CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

importance of HIV on HCC incidence, few studies have evaluated patients with biopsy-proven HCC to examine differences in background hepatic fibrosis, tumor differentiation, HCC stage, and survival between HIV+ and uninfected persons. Methods: We performed a cohort study of HIV+ and uninfected patients with HCC diagnosed between 2000 and 2016 within the Veterans Aging Cohort Study (VACS). We identified patients who underwent liver biopsy and collected the reports of pathologists’ reviews of these specimens. Demographics, comorbidities, background liver fibrosis on biopsy, tumor characteristics, American Joint Committee on Cancer (AJCC) stage, and survival after biopsy was compared by HIV status. Hepatitis virus co-infection status was determined by laboratory test results. Cox proportional hazards regression was used to determine hazard ratios (HRs) of factors associated with death after biopsy- proven HCC, including HIV, hepatitis B virus (HBV) coinfection, hepatitis C virus (HCV) coinfection, level of alcohol consumption, and advanced hepatic fibrosis present on biopsy. Results: Among 304 patients (median age, 58.4 years; 99%male; 59% black) with biopsy-proven HCC, 134 (44%) were HIV+ with no significant differences in demographics by HIV status. HIV+ patients more commonly were infected with HBV (14% vs. 3%; p<0.001), but not RNA-confirmed HCV infection (78% vs. 83%; p=0.18). There were no differences in tumor differentiation, prevalence of advanced hepatic fibrosis/cirrhosis, or HCC stage by HIV status. Median survival after diagnosis was shorter for HIV+ than uninfected patients (397 days [IQR: 127, 1127] vs. 565 days [IQR: 202, 1522]; log-rank: p=0.05). After adjustment for age, race, alcohol use, HBV, HCV, advanced hepatic fibrosis, and HCC stage, the risk of death after diagnosis was higher for HIV+ than uninfected persons (HR, 1.33 [95% CI, 1.00-1.78]). Conclusion: In this sample of patients with biopsy-proven HCC, there were no differences in tumor differentiation, background hepatic fibrosis, or HCC stage between HIV+ and uninfected persons. However, HIV was associated with poorer survival after diagnosis. Future studies should evaluate the impact of HCC treatments by HIV status. 670 ANTIRETROVIRAL DRUGS ASSOCIATED WITH SUBCLINICAL CORONARY ARTERY DISEASE Helen Kovari 1 , Alexandra Calmy 2 , Thanh Doco-Lecompte 2 , René Nkoulou 2 , Alex Marzel 1 , Rainer Weber 1 , Philipp A. Kaufmann 1 , Ronny R. Buechel 1 , Bruno Ledergerber 1 , Philip E. Tarr 3 1 University Hospital Zurich, Zurich, Switzerland, 2 University Hospitals of Geneva, Geneva, Switzerland, 3 University of Basel, Basel, Switzerland Background: Definite and validated coronary artery disease (CAD) events have been associated with certain antiretroviral therapy (ART) agents. In contrast, the influence of ART drugs on early, subclinical atherosclerosis as determined by coronary artery calcium (CAC) scoring and coronary CT angiography (CCTA) is yet to be elucidated. Methods: In this prospective study of ≥45 year old Swiss HIV Cohort Study participants, CAC scoring and CCTA were performed. The following subclinical CAD endpoints were analysed separately: CAC score >0, any plaque, calcified plaque, and non-calcified/mixed plaque on CCTA. Logistic regression was used to explore associations between the different CAD endpoints and cumulative exposure to ART, ART classes, and the ten most often used individual drugs. Covariables included sex, age, smoking, hypertension, dyslipidemia, diabetes, CD4 nadir <50 cells/µL, and peak HIV-1 RNA >100’000 copies/mL. Results: We included 428 participants (mean age 52 years, 86%men, 91% Caucasian, 35% current smokers, 60% homosexual, median CD4 cell count at cardiac imaging 598 cells/μL, 93% on ART, 87%with undetectable HIV-1 RNA). CAC score >0 was recorded in 227 (53%) patients, any plaque in 226 (53%), calcified plaque in 158 (37%), and non-calcified/mixed plaque in 158 (37%) participants, respectively. Cumulative exposure to PIs was associated with calcified plaque (adjusted odds ratio (aOR) per 5 years 1.21 [95% confidence interval, 1.01-1.46]). The associations between cumulative exposure to individual ART drugs and different CAD outcomes are shown in the figure. Adjustment for covariables did only marginally affect point estimates. CAC score >0 was not associated with any individual ART drug. Any plaque was associated with cumulative exposure to regimens containing atazanavir (aOR per 5 years 1.66 [1.11-2.47]), and abacavir (1.38 [1.03-1.85]). Calcified plaque was associated with exposure to atazanavir (1.47 [1.01-2.14]). Non-calcified/mixed plaque was associated with exposure to abacavir (1.45 [1.10-1.91]). Conclusion: We found an increased risk of coronary artery plaque in patients exposed to regimens containing atazanavir or abacavir independent of CAD

risk factors. Atazanavir was associated with calcified plaque whereas abacavir was associated with non-calcified/mixed plaques. Although adjustment for traditional cardiovascular risk factors only marginally affected associations with individual drugs, we cannot fully exclude residual confounding.

Poster Abstracts

671 CAROTID IMT PROGRESSION IS MOST STRIKING IN FIRST 48 WEEKS OF ANTIRETROVIRAL THERAPY Corrilynn O. Hileman 1 , Grace A. McComsey 2 1 MetroHealth Medical Center, Cleveland, OH, USA, 2 University Hospitals Cleveland Medical Center, Cleveland, OH, USA Background: Studies are mixed as to whether common carotid artery intima media thickness (CCA IMT) progression is more accelerated in HIV+ vs. HIV- and the role of contemporary ART and systemic inflammation remain unclear. Methods: This is a 4-year prospective, observational, cohort study of ART-naïve HIV+ adults and age- and sex-matched HIV- healthy controls designed to evaluate CCA IMT progression measured by high resolution ultrasound annually. Subjects were followed for at least 96 weeks after the last subject enrolled. Mixed effects linear modeling was used to compare CCA IMT measurements over time among HIV+ who remained ART-naïve, HIV+ who initiated ART and HIV-. Potential contributors to CCA IMT progression including traditional cardiovascular disease (CVD) risk factors and soluble markers of systemic inflammation were evaluated. Results: 130 adults were enrolled (85 HIV+; 45 HIV-). Among HIV+, 44 initiated ART (45% NNRTI; 39% PI; 9% RAL) and 41 remained ART-naïve. Overall, mean age was 39 with 74%men. There were more smokers in HIV+ (56% vs. 18% in HIV-) and higher Framingham risk score (FRS) of 4% vs 3% in HIV-. None had previously diagnosed CVD or diabetes. Mean duration of HIV (5.6 years) and baseline HIV-1 RNA (15,473 copies/ml) were similar among HIV+, but nadir CD4+ was slightly lower in the HIV+ who initiated ART (453 vs. 525 cells/mm 3 ; p=0.04). Baseline CCA IMT was similar between groups (0.653 ± 0.109 mm). Over time, CCA IMT progressed similarly in HIV+ ART-naïve and HIV- groups (p=0.55); however, there was a trend towards ART-initiation leading to greater CCA IMT over time compared to HIV- (p=0.05)(see figure). Adding baseline FRS, sCD163, IL-6 and hsCRP each in turn attenuated this association. In the HIV+ ART-initiated group, mean absolute change over 48 weeks before and after ART were -0.003 mm (p=0.75 within-group) and 0.039 mm (p=0.01), respectively (p=0.04 between pre-and post-ART 48 week changes). Interestingly, from 48 weeks to 96 weeks post-ART the change declined to 0.013 mm (p=0.52 within- group and p=0.63 between pre- and post-). The only predictor of CCA IMT increase in the first 48 weeks after ART was higher baseline sVCAM-1 (p=0.01). Conclusion: CCA IMT progressed significantly in the first 48 weeks after ART initiation and the HIV+ ART-initiated group had significantly greater CCA IMT over time when compared to HIV-. Both traditional CVD risk factors and systemic inflammation contributed to this difference.

CROI 2018 249