Henry Jackson Foundation MRI, Nigeria, ³Henry Jackson Foundation MRI, Tanzania, United Republic of, ⁴Makerere University Walter Reed Project, Kampala, Uganda, ⁵Henry Jackson Foundation MRI, Kenya

Background: Transactional sex can put vulnerable groups at increased risk for HIV acquisition and transmission. Better understanding the characteristics of those engaging in transactional sex can help prevention programs identify and serve at‐risk groups. We identified factors associated with transactional sex in the African Cohort Study (AFRICOS).

Methods: AFRICOS prospectively enrolls adults at risk for HIV and persons living with HIV (PLWH) at 12 PEPFAR‐supported clinics in Uganda, Kenya, Tanzania, and Nigeria. At twice‐yearly study visits, questionnaires are administered and clinical outcomes assessed. Multivariate logistic regression with generalized estimating equations were used to estimate odds ratios (ORs) and 95% confidence intervals (95% CIs) for associations between factors identified a priori and engaging in transactional sex, defined as having received money, shelter, food, drugs, favors, or gifts in exchange for sex in the past 6 months. To evaluate HIV‐specific factors, a subgroup analysis was performed among PLWH.

Results: Between January 2013 and December 2019, 3466 participants were enrolled, of whom 83.1% were PLWH. At enrollment, 365 participants (10.3%), of whom 80.5% (n = 294) were PLWH and 71.2% (n = 260) female, reported transactional sex with a median of 1 transactional sex event per month (interquartile range 1 to 3). Factors independently associated with increased likelihood of transactional sex included female sex, younger age, clinical site, being unmarried, having less education, consuming alcohol, having experienced physical harm, having had forced sex, and not having enough food to eat in the past year (Table 1). Among PLWH, additional factors associated with transactional sex included status disclosure, experiencing HIV‐related stigma, and not taking antiretroviral therapy (ART).

Conclusions: Multidimensional socio‐behavioral and economic factors were associated with transactional sex. Among PLWH, those engaging in transactional sex were less likely to be taking ART, potentially amplifying transmission risk in this subgroup. Comprehensive public health programming that addresses these factors is essential to curbing the HIV epidemic.

Using emoji stickers to understand opinions of the dapivirine vaginal ring for HIV prevention among female end‐users and their male partners

A.W.K. Katz¹, L.E. Mansoor², M. Tsidya³, F. Mathebula⁴, D. Singh⁵, S. Siva⁶, C. Akello⁷, T.H. Chitowa⁸, M. Garcia⁹, L. Soto-Torres¹⁰, E.T. Montgomery¹

¹RTI International, Women's Global Health Imperative, Berkeley, United States, ²Centre for the AIDS Programme of Research in South Africa (CAPRISA), Durban, South Africa, ³UNC‐Lilongwe, Lilongwe, Malawi, ⁴Wits RHI, Johannesburg, South Africa, ⁵International Clinical Research Center, Department of Global Health, Seattle, United States, ⁶South Africa Medical Research Council, HIV Prevention Research Unit, Durban, South Africa, ⁷Makerere University‐Johns Hopkins University Research Collaboration, Kampala, Uganda, ⁸University of Zimbabwe College of Health Sciences Clinical Trials Research Centre (UZCHS‐CTRC), Harare, Zimbabwe, ⁹FHI 360, Durham, United States, ¹⁰NIH, Division of AIDS, Bethesda, United States

Background: The monthly dapivirine vaginal ring (VR) is a promising female‐initiated prevention method that, with high adherence, has shown reduced risk of HIV acquisition in clinical trials. Understanding VR acceptability among experienced end‐users and their male partners (MP) offers insight into factors that might impact its effective use at a population level.

Methods: To explore VR acceptability, following the Microbicide Trials Network (MTN) HIV Open‐label Prevention Extension (HOPE) trial, former HOPE participants and MP were recruited from 6 of 14 sites, representing all trial countries (Malawi, South Africa, Uganda, Zimbabwe). During in‐depth interviews (IDI) or focus‐group discussions (FGD), participants selected and placed emoji stickers on an “opinion tool” to stimulate discussion of attitudes towards the VR over time. Emoji use was tabulated; qualitative data were transcribed, translated, and coded using Dedoose software.

Results: Fifty‐eight women and one MP had an IDI; 53 men participated in one of 11 FGDs. 12 of 22 possible emojis (55%) were selected by at least 10% of women or MP (Table 1). Although emoji interpretation varied widely, the activity facilitated detailed discussions about VR opinions. Both groups felt primarily positive about the VR because it provided HIV protection, was easy to use, lacked side effects, improved or did not change sex, and enabled involvement in HIV prevention (MP only). Common negative opinions were shock or fear due to a lack of understanding (MP only), suspected lack of HIV protection, and fear of or actual side effects. Many women reported consistently positive opinions throughout HOPE. Some women and men's opinions improved over time with increased familiarity and receipt of more information about the VR, or counselor and staff encouragement.

Abstract PE01.26‐Table 1.

Conclusions: An emoji activity stimulated MP and women to provide multiple VR opinions, predominantly positive. Opinions pointed to educational messages that could enhance future understanding, acceptance and adjustment of ring use.

When to put it on: decision‐making regarding condom use among MSM PrEP‐users

H. Zimmermann, V. Jongen, A. Boyd, E. Hoornenborg, M. Prins, H. de Vries, M. Schim van der Loeff, U. Davidovich

Public Health Service Amsterdam, Department of Infectious Diseases, Amsterdam, Netherlands

Background: It is feared that pre‐exposure prophylaxis (PrEP) will result in decreased condom use and increased STI incidence. In this multi‐method study, we explored per‐sex act condom‐use decisions among PrEP‐users.

Methods: We analyzed data acquired with a mobile application of the Amsterdam PrEP project (AMPrEP) (August 2015‐February 2019) among men who have sex with men who used daily (dPrEP) or event‐driven PrEP (edPrEP). The app allowed participants to register daily their sex‐acts, PrEP‐use and condom‐use per partner type (steady (SP) and casual partners (CPs)). Additionally, we qualitatively analysed, using 43 in‐depth‐interviews among AMPrEP‐participants, motives for four PrEP‐ and condom‐use strategies:

(1)PrEP‐only;

(2)PrEP and condoms;

(3)condoms‐only; and

(4)no‐PrEP, no‐condom.

Results: 352 AMPrEP‐participants reported 48,949 anal sex‐acts, of which 11,632 were with SPs and 37,317 with CPs. PrEP‐only was the most common prevention‐strategy with any partner type (n = 39,650/48,949, 81%) regardless of PrEP‐regimen or time since PrEP‐initiation (Figure 1). PrEP‐only motivations were primarily based on prospects of more pleasurable sex. Combining PrEP and condoms was more often chosen for sex‐acts with CPs (n = 6,829/37,317; 18%) than for sex‐acts with SPs (n = 614/11,632; 5%) and was linked to higher STI/HIV‐risk‐perceptions, CPs wish for condom‐use, additional safety reassurance, or to avoid PrEP‐use disclosure. Condoms‐only was uncommon but occurred occasionally among edPrEP‐users (n = 379/8,695; 4%). dPrEP‐users only replaced PrEP by condoms if non‐adherent to PrEP. The no‐PrEP, no‐condom strategy occurred mostly among edPrEP‐users in sex‐acts with SPs (n = 538/2,122; 25%) and was motivated by low perceived HIV‐risk. We observed an increasing trend (p < 0.001) in practicing the no‐PrEP, no‐condom strategy in sex‐acts with SPs among edPrEP‐users since PrEP‐initiation, but not among dPrEP‐users (Figure 1).

Conclusions: Our data suggest that condom use remains a viable option among PrEP‐users in certain settings. Condoms were primarily applied in perceived higher risk settings, to avoid PrEP‐use disclosure, or as substitute for PrEP, especially among edPrEP‐users.

Grandmothers as key influencers on pregnant and breastfeeding women's health and HIV prevention related decision making in Johannesburg, South Africa

K. Reddy¹, T. Palanee‐Phillips¹, F. Mathebula¹, S. Tenza¹, J. Ryan², N. Macagna³, P. Musara⁴, A. van der Straten²

¹Wits Reproductive Health and HIV Institute, Research Centre, Hillbrow, Johannesburg, South Africa, ²RTI International, Women's Global Health Imperative, San Francisco, United States, ³FHI 360, Science Facilitation, Durham, United States, ⁴UZ‐UCSF Research Programme, Harare, Zimbabwe

Background: The MTN‐041/MAMMA study explored attitudes of pregnant and breastfeeding (P/BF) women, male partners and grandmothers (mothers/mothers‐in‐law of P/BF women) regarding use of the dapivirine vaginal ring (VR) and oral pre‐exposure prophylaxis (PrEP) during these periods of high HIV risk. Herein, we describe the influence of grandmothers on P/BF women's health and HIV prevention related decision making in Johannesburg, South Africa.

Methods: We collected behavioural data through surveys and conducted focus group discussions (FGDs) with three groups of participants (total N = 47): HIV‐uninfected, currently or recently P/BF women aged 18 to 40 years (median 26; N = 15), male partners aged ≥18 years (median 33; N = 12) and grandmothers (median age 57;N = 20). FGDs were conducted in English and/or Zulu by trained facilitators using semi‐structured guides and included a brief educational video about VR and oral PrEP and an opportunity to handle sample products. Discussions were audio‐recorded and summarized in debrief reports before being transcribed in English, coded using Dedoose software (v7.0.23), and thematically analysed.

Results: In surveys, maternal grandmothers were identified by 40% of P/BF women as key influencers on their decision making even though they considered themselves primary decision makers for medication use during pregnancy (60%) and breastfeeding (70%); a view similarly upheld by male partners (67% during pregnancy, 58% during breastfeeding). Grandmothers’ influence was further explored during FGDs, with maternal grandmothers described by all three groups of participants as important sources of information due to their experience using traditional medicine and their healthy pregnancies and children. Paternal grandmothers were also included as key influencers if the male partner had compensated an unmarried pregnant woman's family for a pregnancy [“paid damages”]. For HIV prevention‐related decision making and oral PrEP/VR use specifically, P/BF women indicated they would make the decision themselves as with medication but would welcome support from their grandmothers.

Conclusions: Grandmothers appeared to make significant contributions to P/BF women's health‐related decision making with potential to play a supportive role in HIV prevention product use. With the right framing and approach, grandmothers’ advice and support could optimise uptake and adherence to biomedical HIV prevention and further contribute to the decline in HIV acquisition among P/BF women in this community.

User recommendations to optimize a mobile phone sexual health risk assessment for HIV prevention clinical research

G. Benadé¹, M. Mulaudzi¹, A. Kagee², S. Hornschuh¹, M.P. Lemos³, E. Lazarus¹, M. Andrasik³, K.J. Horvath⁴, J.J. Dietrich¹

¹Perinatal HIV Research Unit, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa, ²Faculty of Arts and Social Sciences, University of Stellenbosch, Department of Psychology, South Africa, ³Fred Hutchinson Cancer Research Center, Vaccine and Infectious Disease Division, Seattle, United States, ⁴San Diego State University, Department of Psychology, United States

Background: Accurate self‐report of sexual behaviour assists in identifying potential HIV exposure in HIV prevention trials. Brief mobile‐phone assessments, completed daily or after sexual activity, can improve the validity and reliability of self‐reported sexual behaviour and allows for remote completion, outside of the clinic setting. Our study describes user recommendations to adapt and optimize an existing mobile phone sexual risk assessment.

Methods: We conducted four age‐stratified focus group discussions (FGDs) and analysed a brief socio‐demographics and mobile phone access questionnaire. All participants completed the existing sexual risk assessment before the FGD. Sexually active, HIV seronegative men (n = 14) and women (n = 15) 19 to 39 years were recruited through a HIV counselling and testing clinic and community outreach in Soweto. Using a framework analytic approach, data were coded with Nvivo software.

Results: All participants had access to mobile phones and internet, and 27 (93.1%) were able to download applications on their personal phones. Analysis of FGDs showed that participants prefer mobile risk assessments to be offered in a choice of South African languages, using formal language (as opposed to emojis), with straight‐forward wording and limited to five to ten questions. Most participants found it acceptable to complete the assessment once a week, on a weekday, while a few were willing to complete it after each sexual encounter. A message reminder to complete the assessment should be sent at least daily until it is done. The majority agreed that a password‐protected application with a discreet logo is ideal for privacy, ease of use and flexibility for completion in any setting. A concern with this format, however, was the potential data requirement. Participants expressed privacy concerns with using SMS, WhatsApp and other social media for the risk assessment. Most agreed on an airtime incentive between ZAR5‐10 (USD 0.29 to 0.58) per assessment. Participants encouraged researchers to provide feedback about their sexual risk with counselling to reduce their risk.

Conclusions: Completion of mobile phone risk assessments can be optimized with minimal incentives by ensuring questionnaires are simple, brief, infrequent and have well‐controlled privacy measures. Further investigation of methods to address risk behaviours identified in the mobile phone assessment responses is required.

Identifying profiles of sexual risk and PrEP initiation among Black sexual minority men in HPTN 073

D. Dangerfield II¹, I. Kuo², M. Magnus², G. Beauchamp³, S. Fields⁴, L. Nelson⁵, S. Shoptaw⁶, L. Wilton⁷, D. Wheeler⁸

¹Johns Hopkins School of Nursing, School of Nursing, Baltimore, United States, ²George Washington University School of Public Health, United States, ³Fred Hutchinson Cancer Research Center, Baltimore, United States, ⁴New York Institute of Technology, Baltimore, United States, ⁵Yale University School of Nursing, Baltimore, United States, ⁶University of California Los Angeles, Baltimore, United States, ⁷State University of New York at Binghamton, United States, ⁸Iono College, Baltimore, United States

Background: Black gay, bisexual, and other Black sexual minority men (BSMM) continue to bear the greatest HIV burden in the U.S. Efforts to successfully engage BSMM to use pre‐exposure prophylaxis (PrEP) are urgently needed. However, additional strategies to understand HIV risk profiles among BSMM are needed. Since few studies show high uptake of PrEP among BSMM, the purpose of this study is to identify sexual risk profiles and PrEP use among BSMM in the vanguard study HPTN 073.

Methods: A total of 226 BSMM were recruited from Los Angeles, CA, Chapel Hill, NC, and Washington D.C. from 2013 to 2015; 79% initiated PrEP. Latent class analysis (LCA) was used to identify sexual risk profiles using baseline data from HPTN 073 study (n = 226). Relationship status, condom use, number of sexual partners, substance use, STI history, and partner HIV status were used as latent class indicators as guided by the CDC PrEP risk behavior assessment. Age and PrEP initiation were used as covariates in a multinomial regression to identify correlates of class membership.

Results: Three latent classes were identified:

1)Single with Condomless Partners (69.4%),

2)Single with Multiple Partners (19.0%), and

3)Serodiscordant Partners (11.5%).

Single with Multiple Partners had the highest conditional probability of having condomless sex (90.5%), having greater than three male partners in the previous six months (93.6%), substance use before sex (58.1%), and receiving an STI diagnosis in the previous six months. Serodiscordant Partners had a 100% conditional probability of both engaging in condomless sex with a male partner and having a male partner who was living with HIV. Relative to BSMM who did not initiate PrEP, BSMM who initiated PrEP had 93% lower odds of being classified as Single with Condomless Partners than Serodiscordant Partners, after adjusting for age (AOR=0.07, 95% CI=0.02, 0.66).

Conclusions: Findings show low odds for PrEP use among single men with condomless partners, an important subgroup of BSMM with high‐transmission risk behaviors who comprise most of this sample. Conversely, serodiscordant partners were more likely to use PrEP. To increase uptake of PrEP, culturally relevant tailored and targeted messaging strategies for BSMM with a combination of high sexual risk indicators are needed.

Adolescent girls and young women's (AGYW) PrEP‐user journey during an implementation science study in South Africa and Kenya

E. Rousseau¹, A.W.K. Katz², S. O'Rourke², L.‐G. Bekker¹, S. Delany-Moretlwe³, E. Bukusi⁴, D. Travill³, V. Omollo⁴, J. Morton⁵, G. O'Malley⁵, R. Johnson⁵, C. Celum⁵, J.M. Baeten⁵, A. van der Straten²

¹Desmond Tutu Health Foundation, University of Cape Town, Cape Town, South Africa, ²RTI International, Research Triangle Park, United States, ³Wits RHI, University of Witswatersand, Johannesburg, South Africa, ⁴Kenya Medical Research Institute, Nairobi, Kenya, ⁵University of Washington, Seattle, United States

Background: Oral pre‐exposure prophylaxis (PrEP) efficacy in preventing HIV is well established. However, adhering to a daily regimen can be challenging, especially for adolescent girls and young women (AGYW) in sub‐Sahara Africa. With PrEP scale‐up ongoing, it is important to understand AGYW's motivations and lived experiences in the uptake and use of PrEP.

Methods: In‐depth interviews were conducted with a purposive sample of 91 AGYW (ages 16 to 25) in the POWER implementation project offering PrEP for up to 36 months from adolescent‐friendly, mobile, and family planning clinics, in Johannesburg and Cape Town, South Africa and Kisumu, Kenya, respectively. A rapid analysis of coded transcripts and interview summaries was conducted to explore AGYW's PrEP‐user journey from motivation to initiate, persistence (sustained PrEP use), discontinuation, and restarting PrEP.

Results: Motivations to initiate PrEP included partner infidelity, a current STI, history or fear of sexual violence, and having an HIV positive family member or partner. AGYW who initiated PrEP early in study displayed high awareness of HIV vulnerability, attributed to their intimate relationship dynamics or personal lifestyle, and frequently shared that starting PrEP empowered them to take control of their sexual health. Early PrEP use challenges and delayed uptake were often due to community stigma and PrEP misconceptions and AGYW prioritizing family, peers, and/or partners social approval (despite HIV vulnerability awareness). Disclosure to family and/or partners occurred early for AGYW who persisted with PrEP; conversely, most non‐persistors disclosed use later in the journey or not at all. Unplanned PrEP pauses occurred due to PrEP access problems when traveling to rural areas or attending school/work. Planned PrEP pauses occurred during periods of no sexual activity (e.g. when partners travel). Many AGYW who had PrEP interruptions restarted PrEP. PrEP discontinuation was often due to perceived side effects, low social support (particularly from mothers), perceived change in HIV risk or logistical PrEP access barriers.

Conclusions: AGYW in South Africa and Kenya recognize their HIV vulnerabilities and the benefits of PrEP, however implementing use is impacted by their social relationships and circumstances. Tailored flexible interventions are needed to address young women's diverse PrEP motivations, social contexts and understandings of prevention‐effective adherence.

Association between DREAMS invitation and attitudes towards gender norms amongst young women in urban and rural Kenya, measured using an adapted and validated version of the GEM Scale

K. Nelson¹, F. Magut², S. Mulwa¹, S. Khagayi², A. Ziraba³, D. Kwaro², S. Floyd¹, I. Birdthistle¹

¹London School of Hygiene and Tropical Medicine, Faculty of Epidemiology and Population Health, London, United Kingdom, ²Kenya Medical Research Institute, Centre for Global Health Research, Kisumu, Kenya, ³African Population and Health Research Center, Nairobi, Kenya

Background: Gender norms play an important role in shaping health‐related behaviours, experiences, and expectations from an early age. Shifting inequitable norms therefore has the potential to improve sexual and reproductive health (SRH) outcomes, including HIV risk. We evaluated the impact of DREAMS, a multi‐component HIV prevention programme, on individual attitudes towards gender norms amongst adolescent girls and young women (AGYW) in two Kenyan settings.

Methods: AGYW aged 15 to 22 in Gem (n = 888) and Nairobi (n = 1081), Kenya, were randomly selected for enrolment into survey cohorts (in 2017 and 2018, respectively) and followed‐up to 2019. We validated and adapted the Gender Equitable Men (GEM) scale using factor analysis, and used it to describe attitudes towards gender norms stratified by age group and setting. Logistic regression was used to estimate the association between being a DREAMS beneficiary by 2018 and GEM scores in 2019, adjusting for confounders.

Results: We identified four factors measured by the GEM Scale items: equitable norms around SRH decision‐making (n = 5 items; alpha=0.86), and inequitable norms around violence (n = 6 items; alpha=0.76), need for sex (n = 3 items; alpha=0.67), and initiation of sex (n = 2 items; alpha=0.62). The first two factors were considered as primary outcomes.

Support for equitable SRH decision‐making was high in both settings: 89% of AGYW in Gem and 90% in Nairobi had a mean score equivalent to ‘agree’ or ‘strongly agree.’ For inequitable norms around violence, only 47% of AGYW in Gem had a score representing mean disagreement, whereas in Nairobi, 82% of AGYW disagreed. There was no evidence of an association between DREAMS invitation and SRH decision‐making nor violence attitudes, in both Gem (SRH decision‐making: adjusted OR=1.13; 95% CI 0.69 to 1.87; Violence: aOR=0.84; 95% CI 0.62 to 1.08) and Nairobi (SRH decision‐making: aOR=1.28; (0.76 to 2.15); violence: aOR=1.08; 95% CI 0.72 to 1.64).

Conclusions: We found no evidence of an effect of DREAMS on individual attitudes towards gender norms amongst AGYW, after three years of DREAMS implementation. It is possible that shifts in norms at the community level are not immediately reflected in individual attitudes. While support for equitable norms was generally high, our findings highlight a need to understand and address inequitable norms around intimate partner violence, especially in rural Kenya.

Preferences for PrEP formulations and service delivery: Results from qualitative in‐depth interviews with transgender women in three South African cities

T. Poteat¹, L.L.A. van der Merwe², A. Cloete³, D. Adams⁴, M. Malik⁴, A. Wirtz⁴

¹University of North Carolina Chapel Hill, Department of Social Medicine, Chapel Hill, United States, ²Social Health Empowerment Feminist Collective of Transgender Women of Africa, South Africa, ³Human Sciences Research Council, South Africa, ⁴Johns Hopkins Bloomberg School of Public Health, Baltimore, United States

Background: Transgender women (TW) in South Africa have an HIV prevalence over 60% in some cities. The National Strategic Plan recommends pre‐exposure prophylaxis (PrEP) for transgender people. However, data on TW's preferences for PrEP formulations and service delivery is absent.

Methods: We purposively sampled 36 TW in Cape Town, East London, and Johannesburg between 06/01/2018 and 11/30/2018. Following informed consent, we conducted qualitative in‐depth interviews with 12 TW in each city in English, Xhosa, or Afrikaans. We asked about perspectives on oral, injectable, and topical PrEP; and preferences for PrEP service delivery. Interviews were audio‐recorded, transcribed verbatim, translated, and coded by two coders with discrepancies resolved by consensus. We grouped codes by topic and analyzed for salient recurrent themes.

Results: Table 1 lists participant characteristics and exemplar quotes for salient themes. Of the 36 participants, 32 had heard of PrEP, 6 had ever taken PrEP, and 3 were currently on PrEP. Dislike of daily dosing was a common theme and the most common reason for preferring injectable or topical to oral formulations of PrEP. Many TW expressed a desire for an injectable option; however, some participants preferred a topical formulation. Overall, the need for choices in prevention technologies emerged as salient theme. Reports of discrimination were ubiquitous, including mistreatment by health care providers. In fact, TW mentioned a reduction in health facility visits as an advantage of injectable PrEP over oral PrEP, which required monthly facility visits for medication refills. Throughout all interviews, the need for de‐stigmatizing services was a powerful and consistent theme.

Abstract PE01.34‐Table 1.

Conclusions: TW are eager for alternatives to daily pill taking; and provision of respectful transgender‐competent services will be key to PrEP engagement, regardless of PrEP formulation. This study contributes important data to inform implementation of PrEP services that meet the needs of TW.

Factors associated to the non‐intention to use PrEP in men who have sex with men in France: Results from the European MSM Internet Survey (EMIS‐2017)

M. Di Ciaccio¹, V. Villes¹, R. Delabre¹, T. Alain², D. Michels², A.J. Schmidt³, D. Rojas Castro¹, A. Velter⁴

¹Coalition PLUS, Community‐Based Research Laboratory, Pantin, France, ²AIDES, Pantin, France, ³Sigma Research, Department of Public Health, Environments and Society, London School of Hygiene and Tropical Medicine, London, United Kingdom, ⁴Santé Publique France, Saint‐Maurice, France

Background: PrEP (on‐demand or daily) has been fully reimbursed by the French national health insurance system since 2016. 50,000 Men who have Sex with Men (MSM) were estimated to be eligible for PrEP. However, there were an estimated 10 000 users in 2018. To increase PrEP coverage, we need to better characterize MSM who may benefit from PrEP but are not using it. The objective of this study was to identify factors associated with non‐intention to use PrEP among MSM who are (1) eligible according to French guidelines (based on sexual behaviours), and (2) aware of PrEP (‘non‐intenders’).

Methods: Data comes from EMIS‐2017, an anonymous, 33‐language, cross‐sectional online survey among MSM, conducted across 50 countries between October 2017 and January 2018. We assessed the percentage of PrEP users and ‘non‐intenders’ amongst respondents who have not been diagnosed with HIV living in France. Socio‐demographic characteristics, sexual behaviours and HIV knowledge were compared between PrEP users and ‘non‐intenders’ using logistic regression models.

Results: Among 8,009 participants, 734 (9.2%) were using PrEP and 1,098 (13.7%) were ‘non‐intenders’, with a median age of 38[IQR 31–46], and 35[26–45], respectively. ‘Non‐intenders’ were more likely to live in smaller towns, be a student, identify as bisexual, report fewer non‐steady intercourse partners and higher frequency of condom use with them. In the multi‐variable model, ‘non‐intenders’ were younger (<25yrs) (aOR[95% CI]= 3.68[2.14;6.33]), more likely to live in locations with few PrEP access points (1.74[1.13;2.68]), not have been spoken to about PrEP at a hospital (12.34[8.87;17.18]) or at a community service/drop‐in (4.88[3.45;6.89]), not report hepatitis B vaccination (2.20[1.45;3.35]), and have lower knowledge about on‐demand (11.62[7.45;18.11]) and daily PrEP (2.53[1.25;5.15]). They also reported greater self‐efficacy regarding safer sex (1.62[1.10;2.38]).

Conclusions: “Non‐intenders” were less engaged with the healthcare system and despite lower HIV‐related risk, were eligible for PrEP. While PrEP may not meet the prevention needs for all “non‐intenders” (preference for/use of TasP, self‐testing, etc), on‐demand PrEP may be a suitable HIV prevention tool for others. Therefore, health and community workers should increase their efforts to identify “non‐intenders” and talk about on‐demand PrEP in addition to other prevention tools.

Transgender women's experiences using a blood‐based, combination HIV/syphilis at‐home rapid test kit that delivers Results in 60‐seconds (INSTI Multiplex) for self‐ and partner‐testing

C. Tagliaferri Rael, J. Lopez-Ríos, C. Lentz, C. Dolezal, B. Kutner, R. Giguere, A. Carballo-Diéguez, I. Balán

NYSPI/Columbia University, HIV Center for Clinical and Behavioral Studies, New York, United States

Background: We report on the self‐ and partner‐testing experiences of N = 11 transgender women (TW) in New York City who used the INSTI Multiplex^Ò, an at‐home, blood‐based, combination HIV/syphilis rapid test that delivers results in 60‐seconds, and a corresponding smartphone app to scan test results.

Methods: TW participants were given six INSTI Multiplex^Ò tests to take home and use on themselves or with potential sexual partners. Participants were also oriented to the corresponding smartphone app used to scan test results. After one month, 11 participants returned for an interview on their experiences with the INSTI Multiplex^Ò, and the app. Themes discussed in the interviews included, what it was like to use the test, with whom participants used tests (including decision‐making around testing), partners’ reactions to the idea of testing/subsequent test results. Data were independently analyzed by two coders.

Results: Participants (N = 11) had a median age of 40 years. Roughly half (n = 6) reported Latinx ethnicity, and nearly all (n = 10) reported being women of color. Participants had a median of 10 sexual partners during the one‐month intervention. They used self‐test kits to test themselves, or a partner a median of 3.5 and 1.5 times, respectively. Some participants reported that remembering the test sequence of the INSTI Multiplex^Ò could be complicated, and that not all partners were willing to give blood through the finger prick procedure. However, the speed of test results (e.g., 60‐seconds), and the simultaneous syphilis test was motivating for participants and partners to overcome this in some cases. Participants reported that most partners who were offered the test were receptive, and few became angry or refused. Most of the time, partners who were offered the test were already known to the participant.

Conclusions: Overwhelmingly, participants’ experiences with the blood‐based INSTI Multiplex^Ò were positive. Most partners who were presented with the test were willing to use it. Though the finger prick procedure was not favorable, the fact that results appear in 60‐seconds, alongside a syphilis test result may have helped to overcome this. This suggests that using a rapid, blood‐based HIV/syphilis test could be a viable harm reduction strategy for TW.

Perspectives on use of PrEP from Black and Latinx sexual and gender minority youth

F. Shorrock¹, A. Alvarenga², K. Hailey-Fair², D. Celentano³, R. Arrington-Sanders²

¹Johns Hopkins School of Medicine, General Pediatrics and Adolescent Medicine, Baltimore, United States, ²Johns Hopkins School of Medicine, Pediatrics, Baltimore, United States, ³Johns Hopkins Bloomberg School of Public Health, Epidemiology, Baltimore, United States

Background: Pre‐exposure prophylaxis (PrEP) effectively protects against HIV, yet uptake has been low among sexual and racial minorities. We sought to understand sex‐specific PrEP use in a sample of Black and Latinx sexual and gender minority youth (BLSGMY) at‐risk for or recently diagnosed with HIV.

Methods: In‐depth interviews (IDIs) were conducted with 31 BLSGMY (19 at risk for and 12 living with HIV and young transgender women) age 15 to 24 who were recruited into randomized control trials aimed at increasing PrEP uptake and ART adherence. Interviews lasted 45 to 60 minutes and focused on barriers and facilitators of prevention and treatment. Interviews were transcribed verbatim. Inductive and deductive coding was used to organize excerpts. Coded transcripts were organized into individual and partner‐specific categories and then by HIV status. Emergent themes were grouped and categorized using a grounded theory approach.

Results: Nearly one‐half of BLSGMY at risk for HIV reported having been on PrEP (n = 9/19), with 5 currently taking or having recently taken PrEP. Most BLSGMY living with HIV (n = 11/12) reported never having heard of or taken PrEP prior to their diagnosis. For both groups, a desire to engage in protected condomless anal sex was the major theme that emerged around sex‐specific PrEP‐use. Non‐use themes included: 1) low perceived risk for HIV because of few concurrent or random sex partners; 2) feeling that condoms were more effective because they also protect against other STIs; and 3) feelings of trust, seriousness, and/or perceived monogamy in their partnership. Both at‐risk for and living with HIV respondents shared that discussions of PrEP eased ‘status disclosure conversations’ by promoting those at risk to ask about the status of their partners and youth living with HIV to disclose one's HIV status.

Conclusions: Sex positive communication with partners (particularly in serodiscordant relationships) and promotion of protective condomless sex may be a key factors in increasing BLSGMY to engage with PrEP. To effectively engage youth in PrEP, efforts will need to focus around perceptions of risk, condom efficacy and partner‐specific factors like trust, which may be key drivers of PrEP non‐use in this population.

Assessing longitudinal patterns of depressive symptoms and the influence of symptom trajectories on PrEP persistence among adolescent girls in HPTN 082

J. Velloza¹, S. Hosek², D. Donnell³, P. Anderson⁴, M. Chirenje⁵, N. Mgodi⁵, L.‐G. Bekker⁶, S. Delany-Moretlwe⁷, C. Celum³

¹University of Washington, Department of Global Health, Seattle, United States, ²Stroger Hospital of Cook County, Department of Psychiatry, United States, ³University of Washington, Seattle, United States, ⁴University of Colorado, United States, ⁵University of Zimbabwe, Harare, Zimbabwe, ⁶The Desmond Tutu HIV Centre, Cape Town, South Africa, ⁷University of the Witwatersrand, Johannesburg, South Africa

Background: African adolescent girls and young women (AGYW) who may benefit from HIV pre‐exposure prophylaxis (PrEP) face high levels of depression and related psychosocial issues, including PrEP stigma and intimate partner violence (IPV). Depressive symptoms at PrEP initiation may reduce PrEP adherence. However, the proportion of AGYW initiating PrEP with persistent depressive symptoms, the relationship between symptoms and other psychosocial factors, and the impact of persistent symptoms on PrEP adherence have not been studied.

Methods: HPTN 082 was an open‐label PrEP study among AGYW (ages 16 to 24) in Harare, Zimbabwe and Cape Town and Johannesburg, South Africa from 2016 to 2018. Depressive symptoms were measured at enrollment and months 3, 6, and 12, using the 10‐item Center for Epidemiologic Studies scale; a score >10 is correlated with clinical depression. PrEP stigma, IPV, sexual behavior, and PrEP adherence were also assessed at these visits. High PrEP adherence was defined as tenofovir diphosphate levels >=700 fmol/DBS punch. Group‐based trajectory modeling was used to model longitudinal patterns of depressive symptoms, assigning participants to a trajectory (e.g., persistent, sporadic, no symptoms). We assessed psychosocial predictors of trajectories to understand factors associated with symptom patterns and used generalized estimating equations to model associations between group trajectory membership and 12‐month PrEP adherence.

Results: At enrollment, 179 (41.9%) of 427 participants had elevated depressive symptoms consistent with clinical depression. 33.0%, 36.7%, and 36.9% had elevated symptoms at months 3, 6, and 12, respectively. Group‐based trajectory models revealed persistent elevated symptoms in 49.2%, declining symptoms in 13.5%, and steady low/no symptoms in 37.3%. AGYW who engaged in transactional sex, reported IPV, or had traumatic stress symptoms were more likely to be assigned to the persistent elevated symptom group compared with the steady low/no symptom group (Wald test p‐value all < 0.01). Participants assigned to the persistent depressive symptom trajectory had significantly lower odds of high PrEP adherence than those in the low/no symptom trajectory (aOR=0.73; 95% CI: 0.56 to 0.96).

Conclusions: Depressive symptoms were common and persistent among AGYW seeking PrEP in sub‐Saharan Africa. Integration of depression screening and treatment into HIV prevention programs may improve PrEP effectiveness among African women.

Attitudes towards two biomedical HIV prevention strategies among HIV‐negative men who have sex with men: an attitude network analysis in the Amsterdam Cohort Study

H. Zimmermann¹, U. Davidovich¹, F. van Harreveld²

¹Public Health Service Amsterdam, Department of Infectious Diseases, Amsterdam, Netherlands, ²University of Amsterdam, Department of Social Psychology, Amsterdam, Netherlands

Background: The impact of biomedical HIV‐prevention strategies among men who have sex with men (MSM) depends on uptake. Using an attitude network analysis, we aimed to investigate which beliefs are associated with the uptake of pre‐exposure prophylaxis (PrEP) and viral load sorting (VLS, i.e. treatment as prevention), which can be targeted in behavior change interventions.

Methods: HIV‐negative MSM reporting anal sex during the previous six months were drawn from one six‐monthly data wave (January‐June 2019) of the Amsterdam Cohort Study. We estimated a weighted, undirected, partial correlation network (Figure) where we included condom, PrEP and VLS use, HIV risk perception, 10 VLS‐beliefs and 23 PrEP‐beliefs. We report on (1) partial correlations to identify direct correlates of condom, PrEP and VLS use, (2) community detection to identify clustering beliefs, and (3) node centrality to identify the nodes most influential in the network.

Results: We included 532 HIV‐negative MSM, of whom 246 (46%) reported condom use, 131 (25%) reported PrEP use, and 39 (7%) reported VLS in the previous six months. Strongest correlates of condom use were no PrEP use (r = ‐0.21, p < 0.001) and lower HIV risk perception (r = ‐0.18, p < 0.001). Strongest correlates of PrEP use were lower expected burden of PrEP procedures (r = 0.12, p < 0.001) and PrEP's perceived positive impact on quality of sex life (r = ‐0.10, p < 0.001), the latter was also the strongest correlate of VLS use (r = 0.05, p =0.018). Community detection suggests that PrEP uptake clusters with several practical considerations such as PrEP's anticipated affordability (green in Figure). PrEP's expected impact on quality of sex life was the most central node within the network.

Conclusions: Our findings suggest that PrEP and VLS use are perceived as having a positive impact on one's quality of sex life which may be used in communication to improve uptake among non‐users. Addressing specific practical considerations of PrEP may also improve PrEP uptake.

Changes in risky sexual behaviour among adult men and women receiving regular personalised counseling in an HIV vaccine preparedness study in south‐western Uganda

J. Kitonsa¹, S. Kansiime², M. Onyango², K. Safina², D. Asio², A. Kabarambi², S. Kusemererwa², E. Ruzagira², P. Kaleebu²

¹MRC/UVRI & LSHTM Uganda Research Unit, Research, Entebbe, Uganda, ²MRC/UVRI & LSHTM Uganda Research Unit, Entebbe, Uganda

Background: Changes in sexual behaviour may occur over time when individuals receive regular HIV risk reduction counselling; while concerns about risk compensation among those initiating pre‐exposure prophylaxis (PrEP) remain. We assessed changes in sexual behaviour among adults in an HIV vaccine preparedness cohort study.

Methods: In preparation for an HIV vaccine efficacy trial, adults (18 to 45 years) at high risk of HIV infection were recruited from July 2018 to February 2020 in Masaka, south‐western Uganda. Volunteers were provided quarterly risk‐reduction counselling and those willing to start PrEP referred to a provider. Self‐reported sexual behaviour data were collected at baseline and every six months using standardised questionnaires. We assessed changes in HIV risk indicators between baseline and one year, using proportion differences, McNemar chi‐square test for binary predictors and ordinary chi‐square for other categorical predictors.

Results: Four hundred and forty‐one volunteers were enrolled of whom 184 (42%) had completed their one year assessments. Of the 184, 80 (43%) were female and 91 (49%) ≤25 years. The following differences in the prevalence of reported high‐risk sexual behaviour were observed between baseline and one year of follow‐up: unprotected sex with ≥3 partners in the last 3 months (47% vs. 17%, p < 0.001); transactional sex in the last 3 months (60% vs. 46%, p =0.004); sex when drunk (18% vs. 15%, p =0.04); ≥6 sexual partners in the last 3 months (23% vs. 16%, p =0.02). 38 (21%) of the participants had initiated PrEP at one year. Absolute differences in the prevalence of reported high‐risk sexual behaviour between baseline and one year were larger among participants who started PrEP: unprotected sex with ≥3 sexual partners in the last 3 months (32% vs. 30%); transactional sex in the last 3 months (21% vs. 11%); ≥6 sexual partners in the last 3 months (16% vs. 6%). No difference was observed in the frequency of drunken sex among participants who initiated PrEP.

Conclusions: We observed a reduction in reported high risk sexual behaviour in this cohort including in participants who initiated PrEP. Regular risk‐reduction counselling should be considered for individuals at high‐risk of HIV acquisition in addition to PrEP and other HIV prevention interventions.

Service providers perception of a home‐based intervention to test and start (HITS) in rural KwaZulu‐Natal, South Africa

O. Adeagbo¹, F. Tanser², K. Hae‐Young³, T. Mathenjwa³, T. Barnighausen³, N. McGrath⁴, A. Blandford⁵, M. Shahmanesh⁶, J. Seeley⁷

¹Africa Health Research Institute, Social Science & Research Ethics, Durban, South Africa, ²University of Lincoln, United Kingdom, ³Africa Health Research Institute, Durban, South Africa, ⁴University of Southampton, Faculty of Medicine, Southampton, United Kingdom, ⁵University College London, UCL Institute of Healthcare Engineering, United Kingdom, ⁶University College London, Institute for Global Health, United Kingdom, ⁷London School of Hygiene and Tropical Medicine, Global Health & Development, London, United Kingdom

Background: Men are missing from the HIV treatment cascade in South Africa, contributing to higher HIV cause‐specific mortality in men and onward transmission to their female partners. Home‐based Intervention to Test and Start (HITS), a factorial design randomised controlled trial (#NCT03757104) was designed to assess the effectiveness of financial micro‐incentives (R50[$3] food vouchers) and/or a male‐targeted tablet‐based counselling application (Empowering People through Informed Choice for HIV [EPIC‐HIV]) to support home‐based testing and linking men to care in rural South Africa. The use of a once‐off voucher increased the uptake of home‐based testing by more than 50%. We report on service providers’ (fieldworkers and clinical staff) perceptions of HITS study.

Methods: Ten in‐depth interviews and one group discussion were conducted with a purposive sample of fieldworkers who offered home‐based HIV testing [n = 10] and clinical staff [n = 4] providing HIV treatment and prevention services between August 2018 and February 2019. Transcripts were coded and categorised using NVIVO while identified themes were thematically analysed.

Results: Service providers reported that the intervention was delivered as planned and the voucher acted as a powerful ‘catalyst’ for, whilst EPIC‐HIV information “nudged” men towards, HIV testing and linkage to care. However, they were concerned that some participants with prior knowledge of their HIV status tested because of the voucher; sustainability of voucher provision; and poor linkage to care in men. Participants in non‐financial arms resented missing out on the voucher and fieldworkers sometimes felt exhausted explaining why certain participants and communities were ineligible for vouchers. They felt the training received was adequate, but the time allocated was too short to absorb the information before implementation.

Conclusions: Home‐based HIV service delivery and financial incentives have been advocated as tools to improve HIV outcomes and the HITS trial demonstrated that provision of a small once‐off voucher substantially increased the uptake of HIV testing. Fieldworkers and clinicians interviewed in this sample felt that whilst the vouchers had acted as a powerful catalyst for HIV testing, they were unsure whether such a strategy would be sustainable in the long term.

MTV‐SHUGA: how did MTV‐Shuga improve HIV prevention and sexual health outcomes among adolescents and young people in rural KwaZulu‐Natal?

T. Zuma¹, N. Kyegombe², S. Hlongwane³, M. Nhlenyama³, N. Chimbindi³, J. Seeley², M. Shahmanesh⁴

¹Africa Health Research Institute, Social Science and Research Ethics, Durban, South Africa, ²London School of Hygiene and Tropical Medicine, London, United Kingdom, ³Africa Health Research Institute, Durban, South Africa, ⁴Institute for Global Health, University College London, London, United Kingdom

Background: MTV‐Shuga is a mass‐media behaviour change intervention that seeks to address social and structural drivers of HIV infection amongst young people in South Africa. It is delivered through television storylines related to sexual and reproductive health, HIV/AIDS, abortion, transactional sex and intimate partner violence. We explore ways in which MTV‐Shuga could improve HIV prevention and sexual health outcomes among adolescents and young people in rural KwaZulu‐Natal.

Methods: Between May and November 2019, we conducted eight (n = 4 in schools and n = 4 in community settings) 22‐minute screenings of episodes of MTV‐Shuga Down South in five communities in rural KZN. Following these, and using a semi‐structured topic guide, we conducted 13 FGDs, 25 IDIs, and structured observations with female and male participants aged 15 to 30 years. IDIs and FGDs were conducted in isiZulu, audio recorded and transcribed verbatim. Data analysis was thematic.

Results: MTV‐Shuga may have had an impact through:

1)valuable access to information on sexual and reproductive health to adolescents;

2)positive storylines, through which adolescents could model positive and aspirational behaviours;

3)negative storylines including about physical abuse, forced sex, partying, abortion and dating sugar daddies (Blessers) may have offered young people an example of how their decisions could lead to adverse outcomes, although some participants believed that these storylines could influence young people to manifest these negative behaviours in their own lives;

4)participants identified with the characters or knew someone in their community who identified with the characters, and thus, easily understood, and examined characters’ choices. For example, participants critically engaged with scenes which encouraged submission and silence in females and control and coercion in males;

5)participants valued the space the viewings provided to discuss topics including, HIV, ART, contraceptives, and other sexual and reproductive health issues “alone” as young people. They juxtaposed this with the awkwardness of discussing these topics with parents or older adults in their households.

Conclusions: Edutainment interventions such as MTV‐Shuga positively impact young people's sexual and reproductive health through different mechanisms and could usefully complement other interventions designed for this purpose.

PrEP persistence among Black ciswomen in Chicago, USA

M. Pyra¹, L. Rusie², C. Blum², S. Irby², J. Ridgway³

¹Howard Brown Health, EEE, United States, ²Howard Brown Health, United States, ³University of Chicago, Medicine, United States

Background: While 11% of new HIV infections in the US are in Black ciswomen, few Black ciswomen are using PrEP and little is known regarding how they use PrEP.

Methods: We used prescription data from electronic medical records from 2015 to 2019 at an urban federally qualified health center in Chicago, IL, USA, to examine PrEP persistence (having≥6 PrEP pills per week) and retention (having≥1 quarterly HIV test) over the first 6months of PrEP use for all Black ciswomen. We used logistic regressions to examine demographic, clinical, and social factors associated with PrEP persistence and retention.

Results: Among 142 women, 74% were≤35 and 61% identified as straight; more than half used public insurance. Most PrEP initiation visits (49%) specifically mentioned PrEP as the chief complaint. The average zipcode HIV prevalence was 1.2%. Overall, 19% of ciswomen persisted on PrEP for 6months and 32% continued in care for HIV prevention. In the adjusted models for persistence, ciswomen with chief complaint of PrEP/PEP‐to‐PrEP were more likely to achieve persistence (aOR 7.64 [95% CI 1.29, 45.1]) compared to those with non‐STI/HIV chief complaints; neighborhood of residence (North, West, South or outside Chicago) and earlier year of PrEP initiation were also significantly associated with higher odds of persistence. Age, insurance status, provider race, and provider gender were non‐significant. In the adjusted models of retention, there were no significant associations; however neighborhood residence and higher HIV prevalence among females (by zipcode) were associated with lower retention, at p=0.07.

Conclusions: PrEP persistence and retention over the first six months were low among Black ciswomen in this sample; efforts are needed to better understand the reasons that Black ciswomen may step away from PrEP, and explore what support could be helpful for them to sustain PrEP when desired. Persistence was significantly higher among ciswomen who sought PrEP during their medical visit, suggesting that pre‐existing knowledge of PrEP may be an important factor. Neighborhood was significantly associated with persistence and almost significantly associated with retention, suggesting the importance of structural factors. Community‐level awareness of and access to PrEP may be important avenues to improve PrEP use among Black ciswomen.

Examining how support influences the association between relationship control and postpartum STI amongst South African adolescent mothers

L. Gebrekristos¹, A. Groves¹, H.L. McNaughton Reyes², S. Maman², D. Moodley³

¹Drexel University Dornsife School of Public Health, Community Health and Prevention, United States, ²University of North Carolina Gillings School of Global Public Health, Health Behavior, Chapel Hill, United States, ³University of KwaZulu‐Natal, Department of Obstetrics and Gynaecology, South Africa

Background: Adolescent girls, and particularly adolescent mothers, are disproportionately impacted by HIV infection in sub‐Saharan Africa. Recent studies have focused on the impact of sexual relationship dynamics (i.e., control) on adolescent mothers’ HIV/STI risk. Yet, the protective effects of familial and peer support, has been understudied. Therefore, this study examines whether familial and peer support buffers the association between adolescent mothers’ relationship control in pregnancy and STIs in the first 6months postpartum.

Methods: Adolescent mothers (<20years) were recruited at a hospital's maternity ward near Durban between July 2017 and April 2018. Participants completed biological and behavioral assessments at 6weeks (baseline) and 6months postpartum (end line). Adolescents who were HIV‐negative and had no STIs at baseline (n=61) were included in the analyses. Relationship control during pregnancy was measured using the Sexual Relationship Power Scale. Higher scores indicated higher relationship control. We used a modified Poisson regression with robust standard errors to test whether familial and peer support weaken the association between relationship control during pregnancy and STI risk at 6months postpartum, controlling for covariates.

Results: 13% of adolescent mothers had at least one incident STI at end line (Table 1). The median relationship control score was 36 (IQR: 32 to 39). In multivariate analysis, there was a significant interaction between relationship control and familial support. For adolescent mothers with high familial support (85th percentile), post‐hoc analysis indicated that relationship control was negatively associated with incident STI (simple slope: b=−0.17; p=0.01) (Figure 1). However, there was no association between relationship control and familial support for adolescent mothers with average (50th percentile) or low (15th percentile) familial support. Interaction between relationship control and peer support was not significant.

Conclusions: Our findings suggest relationship control is negatively associated with STI risk for adolescent mothers with high familial support. Future research is needed to identify other social factors that mitigate adolescent mothers’ HIV/STI risk. Interventions that increase familial support during the perinatal period may reduce HIV/STI risk, particularly for those women in highly controlling relationships. Targeted interventions aimed to reduce STIs among this particularly vulnerable population can improve adolescent mother and her baby's health.

“What the %*^! is PrEP?”: what social media engagement with MTV Shuga reveals about knowledge and attitudes to PrEP

V. Baker¹, I. Birdthistle¹, S. Cousens², S. Sarrassat², C. Cawood³, D. Khanyile³

¹London School of Hygiene and Tropical Medicine, Department of Population Health, London, United Kingdom, ²London School of Hygiene and Tropical Medicine, Department of Infectious Disease Epidemiology, London, United Kingdom, ³Epicentre Health Research, Durban KwaZulu‐Natal, South Africa

Background: ‘MTV Shuga’ is a multi‐media campaign designed to equip young people with knowledge and resources to make informed choices about their sexual and reproductive health and avert HIV. It is centered around a popular edutainment TV drama, with social media activities to boost participatory engagement. Recent series have aimed to increase PrEP awareness and we examine viewers’ knowledge and attitudes about PrEP, as they engage with the show via social media.

Methods: We extracted 2,296 YouTube comments from MTV Shuga (‘Down South 2) episodes featuring PrEP storylines. Posts were downloaded and transcripts analyzed using an inductive thematic approach. Emerging themes on PrEP were discussed and refined in meetings with the production and research teams.

Results: Viewers’ comments reflected mixed levels of knowledge about PrEP. Some were learning of PrEP for the first time (“What the hell is prep? I never heard of it”) while others willingly shared knowledge (“PrEP has some side effects when you first use, but once your body is used to it you're fine”). People posted questions to learn more about PrEP (“Does PrEP alter your body functions?”) and received advice and resources from online peers or MTV Shuga staff. One person was critical about the PrEP messaging in the show saying it felt “forced and unnatural” and some shared false, negative information (“Dat prep sh*t is killing ppl and causing castration”) to which production staff replied with corrections (“You are incorrect, fam”).

Conclusions: MTV Shuga raised audience members’ interest and awareness of PrEP on social media. PrEP was a topic that viewers were motivated and comfortable discussing online. Social media was a useful tool for correcting misinformation, reinforcing accurate messages around PrEP and disseminating additional resources. It revealed both support and resistance to PrEP, offering insights to improve PrEP uptake among young people.

The impact of PrEP intent and use disclosure on PrEP adherence and persistence among South African adolescent girls and young women

D. Giovenco¹, A. Pettifor¹, K. Gill², J. Morton³, A. van der Straten⁴, C. Celum³, L.‐G. Bekker²

¹University of North Carolina at Chapel Hill, Epidemiology, Chapel Hill, United States, ²Desmond Tutu HIV Centre, Cape Town, South Africa, ³University of Washington, Global Health, Seattle, United States, ⁴RTI International, Global Health Imperative, Research Triangle Park, United States

Background: Despite knowledge of the important roles various social groups play in the lives of most young people, there is a lack of research exploring social influences on PrEP use. The objective of this analysis was to determine if PrEP disclosure was associated with adherence and persistence among adolescent girls and young women (AGYW).

Methods: The 3Ps for prevention study (3P) was a prospective cohort study conducted among 200 HIV‐uninfected AGYW in Cape Town, South Africa using daily oral PrEP for 12months. Participants were asked if they disclosed their intent to use PrEP at enrollment and their PrEP use during follow‐up separately for partners, parents, and peers. High PrEP adherence was defined as tenofovir‐diphosphate (TFV‐DP)≥700 fmol/punch and persistence as detectable TFV‐DP (≥16 fmol/punch) in dried blood spots. Modified Poisson regression was used to determine if disclosure at enrollment or during follow‐up was associated with adherence and persistence through month 3. Findings were stratified by age and assessed for effect measure modification.

Results: Participants had a median of 19years (range=16 to 25). At enrollment, most (91%) had disclosed their plans to use PrEP–52% to a partner, 62% to a parent, and 79% to a friend. Similar proportions disclosed their PrEP use to at least one partner (56%), parent (57%), and friend (82%) during the first three months of follow‐up. Almost half (48%) of participants had high adherence and 83% were persistent with PrEP through month 3. Among AG (≤18years), those who disclosed their intent to use PrEP to a parent at enrollment were more likely to persist (aRR=1.41, 95% CI: 1.04 to 1.92) and those who disclosed their PrEP use to a friend during follow‐up were less likely to persist (aRR=0.88, 95% CI:0.81 to 0.97) with PrEP through month 3. These associations were not observed among YW (>18years). There was evidence of effect measure modification by age (Wald p‐value for interaction=.041 and .039, respectively).

Conclusions: PrEP intent disclosure to parents may facilitate 3‐month persistence among AG. Further, PrEP disclosure to peers was frequently reported but negatively associated with persistence. PrEP programs for AG that involve peer support should find ways that peers can motivate PrEP use.

Engaging female caregivers to improve South African girls’ and young women's sexual and reproductive health outcomes

M. Atujuna¹, K. Merill², S. Ndwayana¹, E. Emerson², L. Fynn¹, L.‐G. Bekker¹, G. Donenberg²

¹Desmond Tutu Health Centre, Department of Medicine, Health sciences Faculty, UCT, Cape Town, South Africa, ²Centre for Dissemination and IS, University of Illinois Chicago, Chicago, United States

Background: Despite global reductions in incident HIV infections, South African adolescent girls and young women (AGYW) continue to acquire HIV at twice the rate and much earlier than their male counterparts (Shisana et al 2012; UNAIDS,2015). Innovative approaches are needed to enhance the prevention toolbox for AGYW. An underutilized yet potentially important strategy is to capitalize on family strengths to reduce sexual risk in AGYW. In particular, female caregiver‐adolescent girl relationships may be critical to fostering supportive environments for HIV risk reduction among adolescent girls and young women in South Africa.

Methods: The study reports the steps taken to systematically contextualize and adapt an evidence‐based mother‐daughter HIV/STI prevention program (IMARA) for 15 to 19‐year‐old South African AGYW and their mothers/female caregivers. IMARA consists of joint and separate mother‐daughter activities designed to strengthen family relationships and communication, increase condom use self‐efficacy, improve maternal monitoring, and promote gender and racial/ethnic pride. Consistent with the ADAPT‐ITT model (Wingwood & Diclemente, 2008), we implemented several stakeholder activities, including community advisory board (CAB) meetings (youth n=10; adults N=9). Next, mother‐daughter dyads were recruited through street outreach, libraries, and schools to participate in theatre testing (N=76) where facilitators presented the IMARA curriculum and requested feedback. We analyzed data using the framework analysis approach.

Results: CAB members, AGYW, and mothers uniformly reported that the intervention was acceptable, timely, suitable, and should ideally be offered at schools, churches, clinics, and anywhere mothers and daughters congregate, either together or separately. Participants placed importance on balancing long‐held cultural traditions with modern lifestyle without disempowering and reducing mothers’ role in the upbringing of their daughters. Mothers and daughters embraced the effective communication module as essential, facilitating close mother‐daughter relationships and easing discussions around sensitive topics. Families underscored the urgent need for accurate information about sexual and mental health, HIV, STIs, and the different prevention tools including PrEP.

Conclusions: Family relationships are an important yet underused strategy to enhance prevention efforts for AGYW. IMARA appears to be acceptable and promising in the South African context with adaptations and the next steps include testing effectiveness in this setting.

Multi‐level factors influencing temporary and permanent interruptions in PrEP use among young women in Siaya County, Kenya

B. Perry¹, K. Agot², D. Ochieng Ngoje², A. Corneli¹

¹Duke University, Population Health Sciences, Durham, United States, ²Impact Research and Development Organization, Kisumu, Kenya

Background: Given the dynamic nature of sexual behavior, people may start and stop PrEP as their perceived HIV risk fluctuates, referred to as ‘seasons of risk’. Understanding the multi‐level factors influencing short‐ and long‐term interruptions in young women's (YW) PrEP use is vital to inform programs supporting YW through these ‘seasons’.

Methods: As part of a larger photovoice project on PrEP adherence and persistence among YW in Kenya's Siaya County, we explored temporary and permanent interruptions in YW's PrEP use. YW taking PrEP (n=18) took photographs depicting reasons YW 1) stop taking PrEP for short periods of time, such as a few days or weeks, and 2) stop taking PrEP permanently. YW discussed their photographs in a group with other YW. All discussions were conducted in Dholuo or Kiswahili, audio recorded, simultaneously translated and transcribed, and analyzed using applied thematic analysis.

Results: YW described that interpersonal factors often led to temporary interruptions in PrEP. Participants reported that YW stop PrEP temporarily 1) if they perceived their live‐in male partner would disagree with their PrEP use, and 2) when temporarily living apart from their partner (e.g., when he travelled for work). Permanent PrEP interruption was primarily related to intrapersonal and community factors. Intrapersonal factors included lower perceived risk for HIV due to changes in YW's known risk factors, such as changes in their own sexual behaviors, or assumed changes in partner's behaviors. Community factors focused on the influence of church/religious teachings, such as promotion of monogamous marital relationships, sexual exclusivity, rejecting Luo inheritance rituals, and relying on faith and prayer to preserve and protect health over medication. Negative rumors about PrEP's side effects and disparaging remarks about PrEP users, could also influence permanent PrEP use interruption, as could relational and structural factors, such as negative pressure to stop taking PrEP from partners, doctors, and peers, and frequent clinic stock outs.

Conclusions: Temporary and permanent PrEP interruptions is impacted by intrapersonal, interpersonal, and community influences. PrEP counseling should focus on reasons YW chose to stop taking PrEP to ensure YW remain safe when stopping PrEP.

Impact of the DREAMS Partnership on educational attainment among adolescent girls and young women: Causal analysis of a prospective cohort in urban Kenya

S. Mulwa¹, J. Osindo², E.O. Wambiya², A. Gourlay³, I. Birdthistle³, A. Ziraba², S. Floyd³

¹London School of Hygiene & Tropical Medicine, Department of Population Health, London, United Kingdom, ²Africa Population and Health Research Center, Nairobi, Kenya, ³London School of Hygiene & Tropical Medicine, London, United Kingdom

Background: DREAMS promotes a comprehensive HIV prevention approach, including supporting adolescent girls and young women (AGYW) to stay in school and achieve secondary education. Educating girls has a multiplier effect, with the potential to reduce HIV risk both directly and indirectly.

Methods: In two informal settlements in Nairobi, a cohort of 1081 AGYW aged 15 to 22years was randomly selected in 2017 and followed‐up to 2019. AGYW who reported invitation to participate in DREAMS during 2017 to 2018 were classified as “DREAMS beneficiaries”. We used current schooling status and highest level of education in 2019 to create composite measures of educational attainment. Our main outcome was being in school and/or having completed at least secondary form two. Using causal inference framework, we estimated the proportions achieving this outcome if all AGYW, versus no AGYW, were DREAMS beneficiaries, adjusting for the propensity to be a DREAMS beneficiary.

Results: Of 852 AGYW followed‐up in 2019, 63% and 45% were in school at cohort enrolment and endline, respectively, and 5% re‐enrolled in 2018 or 2019 (Table 1). Proportions in each education category were higher among DREAMS beneficiaries than non‐DREAMS beneficiaries (Figure 1:Panel A). DREAMS was estimated to increase proportions in school or completing minimum secondary form two from 82% if none were DREAMS beneficiaries to 86% if all were beneficiaries (+4% [95% CI −2,10%], with stronger evidence of effect among younger versus older AGYW (Figure 1:Panel B).

Conclusions: We found some evidence that DREAMS improved educational attainment among AGYW living in the Nairobi informal settlement areas, facilitating both retention and re‐enrolment.

Willingness to use HIV self‐tests among female sex workers from diverse sex work contexts: A qualitative study in São Paulo, Brazil

D. Ferraz¹, L. Murray², I. Sorrentino³, E. Miura Zucchi⁴, A. Grangeiro³

¹Fundação Oswaldo Cruz, Escola FIOCRUZ de Governo, Brazil, ²Universidade Federal do Rio de Janeiro, Brazil, ³Universidade de São Paulo, Faculdade de Medicina, Depto de Medicina Preventiva, Brazil, ⁴Universidade Católica de Santos, Programa de Pós‐Graduação em Saúde Coletiva, Brazil

Background: The introduction of new HIV prevention technologies, such as HIV self‐test (HIV‐ST), demands understanding the possible repercussions of their use. This is especially important in sex work (SW) contexts as environmental and structural factors, like working conditions and stigma, have been shown to influence HIV prevention. We analyzed female sex workers’ (FSW) willingness to use HIV‐ST considering their perception of their work environments.

Methods: Data was collected as part of Combina, a demonstration study on HIV combination prevention conducted in five Brazilian cities. From October 2018 to February 2019, we interviewed 12 FSW in São Paulo from diverse SW contexts: street, massage parlors, night clubs, and apartment buildings. Interviews focused on their perceptions of their work environments, HIV prevention practices, and willingness to use HIV‐ST. Data were analyzed using thematic analysis.

Results: Participants’ ages ranged from 21 to 38. Nine had a high school degree or higher. All but one reported skin color as black or brown. Nearly all evaluated their workplaces positively, valuing autonomy, fair payment, and safety. All but one had previously been tested for HIV, but only three had heard of HIV‐ST. None had previously used it. Although time was a concern for use with clients, all were interested in using HIV‐ST in their workplace and believed their colleagues would also be. Forms of possible use varied from personal to a way to earn more money from condomless sex, especially in oral sex. Perceived risks of use included: being coerced into testing or condomless sex, clients feeling offended or suspecting the FSW had HIV, manager/pimp reprimands, and fear of not knowing how to react if faced with a positive result (for themselves or clients).

Conclusions: Knowledge of HIV‐ST was low among participants. FSW believed its use would be feasible across all SW contexts, which might be related to their broader positive evaluation of their workplaces. The identified risks indicate that HIV prevention programs need to promote HIV‐ST appropriate use, ensure strong connections to HIV treatment and support services, and educate clients and /pimps, in addition to FSW, on the technology. HIV‐related stigma also must be addressed as a cross‐cutting issue.

South African mothers and daughters perspectives on a family‐based intervention to improve girls’ sexual and mental health

L. Fynn¹, M. Atujuna¹, K. Merrill², S. Ndwayana¹, E. Emerson², L.‐G. Bekker¹, G. Donenberg²

¹Desmond Tutu Health Centre, Department of Medicine, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa, ²Centre for Dissemination and IS, University of Illinois, Chicago, United States

Background: South African adolescent girls and young women (AGYW) have one of the highest HIV incidence rates globally and twice the rate of their male peers. Addressing their needs is essential to achieve an AIDS‐free generation. Group‐based mother–daughter interventions are a promising strategy to promote positive attitudes and behaviours related to mental health and HIV prevention. This study reports on the acceptability and feasibility of IMARA (Informed, Motivated, Aware and Responsible Adolescents and Adults), an evidence‐based mother‐daughter HIV prevention program for the South African context.

Methods: A feasibility study was conducted at the Desmond Tutu HIV Foundation Youth Center (DTHFYC) in a township on South Africa's Cape Peninsula. Black mothers and daughters, 15 to 19years‐old, were recruited over one month using street outreach and clinic‐based recruitment. Interested AGYW (n=146) provided contact data to obtain additional information, and 67% (n=43) of mother‐daughter dyads attended at least one of two sessions and provided feedback on IMARA's curriculum and its acceptability and feasibility for HIV/STIs prevention efforts.

Results: AGYW (M=17years‐old) and mothers (M=36years‐old) provided recommendations to improve attendance and intervention implementation. Both noted that attendance by biological mothers would be difficult because they cannot miss work as the primary breadwinners. Also, South African culture typically relies on maternal aunts to provide sexual and reproductive health guidance to AGYW. Participants suggested including alternative female caregivers to address this concern. Participants also recommended limiting the intervention duration to fit family needs. In terms of acceptability, mothers and daughters actively participated in the sessions and offered recommendations to engage future participants. Mothers welcomed the opportunity to engage with daughters on sensitive health topics and form relationships with other mothers who differ in parenting styles and HIV prevention strategies. Mothers and daughters expressed satisfaction and excitement about the IMARA content and its potential impact on reducing HIV infections among AGYW in their community.

Conclusions: This study supports the feasibility and acceptability of a group‐based mother‐daughter HIV prevention intervention in South Africa with careful consideration of innovative and creative recruitment and retention strategies.

Attitudes towards PrEP and correlates of common mental disorder symptoms and substance use among young people

T. Nematadzira¹, H. Weiss², L. Stranix-Chibanda¹, T. Bere³, A. Ssemata⁴, R. Muhumuza⁴, J. Dietrich⁵, S. Vermaak⁶, G. Tshabalala⁶, N. Ahmed⁷, M. Atujuna⁷, J. Seeley⁴, J. Fox⁸

¹University of Zimbabwe College of Health Sciences Clinical Trial Research Centre, College of Health Sciences, Harare, Zimbabwe, ²MRC Tropical Epidemiology Group London School of Hygiene and Tropical Medicine, London, United Kingdom, ³University of Zimbabwe College of Health Sciences, Department of Psychiatry, Harare, Zimbabwe, ⁴Medical Research Council/Uganda Virus Research Institute and London School of Hygiene &Tropical Medicine Uganda Research Unit, Entebbe, Uganda, ⁵University of Witswatersrand, Perinatal HIV Research Unit/Health Systems Research Unit/South African Medical Research Council, South Africa, ⁶University of Witswatersrand, Perinatal HIV Research Unit, South Africa, ⁷Desmond Tutu HIV Foundation, University of Cape Town, Cape Town, South Africa, ⁸King's College London, London, United Kingdom

Background: Poor mental health can increase HIV risk by influencing sexual behavior and hindering uptake of HIV prevention strategies. We assessed the associations of attitudes to HIV prevention strategies with common mental health symptoms and substance use, among young people (YP) in sub‐Saharan Africa.

Methods: We recruited 673 male and 657 female HIV‐uninfected YP aged 13 to 24years in South Africa, Uganda and Zimbabwe as part of the Combined HIV Adolescent PrEP and Prevention Study (CHAPS) in 2019. Participants completed a local language, tablet‐based, quantitative survey evaluating PrEP attitudes/preference, HIV risk behavior and alcohol/substance use. Outcomes included depression (PHQ‐9>10), anxiety (GAD‐2>3), self‐reported binge drinking (>6 drinks on one occasion at least once a month) and self‐reported drug use in the past month. Multivariable logistic regression assessed factors associated with screening positive for each condition, adjusting for site, age and sex.

Results: 40.7% (540/1328) participants screened positive for at least one condition (depression:17.4%, anxiety:15.7%, binge drinking:19.2%, and drug use:12.6%). Poor mental health was more common in South Africa than in Zimbabwe and Uganda. Probable depression was associated with female sex (aOR=2.02; 95% CI 1.49 to 2.77) and age 18 to 24 vs 13 to 15years (aOR=2.04; 95% CI 1.17 to 3.55). Only 25% of YP had heard about PrEP. Binge drinking was associated with concern about PrEP drug side‐effects (aOR=2.09, 95% CI 1.16 to 3.77), forgetting to take pills (aOR=2.15, 95% CI 1.08 to 4.29) and sexual risk disinhibition (aOR=2.68, 95% CI 1.34 to 5.36). Anxiety disorder symptoms were associated with concerns about the cost of paying for PrEP (aOR=2.15, 95% CI 0.99 to 4.69). There was no effect of depression, anxiety, binge drinking and drug use with potential use of condoms, injectable PrEP as HIV prevention modalities. However binge drinking and drug use were associated with less willingness to use HIV testing and counselling services (binge drinking: aOR=0.69, 95% CI 0.49 to 0.97; drug us; aOR=0.73; 95% CI 0.49 to 1.09).

Conclusions: There was some variation in attitudes to HIV prevention strategies by mental health and substance use behaviors integrating mental health and substance use screening into HIV prevention services and specific HIV prevention messaging depending on mental health symptoms could be an effective strategy to promote in PrEP use as an HIV prevention strategy among young people.

Challenges in the Implementation of PrEP in Peru: a qualitative perspective from the experience of MSM continuers and discontinuers in the ImPrEP study

J.P. Jiron Sosa¹, K.A. Konda¹, C. Sandoval Figueroa¹, J.V. Guanira², E.H. Vega Ramirez³, V. Veloso⁴, C.F. Cáceres Palacios¹

¹Universidad Peruana Cayetano Heredia, Lima, Miraflores, Peru, ²Investigaciones Médicas en Salud (INMENSA), Lima, Lima, Peru, ³Instituto Nacional de Psiquiatría Ramón de la Fuente Muñiz, Ciudad de México, Mexico, ⁴Instituto Nacional de Infectologia Evandro Chagas, Fundação Oswaldo Cruz, Rio de Janeiro, Brazil

Background: In 2018, when the PrEP Implementation project, ImPrEP, started in Peru, PrEP became a reality in HIV prevention for many MSM at very high risk for HIV in the country. While there were many enrollees searching to use PrEP and succeeded to adapt, others faced various challenges that prevented adherence and led them to discontinue PrEP. This study explored the experiences, perceptions and challenges faced by continuers and discontinuers from ImPrEP.

Methods: We conducted semi‐structured interviews with 27 MSM (13 continuers; 14 discontinuers) from 3 regions of Peru. The interviews explored five dimensions: cognitive/affective, sexual behavior and preventive practices, interactions with the community, care seeking and provider interactions, and Structural. The interviews were analyzed to produce recommendations to improve various aspects of the ImPrEP Project.

Results: Cognitive/affective dimension: Continuers possess more precise information on the effectiveness, use, and indications of PrEP, and feel trust, security, greater freedom and satisfaction in their sexual encounters; they see PrEP as a need in terms of their wellbeing and ability to control their sexual health; in turn, discontinuers explain their dropping out based on a low perceived risk, side effects, and difficulty in adhering to the daily PrEP regimen. Sexual behavior and preventive practices: A key difference was motivation prior to recruitment among continuers, while discontinuers were mostly offered PrEP at enrollment. Interaction with the community: both groups perceived positive reactions from their family and friends and negative reactions from sexual partners. Care seeking and provider interactions: For both groups, being able to trust their providers was essential. Structural: both groups perceived health facilities as offering little privacy; operating at inflexible hours, and exposing them to situations of discrimination based on sexual orientation and gender identity.

Conclusions: Key aspects in PrEP continuation included: appropriate background awareness; prior motivation to use PrEP due to reasons of safety, control and personal satisfaction. Conversely, discontinuers perceived themselves to be at low risk, and were less tolerant to side effects and to the adherence and regular check‐up requirements of the program.

‘When you are old like me, then you get circumcised, you will become like a castrated bull’: Barriers to VMMC uptake in Eswatini

A. Adams

Center for HIV/AIDS Prevention Studies (CHAPS), Health, Mbabane, Eswatini

Background: Compelling evidence from three randomized controlled trials which showed that voluntary medical male circumcision (VMMC) reduces HIV infection from women to men by up to 60% led to the WHO recommending that VMMC be implemented in 14 priority countries. As one of the priority countries, Eswatini aimed to reach 80% VMMC coverage among males 10 to 49years since program inception in 2009. By the end of 2019, the country had reached a modest 40%. With the highest HIV prevalence in the world at 27% among those aged 15 and above, VMMC remains a crucial HIV prevention strategy in the country. This study aimed to explore the reasons why there is limited VMMC uptake among Swazi men as well as strategies improve to the program.

Methods: Six focus group discussions were used to collect data from males aged 15 to 49years in Mbabane East in 2020 and analyzed using thematic content analysis.

Results: There are significant barriers among Swazi men which lead to the continued slow uptake of VMMC services in Eswatini. The study found that VMMC is symbolically linked, especially for uncircumcised men, to perceptions of masculinity. VMMC is constructed as a threat to masculinity at multiple levels, including reduced sexual performance. Perceptions of masculinity are also intrinsically linked to material realities around income provision for families. As many men are breadwinners within their family, men were concerned about time lost while healing from a circumcision wound and consequently being unable to provide for their dependents. Respondents also report that ambiguous information available on VMMC is a barrier to uptake. Finally, Participants also assumed that HIV testing is essential for circumcision. The fear of taking an HIV test may stop men from getting circumcised.

Conclusions: VMMC programmes need to consider the material realities of the men who are targeted by the programme. These include clients compensated for time lost after circumcision, mobilizers should target families, and sensitizing political and traditional leaders as gate keepers.

HIV, risk and time preferences: evidence from a general population sample in Lesotho

M. Björkman Nyqvist¹, L. Corno², D. De walque³, J. Svensson⁴

¹Stockholm School of Economics, Sweden, ²Universita’ Cattolica del Sacro Cuore, Italy, ³The World Bank, Development Research Group, Washington, United States, ⁴Stockholm University, Institute for International Economic Studies, Sweden

Background: Identifying individuals most at risk of HIV infection is a priority for policy makers. Apart from specific groups, however, little is known about how to identify and target those at high risk in a population at large. Research in psychology and behavioral economics suggests that attitudes towards risk (risk preferences or risk sensitivity) and how individuals’ trade off costs and rewards over time (time preferences or delayed gratification) may influence risky sexual behavior, but no studies have so far investigated the interplay between risk attitudes, time preference and HIV infection. Using data from a two‐year trial, we assess the correlation between baseline measures of risk and time preferences, and the subsequent risk of becoming infected by HIV.

Methods: We collected data on risk and time preferences using hypothetical games (multiple price list (MPL) method) at baseline, and HIV prevalence over a two‐year period (2010 to 2012), among 675 participants, males and females 18 to 32years old drawn from 29 rural and peri‐urban villages in Lesotho. We report unadjusted and adjusted odds ratios between HIV prevalence at endline and the measures of attitudes towards risk and time preference, measured at baseline. Adjusted OR are adjusted for gender, age, marital status, education and wealth.

Results: The positive association between HIV prevalence at endline and risk‐loving attitudes at baseline is statistically significant and robust, while the association with risky behavior as measured by either alcohol or tobacco consumption or elicited measures of time preferences (present‐orientation and hyperbolic discounting) is positive but not statistically significant.

Conclusions: A measure of attitude towards risk which is relatively easy to administer to individuals in a survey is highly predictive of future HIV status. This is an important finding for policy makers and suggests the importance of targeting HIV prevention programs to risk‐loving individuals and therefore improving program efficiency.

Lessons learned from rolling out the Stepping Stones program in informal settlements in South Africa

C. Milford¹, J. Pulerwitz², M. Mtshali¹, S. Psaki², B. Zieman², M. Beksinska¹

¹MRU (MatCH Research Unit), Obstetrics and Gynaecology, Faculty of Health Sciences, University of the Witwatersrand, Durban, South Africa, ²Population Council, New York, United States

Background: The community‐based Stepping Stones program – designed to promote HIV prevention through addressing HIV risk, violence, relationship skills, and gender – has been implemented widely. Yet, limited evidence is available about the challenges/successes of program scale‐up – and program effects ‐ when adapting to new settings.

Methods: Qualitative research was used to explore the experiences and effects of implementing Stepping Stones (as a component of USAID/CCI's Community Responses program) in informal settlements in KwaZulu‐Natal, South Africa. Six focus group discussions (n=52) and 14 in‐depth interviews were held with community participants (34 males, 32 females), plus ten IDIs with program staff. Eight Stepping Stones sessions were directly observed (different session topics/facilitators). Interviews were transcribed and coded, and analyzed via NVivo v10.

Results: Male and female participants expressed positive reactions to Stepping Stones, both in terms of HIV and GBV‐related knowledge gained, and the opportunity to discuss sensitive issues in a safe and supportive environment –“they are able to help you talk […] and feel comfortable”. They also reported improvements in couple communication, HIV service use, and HIV prevention behaviors – “I now test every month and I use the condom”. They preferred mixed sex groups. However, there were some challenges – sourcing an appropriate venue was difficult – “we don't have venues”, and competing priorities (such as work commitments) made it difficult for community members to stick to session times. In addition, some facilitators found it difficult to follow the session manuals, and often skipped some program content.

Conclusions: Continued implementation of Stepping Stones in informal settlements of South Africa, for HIV and violence prevention, is feasible and acceptable. The program was well received, and positive effects were reported. Yet there were also challenges with implementation in the informal settlement setting. Recommendations include: shorter sessions as well as catch‐up sessions, balance of sexes in groups, careful venue selection (e.g., indoors), and ongoing mentorship of facilitators.

Using social maps to explore young women's experiences with social support of their oral PrEP use in Kenya and South Africa

A.W.K. Katz¹, E. Rousseau², N. Khoza³, F. Mogaka⁴, E. Bukusi⁴, S. Delany-Moretlwe³, L.‐G. Bekker², J. Morton⁵, R. Johnson⁵, C. Celum⁵, J. Baeten⁵, A. van der Straten¹

¹RTI International, Women's Global Health Imperative, Berkeley, United States, ²Desmond Tutu Health Foundation, University of Cape Town, Cape Town, South Africa, ³Wits RHI, University of the Witwatersrand, Faculty of Health Sciences, Johannesburg, South Africa, ⁴Kenya Medical Research Institute, Nairobi, Kenya, ⁵University of Washington, Seattle, United States

Background: Oral PrEP adherence is challenging for adolescent girls and young women (AGYW) in sub‐Saharan Africa, despite their desire to stay HIV‐free. We explored AGYW's views on social influencers of PrEP use and AGYW's perception of those influencers’ PrEP knowledge and support.

Abstract PE01.59‐Table 1.

Methods: Focus group discussions (FGD) were conducted with a purposive sample of AGYW during the POWER PrEP demonstration project (South Africa and Kenya). Participants completed a social mapping exercise by placing pre‐labeled stickers of PrEP influencers (e.g. mother, sex partner, clinic counselor) on an egocentric circle map, representing their level of influence (from inner/most influential circle to outer/least influential circle). Participants color marked the influence as positive (e.g. encouragement and reminders to use PrEP; advice and guidance), negative (e.g. discouragement/disapproval, judgement), or both. AGYW then discussed their map with the group.

Results: Six FGDs occurred with 33 AGYW. Mothers and counselors were labeled mostly as positive influencers and placed in the inner circle by >50% of participants; sex partners were also placed in the inner circle by a majority but were either labeled negative influencers or both (Table 1). Regarding peers, best friends (41% inner circle) were mostly positive influencers whereas “friend groups” (25% inner circle) were negative or both. For the inner circle, AGYW mentioned both direct (e.g. criticism or praise) and indirect influence (e.g. partner's behavior or knowing a person living with HIV). Participants labeled some outer circle influencers as uninterested or unwilling to learn about PrEP. Participants wanted all levels and types of influencers to be better educated about PrEP and ultimately to accept and support their PrEP use.

Conclusions: Through social mapping, AGYW described key supporters and detractors, with mothers, counselors, and best friends emerging as important supporters of AGYW's PrEP use. To improve PrEP outcomes, community‐ and peer‐based PrEP sensitization and delivery programs should be evaluated.

What do HIV clinic administrators think about PrEP implementation in Colombia? A qualitative CFIR guided study

S.A. Gómez¹, J.L. Martínez-Cajas², H.F. Muéses^3,4, B. Alvarado-Llano², X. Galindo^3,4, P. Camargo², E. Martínez^5,4, J. Torres⁶, M. Arrivillaga¹

¹Pontificia Universidad Javeriana, Cali, Colombia²Queen's University, Kingston, Canada³Corporación de Lucha contra el Sida, Cali, Colombia⁴Red VIH‐Col, Cali, Colombia⁵Universidad del Valle, Cali, Colombia⁶Montefiore Medical Center, New York, United States

Background: PrEP implementation needs to be informed by the perceptions of health service administrators. This study uses the Consolidated Framework for Implementation Research (CFIR) to gain insight into the perspectives of managers of health care services of HIV clinics in Colombia given eventual attempts to implement PrEP nationally.

Methods: This qualitative study conducted semi‐structured interviews in a purposive sample of HIV clinic administrators. The CFIR was used to guide data collection and analysis. Content Analysis was used to determine the relevance of one of the CFIR constructs, namely characteristics of the intervention on PrEP implementation in HIV clinics in Colombia.

Results: Twenty interviews were conducted representing eight HIV organizations (IPS) at eight Colombian cities. Two major themes emerged: PrEP acceptability and PrEP Complexity. Regarding acceptability, PrEP was perceived as an effective intervention, having a relative advantage when compared to other prevention strategies, amenable to pilot implementation in a setting with capacity for this. Notable barriers were uncertainty about how the intervention would be covered by the health care system, the misinformation in health care providers, and issues of compatibility e.g. moral values. Regarding complexity, managers expressed barriers related to the lack of guidelines, sociocultural diversity that may require adaptations (e.g. rural vs. urban; trans vs. gay men), and the need to adjust existing institutional resources (e.g. space allocation, training of personnel). Administrators worried about medication adherence and misuse, stigma, and risk compensation.

Conclusions: Structuring this situational analysis around the CFIR was instrumental in identifying multi‐level factors that will affect the large‐scale adoption of PrEP. Future research is encouraged to identify among all the barriers the one to target and propose some implementation strategies.

MRSI evaluation in the whole brain of HIV‐1 clade C infected treatment naïve individuals

D. Aggarwal¹, G. Garg¹, A. Sharma¹, S. Vyas¹, A. Sharma¹, M. Mohanty¹, C. Ahuja¹, D.J. Weiss², S. Weiss², M. Kumar², V. Govind², P. Singh¹

¹Postgraduate Institute of Medical Education and Research, Radiodiagnosis and Imaging, Chandigarh, India, ²University of Miami, Miller School of Medicine, Miami, United States

Background: HIV can enter into the central nervous system in the initial stages and over time can cause HIV‐Associated Neurocognitive Disorders (HAND). HIV viral density and infection status vary across brain anatomical regions. Also, there is an ambiguity in the association of different type of HIV clades with HAND. A single centre blinded study was conducted to evaluate the HIV‐1 clade C infection in the whole brain of treatment naïve people living with HIV through MRSI.

Methods: Treatment naïve HIV+ (age between 18 to 45years) were recruited from the HIV clinic of a teaching research institute in North India (2017 to 2020). Age and sex matched healthy controls were also recruited. Subjects with history of cerebral vascular diseases, brain injury, neurological or psychiatric illness, or any other CNS disease were excluded. Demographic details and clinical characteristics were recorded. CD4, viral load and clade subtype were investigated. Routine and whole brain MRI and MRSI evaluation were performed. Average metabolite values (NAA and Cho) and ratios (NAA/Cre, Cho/Cre, and Cho/NAA) over each lobar region were evaluated.

Results: 147 subjects were enrolled (HIV+: n=79; HIV‐: n=68). The median age was 30years with 67% male in HIV+ and 61% male in HIV‐ group. Average number of years of education was similar in the two groups (10.35years in HIV+ vs. 11.15years in HIV‐ group). Most of the HIV+ individuals were recently diagnosed and asymptomatic. Median CD4 count and viral load were 346 (range 14 to 938) and 20300 copies/mL (441‐6120561), respectively (Table 1). MRSI data showed altered metabolite levels throughout the brain of the subjects in the HIV+ group. There was a significant decrease of NAA and NAA/CR (p <0.05) and a significant increase of Cho/NAA (p <0.05), within all lobes of the HIV+ as compared to HIV‐ individual. Also, the levels of MI increased but NAA decreased significantly in HIV+ group (p <0.05).

Conclusions: The altered metabolites indicate neuronal dysfunction or loss (NAA) and inflammation (MI and Cho) in HIV+ treatment naïve individuals as compared to HIV‐ cohort. Since, the findings were consistent across different regions, the involvement of whole cerebrum is suspected.

MyPrEP Decision Support Tool increases PrEP persistence in adolescent girls and young women attending an urban primary health care clinic in South Africa

D. Seidman¹, D. Travill², C. Celum³, C. Dehlendorf⁴, M. Mdlovu², K. Zewdie³, D. Donnell³, J. Morton³, J. Baeten³, S. Delany-Moretlwe²

¹University of California San Francisco, Obstetrics, Gynecology & Reproductive Sciences, San Francisco, United States, ²Wits Reproductive Health and HIV Institute, University of the Witwatersrand, Johannesburg, South Africa, ³University of Washington, Seattle, United States, ⁴University of California San Francisco, Family & Community Medicine, San Francisco, United States

Background: Meeting the HIV prevention needs for adolescent girls and young women in Southern Africa, where the HIV incidence remains high, includes efficiently delivering PrEP services using person‐centered care approaches. Decision support tools (DSTs) have been successfully used in healthcare to improve patients’ risk perception and knowledge of options, and facilitate informed decision‐making. Patient‐facing DSTs may be particularly beneficial in busy clinical environments, or when new services are integrated and clinicians may be less consistent with counseling, such as the integration of PrEP into primary health clinics (PHC).

Methods: We developed a tablet‐based, patient‐facing DST, MyPrEP, which included youth‐friendly images, young women's narratives about HIV prevention decision‐making, and information about PrEP. Women ages 18 to 25 and HIV‐negative, presenting to a PHC in Johannesburg South Africa in 2019 to 2020, were randomized by day to either the DST or a general health website on a tablet (control). Participants were then seen by study clinicians blinded to DST vs. control allocation, who provided standard counseling and offered PrEP. Participants completed a survey and laboratory‐based STI testing. We assessed PrEP initiation and PrEP persistence at 1month by pharmacy records.

Results: Of 386 women screened, 353 were randomized. Participants’ median age was 21years, 98% were unmarried, 100% were sexually active in the prior 3months, 85% used condoms sometimes or never, and 37% were diagnosed with gonorrhea or chlamydia. Participants’ reasons for attending the PHC that day included 57% for an HIV test and 29% for family planning services. All characteristics were similar between the DST and control groups. PrEP initiation was high in both groups (97% DST vs. 94% control, p =0.2). At 1month, PrEP persistence was higher in the DST compared to the control group (20% vs. 11%, OR 1.97, 95% CI 1.08 to 3.69, p =0.03).

Conclusions: A patient‐facing DST offered at a South African PHC resulted in two‐fold higher PrEP persistence at 1month among young women. While overall PrEP persistence was low in this population not specifically presenting for HIV prevention services, findings suggest a brief, one‐time intervention at PrEP initiation, such as a PrEP DST, may support young African women's informed decision‐making about PrEP.

“…I am here going strong”: HIV‐positive adolescents’ perspectives on the influence of Operation Triple Zero initiative on HIV adherence and viral load suppression in Kisumu, Kenya

E. Ododa^1,2, G. Owino³, F. Odhiambo^2,4, M. Moghadassi⁵, N. Okoko^2,4, C.R. Cohen⁵, E.A. Bukusi^2,1, J.L. Kulzer⁵

¹Family AIDS Care and Education Services, (FACES), Kisumu, Kenya, ²Center for Microbiology Research, Kenya Medical Research Institute, Kisumu, Kenya, ³Family AIDS Care and Education, HIV Testing Services, Kisumu, Kenya, ⁴Family AIDS Care and Education Services, Clinical Services, (FACES), Kisumu, Kenya, ⁵University of California San Francisco, Department of Obstetrics, Gynecology & Reproductive Sciences, San Francisco, United States

Background: Adolescents living with HIV (ALHIV) in sub‐Saharan Africa have higher rates of virologic failure compared to adults. Operation Triple Zero (OTZ) aims to empower ALHIV to achieve: zero missed appointments, zero missed drugs, and zero viral load with emphasis on treatment literacy and psycho‐social support. This study explored ALHIV perspectives on OTZ and its influence on viral suppression.

Methods: In September 2019, eight focus group discussions (FGDs) were purposively sampled to recruit 80 ALHIV (10 to 21years) in eight government health facilities implementing OTZ in Kisumu County, Kenya. FGDs were conducted in the local language using semi‐structured guides, audio‐recorded, transcribed, and analyzed using Dedoose software.

Results: Of 80 participants, 42 (53%) were female, median age was 13years (range: 10 to 21years). Themes encompassed OTZ influence on health ownership, social support and long‐term health impact. ALHIV exhibited self‐drive to achieve the three zeros for a healthy life. “The OTZ pledge enlightens and motivates me to commit myself to zero missed appointments, zero missed drugs and reminds me to take full charge of my life” and “I can say OTZ has helped me because I now have a suppressed virus”. Social support garnered from OTZ helped ALHIV gain confidence, develop friendship bonds, and value health. ‘’Before I joined OTZ, I never knew other HIV positive adolescents existed. In OTZ I found friendship which has encouraged me to take my life seriously and care for my health so I can succeed in life…”. Participants expressed increased self‐acceptance, ability to manage discrimination, and share and address challenges. “OTZ has helped me a lot. I was isolated by friends at school, they talked ill about me, discriminated me until I stopped taking my medication but due to OTZ, I am here going strong.” Sustaining viral suppression in their adult lives will be aided by OTZ: “…all that we are taught in OTZ will help throughout one's life”.

Conclusions: Through OTZ, ALHIV feel motivated and capable of sustaining viral suppression now and in their adult lives. Sense of belonging and confidence gained through OTZ helps ALHIV manage discrimination, secure social support systems, and navigate obstacles.

Structure of a CD4 binding site directed antibody in a donor with broadly neutralizing antibodies

T. Moyo¹, C. Scheepers¹, T. Sicard², Z. Makhado¹, F. Ayres¹, R.E. Mapengo¹, J.‐P. Julien², L. Morris¹, P. Moore¹

¹National Institute for Communicable Diseases, HIV Virology Section, Johannesburg, South Africa, ²The Hospital for Sick Children Research Institute, University of Toronto, Canada

Background: HIV broadly neutralizing antibodies (bNAbs), required for an HIV vaccine, develop in only a subset of chronically HIV‐infected individuals. In some infected donors, a single dominant bNAb lineage confers breadth. In contrast, in other donors, multiple antibodies with moderate neutralizing activity contribute to breadth, a developmental pathway that may be easier to achieve by vaccination.

Methods: Here, we studied an HIV‐infected donor, CAP314, who developed at least three distinct antibody specificities that contributed to breadth within 2years of HIV infection. We isolated several monoclonal antibodies (mAbs) from this participant at multiple weeks post‐infection (wpi) including mAb CAP314_52 from 115 wpi which exhibited 15% breadth. To define its epitope, we crystallized the antibody in complex with a gp120 envelope protein from the CAP314 transmitted/founder (TF) virus and obtained a structure at 3.3 Å resolution.

Results: CAP314_52 has a uniquely long CDRL3 loop insertion of 29 amino acids, supported by a disulphide bond. This is unlike most HIV‐1 bNAbs which have CDRL3 loops ranging from 5 to 12 amino acids. Sequence analysis of the CAP314_52 lineage implicated the CDRL3 in neutralization breadth. Early lineage members from 24 wpi lacked the insertion, having a CDRL3 of only 9 amino acids, whereas antibodies from 90 wpi which exhibited increased neutralizing activity had the extended CDRL3. Structural analysis showed that the CAP314_52 CDRL3 loop interacts with residues in the CD4bs of the HIV envelope, and is responsible for>50% of the interactions between this antibody and its epitope. The CDRH3 also mediates epitope contacts within the CD4bs, though to a lesser extent. The angle of approach of CAP314_52 is more similar to HJ16 (with 30% breadth) than the exceptionally broad CD4bs bNAb, VRC01 (91% breadth).

Conclusions: Sequencing and structural analysis of a novel CD4bs antibody lineage suggests that a CDRL3 insertion contributes substantially to the binding and neutralization breadth of mAb CAP314_52. The characterization of novel neutralizing antibodies targeting the CD4bs may provide insights into the diverse pathways used by antibodies to target this site.

Antibody‐antigen distance of broadly neutralizing HIV‐1 antibodies correlates with glycan‐shield coverage of HIV‐1 envelope trimer

M. Lee¹, R. Rawi¹, L. Shapiro², P. Kwong¹, G.‐Y. Chuang¹

¹National Institutes of Health, National Institute of Allergy and Infectious Diseases, Bethesda, United States, ²Columbia University, New York, United States

Background: Antibody‐antigen distance (AAD) – the distance between an antibody and a protein antigen – is a fundamental parameter governing antibody recognition. We have recently quantified AAD for over 1000 non‐redundant antibody‐protein antigen complexes in the Protein Data Bank, and found AAD to have a Gaussian distribution, with 2.3 Å standard deviation (SD). We observed a few antibody‐antigen complexes (less than 1%) to have AAD of>5 SDs longer than average. Interestingly, all of the longer AAD were broadly neutralizing antibodies directed against HIV‐1. None of the antibodies against other viral antigens, including those against influenza hemagglutinin and ebolavirus glycoprotein, had an AAD extending>3 SDs longer than average.

Methods: Here, we hypothesized that the very dense glycan shield of HIV‐1 might provide an explanation for the observed extended AADs. To test this hypothesis, we developed a computational method incorporating molecular dynamics simulations to quantify the number of glycan atoms neighboring each surface residue on the HIV‐1 envelope (Env) trimer. Epitope‐glycan coverage was determined by normalizing the number of glycan atoms with surface area of the epitopes on the protein surface and averaging over one microsecond of molecular dynamics simulation.

Results: We found antibodies with high AADs to recognize region of the Env trimer with high glycan shield coverage, whereas, antibodies with average AADs to recognize regions of the Env trimer with lower glycan coverage. High correlation (R²=0.776, p<0.0001) was observed between AAD and glycan coverage for 23 classes of HIV‐1 broadly neutralizing antibodies.

Conclusions: Glycan coverage of the Env‐trimer surface thus correlates with extended AADs for antibodies recognizing these surfaces. We are currently exploring AAD for antibodies recognizing glycan holes, and how an understanding of AAD and its relationship to glycan coverage may provide insights into the parameters governing the elicitation of broadly neutralizing antibodies.

Variation in neutralization susceptibility of HIV‐1 Indian subtype C to potent and broadly neutralizing monoclonal antibodies (bnAbs) having distinct epitope specificities

N. Hingankar¹, R. Mullick¹, J. Sutar¹, S. Deshpande¹, M. Bansal¹, N. Kumar¹, L. Morris², D. Sok³, J. Bhattacharya¹

¹Translational Health Sciences & Technology Institute, HVTR (Infection and Immunology), Faridabad, India, ²National Institute for Communicable Diseases, Centre for HIV & STI's, Johannesburg, South Africa, ³International AIDS Vaccine Initiative, Scripps Research Institute, IAVI's Neutralizing Antibody Center, San Diego, United States

Background: Broadly neutralizing antibodies (bnAbs) have shown promising results in phase I clinical trials as agents for HIV‐1 prevention and treatment. However, given the substantial geographic variation of HIV‐1 subtypes, we aim to assess neutralization resistance and sensitivity of available bnAbs against region‐specific circulating HIV‐1 subtypes/recombinants in India. In the present study, we report neutralization susceptibility of 56 HIV‐1 Indian subtype C Env‐pseudotyped viruses against four potent bnAbs targeting three different epitope specificities on HIV Env (VRC01, PGT121, PGDM1400 and CAP256.VRC26).

Methods: Complete gp160 env genes were amplified from donor plasma (n=24) and cloned into pcDNA3.1TOPO mammalian expression vector. Env‐Pseudotyped viruses were produced in 293T cells and tested in standard neutralization assays using TZM‐bl cells. Basis for neutralization resistance was assessed through sequence analysis and CATNAP database.

Results: At a concentration of 5µg/mL, we identified several strains (36/56, 64.28%) with resistance (IC50>5µg/mL) against at least one of these four bnAbs and some (17/56) with resistance against at least two bnAbs. However, only four viruses were resistant to three of the four bnAbs and none resistant to all four. There are varying degrees of resistance of Env‐pseudoviruses to CAP256.VRC26 (14/56), PGT121 (15/56), PGDM1400 (19/56) and VRC01 (9/56). Detailed sequence analysis reveals that characteristic substitutions at the key epitope residues are associated with neutralization resistance. In some cases, resistant viruses with some intact epitope residues indicates not yet known env sequence features that contribute to neutralization resistance.

Conclusions: Our study indicates that Indian subtype C strains among those chronically infected with HIV‐1 are relatively resistant to individual bnAbs, but are less resistant or sensitive to combinations of bnAbs targeting different epitopes. This resistance profile presents an urgent need for assessing the efficacy of bnAbs that are under clinical development, individually as well as in combinations, against a larger panel of subtype C strains that are currently circulating in different geographical regions and distinct key population in India. This assessment will clarify our understanding as to whether these promising bnAbs will be efficacious individually and/or in combination against Indian subtype C.

Antibody genetic diversity with large structural variation in a South African population

A. Marsden¹, W. Gibson², O. Rodriguez³, A. Ismail⁴, B. Lambson¹, P. Moore¹, C. Watson², L. Morris¹, C. Scheepers¹

¹National Institute for Communicable Diseases, Centre for HIV and STIs, Johannesburg, South Africa, ²University of Louisville School of Medicine, Biochemistry and Molecular Genetics, United States, ³Icahn School of Medicine at Mount Sinai & Icahn Institute for Genomics and Multi‐Scale Biology, New York, United States, ⁴National Institute for Communicable Diseases, Sequencing Core, Johannesburg, South Africa

Background: The immunoglobulin heavy chain variable (IGHV) genes make up the largest portion of the antigen binding site of an antibody. These genes are highly polymorphic and demonstrate copy number variation (CNV) due to structural rearrangement of the immunoglobulin heavy chain (IGH) locus. There is increasing evidence that this genetic variation leads to differential immune responses in infection and vaccination. Previous studies of South African populations have described substantial undocumented allelic diversity within the antigen binding region of IGHV genes. However, the level of structural and regulatory region diversity, including CNV, within this population, is unknown.

Methods: We used two methods to explore IGHV genetic diversity. Targeted amplicon NGS (using PacBio and MiSeq) was used to investigate single nucleotide variation (SNV) across the entire IGHV gene, including the regulatory regions, in 70 CAPRISA participants. Large structural variants (SVs) were studied in two CAPRISA individuals using an IGH hybridization capture assay in which IGH specific probes were used to isolate large DNA fragments before amplification and sequencing using PacBio.

Results: Of the 85 novel alleles that our group has previously reported, 36 were found in new individuals included in this larger cohort. In addition, another seven novel IGHV alleles were discovered in multiple individuals. Furthermore, we observed>3000 SNVs in the regulatory elements and intergenic regions of the IGHV genes, which were detected across half the sample population. CNV was present in ~33% of these participants with one gene being deleted and eight IGHV genes being duplicated; one of which (IGHV1‐69) was duplicated in over 20% of participants sequenced by targeted NGS. Further, IGH capture identified 5 large SVs (an insertion or deletion greater than 50bp) that led to the insertion of 10 genes, demonstrating that the genetic diversity in this population exists at both nucleotide and structural levels.

Conclusions: This study demonstrates that considerable genetic complexity remains to be uncovered in the African antibodyome. Studying this genetic variation is important for understanding l immune responses to infection and vaccination, particularly as the continent is disproportionately burdened by infectious diseases such as HIV, TB and malaria.

Founder Env‐specific IgM B cell responses during acute HIV‐1 infection associate with the development of broadly neutralizing antibodies

S. Townsley¹, G. Donofrio¹, N. Jian¹, D. Leggat¹, V. Dussupt¹, L. Mendez-Rivera¹, L.A. Eller¹, B. Slike¹, P. Ehrenberg¹, A. Geretz¹, N. Doria‐Rose², V. Polonis³, J. Mascola², M. Rolland¹, S. Tovanabutra¹

¹Walter Reed Army Institute of Research, HJF/U.S. Military HIV Research Program, Silver Spring, United States, ²NIH, Vaccine Research Center, Bethesda, United States, ³Walter Reed Army Institute of Research, U.S. Military HIV Research Program, Silver Spring, United States

Background: Generating broadly neutralizing antibodies (bNAbs) that overcome the sequence diversity of HIV‐1 envelope glycoprotein (Env) is thought to be a critical component toward the design of a protective vaccine, yet no HIV‐1 vaccine candidate to date has successfully elicited bNAbs. Understanding the humoral response to Env during the acute stages of infection may provide insight into the development of bNAbs.

Methods: Longitudinal PBMCs from the RV217 cohort spanning from acute to chronic infection were assessed for founder Env specific‐ B cell phenotyping and binding antibodies in 13 broad and 12 non‐broad neutralizers. B cells that were able to bind to their respective founder Envs were analyzed using flow cytometry and sorted for BCR sequencing. In addition, longitudinal plasma were tested for antibody binding to respective founder Envs using a Luminex assay.

Results: Higher frequencies of founder Env‐specific B cells 1month post‐infection were predictive of the development of neutralization breadth (p=0.035). B cell phenotyping revealed higher frequencies of Env‐specific naïve B cells, harboring both IgD and IgM, starting at 14days post‐infection. Significantly higher frequencies of IgM heavy chains were isolated from broad neutralizers compared to non‐broad neutralizers at month 1, with both groups showing similar levels of somatic hypermutation, CDRH3 lengths, and heavy chain usage. Early engagement of naïve B cells resulted in significantly elevated IgM binding antibodies to respective autologous founder Envs starting at day 14 in broad neutralizers compared to non‐broad neutralizers. Higher levels of longitudinal autologous IgM responses in plasma significantly predicted the development of neutralization breadth (p=0.001), while autologous IgG and IgA responses were similar between the two groups.

Conclusions: These results demonstrate that early engagement of founder Env by naïve B cells and subsequent secretion of IgM to autologous founder Envs associate with the development of broad neutralizing antibodies. Factors that have the potential to enhance initial antigen engagement by naïve B cells should be considered when designing immunogens for HIV‐1 vaccines.

IgG3 HIV broadly neutralizing antibodies show improved neutralization potency and phagocytosis compared to IgG1 variants

S. Richardson, F. Ayres, N. Manamela, B. Oosthuysen, Z. Makhado, B. Lambson, L. Morris, P. Moore

National Institute for Communicable Diseases, Antibody Immunity Research Unit, Johannesburg, South Africa

Background: The ability of broadly neutralizing antibodies (bNAbs) to protect against HIV infection is enhanced through Fc receptor binding. The isotype of the antibody can modulate this effect, with IgG3 associated with improved HIV control and vaccine efficacy. HIV bNAbs are rarely assessed in the context of isotypes other than IgG1. We recently showed that an IgG3 version of CAP256‐VRC26.25 exhibited more potent neutralization and phagocytosis than its IgG1 counterpart. Here, we expanded this analysis to 14 additional bNAbs targeting all major bNAb epitopes.

Methods: The variable regions of antibodies to the V2‐apex, V3‐glycan supersite, CD4 binding site, gp41‐gp120 interface and MPER were cloned into vectors with CH1‐3 Fc regions from both IgG3 and IgG1. Pairs were assessed for neutralization potency against 11 viruses from the global panel. Antibody dependent cellular cytotoxicity (ADCC) and phagocytosis were measured using infectious viruses and global panel SOSIP Env trimers, respectively. Binding to the neonatal Fc receptor (FcRn) and Fcγ receptors were measured using ELISA.

Results: IgG3 bNAbs showed similar or higher (up to 60 fold) neutralization potency than IgG1 versions, though the effect was virus‐specific. This improvement was statistically significant for CAP256‐VRC26.25, 35,022, PGT135 and CAP255.G3. Comparison of antibody combinations as IgG1 or IgG3 indicated that the best triple combination consisted of IgG3 versions of CAP256‐VRC26.25, 35,022 and PGT121. IgG3 bNAbs also showed significantly improved binding to FcγRIIa which correlated with enhanced phagocytosis of all trimeric Env antigens. Isotype‐dependent differences in ADCC were antigen and bNAb‐specific, indicating a role for antibody specificity beyond FcγRIIIa binding alone. Lastly, we explored the pH dependence of variants for FcRn binding; a decrease in which typically impairs the half‐life of IgG3 antibodies. We showed that reduced pH dependence is specific to IgG3 bNAbs with κ‐light chains, whereas IgG3 bNAbs that use λ‐light chains showed similar pH dependence to their IgG1 counterparts. This implicates the κ‐light chain, specifically, in the reduced half‐life of IgG3 compared to IgG1.

Conclusions: This study supports the manipulation of the constant region to improve both the neutralizing and Fc effector activity of bNAbs, and indicates that IgG3 versions may be preferable for passive immunity given their polyfunctionality.

Defining the mechanism for varied neutralization potency of HIV neutralizing antibodies with identical variable regions but differing isotype

Z. Makhado, T. Moyo, C. Scheepers, P. Kgagudi, F. Aryes, N. Manamela, S. Richardson, P. Moore, L. Morris

National Institute for Communicable Diseases (NICD), Centre for HIV and STIs, Johannesburg, South Africa

Background: HIV‐1 neutralizing antibodies (nAbs) are an important focus of HIV‐1 vaccine development due to their ability to inhibit HIV‐1 entry into host cells. NAbs exist as various isotypes with each having unique effector functions, conferred by their respective Fc region. However, isotype can also impact neutralization potency, though the mechanism for this is poorly defined. We have previously isolated a monoclonal antibody (mAb), CAP88‐CH06, as an IgA1. We also showed that multiple co‐circulating IgG1, IgG3 and IgA1 isotypes of the CAP88‐CH06 lineage exist, some of which share identical variable regions, but exhibit markedly different neutralization capacity. Here we have explored the mechanism whereby isotype impacts neutralization potency.

Methods: The CAP88‐CH06 IgA1 mAb was engineered as IgG1 and IgG3 isotypes, with all three mAbs sharing identical variable regions. These were tested against a panel of autologous env‐pseudotyped viruses isolated from the first year of infection. Constant heavy chain (CH1) and hinge region swaps were constructed between the three mAbs, and these variants were assessed for neutralization potency and antibody‐dependent cellular phagocytosis (ADCP).

Results: The CAP88‐CH06 IgG3 (GMT 0.01µg/mL) and IgA1 (GMT 0.01µg/mL) isotypes showed enhanced neutralization of autologous viruses respectively, both in terms of potency and number of viruses neutralized, compared to the CH06 IgG1 (GMT 0.04µg/mL). Hinge swaps indicated that the increased neutralization potency of the IgG3 isotype could be attributed to its longer hinge region (62 amino acids), compared to that of IgG1 (15 amino acids) or IgA1 (22 amino acids). In contrast, in IgA1, we showed that the CH1 was responsible for the enhanced potency of this isotype compared to IgG1. ADCP assays revealed that the increased phagocytic activity of the IgG3 mAb, compared to its IgG1 and IgA1 counterparts, was also attributable to the hinge region, highlighting the contribution of the hinge to the polyfunctionality of IgG3.

Conclusions: Overall, these data show that enhanced neutralization potency of CAP88‐CH06 IgA1 and IgG3 compared to IgG1 is achieved by distinct mechanisms, which differentially impact mAb potency. This interplay between Fc and Fab regions may have broader applicability for the use of nAbs as therapeutics.

35O22 and VRC44 define a new highly glycan‐dependent multidonor class of HIV‐1 gp120:gp41 interface‐directed bnAbs

E. Cale¹, J. Gorman¹, A. Bennici¹, J. Driscoll¹, M. Messina¹, N. Radakovich¹, D. Xie¹, A. Olia¹, C.‐H. Shen¹, R. Rawi¹, R. Verardi¹, Y. Yang¹, B. Zhang¹, E. Crooks², M. Connors³

¹Vaccine Research Center, National Institute of Allergy and Infectious Diseases, NIH, Vaccine Research Center, Bethesda, United States, ²San Diego Biomedical Research Institute, United States, ³Laboratory of Immunoregulation, National Institute for Allergy and Infectious Diseases, NIH, Laboratory of Immunoregulation, Bethesda, United States

Background: At least seven sites on the HIV‐1 trimer are known epitopes for HIV‐1 broadly neutralizing antibodies (bnAbs), and two of these epitopes have been shown to be targeted by bnAbs from multiple individuals using the same germline V‐genes and structural mode of recognition. These are the VH1‐2 “VRC01 class” that targets the CD4 binding site and the VH1‐69 “4E10 class” that targets gp41. Here, we define a new reproducible, multidonor class of antibodies, the “35O22 class”, that targets a highly glycosylated region of the gp41:gp120 interface of the trimer.

Methods: B cells from an HIV‐1 chronic donor were stained with fluorescently labeled HIV‐1 SOSIP trimers and sorted individually into PCR plates by flow cytometry. Cells were subjected to RT‐PCR amplification of variable heavy and light chain genes to enable cloning and expression of antibodies. Antibodies were evaluated for neutralization of a global panel of HIV‐1 Env‐pseudotyped viruses by the TZM‐bl neutralization assay. Cryo‐EM of antibodies in complex with trimer was performed with strain C1080 SOSIP trimer stabilized by RnS mutations.

Results: Two clonally related antibodies, VRC44.01 and VRC44.02, demonstrated a combined neutralization breadth of 57% of 208 Env‐pseudotyped viruses that mapped to the same highly glycosylated gp41:gp120 interface site that is targeted by the previously characterized bnAb 35O22. Cryo‐EM indicated that VRC44.01 and 35O22 binding modes closely overlap in their angle of approach to the HIV‐1 Env trimer and contacts with glycans N88 and N625. Strikingly, both lineages are able to achieve 100% neutralization breadth of 31 Env‐pseudotyped viruses produced under conditions where glycan processing is arrested at the oligomannose state.

Conclusions: Our results define a new multidonor “class” of gp120‐gp41 interface bnAbs that use the same germline heavy chain variable genes and mode of trimer recognition. Moreover, this class exhibits extensive glycan dependence, with approximately 70% of the binding epitope composed of glycans. These findings highlight the potential interest for the 35O22 epitope as a template for vaccine design, as it is targeted reproducibly by multiple HIV‐1‐infected individuals, and that priming immunogens with oligomannose glycans may be a potential strategy for eliciting such lineages.

Precursor frequencies of naïve B cells targeting HIV candidate immunogens

M. Prabhakaran¹, A. Ruppel¹, D. Leggat¹, J. Plyler¹, J. Brand¹, X. Chen¹, H. Holdsworth¹, Y.‐T. Lai¹, R. Derking², R. Sanders², A. Stuart³, L. Stamatatos³, P. Kwong¹, A. McDermott¹

¹National Institutes of Health, Vaccine Research Center, Bethesda, United States, ²Amsterdam UMC, Amsterdam, Netherlands, ³Fred Hutchinson Cancer Research Center, Seattle, United States

Background: Lineage‐based vaccine design strategies engineer immunogens capable of engaging broadly neutralizing antibody (bNab)‐precursor B cells (BP) from the exceptionally diverse naïve B cell (NB) repertoire and subsequently both expanding and driving high somatic hypermutation levels to achieve neutralization breadth. Pre‐clinical assessment and validation of candidate immunigens (CI) is paramount to predicting their function in clinical trials. The selection of CI for vaccine development has been guided by their ability to bind predicted germline‐reverted or intermediate forms of bNabs. We set out to calculate the frequencies of BP in the NB compartment capable of binding to HIV CI; GT1.1.v4.1_BG505 SOSIP (CD4‐binding site – and apex), 426C DM RS CORE (CD4‐binding site), Mut16 (CD4‐binding site), Mut49 (CD4‐binding site), T117Fv3 (MPER), FP9‐PEG12 (Fusion peptide), designed to engage with lineages specific to supersites on gp140.

Methods: To assess the proportion of BP capable of binding to CI in the NB compartment, we enriched for B cells from 100 million PBMCs from multiple individuals and sorted naïve B cells using fluorescently labelled CI and any associated epitope‐specific knock‐out proteins. We amplified B cell receptor heavy and light chains from sorted cells, sequenced them and queried the sequences for the presence of known bNab signatures.

Results: The precursor frequencies of B cells capable of engaging CIs range from 16 – 65 NB per million B cells. We calculated the percentage of B cells containing signatures associated with CD4‐binding site targeting bNab classes. We observed Mut49 exhibited a high degree (14.5% of sorted cells) of selection for VRC01‐class precursor B cells, followed by Mut16 (4.3%), within the same individual. GT1.1 BG505 SOSIP was able to engage NIH45‐46 lineage precursors, while 426C DM RS CORE selected precursors of NIH45‐46, VRC16.01 and CH103 lineages. T117Fv3 did not identify signatures associated with an MPER bNAb. No NB were observed binding to FP9‐PEG12.

Conclusions: Studying the interaction of the NB repertoire with lineage‐based vaccine candidates is paramount to understanding if germline targeting potential of immunogens is physiologically relevant. This reveals how frequent precursors to bNAb lineages are in the NB repertoire and aids iterative immunogen design for better vaccine trial outcomes.

Low dose administration of unmodified 10E8v4, but not FcγR/complement dual enhanced or dual ablated 10E8v4 variants, decreases viremia in SHIV challenged macaques

D. Spencer¹, B. Goldberg², J. Dufloo³, T. Bruel³, O. Schwartz³, M. Ackerman², A. Hessell¹

¹Oregon Health and Science University, Pathobiology, Beaverton, United States, ²Dartmouth College, United States, ³Institut Pasteur, Paris, France

Background: Optimizing broadly neutralizing antibodies (bNAbs) for prevention of HIV infection is critical. Numerous studies affirm an Fc‐mediated contribution to protection, particularly as neutralization titers wane, but little is known about the specific contribution of complement activity. We hypothesized that bNAb dually enhanced for FcγR interaction and complement could mediate better therapeutic outcomes than the unmodified equivalent.

Methods: A panel of 10 bNAbs with Fc mutations altering complement and/or FcγR interaction was generated and screened in vitro. We selected 10E8v4 bNAb, and EFTAE (dual enhanced) and LALA (dual knockdown) mutations, because 10E8v4 has weak neutralization against the challenge virus (IC50 30mg/mL), no ADCC activity, and high complement mediated viral lysis. Rhesus macaques were pretreated with 5mg/kg unmodified, 10E8v4 EFTAE, 10E8v4 LALA, or an irrelevant Ab (N=6/group) prior to SHIVSF162P3 challenge. A separate cohort was treated with 10 or 20mg/kg unmodified 10E8v4 or 10E8v4 EFTAE (N=2/dose/group). Viremia and 10E8v4 dynamics, immunogenicity, and impact on subsequent SHIV‐specific humoral responses were monitored.

Results: 10E8v4 plasma t1/2 in unmodified, LALA, and EFTAE groups was 5.31, 5.27, and 2.21days, respectively. In high dose animals, a median of 0.010µg/mg unmodified 10E8v4 was measured in rectal tissue 7days after injection but not detected in EFTAE treated animals. Anti‐drug antibody (ADA) responses were weak or undetectable except in the EFTAE groups, where 33.3% (2/6 low dose) and 75% (3/4 high dose) displayed ADA. Compared to the control group, plasma viremia was reduced in 10E8v4 unmodified groups (p=0.022 low dose, p=0.0285 high dose) but not in LALA or EFTAE groups. Consistently, tissue viral DNA was lower in the unmodified group (p=0.0049) at necropsy. Flow analysis suggested distinct effector cell differentiation between groups, while in vitro studies showed splenocytes pretreated with sub‐neutralizing 10E8v4‐EFTAE increased infection over controls dependent on the presence of monocyte derived dendritic cells.

Conclusions: Non‐ADCC effector functions reduce viremia with 10E8v4 in the absence of sterilizing neutralization. The EFTAE mutation reduced efficacy by altering pharmacokinetics and/or through complement‐mediated infection enhancement. Thus, complement enhancement may reduce antibody therapeutic efficacy and warrants further study.

Functional convergence of clonally related but genetically distinct bNAbs that target the V3‐glycan epitope

D. Kitchin¹, J. Bhiman², T. Moyo¹, F. Ayres¹, Z. Molaudzi¹, B. Oosthuysen¹, B. Lambson¹, S.S. Abdool Karim³, N.J. Garrett³, N.A. Doria‐Rose⁴, J.R. Mascola⁴, L. Morris¹, P. Moore¹

¹National Institute for Communicable Diseases, Centre for HIV and STIs, Johannesburg, South Africa, ²National Institute for Communicable Diseases, Centre for Respiratory Diseases and Meningitis, Johannesburg, South Africa, ³Centre for the AIDS Programme of Research in South Africa, Durban, South Africa, ⁴Vaccine Research Center, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, United States

Background: Several broadly neutralizing antibody (bNAb) lineages directed at the HIV‐1 envelope (Env) V3/glycan epitope have genetically divergent branches that are all able to mature to breadth using different strategies. Delineating the ontogeny of these complex lineages may provide insights into how immunogens could be designed to elicit V3/glycan responses.

Methods: Sequencing and phylogenetic methods were used to establish the ontogeny of two clonally related V3/glycan bNAbs, CAP255.C5 (56% breadth) and CAP255.G3 (55% breadth), isolated at 149weeks post infection (wpi) from donor CAP255. Breadth and potency of longitudinal intermediates were assessed in neutralization assays against a panel of Env‐pseudotyped viruses. Through mutagenesis and by exchanging heavy and light chains, or complementary determining regions (CDRs), between broad and non‐broad lineage intermediates, residues essential for the maturation of neutralization breadth were identified.

Results: CAP255.C5 and CAP255.G3 were shown to belong to distinct branches of a single lineage that diverged shortly after the emergence of the precursor B cell. Although rapid CDRH3 maturation occurred in the CAP255.C5 branch, with the mature 149wk CDRH3 sequence observed as early as 47 wpi, this was not responsible for neutralization breadth. Rather, the light chain and mutations in heavy chain regions outside the CDRH3, especially a Y34F mutation in the CDRH1 and a 60VYG62 insertion in the CDRH2, were essential for breadth. The CAP255.G3 branch acquired breadth differently, with neutralization largely mediated by the heavy chain, which lacked indels but was more somatically mutated. Despite following divergent affinity maturation pathways, CAP255.C5 and CAP255.G3 shared an epitope footprint and showed remarkable functional convergence. Among a 208‐virus panel, 108 viruses were neutralized by both mAbs, with only 8 and 6 viruses exclusively neutralized by CAP255.C5 and CAP255.G3, respectively.

Conclusions: The functional convergence of genetically distinct CAP255 clonal relatives confirms the multiplicity of options available to V3/glycan bNAb lineages. This includes a largely CDRH3‐independent CAP255.C5 approach, different from most reported V3/glycan lineages. This has implications for HIV vaccine design as immunogens will ideally need to drive maturation of breadth in V3/glycan‐directed lineages with multiple, diverse branches.

Enhanced protection at the site of challenge of rhesus macaques that receive PGT121 one week prior to intravaginal challenge with SHIV‐SF162P3

J. Schneider¹, F. Engelmann², A. Carias², D. Barouch³, R. Veazey⁴

¹Rush University, Microbial Pathogens and Immunity, Chicago, United States, ²Northwestern University, Evanston, United States, ³Harvard University, Boston, United States, ⁴Tulane University, United States

Background: In a recent study, rhesus macaques (RM) that got an intravenous (IV) infusion of the broadly neutralizing antibody (bNAb) PGT121, 24hrs prior to intravaginal challenge with SHIV‐SF162P3, had distal site accumulation of virus one to three days after challenge. Using Cy5‐labeled VRC01 IV‐injected into RMs we found that it takes antibodies ~1week to achieve peak anatomical distribution in mucosal tissues. The aim of this study is to determine if giving antibodies more time to fully distribute can block distal site accumulation of virus following intravaginal challenge.

Methods: Utilizing Cy5‐labeled PGT121 and sham antibody DEN3, we compared −7days (n=5) and −1days (n=5) IV infusion prior to intravaginal challenge with SHIV‐SF162P3 in RM and measured virus 48hrs after challenge. Tissue and plasma levels of viral RNA and DNA were detected using gag qPCR and antibody levels were measured through Cy5 fluorescence using deconvolution microscopy and a fluorometer. Transcriptomics in these tissues was assessed through RNA‐seq.

Results: Whereas we detected viral RNA and DNA at the site of challenge in all the DEN3 and −1day PGT121 RMs, we only detected viral DNA in 1/5 RMs in the −7day PGT121 group. In a small subset of RMs in both the −7 and −1day PGT121 groups, we detected viral DNA in the lymph nodes (LN) and viral DNA and RNA in the brain. In these tissues that were qPCR positive, PGT121 was also present. RMs that received DEN3 had no distal site accumulation of viral RNA or DNA. In the −7day PGT121 group, there was an anti‐viral signature detected via RNA‐Seq in the vaginal epithelium at 48hrs post challenge.

Conclusions: We have found that giving antibodies more time to distribute enhances protection at the site of challenge as shown both through a decrease in viral RNA and DNA as well as changes in anti‐viral transcriptomics. However there was still distal site accumulation of viral DNA and RNA 48hours after challenge in a small subset of animals. Since this does not occur in DEN3‐injected RMs, the early distal site accumulation of viral RNA and DNA in the LN and brain appears to be PGT121 dependent.

Functional barriers in the elicitation of broadly neutralizing antibodies against the glycan‐V3 site of Env by vaccination

O. Swanson¹, B. Rhodes¹, A. Wang¹, A. Sanzone¹, M.K. Lauder², J. Mascola², K.O. Saunders¹, M. Bonsignori¹, K. Wiehe¹, B.F. Haynes¹, M.L. Azoitei¹

¹Duke University, Duke Human Vaccine Institute, Durham, United States, ²National Institute of Allergy and Infectious Diseases, Vaccine Research Center, Bethesda, United States

Background: Induction of broadly neutralizing antibodies (bnAbs) as part of an HIV vaccine will likely require “priming” the immune system to robustly activate antibody precursors, followed by a series of “boosts” to mature these initial responses to neutralization breadth. BnAbs that target the glycan‐V3 epitope on Env, such as DH270.6, are of great interest for vaccine development. However, these bnAbs contain rare sequence and structural features, like long CDR‐H3 loops and high levels of somatic mutations, that make their elicitation challenging. Here we determined key functional barriers for the induction of DH270.6‐like bnAbs by vaccination and identified immunogen properties required to overcome these hurdles.

Methods: High throughput library screening guided by computational and structural analysis, together with recombinant antibody production and neutralization assays were employed: 1) to identify the ability of existing germline targeting immunogens to engage potential DH270.6 antibody precursors with diverse CDR‐H3 loops and 2) to determine the key acquired mutations required for antibody maturation to DH270.6‐like neutralization breadth.

Results: Current DH270.6 germline‐targeting immunogens were shown to be highly sensitive to the CDR‐H3 sequence of potential precursors and to have limited recognition of such loops present in the natural human antibody repertoire. Beyond precursor engagement, we determined a subset of 12 out of the 42 mutations acquired by DH270.6 that is sufficient to provide approximately 90% of the bnAb breadth and potency. The majority of these mutations are predicted to occur with low probability in vivo. However, using high throughput screening, we identified alternative amino acids at these key sites that are more likely to emerge upon lineage activation. Based on these results, optimized “priming” and “boosting” immunogens to elicit DH270.6‐like antibodies were developed and will be discussed.

Conclusions: Our studies demonstrate that for robust “priming”, immunogens need to engage DH270.6 precursors with discrete CDR‐H3 loops through regions encoded by particular D genes. Once the lineage has been activated, only a few acquired mutations are sufficient to mature antibody responses to DH270.6‐like breadth and potency. Multiple amino acid variants are possible at the key sites that control DH270.6 function, revealing that diverse evolution pathways can be pursued to develop breadth by “boosting”.

The near‐pan‐neutralizing, plasma deconvoluted antibody N49P6 mimics CD4 in its quaternary interactions with the HIV‐1 envelope trimer

W. Tolbert¹, D. Nguyen¹, Z. Tehrani², M. Sajadi^2,3, M. Pazgier¹

¹Uniformed Services University of the Health Sciences, Infectious Disease Division, Department of Medicine, Bethesda, United States, ²Institute of Human Virology, University of Maryland School of Medicine, Divisions of Vaccine Research and Clinical Care and Research, Baltimore, United States, ³Baltimore VA Medical Center, Department of Medicine, Baltimore, United States

Background: The first step in the HIV‐1 entry process is the attachment of the Envelope (Env) trimer to target cell CD4. The CD4 binding site (CD4bs) therefore remains one of the only universally accessible sites on the Env trimer. Few antibodies (Abs) are able to capitalize on this however, due the steric constraints involved in accessing the CD4bs. N49P7 is a VRC01‐like CD4bs Ab isolated from the plasma of a HIV‐1 “elite neutralizer” that achieves near pan neutralizing breadth against HIV‐1. It does this in part through unique interactions to the highly conserved gp120 inner domain layer 3. We determined the structure of N49P6, another CD4bs Ab isolated from the same donor, in complex with gp120 and BG505 SOSIP to better understand the breadth and potency of this class of CD4bs Abs.

Methods: N49P6 IgG was expressed in HEK 293 cells and purified by protein A affinity. Clade A/E 93TH057gp120coree and clade A BG505 SOSIP.664 were expressed in GnT1‐ cells and purified with antibody affinity columns, 17b and 2G12 respectively. 93TH057gp120 was deglycosylated with EndoHf and N49P6 Fab was generated from IgG by papain digest. Complexes were made and used to grow protein crystals for x‐ray structure determination.

Results: We determined the structures of the N49P6 Fab‐93TH057gp120coree and N49P6 Fab‐BG505 SOSIP.664 complexes to 2.55 Å and 4.05 Å resolution respectively. N49P6 similar to N49P7 omits the Phe⁴³ cavity and relies instead on the gp120 inner domain layer 3 in binding to the primary gp120 protomer. N49P6 also contacts the inner domain layer 1 on the adjacent gp120 in the trimer mimicking the interprotomer contact of host receptor CD4. The high conservation of these contact residues contributes to N49P6 neutralization breadth and potency.

Conclusions: N49P6 utilizes many of the same unique characteristics used by N49P7 to achieve similar neutralization breadth. Furthermore, when binding the HIV trimer N49P6 mimics CD4 in its initial quaternary contacts with the neighboring gp120 in the trimer. The details of these interactions pave the way to the creation of the next generation of HIV‐1 neutralizing Abs for the use in preformed vaccines and HIV‐1 therapeutics.

Improved potency, breadth, and pharmacokinetics of VRC01‐class antibodies for HIV‐1 prevention and treatment

Y.D. Kwon¹, B. Zhang¹, J. Gorman¹, M. Louder¹, K. Mckee¹, A. Pegu¹, M. Asokan¹, E.S. Yang¹, R. Verardi¹, B. Lin¹, Q. Liu², P. Lusso², N. Doria‐Rose¹, J. Mascola¹, P. Kwong¹

¹Vaccine Research Center, National Institutes of Health, Bethesda, United States, ²Laboratory of Immunoregulation, National Institutes of Health, Bethesda, United States

Background: Passive transfer of broadly neutralizing HIV‐1 antibodies has shown promise in preventing infection from HIV‐1 exposure. The high dose required for this therapy in humans, however, limits the route of administration to infusion, which demands further improvement of the potency of antibodies to circumvent this disadvantage and to reduce the cost of treatment.

Methods: Here we sought to increase the potency and breadth of VRC01‐class antibodies by structure‐based rational design. We generated a matrix of antibody variants with mutations that fill interspatial cavity, extend their contacts to the neighboring Env protomer, or reduce potential steric clashes, and assessed their neutralization potency, autoreactivity, pharmacokinetics, and determined the cryo‐EM structure of BG505 DS‐SOSIP Env trimer in complex with a VRC01 variant.

Results: Against a 208‐virus panel, one variant, VRC01.23LS, showed increased breadth and ~10‐fold improvement in potency, with a geometric mean IC80 of 0.12mg/ml and a breadth of 96%. Moreover, its serum half‐life was maintained at a level comparable to that of VRC01LS. Another variant, VRC07‐523‐F54‐LS.v3, exhibited even greater potency and breadth, with a geometric mean IC80 of 0.072mg/ml and a breadth of 97%. Cryo‐EM structure of BG505 DS‐SOSIP.664 Env trimer in complex with VRC01.23LS Fab determined at 3.4‐Å resolution confirmed the structural basis for the improved potency, revealing that the Trp at position 54 in the heavy chain occupied the Phe43CD4 cavity on gp120, the extended heavy chain framework region 3 contacted the neighboring Env protomer, and the truncated N‐terminal light chain enabled the antibody to better accommodate diverse conformations of the V3 loop of gp120.

Conclusions: Thus, a matrix‐based approach combined with autoreactivity measurements and with serum half‐life assessment in human neonatal Fc receptor mice enabled the engineering of VRC01‐class variants with improved potency, breadth, and pharmacokinetics.

A lineage of exceptionally potent and broad BnAbs with ultralong CDRH3 from vaccinated cows focus low‐mutation rate from germline into a globular knob contacting the CD4bs

N. Salazar Quiroz¹, B. Heydarchi¹, S. Li¹, M. Corcoran², G. Karlsson Hedestam², D. Purcell¹

¹The University of Melbourne, Microbiology and Immunology, Melbourne, Australia, ²Karolinska Institutet, Microbiology, Tumor and Cell Biology, Stockholm, Sweden

Background: Despite the known challenges of producing broadly neutralising antibodies (BnAbs) against HIV‐1 envelope (Env), bovines elicit antibodies with neutralising features only weeks after vaccination with stabilised Env trimers. Previous results in our laboratory showed high neutralising titres in cows vaccinated with KNH1/BG505 SOSIP gp140 and mild neutralisation in cows vaccinated with AD8 uncleaved gp140. Moreover, potent BnAbs were isolated from one animal from KNH1/BG505 SOSIP group (monoclonal antibodies (mAbs) 1842, 1872 and 2129). In order to understand the development of neutralising antibodies in these animals, analysis of IgG sequences by next‐generation sequencing (NGS) was performed.

Methods: Four Holstein Friesian cows were vaccinated in two phases, receiving AD8 Unc gp140 or KNH1/BG505 SOSIP gp140 during phase 1, and AD8 SOSIP gp140 or BG505 SOSIP gp140 during phase 2. Total RNA from PBMCs isolated after phase 1 and phase 2 of vaccination was purified, and Multiplex PCR was used to construct IgM and IgG libraries from cDNA, using specific primers targeting VDJ region for further Illumina paired‐end sequencing. Sequence analysis was performed using programs IgDiscover and AbMining Toolbox. Additionally, sequences from mAbs isolated from same animals were analysed.

Results: IgM library sequences allowed the design of custom databases, using novel V gene alleles discovered in each animal. Using these animal‐specific databases, IgG sequences of KNH1/BG505 SOSIP 100 and AD8 Unc 500 groups showed that KNH1/BG505 SOSIP gp140 induced slightly higher rates of somatic hypermutation (SHM) in germline genes compared to AD8 Unc gp140. BnAbs 1842, 1872 and 2129 showed ultra‐long CDRH3s which concentrated most of the mutations found in the whole variable region, with rates below 7.4% in V segment, in comparison with non‐neutralising antibodies which presented higher SHM rates.

Conclusions: High levels of affinity maturation were not required to obtain BnAbs in bovines. Ultra‐long CDRH3 antibodies, which represented around 8% of total sequences found in libraries, needed less than 7.4% of SHM to achieve broad neutralisation, and in contrast presented a highly variable CDRH3. The study shows that bovines present genetic advantages to produce BnAbs against HIV‐1 rapidly compared to humans, and that KNH1/BG505 SOSIP gp140 induces higher rates of SHM than uncleaved trimers.

Breadth of CD8 T‐cell mediated inhibition of HIV‐1 replication correlates with breadth of epitope recognition mapped with a comprehensive Gag, Nef, Env and Pol potential T‐cell Epitope (PTE) peptide set

N. Fernandez, P. Hayes, J. Dalel, C. Streatfield, G. Macharia, J. Hare, L. Black, D. King, J. Gilmour

IAVI Human Immunology Laboratory, Imperial College London, London, United Kingdom

Background: Full characterisation of CD8 T‐cell functional HIV‐1 responses, including identification of HIV‐1 epitopes recognised and CD8 T‐cell ability to directly inhibit HIV‐1 replication, would assist HIV‐1 vaccine development. Identification of individual epitopes recognised is hampered by immense HIV‐1 diversity. A Potential T‐cell Epitope (PTE) peptide set (NIH AIDS Reagent Program) designed to address HIV‐1 sequence diversity consists of 1408 HIV‐1 Gag, Nef, Pol and Env peptides. Mapping of individual CD8 T‐cell epitopes within such large peptide sets is problematic due to limits in volunteer blood sampling. This was addressed through polyclonal expansion of CD8 T‐cells.

Methods: CD8 T‐cells were polyclonally expanded from PBMC from 13 HIV‐1 seropositive anti‐retroviral naïve subjects using CD3/CD4 bi‐specific antibody.

Identification of individual peptides recognised by CD8 T‐cells was achieved by 2‐step IFN‐γ ELISpot approach utilising peptide matrices following 7‐day expansion followed by single peptide confirmation at expansion day 10. Responses of CD8 T‐cells isolated directly from PBMC or following polyclonal expansion were compared. CD8 T‐cell‐mediated inhibition of HIV‐1 replication of a diverse cross‐clade panel of 10 HIV‐1 isolates in autologous CD4 T‐cells was assessed using a Renilla reniformis luciferase viral inhibition assay (VIA).

Results: Polyclonal expansion from 1 frozen PBMC vial provided sufficient CD8 T‐cells for 2‐step IFN‐γ ELISpot analysis of peptide recognition in 12 of 13 subjects. One subject required an additional vial for both ELISpot steps. Median 36 (5 to 65) individual PTE peptides and median 18 (3 to 41) epitope regions were recognised. The numbers of peptides recognised was positively correlated with HIV‐1 isolate inhibition breadth (r=0.667, p=0.021) as measured in the VIA. There was no significant difference in ELISpot magnitudes between expanded and directly PBMC‐isolated CD8 T‐cells (p=0.832).

Conclusions: A 2 step IFN‐γ ELISpot approach allowed identification of HIV‐1 PTE peptides recognised by CD8 T‐cells polyclonally expanded from 1 frozen PBMC vial. CD8 T‐cell‐mediated inhibition breadth was positively correlated with the number of HIV‐1 PTE peptides recognised. This approach overcomes the problem of fully characterising HIV‐1 specific CD8 T‐cell functional responses in the context of immense HIV‐1 diversity and can be applied to inform on HIV‐1 pathogenesis and vaccine design, with minimal sample requirements.

Incidence, clinical spectrum, risk factors and outcome of immune reconstitution inflammatory syndrome (IRIS) among HIV patients on highly active anti‐retroviral therapy (HAART) in the south west region of Cameroon

M. Julius¹, N.F. Peter², M.C. Kebeya³, N.A. Forbinake⁴, O.C. Stephane⁵

¹Cameroon Baptist Convention Health Services, HIV Free Project, Bamenda, Cameroon, ²Faculty of Health Sciences, University of Buea, Public Health, Cameroon, ³Cameroon Baptist Convention Health Services, Baptist Hospital Mutengene, Bamenda, Cameroon, ⁴Medecins Sans Frontiere, Cameroon, ⁵University of Yaounde 1, Yaounde, Cameroon

Background: Immune reconstitution inflammatory syndrome (IRIS) in HIV‐ infected patients initiating highly active antiretroviral therapy (HAART) results from an exhuberant restoration of pathogen‐specific immunity. The overall incidence of IRIS is unknown and there are still limited data from resource‐limited settings. Given the paucity of data in sub‐saharan Africa, our study characterises a typical Cameroonian cohort, useful to anticipate and prevent this syndrome in patients initiating ART. Our objective was determine the incidence, clinical spectrum, risk factors and outcome of IRIS in HIV‐infected patients initiating ART in South West Region, Cameroon.

Methods: 120 consecutive ART‐naive HIV‐infected adults aged 21years and over, initiating ART at two major treatment centres in the South West Region were prospectively followed for development of IRIS over 16weeks. Following their consent, patients were interviewed on a face to face basis and on their subsequent visits IRIS events were classified using diagnostic criteria by Haddow et al. Data were analyzed using SPSS version 21.0.

Results: During a 16week follow‐up period from ART initiation, the cumulative incidence of IRIS was 10.8% with incidence rate of 43.0/100 person‐years. 13 patients (M:F‐3:10) experienced 15 IRIS clinical events with 9/15 (60%) unmasking and the remaining paradoxical. Diagnoses included: miliary tuberculosis (1/15, 6.7%), psoas abscess (1/15, 6.7%) folliculitis (3/15, 20.0%), herpes zoster (1/15, 6.7%), herpes simplex (1/15, 6.7%), cryptococcal meningitis (2/15, 13.3%), progressive multifocal leucoencephalopathy (PML) (1/15, 6.7%), cellulitis (1/15, 6.7%), lymphadenitis (1/15, 6.7%), genital warts (1/15, 6.7%), and Kaposi's sarcoma (2/15, 13.3%). Median time to IRIS onset was 28days (interquartile range, 14 to 84) from ART initiation. Low CD4 count (<200cells/µL) was the only risk factor (p=0.036) on univariate analysis of the pre‐ART predictors. Most IRIS cases were mild. 11/13 patients recovered, 3/11 of whom required hospitalisation. Two deaths were attributable to IRIS (PML and psoas abscess).

Conclusions: For every 100 HIV patients initiated into HAART in South West Region, Cameroon, 10 to 11 patients may develop IRIS of varied patterns primarily around 28weeks from HAART initiation, particularly those with advanced immunosuppression. However, severe life‐threatening IRIS is uncommon. Close follow‐up of ART naive clients especially those with low initiating CD4 counts is therefore pertinent.

Infection with a highly Gag pre‐adapted virus broadens the proteins targeted by CTLs during acute HIV infection

D. Monaco¹, P. Hayes², D. Dilernia¹, D. Wooding², J. Gillmour², E. Hunter¹

¹Emory Vaccine Center, Emory University, Atlanta, United States, ²Human Immunology Lab, Imperial College London, London, United Kingdom

Background: Pre‐adaptation, the degree to which the polymorphisms in the T/F virus are linked to the HLA allele repertoire of the newly‐infected individual, has been shown to impair virus control and accelerate disease progression. This impact was associated with an reduced ability to target epitopes in the highly pre‐adapted protein and weaker responses towards the adapted epitopes. In this study, we looked to identify the targets of the cellular immune response at the whole‐proteome level during acute infection, comparing individuals infected with a high or low pre‐adapted virus in Gag.

Methods: Gag was sequenced in plasma samples (<45days post‐EDI) from recipients in epidemiologically‐linked heterosexual transmission pairs from Zambia (Zambia Emory Research Project) to determine their level of pre‐adaptation (N=27). PBMCs samples from early in infection (<6months post‐EDI) were polyclonally expanded for 9days using a CD3/CD4 bispecific antibody and IL‐2. IFN‐g responses against pools spanning each HIV protein were tested via ELISpot. Peptides used to construct the pools were 15‐mers, overlapped by 11, representing the consensus of chronic sequences obtained from the same Zambian cohort. Responses higher than 50 SFU/10⁶ cells were considered positive.

Results: Individuals infected with highly pre‐adapted viruses showed a significantly greater number of proteins being targeted outside of Gag (p =0.049) than individuals infected with low pre‐adapted viruses. Pol was significantly more targeted in the high pre‐adaptation group, in which all the individuals showed a positive response against the Pol pool (p =0.03; 10/10 vs. 4/9). Among accessory proteins, in the high pre‐adaptation group, Nef and Vif were the most targeted while Tat was the least. Interestingly, Tat along with Nef were the most targeted accessory proteins in the low pre‐adaptation group.

Conclusions: These results support the hypothesis that a high degree of preadaptation in Gag forces the immune response to redirect against proteins that seem to be less efficient in controlling HIV replication, as evidenced by the faster disease progression observed in this group. Identifying subdominant Gag epitopes, which do not select or transmit escape, in these high Gag pre‐adaptation individuals may be an important strategy to find novel targets for future vaccines.

Correlates of dapivirine vaginal ring uptake among women participating in an open label extension trial‐MTN‐025/HOPE

B. Gati Mirembe¹, M.V. Cabrera², M. Cobbing³, N. Mgodi⁴, T.P. Palanee⁵, A. Mayo⁶, S. Dadabhai⁷, L.E. Mansoor⁸, S. Siva⁹, G. Nair¹⁰, C.A. Akello¹, C. Nakabiito¹, L.E. Soto-Torres¹¹, A. Van der Straten¹², J. Baeten¹³

¹Mujhu Research Collaboration, Kampala, Uganda, ²University of Washington, Biostatistics, Seattle, United States, ³HIV Prevention Research Unit, South Africa Medical Research Council, Pharmacy, Durban, South Africa, ⁴University of Zimbabwe College of Health Sciences Clinical Trials Research Centre, Harare, Zimbabwe, ⁵Wits Reproductive health and HIV Institute (Wits RHI) University of the Witwatersrand, Johannesburg, South Africa, ⁶FHI 360, Durham, United States, ⁷Malawi College of Medicine‐Johns Hopkins University Resreach Project, Queen Elizabeth Central Hospital Blantyre Malawi, Malawi, ⁸Centre for AIDS Programme of Research in South Africa, University of Kwazulu Natal, Durban, South Africa, ⁹HIV Prevention Research Unit, South Africa Medical Research Council, Durban, South Africa, ¹⁰The Desmond Tutu HIV Centre, University of Cape Town, Cape Town, South Africa, ¹¹Division of AIDS, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, United States, ¹²W omens Global Health Imperative,RTI International, Berkerly, Research Triangle Park, United States, ¹³University of Washington, Department of Global Health Medicine, Epidemiology, Seattle, United States

Background: MTN‐025/HOPE was a Phase IIIb open‐label extension trial of the dapivirine vaginal ring (DVR) for former participants of MTN‐020/ASPIRE, one of two phase 3 studies that showed the DVR was well‐tolerated and reduced the risk of HIV infection by approximately 30%. Use of the DVR was not required for participation in HOPE and women could change their mind about ring acceptance during study follow‐up. We explored factors associated with DVR uptake in MTN‐025/HOPE.

Methods: MTN‐025/HOPE was conducted at 14 sites in Malawi, South Africa, Uganda, and Zimbabwe between July 2016 and August 2018. To be eligible for the study, women had to be HIV uninfected, not pregnant, not breastfeeding and willing to provide informed consent. The DVR was offered monthly for the first 3months, then quarterly (3 rings dispensed) thereafter at months 3, 6 and 9. Logistic regression analysis was used to assess correlates of DVR uptake from entry to month 9 adjusting for region, including demographics, participant/partner characteristics, sexual behavior, HIV risk perception, family planning methods, and intention to use the ring in future. Covariates from univariate analyses that were significant at p<0.10 were included in a multivariate analysis.

Results: A total of 1456 women (median age 31years) enrolled in HOPE and 8390 visits were included in the analysis. At baseline, 1342 (92.0%) chose to accept the DVR and 1153 (79.2%) were consistent acceptors (accepted the ring at all visits). Significant correlates of ring uptake (at p<0.05) are included in Table 1 below: