Canada Word Count : 3 , 828 Number of Tables and Figures : 4 Tables and 2 Figures

Background & Aims: Direct-acting antiviral therapies (DAA) are an important tool for hepatitis C virus (HCV) elimination. However, reinfection among people who inject drugs (PWID) may hamper elimination targets. We therefore estimated HCV reinfection rates among DAA-treated individuals, including PWIDs. Methods: We analyzed data from the BC Hepatitis Testers Cohort which included ~1.7 million individuals screened for HCV in British Columbia, Canada. We followed HCV-infected individuals treated with DAAs who achieved a sustained virologic response (SVR) and had ≥1 subsequent HCV RNA measurement to April 22 nd , 2018. Reinfection was defined as a positive RNA measurement after SVR. PWIDs were identified using a validated algorithm and classified based on recent (<3 years) or former (≥3 years before SVR) use. Crude reinfection rates per 100 person-years (PYs) were calculated. Poisson regression was used to model adjusted incidence rate ratios (IRRs) and 95% confidence intervals (CI). Results: Of 4,114 individuals who met inclusion, most were male (n=2,692, 65%), born before 1965 (n=3,411, 83%) and were either recent (n=875, 21%) or former PWIDs (n=1,793, 44%). Opioid-agonist therapy (OAT) was observed in 19% of PWIDs. We identified 40 reinfections during 2,767 PYs. Reinfection rates were higher among recent (3.1/100 PYs; IRR: 6.7, 95% CI: 1.9, 23.5) and former PWIDs (1.4/100 PYs; IRR: 3.7, 95% CI: 1.1, 12.9) than non-PWIDs (0.3/100 PYs). Among recent PWIDs, reinfection rates were higher among individuals born after 1975 (10.2/100 PYs) and those with co-infected with HIV (5.7/100 PYs). Only one PWID receiving daily OAT developed reinfection. 5 Conclusions: Population-level reinfection rates remain elevated after DAA therapy among PWIDs because of ongoing exposure risk. Engagement of PWIDs in harm-reduction and support services is needed to prevent reinfections. Abstract word count: 275 wordsword count: 275 words