A22 Phylogenetic clustering of hepatitis C virus infection among people who inject drugs in Baltimore

Abstract The availability of effective, oral direct acting antivirals for hepatitis C virus (HCV) treatment has fueled optimism for HCV elimination through treatment as prevention (TasP) among people who inject drugs (PWID). Identifying characteristics of individuals in transmission networks would provide critical information for the development and implementation of effective, targeted HCV TasP strategies. The AIDS linked to the IntraVenous Experience (ALIVE) cohort has followed PWID in Baltimore since 1988. Sequencing of the HCV core/E1 region (342 nucleotides) was performed on HCV viremic samples from the most recent study visit attended by ALIVE participants between August, 2005 and December, 2016. Outgroup sequences were retrieved from GenBank through a BLAST search for HCV sequences similar to study sequences to support identification of ‘local clusters’ and were aligned to study sequences using Clustal O. Phylogenetic trees were inferred for each of HCV subtype 1a and 1b separately through maximum likelihood analysis implemented in the MEGA X software using the Tamura-Nei model with gamma distribution and invariant sites. Nucleotide substitution model selection was based on the corrected Akaike information criterion scores of various models in MEGA. Robustness of the resulting tree was assessed by bootstrapping with 1,000 replicates. Clusters were identified using ClusterPicker software (70% bootstrap threshold and 0.05 maximum genetic distance threshold). Sensitivity analyses were performed by varying the genetic distance threshold between 0.025 and 0.05 to determine the effect on identification of factors associated with clustering. HCV infection clustering was defined as > 2 participants with HCV genome sequences satisfying 70 per cent bootstrap and 0.05 genetic threshold distance requirement for sequence similarity. Logistic regression was used to assess sociodemographic factors associated with being in an HCV cluster. Among 512 HCV genotype 1 viremic PWID, HCV subtype prevalence was 83 per cent genotype 1a and 17 per cent genotype 1b. The median age of participants was 54 years, 68 per cent male, 87 per cent Black, and 38 per cent HIV infected. Overall, 9 per cent (n = 44) were grouped into 21 clusters, consisting of 20 pairs and 1 triad. Of the 425 genotype 1a and 87 genotype 1b samples evaluated, 8 per cent (n = 33) and 13 per cent (n = 11) respectively, were in clusters. In unadjusted analyses, membership in a cluster, was associated with younger age (odds ratio (OR) 1.5 [95% confidence interval (CI) 1.1–2.1] per 10 year age decrease); female sex (OR 2.8 [95% CI 1.5–5.3]), HIV infection (OR 4.9 [95% CI 2.5–9.9]), and living in East Baltimore (versus outside East Baltimore, OR 2.0 [95% CI 1.0–3.9]). In adjusted analyses, female sex (OR 2.0 [95% CI 1.0–3.9] and HIV infection (OR 5.4 [95% CI 2.6–11.1] remained independently associated with being in an HCV infection cluster. HIV-infected PWID and their networks should be prioritized for HCV treatment and prevention interventions given an increased likelihood of transmission in these groups.