CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

performance of cross-sectional incidence assays on samples fromWest and Central Africa, though population-based HIV impact assessments, which will use this methodology, are planned in Cameroon and Cote D’Ivoire in 2017. Methods: This study examined plasma samples of participants from the MDC cohort in Cameroon. Samples were collected from 2011-2015 from individuals known to be HIV- infected more than one year (n=129). CD4 counts and HIV-subtype were previously determined for this cohort. All the samples were tested by the Limited Antigen (LAg) Avidity assay and the Johns Hopkins modified (JHU) BioRad-Avidity assay. The proportion of samples infected greater than one year being misclassified as recent were examined using previously determined cut off values, specifically: (1) LAg Avidity at <1.5 normalized optical density (OD-n), (2) LAg Avidity at <1.5 OD-n and >1000 HIV RNA copies/mL, and (3) LAg Avidity at <1.5 OD-n, >1000 HIV RNA copies/mL, and JHU BioRad-Avidity Assay at an avidity index (AI) < 40%. Results: The median CD4 count for this population was 364 (interquartile range, 249-521), and the majority of individuals with known HIV subtype data were infected with an AG recombinant virus (59% [23/39]). Overall, 16.3% (21/129, 95% CI 1.4, 23.8) of samples were misclassified using the LAg-Avidity assay with a cutoff ≤ 1.5 OD-n. This was reduced to 1.71% (2/117, 95% CI 0.19, 5.49) for LAg-Avidity assay <1.5 OD-n and >1000 HIV RNA copies/mL. The two individuals that misclassified had a JHU BioRad HIV Avidity index of 53% and 55%, therefore the false recent rate was reduced to zero for the third algorithm Conclusion: Individuals in the Cameroon cohort were shown to be misclassified when assessing the LAg-Avidity alone. However, the LAg-Avidity assay (<1.5 ODn) with viral load testing algorithm yielded a false recent rate less than 2%, which could be improved by adding the JHU BioRad assay at a cut-off of 40%. Further studies are needed to determine if there is a subtype specific misclassification in West-Central Africa. 882 WITHDRAWN 883 HIGH HIV INCIDENCE AMONG PWID AND MSM ATTENDING INTEGRATED CARE CENTERS IN INDIA Allison M. McFall 1 , Sunil S. Solomon 1 , Gregory M. Lucas 1 , Aylur K. Srikrishnan 2 , Muniratnam S. Kumar 2 , Santhanam Anand 2 , Canjeevaram K. Vasudevan 2 , David D. Celentano 1 , Shruti H. Mehta 1 1 The Johns Hopkins Univ, Baltimore, MD, USA, 2 YR Gaitonde Cntr for AIDS Rsr and Educ, Chennai, India Background: In India, similar to other lower- and middle-income countries, HIV incidence has declined over the past decade following scale-up of HIV prevention and treatment services for heterosexual populations. While prevalence data among people who inject drugs (PWID) and men who have sex with men (MSM) suggest increasing burden, HIV incidence data among these groups are sparse. Methods: As part of a cluster-randomized trial among PWID and MSM in India, integrated care centers (ICCs) were established in 11 cities (6 PWID and 5 MSM) and have been running for nearly two years. ICCs provide core and PWID- or MSM-focused HIV prevention and treatment services, including HIV counseling and testing, in a single venue. HIV negative clients are actively tracked to promote annual HIV testing. HIV incidence rates were calculated for clients with ≥2 HIV tests and negative on the first test. Multi-level Poisson regression models were used to explore correlates of HIV incidence. Results: 5,012 ICC clients (3,430 PWID and 1,582 MSM) who were initially HIV negative were included. Median age was 28 years and 8.9% of PWID were women. There were 48 PWID and 13 MSM seroconverters resulting in HIV incidence rates of 1.30 per 100 person-years (PY) (95% confidence interval [CI]: 0.98 - 1.73) and 0.99 per 100 PY (95% CI: 0.58 - 1.71), respectively. There was considerable variability across cities with a range of 0 - 6.71 for PWID and 0 - 1.99 for MSM (Figure). Among PWID, HIV incidence was higher among women (adjusted incidence rate ratio [aIRR]: 2.54) and those with traditional risk factors - recent injection drug use (aIRR: 2.77), sharing needles/syringes (aIRR: 17.7), and a higher number of sexual partners (aIRR for ≥3 partners vs. none: 3.05). Lower incidence was observed among those receiving opioid substitution therapy >2 times/week (aIRR: 0.21) and receiving at least one session of safe sex counseling (aIRR: 0.23); however, those using needle/syringe exchange were at higher risk (aIRR: 2.41). Among MSM, the only factor significantly associated with HIV incidence was a recent sexually transmitted infection diagnosis (aIRR: 9.51). Conclusion: PWID and MSM attending HIV-focused care centers in India experience high HIV incidence. Specific sub-groups of clients continue to engage in high-risk behaviors and should be targeted for additional harm reduction services and biomedical prevention approaches such as pre-exposure prophylaxis (PrEP), particularly PWID who have not been the focus of PrEP programs thus far.

Poster and Themed Discussion Abstracts

884 TRENDS IN HIV INCIDENCE AMONG PWID, MSM AND CSW USING PREVENTION SERVICES IN UKRAINE

Kostyantyn Dumchev 1 , Olga Varetska 2 , Yana Sazonova 2 , Pavlo Smyrnov 2 1 Ukrainian Inst on Pub Hlth Policy, Kyiv, Ukraine, 2 Alliance for Pub Hlth, Kyiv, Ukraine

Background: HIV epidemic in Ukraine is one of the largest in Europe, with estimated 223,000 people living with HIV and 10,500 new HIV cases annually. The epidemic was mainly driven by people who inject drugs (PWID), 21.9% of whom are HIV infected, and to lesser extent by other key groups such as men who have sex with men (MSM, 8.5% prevalence) and commercial sex workers (CSW, 7% prevalence). Prevention programs supported by the Global Fund through Alliance for Public Health in Ukraine rapidly grew since 2004 to

CROI 2017 382