CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

may indicate that the elimination of HIV-1 reservoir in HIV+/HCV+ subjects might be even harder than in HIV+ patients.

616 LIVER INFLAMMATION IS COMMON AND LINKED TO METABOLIC DERANGEMENTS IN TREATED HIV Kara W. Chew 1 , Katherine Tassiopoulos 2 , Frank J. Palella 3 , Kunling Wu 2 , Susanna Naggie 4 , Netanya S. Utay 5 , Edgar T. Overton 6 , Mark Sulkowski 7 1 University of California Los Angeles, Los Angeles, CA, USA, 2 Harvard University, Boston, MA, USA, 3 Northwestern University, Chicago, IL, USA, 4 Duke University, Durham, NC, USA, 5 University of Texas at Houston, Houston, TX, USA, 6 University of Alabama at Birmingham, Birmingham, AL, USA, 7 Johns Hopkins University, Baltimore, MD, USA Background: Abnormal serum liver enzymes in people with HIV (PWH) are common and often unexplained. We sought to identify the prevalence of and underlying reasons for aspartate and alanine aminotransferase (AST and ALT) elevation in a well-characterized cohort of adults with treated HIV without hepatitis C or B virus (HCV or HBV) infection or heavy alcohol use. Methods: Participants from the longitudinal observational AIDS Clinical Trials Group HAILO cohort who did not report heavy alcohol use, were negative for anti-HCV and hepatitis B surface antigen, and with at least 2 AST and ALT measures between 11/2013–2/2018 were included. Clinical and demographic characteristics, including the Hepatic Steatosis Index (HSI = 8x(ALT/AST)+BMI (+2 female, +2 diabetes)), FIB-4 score and metabolic syndrome (MetS) were compared between persons with and without ≥1 elevated AST or ALT (defined as AST >36 and ALT >30 U/L for men and AST >30 and ALT >19 U/L for women), using chi-square and Wilcoxon tests and multiple logistic regression models. Covariates with p<0.10 in univariate analysis were included in the multivariable models. Results: Of 1035 participants, 662 met criteria for inclusion; 456 (69%) had ≥ 1 and 236 (36%) ≥ 2 elevated AST/ALT during a median of 4.0 years of follow-up. Median age at entry was 51 years; 138 (21%) female; 184 (28%) black and 122 (18%) Hispanic; median entry and nadir CD4 cell counts/mm 3 (CD4) 621 and 195, respectively; and 627 (95%) had plasma HIV RNA <200 copies/mL at entry. In univariate analysis, the elevated liver enzyme group was younger, had a higher proportion of Hispanic and female participants, higher entry CD4 without differences in nadir CD4, higher HSI score, and a higher proportion with MetS and HSI ≥ 36 (p<0.05 for all). There were no differences in the proportions with HIV RNA suppression or antiretroviral use (current or previous); FIB-4 score was similar in each group. The Table summarizes the results frommultiple logistic regression models. Conclusion: After exclusion of HCV, HBV and alcohol, liver enzyme elevation was remarkably common in this cohort and independently associated with metabolic disease, presence of hepatic steatosis by HSI, Hispanic ethnicity, and lower CD4 at entry. These findings suggest that NAFLD may be a common cause of liver inflammation in PWH receiving suppressive antiretroviral therapy (ART). Further research is needed to understand the contribution of NAFLD and other mechanisms of liver injury in PWH on suppressive ART.

615 PREVALENCE OF FATTY LIVER DISEASE IN INDIVIDUALS WITH AND WITHOUT HIV INFECTION Ditte M. Kirkegaard-Klitbo 1 , Flemming Bendtsen 1 , Stefan Stender 2 , Klaus F. Kofoed 2 , Børge Nordestgaard 3 , Jens D. Lundgren 4 , Susanne D. Nielsen 2 , Thomas Benfield 1 1 Hvidovre Hospital, Hvidovre, Denmark, 2 Rigshospitalet, Copenhagen, Denmark, 3 University of Copenhagen, Copenhagen, Denmark, 4 Copenhagen HIV Program, Copenhagen, Denmark Background: Fatty liver disease (FLD) is a growing cause of chronic liver disease. People living with HIV (PLWH) may be at a higher risk of FLD due to life style and antiretroviral medication. Here we assessed the prevalence of FLD in PLWH and matched HIV uninfected controls by unenhanced CT liver scan. Further, factors associated with hepatic steatosis were assessed and the effect of HIV per se evaluated. Methods: PLWH (n=765) aged 40 years and above were recruited from the Copenhagen Co-morbidity (COCOMO) in HIV Infection study. Uninfected controls (n=1192), matched on gender and 5-years age strata, were recruited from the Copenhagen General Population Study (CGPS). Unenhanced CT liver scan was performed on all participants and liver attenuation measured. We defined FLD as a liver attenuation ≤ 48 Hounsfield Units (HU) equivalent with moderate to severe steatosis. Unadjusted and adjusted logistic regression analyses were performed. Sensitivity analyses were performed with exclusion of individuals with excessive alcohol intake (def.: 20 g/wk for females and 30 g/wk for males) and significant liver fibrosis (def.: Fibroscan ≥ 12kPa). Results: Of PLWH, FLD was detected in 8.5 % compared to 17.4% of uninfected controls. After adjustment, 1 unit increase in BMI or waist circumference was associated with FLD in PLWH (OR (95% CI): 1.22 (1.03;1.45)) and 1.08 (1.01;1.14)) and (OR (95% CI): 1.12 (1.01;1.24) and 1.12 (1.08;1.17)) in uninfected controls. In PLWH, male sex was associated with FLD (OR (95% CI): 7.93 (1.01;61.97). A 1 unit increase in triglycerides was associated with FLD in uninfected controls (OR (95% CI):1.44 (1.24;1.69)), while there was no association for PLWH. HIV was significantly associated with lower odds of FLD (OR (95% CI): 0.30 (0.19;0.49)) and the association persisted after further adjustments for e.g. current antiretroviral treatment, lipids, use of statin, injection drug use and ethnicity. Sensitivity analyses did not change these results. Conclusion: The prevalence of FLD was lower in well-treated PLWH compared to age and sex matched HIV uninfected controls. HIV per se was associated with lower odds of FLD. Higher BMI and waist circumference were associated with higher odds of FLD in both cohorts. These results may be explained by a more healthy lifestyle in PLWH or unmeasured residual confounding.

Poster Abstracts

CROI 2019 233