CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

sampled population (median 269 km; p<0.001) (Figure 1B). This suggests that genetically linked individuals tended to be linked with partners within about an hour’s travel time. Conclusion: We found evidence that HIV-1 transmission in Germany is between younger MSM living in proximity to each other. This provides further support for real-time monitoring of HIV transmission using molecular epidemiology, which can be leveraged to target specific geographic areas where new transmissions are being observed. Furthermore, this work suggest that HIV prevention efforts targeted towards young MSMmay provide the most impact on the German epidemic.

864 PRETREATMENT HIV DRUG RESISTANCE SPREAD IN 6 GERMAN METROPOLITAN REGIONS

Melanie Stecher 1 , Martin Hoenigl 2 , Philipp De Leuw 3 , Josef Eberle 4 , Jan- Christian Wasmuth 5 , Georg Behrens 6 , Guido Schäfer 7 , Christoph D. Spinner 8 , Christoph Stephan 9 , Elena Knops 1 , Sanjay R. Mehta 2 , Joerg Janne Vehreschild 1 , Antoine Chaillon 2 , for the Translational Platform HIV (TP-HIV) Molecular Epidemiology Team 1 Cologne University Hospital, Cologne, Germany, 2 University of California San Diego, San Diego, CA, USA, 3 University Hospital Frankfurt, Frankfurt, Germany, 4 Max von Pettenkofer-Institute, Munich, Germany, 5 Bonn University Hospital, Bonn, Germany, 6 Medizinische Hochschule Hannover, Hannover, Germany, 7 University Hospital Hamburg–Eppendorf, Hamburg, Germany, 8 Technical University of Munich, Munich, Germany, 9 University Hospital Klinikum, Munich, Germany Background: Transmitted HIV drug resistance impacts the choice and success of antiretroviral therapy (ART). In Germany, the reported prevalence of transmitted drug resistance mutations (DRM) differs between regions with a maximum of 12.4%. Here, we sought to understand the molecular epidemiology of DRM transmission in Germany, by analyzing data from six major cities with high rates of new HIV diagnoses (ranging from 5.6 to 17.3 per 100,000 habitants). Methods: Phylogenetic and genetic network analyses were performed to infer putative transmission links and shared DRMs. Screening for DRMwas performed according to the Stanford University Genotypic Resistance Interpretation. We defined a shared DRM as any DRM present in two or more genetically linked individuals (<1.5% genetic distance). Results: We obtained 1,397 HIV pol sequences from HIV-1 ART naïve individuals. The prevalence of any DRM at time of diagnosis was 17.8% (248/1,397 individuals; 31.4% in Hannover, 18.9% in Cologne, 18.8% in Hamburg, 15.3% in Frankfurt, 14.5% in Bonn, 9.1% in Munich). The frequency of any DRMwas comparable among risk groups but was highest among men having sex with men (MSM) (138/792, 17.4%). Genetic transmission network analyses showed comparable frequencies of DRM in clustering and not-clustering individuals (16.3% versus 18.1%; p= 0.46). Of the 47 sequences harboring DRM in the inferred transmission network, 30 (63.8%) were shared by HIV genetically linked partners, predominantly among residents of Cologne (19/30, 63.6%) and Bonn (7/30, 23.3%). Clustering individuals harboring shared DRMs were more likely than non-clustering individuals to be infected with HIV-1 subtype B (p<0.001), < 25 years of age (p<0.001), and living in Cologne (19/30, 63.3%) and Bonn (7/30, 23.3%) (p<0.001 each) (Figure 1A). The most frequently transmitted DRMs were E138A (11/30 in individuals sharing DRMs, 5 clusters) and K103N (9/30, 4 clusters) (Figure 1B). Conclusion: We observed very high rates of DRMs in newly diagnosed individuals in Hannover, Cologne and Hamburg. Network analysis also revealed frequent cases of shared DRMs among genetically-linked individuals, especially K103N and E138A. Our findings highlight regional differences in transmitted drug resistance, and the necessity to focus prevention efforts on specific areas and risk groups to prevent onward transmission across Germany.

Poster Abstracts

863 GEOSPATIAL DISPERSAL OF HIV-1 TRANSMISSION IN 6 MAJOR CITIES IN GERMANY Melanie Stecher 1 , Antoine Chaillon 2 , Philipp De Leuw 3 , Johannes Bogner 4 , Anna M. Eis-Hübinger 5 , Clara Lehmann 1 , Georg Behrens 6 , Olaf Degen 7 , Christoph D. Spinner 8 , Christoph Stephan 3 , Joerg Janne Vehreschild 9 , Sanjay R. Mehta 2 , Martin Hoenigl 2 , for the The Translational Platform HIV (TP-HIV) Molecular Epidemiology Team 1 Cologne University Hospital, Cologne, Germany, 2 University of California San Diego, San Diego, CA, USA, 3 University Hospital Frankfurt, Frankfurt, Germany, 4 Klinikum der Universität München, Munich, Germany, 5 Bonn University Hospital, Bonn, Germany, 6 Medizinische Hochschule Hannover, Hannover, Germany, 7 University Hospital Hamburg–Eppendorf, Hamburg, Germany, 8 Technical University of Munich, Munich, Germany, 9 University of Cologne, Cologne, Germany Background: Geographic targeting of HIV prevention interventions to the highest risk individuals will increase the impact of these efforts. We combined molecular epidemiology and geospatial analyses to provide insights into the drivers of HIV-1 transmission in the German epidemic. Methods: Sociodemographic, geographic and HIV-1 pol sequence data were collected from newly diagnosed individuals in six cities across Germany between 2001 and 2018. Genetic-distance based molecular network analyses were performed to infer putative transmission links. Similarity between genetically-linked individuals was assessed using the Assortativity Index (AI, i.e. shared attributes). Geospatial dispersal was determined by calculating the average distance between the residences of genetically linked individuals (centroids of 3-digit zip codes). Results: We included data from 1,397 HIV-1 infected ART naïve individuals, of which 289/1,397 (20.7%) were putatively linked, forming 102 transmission clusters (size range: 2-12). The largest clusters (more than 10 individuals) consisted mainly of men having sex with men (MSM) from Cologne, Bonn and Frankfurt. Genetically linked individuals were significantly younger (<25 years of age, p=<0.001), more likely to reside in Bonn and Cologne (p<0.001 and p= 0.044, respectively), and more likely to report MSM as a risk factor (p=0.015). Genetically linked individuals were highly assortative by risk group (AI=0.08, p=0.006), age (AI= 0.06, p<0.001), and location (AI= 0.08, p<0.001) (Figure 1A), indicating that individuals tended to cluster with other persons of the same age, risk group, and from the same area. Geospatial analyses revealed that the median distance between residences of genetically linked individuals was 78 kilometers (km), significantly lower than the distance of the random sub-

CROI 2019 336