CROI 2019 Abstract eBook

Abstract eBook

Oral Abstracts

in MSM aged ≥16 years attending SHCs. Period prevalence and 95% confidence intervals (CIs) for HIV among all attendees and bacterial STIs (at least one diagnosis in calendar period) among HIV negative attendees in 2017 were calculated. Annual HIV incidence per 100 person-years (PY) and 95% CIs in MSM who tested for HIV at least twice in the same SHC from 2012 to 2016 were determined. As a proxy measure of high risk, HIV incidence in a subset of MSM with a history of a negative HIV test and an ano-genital bacterial STI in the preceding year was also examined. Results: In the 159,368 MSM attending SHCs in 2017, HIV period prevalence was 20.0% (95% CI 19.8-20.2%). In MSM not known to be HIV positive (n=128,772), gonorrhoea, chlamydia, and syphilis period prevalence in 2017 was 12.1% (11.9-12.2%), 9.0% (8.9-9.2%), and 2.7% (2.6-2.8%), respectively. The number of MSM not known to be HIV positive (% tested for HIV at least twice) increased from 85,500 (31.0%) in 2012 to 120,606 (36.2%) in 2016. The annual incidence of HIV in MSM decreased 60.5% from 2.0 per 100 PY (95% CI: 1.8-2.2) in 2012 to 0.79 per 100 PY (0.69-0.89) in 2016; compared to the latter, MSMmeeting proxy high risk criteria in 2016 had a two-fold higher HIV incidence [1.58 (1.25-1.99) per 100 PY]. Conclusion: While there is a high prevalence of bacterial STIs, there has been a sharp decrease in the incidence of HIV in MSM regularly attending SHCs. The fall in HIV incidence coincides with further intensification of HIV testing, especially repeat testing, and earlier initiation of HIV treatment and, more recently, the scale up of privately purchased generic PrEP in England from late 2017. The PrEP Impact trial, which aims to enrol 13,000 participants from communities most affected by HIV, is likely to have an additional effect on the incidence of HIV. Delphine Rahib 1 , Heloise M. Delagreverie 2 , Iris Bichard 3 , Audrey Gabassi 2 , Nicolas Guigue 2 , Marie-Laure Chaix Baudier 2 , Beatrice Bercot 2 , Constance Delaugerre 2 , Nathalie Lydié 1 1 Santé Publique France, Saint-Maurice, France, 2 Hôpital Saint-Louis, Paris, France, 3 Le Kiosque Information Sida Toxicomanie, Paris, France Background: Testing is critical to HIV and other sexually transmitted infections (STI) treatment and prevention. The MemoDepistages multicenter study is a novel French government-based comprehensive STI testing program, providing free postal self-sampling kits to high-risk MSM. We present here the baseline results from the Paris area, accounting for 60% of all participants. Methods: The programwas advertised by dating apps and social media. Inclusion criteria were: MSM aged over 18, ≥2 male sex partners in the last year, HIV-seronegative without PrEP. Test kits included 1 Microtainer Serum Separator Tube, lancets for the collection of 600µl capillary blood, 1 UriSwab collection tube, 2 PCR Dual Swab kits and pre-paid packaging to return samples. Serumwas tested for HIV-1/2 EIA 4G serology, anti-HCV antibodies, HBs antigen (Architect, Abbott) and syphilis Tp and RPR (Bioplex, Biorad). Urine, throat and anal swabs were tested for Chlamydiae trachomatis (CT) and Neisseria gonorrhoeae (NG) DNA (Cobas 6800, Roche). Results were provided to the participants, as they chose, by community-based workers following text, phone or e-mail contact or by their family physician. Results: From April, 10th to June, 11th 2018, 4419 applicant men from the Paris area were eligible. Median age was 30 years, 13.0% had never been tested for HIV, they reported a median of 10 partners/year. 48.5% confirmed their inclusion and ordered the kit. As of August, 31st 2018, 1238 kits were returned to the core lab. 1.3% (13) of tested samples were positive for HIV, 0.5% (6) for anti-HCV antibodies, 0.3% (3) for HBs antigen and 1.8% (11) for active syphilis. Full serology testing was hindered by hemolysis and low blood volume (Table 1). Six HIV-positive participants were unaware of their infection (0.6%) and one sample revealed an acute primary infection, as confirmed by further sampling. NG was found in 11.2% of participants and CT in 9.4%. The most common bacterial STI was throat NG (8.2%). Community-based linkage to care was excellent and newly diagnosed participants sought counseling and treatment. Conclusion: This is the first study of a postal viral and bacterial STI test in France. Enrollment was fast, highlighting the need for access to testing in the target population. Blood collection was feasible in most participants and laboratory-based serology allowed for the detection of acute and chronic HIV infections, HCV, HBV and syphilis. CT and NG were frequent and would not have been detected by a typical urine-only strategy.

Oral Abstracts

50 RANDOMIZED CONTROLLED TRIAL OF INTRAUTERINE DEVICE SAFETY IN WOMEN LIVING WITH HIV Catherine S. Todd 1 , Heidi Jones 2 , Nontokozo Langwenya 3 , Donald R. Hoover 4 , Greg Petro 3 , Landon Myer 3 1 FHI 360, Durham, NC, USA, 2 City University of New York, New York, NY, USA, 3 University of Cape Town, Cape Town, South Africa, 4 Rutgers University, Piscataway, NJ, USA Background: Few data exist regarding impact of exogenous progestins on genital tract HIV viral load (gVL) in women living with HIV (WLHIV). We compared safety via gVL and plasma HIV VL (pVL) and acceptability via continuation of copper (C-IUD) compared to levonorgestrel (LNG IUD) intrauterine devices among WLHIV stratified by ART use in Cape Town, South Africa. Methods: This double-blind trial allocated consenting WLHIV to C-IUD or LNG IUD 1:1 between October 2015 and December 2016. Eligibility included screening and treatment for reproductive tract infections (RTIs) within the past 1m, not desiring pregnancy within 2y, and either viral suppression (pVL<1000 c/mL) in the last 6m (on ART) or CD4 count above ART initiation threshold (non- ART). We tested genital tract menstrual cup samples for gVL and swabs for RTIs, and pVL at enrollment and 3, 6, 12, 18, and 24m follow-up visits. We compared detectable gVL at 6m (primary outcome) and 24m by armwith intent-to-treat (IT) and as-treated (AT) Mantel-Haenszel Odds Ratios (OR), stratified by baseline gVL and ART use. We reported serious adverse events (SAEs) related to IUD use and compared acceptability via IUD removal rates over 24 months by arm using proportional hazards models. Results: We enrolled 199 WLHIV (134 on ART/65 non-ART users; median age of 31y and 95% had >1 prior pregnancy). 62% of non-ART users and 15% of ART users had detectable gVL at enrollment with no differences by IUD arm. There were no significant differences in detectable gVL between arms adjusting for baseline gVL and ART group at 6m (IT OR=1.03, 95%CI 0.53-2.02, p=0.92; AT OR=1.01, 95%CI 0.51-2.01, p=0.98) and 24m (Table). Over 24m, there were 39 SAEs (18%, n=7 related to IUD). IUD continuation was 75% overall, with 3 partial and 7 complete expulsions and 34 elective and 5 non-elective (for PID, colposcopy and pregnancy) removals by 24m. Expulsion (8% vs. 2%, p<0.001) and elective discontinuation (7.1m vs. 10.9mmedian time to removal, Hazard Ratio=9.00, 95% CI 3.17-25.5) were higher for C-IUD users. Common elective discontinuation reasons were dysmenorrhea/pain (40%, C-IUD and 75%, LNG IUD) and heavy bleeding (33%, C-IUD). Conclusion: This first randomized study comparing local progestin effect (LNG IUD) to a non-hormonal method (C-IUD) indicates no increase in gVL shedding, a proxy for sexual transmission risk, or pVL between IUDs, with or without ART use. The LNG IUD had low discontinuation rates, reflecting its value in broadening the contraceptive method mix for WLHIV.

49 EXPANDING TESTING STRATEGIES IN PARIS: A FREE POSTAL COMPREHENSIVE STI TEST KIT

21

CROI 2019