CROI 2017 Abstract e-Book

Abstract eBook

Poster and Themed Discussion Abstracts

516 HEPATITIS C VIRUS GENOTYPE AND IG-G AVIDITY RESPONSE DURING RECENT INFECTION Veronica Bruce 1 , Eshan Patel 2 , Andrea Cox 2 , Shruti H. Mehta 2 , Denali Boon 2 , Jacob Konikoff 2 , David Thomas 2 , Thomas C. Quinn 3 , Kimberly Page 1 , Oliver Laeyendecker 3 1 Univ of NewMexico, Albuquerque, NM, USA, 2 The Johns Hopkins Univ, Baltimore, MD, USA, 3 NIAID, Bethesda, MD, USA Background: Recently declared WHO hepatitis C virus (HCV) elimination goals include a 90% reduction in HCV infection by 2030. However, accurate tools to monitor reduced HCV incidence are lacking. We assessed the association of HCV genotype and the antibody avidity response during recent infection, and evaluated the impact on parameters for cross- sectional incidence estimation. Methods: Serum or plasma samples from participants with known duration of HCV infection were obtained from participants enrolled in prospective cohort studies: The U Find Out (UFO) Study and the Before and After Acute Study of Hepatitis (BBAASH) Study. We tested HCV IgG antibody positive and HCV RNA positive samples (n=246) from HIV negative participants (n=93) with the HCV Ortho Avidity Assay. The time-scale origin of this analysis was the estimated Ab seroconversion date. The mean duration of recent infection (MDRI) was calculated at varied thresholds of the Ortho Avidity Assay and by the infected visit-specific HCV genotype. The MDRI estimate, indicative of how long an individual appears recently infected within the first two years, was calculated by binomial regression with a logit cubic functional link and a maximum likelihood approach. Subject-level bootstrapping was used to calculate 95% confidence intervals (10,000 replications). Results: Among 246 samples from HCV infected persons: 173 (70.3%) were genotype 1; 15 (6.1%) were genotype 2; 42 (17.1%) were genotype 3; 3 (1.2%) were genotype 4; and 13 (5.3%) were of an unspecified genotype. Increases in HCV IgG antibody avidity correlated with the days post-seroconversion for both genotype 1 and non-genotype 1 infections (Figure). At an Ortho HCV Avidity Index cut-off of 30%, the MDRI was lower for genotype 1 infections (155 days [95% CI, 127-185]) than for non-genotype 1 infections (287 days [95% CI, 139-415]). Using the same cut-off, the MDRI for all available data was 199 days [95% CI, 146-256]. When reducing the Ortho HCV Avidity Index cut-off to 20%, all MDRI point estimates decreased but the lower limit remained above 100 days. Similar estimates were obtained in a sensitivity analysis using a log-log link function. Conclusion: When assessed with the HCV Ortho Avidity Assay, we detected shorter MDRI in persons with genotype 1 infection. If confirmed with a larger sample, these results indicate more work is needed to ascertain how to apply MDRI to mixed genotype populations for HCV incidence estimation.

Poster and Themed Discussion Abstracts

517 IDENTIFICATION OF RISK FACTORS FOR HEPATITIS-C TESTING IN NON-BIRTH COHORT PATIENTS Amanda E. Smart 1 , Alexander Geboy 2 , Peter Basch 2 , Whitney Nichols 2 , Alexander Zeymo 2 , Idene Perez 2 , Maria Hafeez 2 , Ilan Fleisher 2 , Stephen Fernandez 2 , Dawn Fishbein 2 1 Georgetown Univ, Washington, DC, USA, 2 MedStar Hlth Rsr Inst, Hyattsville, MD, USA Background: CDC data from 2012 indicated that 45% of HCV-infected people reported no known risk factors (RFs). However, initiating widespread, automated, RF-based screening outside the Birth Cohort (BC) (b. 1945-1965) is challenging as RFs are often unstructured, not searchable data within Electronic Health Records (EHR). Therefore, testing non-BC patients solely based on RFs has the potential to miss a substantial number of HCV infected patients. Methods: In July 2015 HCV testing data was collected on non-BC patients who were HCV tested across MedStar Health, as a presumptive marker for high risk. A 1:3 case-control retrospective nested chart review was conducted. HCV RFs and opiate prescriptions were manually abstracted from the EHR; other variables were collected using Explorys. Univariate and multivariate logistic regression models were utilized to determine HCV Ab positive (Ab+) predictors.

CROI 2017 215