CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

Northwestern Univ, Chicago, IL, USA Background: Black young men who have sex with men (YMSM) experience substantially higher incidence rates of HIV despite reporting fewer risk behaviors. Concurrency (i.e., overlapping sexual partnerships) is a feature of sexual networks that could explain this contradiction, given that reporting more concurrent partners has been shown to increase the spread of disease through a sexual network, even with the same levels of contact. Accordingly, this study leverages recent advances in modeling sexual networks to a) model variation in concurrency across race among YMSM and b) examine unobserved ‘whole network’ structural features based on these models. Methods: Data for this analysis come from an ongoing cohort study of YMSM in Chicago (RADAR). Participants (n = 574) completed a comprehensive inventory of their sexual partners. Using a series of egocentric exponential random graph models, we examined if the tendency to have fewer than 2 current partners varies across race, while controlling for other features of sexual networks such as racial homophily. Finally, we simulated sexual networks using these models in an effort to estimate risk metrics that cannot be observed in egocentric network data. Results: Results suggest that YMSM have a disinclination for concurrency. However, providing separate estimates for concurrency for each racial group did not improve the fit of the model, and this tendency did not significantly vary between Black and White YMSM (z = 1.9, p = 0.051). Contrastingly, the tendency for same-race sexual partners was significantly greater among Black YMSM (z = 3.2, p = 0.001). Sexual networks simulated from these models suggest that Black YMSM are more closely connected to individuals perceived to be HIV-positive but are less likely to be in a large connected component, relative to their White peers. Conclusion: These results suggest racial differences in concurrency are not observed among YMSM in Chicago, and therefore differences in concurrency are unlikely to explain racial disparities in HIV. Yet, existing racial disparities may be partially maintained through higher levels of racial homophily among Black YMSM. Furthermore, HIV prevalence in the sexual networks of Black YMSM place them at risk for HIV acquisition. Yet, White YMSM are more likely to be in connected components, suggesting a similar paradox to individual level risk behavior. 844 HIV TRANSMISSION IN TIJUANA, MEXICO: REAL-TIME IDENTIFICATION OF A NEW CLUSTER Antoine Chaillon 1 , Patricia Gonzalez-Zuniga 1 , Tommi O. Gaines 1 , Horacio Almanza 2 , Jose R. Chavez 2 , Jesus Vera 1 , Thomas L. Patterson 1 , Davey M. Smith 3 , Steffanie A. Strathdee 1 , Sanjay R. Mehta 1 1 Univ of California San Diego, San Diego, CA, USA, 2 Autonomous Univ of Baja California, Tijuana, Mexico, 3 Univ of California San Diego, La Jolla, CA, USA Background: Tijuana’s proximity to the border and a large transient population has led to a thriving red light district and illicit drug market, creating populations at high risk for acquiring and transmitting HIV. Policy decisions designed to clear homeless individuals from the canal region along the US border have driven some of these behaviors “underground”. Benefiting from continued surveillance in Tijuana between 2004 and 2016, we applied molecular epidemiologic techniques to infer the presence and monitor the development of putative transmission clusters to help target HIV prevention efforts. Methods: Participants from eight research studies in Tijuana were screened for HIV infection at baseline, and in two of these studies participants were screened longitudinally on a six-month basis in follow-up. After quality filtering, a total of 288 sequences were obtained from unique individuals sampled between 2004-2016. Phylogenetic and genetic network analyses were performed to infer putative relationships between these HIV sequences. Correlates of identified clusters were evaluated. Results: Among individuals with reported demographic data, 46.4% (116/250) were MSM, 42.1% (99/235) reported transactional sex, and 27.8% (65/234) reported injection drug use (PWID). 32 individuals seroconverted during the observation period from 2011-2016, of which 21 had available sequence data and were included in the analysis. 123/288 (42.7%) of sequences were linked with at least one other sequence, forming 37 transmission clusters (ranging size: 2-14) (Fig.1). 27/123 (22.8%) of the clustering subjects were FSWs, and these individuals were found in four out of the five clusters linking to individuals residing in San Diego. We identified one cluster of five seroconverters who were all PWID. Two were diagnosed in 2012, and three were diagnosed April-May, 2016. Another three individuals with no sequence data seroconverted during this period. Notably, nearly all individuals diagnosed in April-May 2016 reported moving to a new “underground” shooting gallery away from the canal region, and 3 reported an interaction with police one month prior to their HIV diagnosis. These results were shared with the Mexican authorities. Conclusion: These data suggest a dynamic HIV transmission pattern reflecting transmissions across risk groups. However, our surveillance also allowed us to identify a rapidly growing transmission cluster in a demographically and geographically focused population impacted by policy change. 845LB IDENTIFYING AND INVESTIGATING A RAPIDLY GROWING HIV TRANSMISSION CLUSTER IN TEXAS Analise C. Monterosso 1 , Sydney Minnerly 1 , Samuel Goings 1 , Allyson Morris 1 , Anne Marie France 2 , Sharoda Dasgupta 2 , Alexandra Oster 2 , Miranda Fanning 1 1 Texas Dept of State Hlth Services, Austin, TX, USA, 2 CDC, Atlanta, GA, USA Background: Analysis of HIV nucleotide sequence data collected through the National HIV Surveillance System can identify rapidly growing transmission clusters. The Centers for Disease Control and Prevention (CDC) identified a molecular cluster in Texas that grew substantially during July 2015–June 2016. CDC and the Texas Department of State Health Services investigated to define the extent of the cluster, identify relationships between cases, characterize the epidemiology and factors facilitating transmission, characterize timing of viral suppression, and prioritize intervention. Methods: Persons in the molecular cluster were considered confirmed cluster cases. Based on partner services interview records, we identified HIV-infected persons without nucleotide sequences available for analysis who were named sex or needle sharing partners of confirmed cases, partners of sex or needle sharing partners, or social network contacts of confirmed cases. During August–October 2016, we reviewed medical records and partner services interview records to collect data on demographics, risk behaviors, partner meeting sites, and time to achieve viral suppression. Results: From 27 confirmed cluster cases, we identified 112 additional cluster cases. Of 27 confirmed cases, 12 (44%) were connected through named partners or social contacts into one large cluster; no links were identified for 15. Of 76 confirmed and other cluster cases with records available, 76 (100%) were male at birth, 59 (78%) were aged 13–29 y, 66 (87%) were Hispanic, and 68 (89%) reported sex with men. Reported lifetime sex partners ranged from 2–300; 18 (24%) reported anonymous partners and 18 (13%) had an STD diagnosis within 12 months before HIV; none were on PrEP. In all, 31 (41%) had evidence of viral suppression within 6 months of diagnosis (figure); 10 (13%) have never had a viral load test. After this investigation, 25 people were initiated for reengagement in care. Conclusion: Our investigation identified an actively growing transmission cluster of primarily young Hispanic MSM that was substantially larger than the molecular cluster; this network is likely even larger, given the large number with anonymous partners or without identified links to the cluster. High risk sexual behavior coupled with delays in achieving viral suppression among some cluster cases likely contributed to rapid growth. These findings reveal opportunities for prioritization of persons associated with this cluster for linkage to care and PrEP referral. 846 SEXUAL NETWORKS OF CIS AND TRANSGENDER PEOPLE WITH & AT-RISK FOR HIV, KING COUNTY, WA Background: In King County (KC), Washington, HIV is concentrated among cisgender (“cis”) men who have sex with men (MSM). The number of transgender (“trans”) people in KC is unknown, complicating estimates of disease burden. We triangulated several health department data sources to describe characteristics of trans people known to be HIV- positive and to assess the extent to which trans sexual networks overlap with HIV-negative cis MSM and people living with diagnosed HIV (PLWDH). Methods: Data sources are outlined in the Table. We describe the number and characteristics of PLWDH known to be trans; the percent of HIV-negative trans men and trans women who reported cis male sex partners; and the percent of PLWDH and HIV-negative cis MSM who reported trans sex partners. Unless otherwise indicated, questions about sex partners referred to the prior 12 months. Julia Hood 1 , David A. Katz 2 , Lindley A. Barbee 2 , Sara N. Glick 2 , Amy Bennett 1 , Susan Buskin 1 1 Pub Hlth–Seattle & King County, Seattle, WA, USA, 2 Univ of Washington, Seattle, WA, USA

Poster and Themed Discussion Abstracts

CROI 2017 365