CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

566 Tenofovir and the Incidence of Hepatocellular Carcinoma in HIV/HBV-Coinfected Persons Gilles Wandeler 1 ; David Kraus 2 ; Peter Reiss 3 ; Lars Peters 4 ; Francois Dabis 5 ; Jan Fehr 6 ; Marc van derValk 7 ; Lars Gjaerde 4 ; Fabrice Bonnet 8 ; Andri Rauch 9

1 Univ Hosp Bern, Bern, Switzerland; 2 Inst of Social and Preventive Med, Univ of Bern, Bern, Switzerland; 3 Amsterdam Inst for Global Hlth and Develop, Amsterdam, Netherlands; 4 Copenhagen HIV Prog, Copenhagen, Denmark; 5 INSERM U897, ISPED, Univ de Bordeaux, Bordeaux, France; 6 Univ Hosp Zurich, Zurich, Switzerland; 7 Academic Med Cntr, Amsterdam, Netherlands; 8 Univ Hosp Bordeaux, Bordeaux, France; 9 Bern Univ Hosp and Univ of Bern, Bern, Switzerland Background: Hepatocellular carcinoma (HCC) is a leading cause of death in HIV/hepatitis B virus (HBV)-coinfected patients. Current screening recommendations are based on incidence estimates in untreated HBV-infected patients and might be inadequate for HIV/HBV-coinfected individuals on antiretroviral therapy (ART). We explored the impact of tenofovir (TDF) on HCC incidence in a large collaboration of HIV cohorts including the Swiss HIV Cohort Study, Athena, EuroSIDA and ANRS CO3 Aquitaine. Methods: We included all HBsAg-positive adults with complete ART history available. HCC incidence was described for the full population and compared between sub-groups according to the main demographic and clinical characteristics. We defined the cumulative time off TDF (either without any HBV-active ART or including only lamivudine [3TC]) as the main HBV therapy exposure variable. A binary variable was created according to the median follow-up (FUP) time on TDF (4 years). Liver cirrhosis was defined according to histology or as an AST-to-platelet ratio index (APRI) >1.5. We evaluated the association between cumulative time off TDF and the incidence of HCC using multivariable Poisson regression, adjusted for sex, ethnicity, hepatitis C virus (HCV) infection and liver cirrhosis. Results: Of 3,593 HIV/HBV-coinfected patients included, 587 (16.3%) were female, 1,803 (50.2%) men who have sex with men, 2,876 (80.0%) Caucasians and 835 (23.2%) HCV- coinfected. Overall, 40.3% of the total FUP time was spent on TDF, 30.6% on 3TC only and 29.1% on ART without HBV-activity. Over 32,644 patient-years (py), 60 individuals (1.7%) developed an HCC, resulting in an overall incidence of 1.84 per 1,000 py (95% confidence interval [CI] 1.40-2.37). The incidence of HCC was highest in patients with >4 years of FUP off TDF (incidence rate ratio [IRR] 4.04, 95% CI 2.10-7.70) and in those with liver cirrhosis (IRR 3.04, 95% CI 1.83-5.04) (Figure). In adjusted analyses, there was a significant increase in the incidence of HCC per year off TDF (adjusted IRR [aIRR] 1.12, 95% CI 1.07-1.17), and patients with cirrhosis remained at higher risk of HCC (aIRR 2.85, 95% CI 1.70-4.79). During TDF therapy, the risk of HCC remained stable per additional year of FUP (aIRR 0.96, 95% CI 0.86-1.05). Conclusions: Approximately 2% of patients developed an HCC over a median follow-up time of 8.4 years. HCC incidence increased with the length of FUP off TDF and was three times higher in cirrhotic compared to non-cirrhotic patients.

Poster Abstracts

567 Dually Active HIV/HBV Antiretrovirals Protect Against Incident Hepatitis B Infections Mohaned Shilaih 1 ; Alex Marzel 1 ; Alexandra L. Calmy 2 ; Katharine Darling 3 ; Manuel Battegay 4 ; Matthias Hoffmann 5 ; Enos Bernasconi 6 ; Alexandra Scherrer 7 ; Huldrych F. Günthard 7 ; Roger Kouyos 7 ; for the Swiss HIV Cohort Study 1 Univ Hosp Zurich, Zürich, Switzerland; 2 Univ Hosp Geneva, Univ of Geneva, Geneva, Switzerland; 3 CHUV, Lausanne, Switzerland; 4 Univ Hosp Basel, Basel, Switzerland; 5 Kantonsspital St Gallen, St Gallen, Switzerland; 6 Regional Hosp Lugano, Lugano, Switzerland; 7 Univ Hosp Zürich, Zürich, Switzerland Background: The prevention of hepatitis B virus (HBV) transmission in HIV infected individuals is important as both viruses share common transmission modes and both have detrimental effects on each other’s natural course. Vaccination against HBV remains the mainstay of preventing HBV acquisition both in HIV infected and uninfected patients. However owing to HIV’s effect on the immune system, mounting and maintaining a protective immune response against HBV is sometimes unattainable with a success rate between 18%-71%. The strength of this study is twofold: (i) the ability to address confounding by condomless sex and (ii) sample size. We hypothesize that dually active antiretroviral therapy (DAART) against HBV and HIV (Tenofovir, Lamivudine, and Emtricitabine) has a protective effect against HBV but that the magnitude of the association could be confounded by behavioral and immune factors. Methods: Patients with at least one negative serological marker (HBsAg, AntiHBC, or HBV-DNA) for HBV infection at baseline were included in the analysis. We excluded patients with a positive AntiHBs response, and their follow-up time after the first positive AntiHBs test. An incident case was then defined to be the presence of any of HBV serological markers following a negative baseline test. Both univariate and multivariate Cox proportional hazard models were utilized, with the outcome variable being an incident case of HBV infection and the explanatory variable being the proportion of observation time on DAART. Results: We included 1826 patients from the Swiss HIV Cohort Study (988 heterosexuals (54%), 226 intravenous drug user (IDU) (12%) and 612 men who have sex with men (MSM) (34%)). The total number of incident HBV cases was 180 of which 49%were in MSM. Most patients had only two tests (Median 2, IQR 2-3), and the median time between tests was 29 months (IQR 12-59). Both univariable and multivariable analysis show a risk reduction of acquiring HBV for patients on DAART. DAART was associated with a lower HBV hazard ratio 0.4 (95% CI 0.2-0.6). DAART association was robust to adjustment (0.3, 0.2-0.5) for condomless sex, √CD4 cell count, and patients’ demographics. Condomless sex (1.4, 1-1.9), belonging to MSM (3.0, 2.0-4.3) or IDU (4, 2.6-6.5) transmission group were all associated with higher HBV incidence. Conclusions: Overall, our study suggests that antiretroviral therapy, regardless of CD4 counts, has a strong beneficial public health impact that includes pre-exposure prophylaxis of an HBV co-infection.

224

CROI 2016