CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

Methods: As a component of the HIV outbreak response, persons at-risk for HIV or HCV (e.g., persons who inject drugs [PWID]) were also tested for HCV. HCV NS5b gene sequencing was performed on HCV antibody-reactive specimens with detectable HCV RNA. HCV clusters were defined when NS5b sequences were the same genotype and highly genetically related (>96% nucleotide identity). Persons infected only with HCV were cross-referenced with HIV contact investigation data to determine if they had been identified as a contact of an HIV-infected case-patient in this outbreak. Results: From October 2014 – August 2015, HCV antibody testing was conducted for 647 at-risk persons, of whom 570 (88.1%) had a reactive HCV result and 126 (19.5%) were also HIV co-infected. The NS5b genomic region was amplified in a subset of 270 specimens (25.0% HIV-coinfected) that had detectable HCV RNA. NS5b subtype 1a (n=192) was most common, followed by 3a (n=56), 2b (n=16), and 1b (n=6). Overall, 132 (48.9%) specimens clustered in one of two large HCV NS5b clusters (Cluster 1, n=89; Cluster 2, n=43) that included persons infected with HCV alone (n=90) and those co-infected with HIV (n=42) [Figure]. All 42 HIV co-infected patients in these clusters were linked to the HIV outbreak. Of the persons infected only with HCV within these clusters, 52 (57.8%) were known contacts and 38 (42.2%) were not known contacts of HIV-infected patients in this outbreak. Of the 38 not identified as contacts, 19 (50%) resided outside of the outbreak county. Compared with HCV-infected persons not in an HCV cluster, a greater proportion of those in HCV clusters 1 and 2 resided within the HIV outbreak county (79.6% vs 42.9%, p<0.01) and were known contacts of HIV-infected persons in the outbreak (71.2% vs 42.9%), p<0.01). Conclusions: In this analysis, many persons infected only with HCV had HCV strains that phylogenetically clustered with those from persons dually infected with HIV/HCV, yet one-half of those infected only with HCV were not identified as needle-sharing contacts in the HIV outbreak investigation. HCV phylogenetic analysis among PWID may provide information complementary to contact-tracing in identifying networks of persons at risk for HIV acquisition.

Poster Abstracts

542 Limited Overlap in Transmission Clusters of HIV and HCV Among MSM in the Netherlands Joost W. Vanhommerig 1 ; Daniela Bezemer 2 ; Ard van Sighem 2 ; Colette Smit 2 ; JanT. van der Meer 3 ; Bart J. Rijnders 4 ; Joop E. Arends 5 ; Maria Prins 1 ;Thijs J. van de Laar 6 ; Janke Schinkel 7 ; for the observational ATHENA and MOSAIC cohort studies 1 PH Service of Amsterdam, Amsterdam, Netherlands; 2 Stichting HIV Monitoring, Amsterdam, Netherlands; 3 Cntr of Infectious Diseases and Immunology Amsterdam, Academic Med Cntr, Amsterdam, Netherlands; 4 Erasmus Univ Med Cntr, Rotterdam, Netherlands; 5 Univ Med Cntr Utrecht, Utrecht, Netherlands; 6 Sanquin Blood Supply Fndn, Amsterdam, Netherlands; 7 Academic Med Cntr, Amsterdam, Netherlands Background: HIV-1 and hepatitis C virus (HCV) circulate among men who have sex with men (MSM). We conducted a study to investigate the overlap of these epidemics among MSM in the Netherlands using a phylogenetic approach. Methods: Data were derived from the observational ATHENA and MOSAIC cohort studies. We included 5,038 MSM who were diagnosed with a HIV subtype B infection between 1981 and 2014. HIV subtype was based on the availability of a pol sequence. Of them, 562 (11.2%) were (ever) coinfected with HCV (until October 2014). Time from HIV diagnosis to HCV infection was calculated using the Kaplan-Meier method. HCV NS5B sequences were available for 126/562 (26.7%) coinfected MSM, allowing phylogenetic analysis of both HIV and HCV. HIV phylogenetic clusters were defined as having ≥10 sequences with a bootstrap value >90 and a median pairwise distance within the clade smaller than the 5 th percentile threshold of the pairwise distances in the whole tree. We investigated the presence of HIV clusters that had an increased risk for HCV infection. Results: In total, 118 HIV phylogenetic clusters were identified. These clusters included 3,084/5,038 (61.2%) MSM infected with HIV subtype B, and 97/118 (82.2%) clusters contained ≥1 HCV infection. Median HIV-1 cluster size of those with past or present HCV infection was comparable to those with no history of HCV. Median time between HIV diagnosis and HCV diagnosis was 3.3 years (IQR: 1.0-7.3), but decreased over time. HCV NS5B sequences were obtained from 150 HCV infections among 126 MSM; 21 MSM had ≥1 reinfection. Among 51 HIV clusters with ≥2 HCV sequences, 14 clusters contained HCV strains of concordant genotypes, but only 8/14 HIV clusters with ≥2 HCV sequences contained two HCV strains of the exact same HCV lineage. Ultimately, 19/150 (12.7%) coinfected MSM clustered on both HCV and HIV phylogeny. Conclusions: In this study, HCV infection was not confined to specific HIV clusters, indicating there are no specific HIV clusters with elevated risk of HCV infection. When multiple HCV infections were present within an HIV cluster, concordance of HCV phylogeny was relatively uncommon, even among those with concordant HCV genotype. One explanation may be that HCV spreads in MSM networks that differ from the HIV transmission networks. The median duration from HIV diagnosis to HCV infection of 3.3 years suggests that these HCV networks are established some time after HIV infection.

212

CROI 2016