CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

serological markers. Likely transmitters to RHI were identified according to an algorithm considering phylogenetic and clinical data at transmission. Chronic infections linked to, but diagnosed after a RHI were considered potential transmitters. Results: 389 RHI were identified, for which a likely transmitter was identified for 186 (48%). 176 (95%) transmitters were male, 168 (90%) were white. 173 (93%) transmissions were between MSM. 110 (59%) infections were acquired from the local population, 75 (40%) from elsewhere in the UK; 30 from the nearest major city (London), and one from the USA. 22 (20%) transmitters were diagnosed shortly after the estimated transmission date. A further 142 RHI (37%) were linked to a potential transmitter, 108 being undiagnosed at the time of transmission. 61 RHI (16%) had no potential source. Conclusion: We identified a transmission source for 48% of RHIs, suggesting phylogenetically guided interventions may be feasible within this population. We are exploring the application of a structured coalescent model to explore the potential of transmission sources as a target for intervention. In parallel we are developing an ethical framework to ensure patient acceptability of phylogeny- based interventions. 945 DYNAMICS OF THE CRF01AE EPIDEMICS IN THE CITIES OF SHANGHAI, SHENZHEN AND SHENYANG Xiaoxu Han 1 , Antoine Chaillon 2 , Ping Zhong 3 , Jin Zhao 4 , Bin Zhao 1 , Haibo Ding 1 , Junjie Xu 1 , Yi Lin 3 , Yile Xue 3 , Xuqin Wang 3 , Min Zhang 1 , Lin Wang 1 , Davey M. Smith 2 , Hong Shang 1 1 China Medical University, Shenyang, China, 2 University of California San Diego, La Jolla, CA, USA, 3 Shanghai Municipal Center for Disease Control and Prevention, Shenzhen, China, 4 Shenzhen Center for Disease Control and Prevention, Shenzhen, China Background: China is experiencing a large increase of HIV infections among men who have sex with men (MSM) and migrants who flow between cities to seek better living conditions. This increase may fuel and reshape local epidemics in large industrialized cities. Here, we evaluated the dynamics and geospatial mixing of rapidly evolving HIV CRF01AE epidemics in three of the largest cities in China. Methods: HIV-1 CRF01AE pol sequences generated from 3,071 individuals diagnosed in Shanghai (n=884), Shenyang (n=1,719) and Shenzhen (n=568) between 2002-2016 were analyzed for clustering. Optimal genetic distance threshold (GDT) were determined by sensitivity analyses designed to provide the highest resolution. Bayesian analyses were performed to assess the dynamics and transmission rates (TR) among clusters. Viral gene flow between the 3 cities was estimated using a Bayesian phylogeographic diffusion model and a Slatkin-Maddison (SM) approach after adjusting for sampling heterogeneity between sites. Results: A total of 3,071 individuals predominantly MSM (85.2%) with a median age of 31 years (IQR:26-40) were included. Individuals from Shanghai were exclusively MSM and significantly younger (median age=28, p<0.001). A similar optimal GDT of 0.5%was determined for all cohorts and revealed a higher clustering rate in Shanghai (42.2%) compared to Shenyang (37.3%) and Shenzhen (36.1%). Overall, 38.4% of the sequences were linked (1,178/3,071) into 276 distinct clusters (range: 2-123 seqs/cluster, 149 non-dyad clusters). Clustering individuals were more likely to be younger MSM and diagnosed in Shanghai. Bayesian methods revealed high TR among the 12 largest clusters (range 11-123 seqs/cluster) with a median TR of 20.6/100 person-years (IQR 17-24). Overall, 87.7% (1,895/2,160) linkages were between individuals diagnosed in the same city and 35/276 (12.7%) clusters were spatially heterogeneous clusters (i.e. individuals from 2 cities) (Fig. 1A). Phylogeographic and SM analyses confirmed high gene flow between the three large cities with predominant migration from the central and most populated city of Shanghai toward Shenyang and Shenzhen (Fig 1B). Conclusion: This study revealed similar dynamics of the CR01AE HIV local epidemics in the 3 cities of China and high clustering rate among young MSM. Network inferences across these cities and spatial dispersal suggest that Shanghai likely serves as hub for HIV dispersal among young MSM. Such results could inform public health efforts among young MSM in large cities.

946 PHYLODYNAMIC FEATURES OF ACTIVE LARGE CLUSTERS FUELING THE HIV EPIDEMIC IN QUEBEC MSM Bluma G. Brenner 1 , Ruxandra-Ilinca Ibanescu 1 , Irene Vrbik 1 , Michel Roger 2 , Isabelle Hardy 2 , David Stephens 1 1 McGill University, Montreal, QC, Canada, 2 Centre de Research du Centre Hospitalier de l’Université de Montreal, Montreal, QC, Canada Background: An understanding of the dynamics of onward spread of HIV is essential to the design and optimization of long-term prevention strategies to control epidemics among Men having Sex with Men (MSM). Our studies used Pol sequence datasets from the Quebec genotyping program and phylogenetic modelling strategies to map transmission networks and deduce factors implicated in the expansion of HIV among MSM in Quebec. Methods: Our data comprise 4051 time-stamped HIV Pol sequences taken from the male subtype B infections excluding mixed gender and IDU/HET clusters. Expanding on previous work, 34 large clusters having 20+ distinct members were identified, having high bootstrap support (>90) and sufficient genetic similarity (<0.05 maximum pairwise patristic distance). We applied a birth-death SIR (BDSIR) model available in the phylodynamic add-on for BEAST2 version 2.4.3 and Richard’s five parameter asymmetric dose response curves to model growth trajectories. Genotyping across the viral integrase and V3 loop was performed on representative infections within clusters. Epidemiological and demographic data from the genotyping and Montreal primary HIV cohort deduced risk correlates implicated in clustering. Results: Phylogenetics revealed two patterns of HIV-1 spread among MSM. While half of the HIV epidemic was ascribed to small self-limiting clusters (size 1-4), thirty-two viral strains contributed to micro-epidemics (cluster size 20-145) disproportionately rising from 13%, 25%, and 42% of new diagnoses in 2004-2007, 2008-2011, and 2012-2015, respectively. BDSIR plots deduced early, active and dying phases of expansion for individual clusters. Ten to twelve 20+ clusters fueled spread of HIV in each quadrennial period. Epidemiological and virological data deduced factors contributing to the expansion of the ten active strains from 2012-2016. Clusters were concentrated in the Montreal area with cluster 67B reflecting a second-wave epidemic in Quebec City. Belonging to 20+ clusters was associated with primary/recent infection and being under 30 years of age (odds ratio 3.7 and 3.3, respectively). Clusters over the 2012-2015 quadrennial period arose in significantly younger populations. The heightened transmissibility of strains belonging to distinct 20+ clusters were related to increased viral replicative fitness and/or dual tropism. Conclusion: HIV-1 continues to spread among MSM with an alarming shift towards large cluster outbreaks, emphasizing the need for improved prevention paradigms.

Poster Abstracts

CROI 2018 361