CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

Methods: Demographic, clinical, and laboratory data were collected between 2011-2016 for HIV-infected patients enrolled in the DC Cohort Study, a multicenter prospective observational study of HIV-infected persons in care in Washington, DC. To determine whether participants ≥21 years old were eligible for statins, we applied ATP III, ACC/AHA, and NLA guidelines to data for participants receiving primary care at their HIV clinic site with ≥1 cholesterol result available. Demographic, behavioral, and clinical predictors of being prescribed statins and of achieving NLA non-high-density lipoprotein cholesterol (non-HDL-C) goals were assessed using multivariable Cox proportional hazards regression. Results: Of 3,312 participants (median age 52; 78%male; 79% Black), 52% were eligible for statin therapy based on ≥1 guideline, including 30% (ATP III), 40% (ACC/AHA), and 45% (NLA). Using each guideline, 73% (ATP III), 56% (ACC/ AHA), and 49% (NLA) of eligible participants were prescribed statins. Predictors of receiving prescriptions were older age (aHR=1.16 [1.07-1.25]/5 years), body mass index ≥30 (aHR=1.50 [1.07-2.11]), and diabetes (aHR=1.37 [1.04-1.82]). Hepatitis C coinfection was associated with a lower likelihood of prescription (aHR=0.66 [0.44-0.98]). Among 216 NLA-eligible participants with available cholesterol results pre-/post-prescription, 53% achieved their non-HDL-C goal within six months. Depression (aHR=0.61 [0.37-0.99]) and nadir CD4 cell count 200-500 (vs. >500) cells/µL (aHR=0.51 [0.27-0.99]) were associated with a lower likelihood of achieving the goal. Conclusion: Approximately half of HIV-infected participants were eligible for statins based on current US guidelines, with the highest proportion eligible based on NLA guidelines, yet statin coverage was substantially lower as measured by prescriptions and achievement of treatment goals. Greater compliance with recommended statin prescribing practices may reduce cardiovascular risk among HIV-infected individuals. 691 HIV+ MEN MATCHED BY AGE AND FRS WITH CONTROLS HAVE MORE CARDIOVASCULAR EVENTS Stefan Esser 1 , Marina Arendt 1 , Christina Schulze 1 , Volker Holzendorf 2 , Norbert H. Brockmeyer 3 , Karl-Heinz Joeckel 1 , Raimund Erbel 1 , Nico Reinsch 4 1 University Hospital Essen, Essen, Germany, 2 University Leipzig, Leipzig, Germany, 3 Ruhr-University Bochum, Bochum, Germany, 4 Alfried Krupp Hospital, Essen, Germany Background: Cardiovascular events (CVE) are more frequent in the HIV- positive patients (HIV+) than in the general population. HIV-infection may be causative for more CVE beside the traditional risk factors. Therefore, we investigated the incidence of CVE in HIV+ and HIV-negative controls leading to more insides of HIV-specific effects on CVE. Methods: We compared CVE of HIV+ outpatients of the HIV HEART study (HIVH) with controls of the population-based Heinz Nixdorf Recall study (HNR) without any CVE at baseline. Both were recruited from the German Ruhr area since 2000. HIVH men with HNR controls were matched in a 1:2 ratio by age and Framingham risk score (FRS). For comparability we limited the follow-up time (FUP) of HNR to the maximum FUP of HIVH of 7.8 years. CVE are defined by myocardial infarction and sudden cardiac death. Hazard Ratios (HR) with corresponding 95%-confidence intervals (CI) were evaluated using Cox proportional Hazard regression with time to CVE and CVE as event. In Kaplan-Meier curves we show CVE-free survival stratified by HIV status. In HIV+ subgroup the CVE-free survival was compared in different ART groups, viral load and clinical HIV-stage. Results: 322 HIVH males were matched with 644 HNR controls by age (54.7 ± 6.5 years) and FRS (14.7 ± 8.5). The mean follow-up time was 7.8 ± 0.8 years in HNR and 4.6 ± 1.7 years in HIVH. The HIV+ were diagnosed for 9.2 ± 6.5 years and 118 (37.0 %) already had AIDS as defined in the CDC classification. At baseline in 245 HIV+ (76.1 %) the viral load was below the level of detection and 302 (93.8 %) received antiretroviral treatment (ART). For HIV+ we achieved HR of 3.9 (CI: 1.7; 8.7) for CVE in comparison to HNR. A Kaplan-Meier curve of CVE-free survival between HIVH and HNR is shown in figure 1. In HIV+, CVE-free survival tended to be different in main ART regimens compared to ART-naïve HIV+ (HR of NNRTI or PI as third agent: 0.3 (CI: 0.03; 3.3) and 1.1 (CI: 0.1; 8.5)), was worse in those having AIDS (HR: 1.6 (0.6; 4.5)) and a viral load above the detection limit (1.2 (CI: 0.4; 3.7)). Conclusion: HIV-infection was associated with a higher CVE incidence. We could show that ART, clinical HIV stage and viral load seem to an effect on CVE-free survival. Other not in the FRS requested CVE risk factors like effects of

the HIV infection itself, drug use and vascular inflammation may have an impact on CVE in HIV+ and must be evaluated in further studies.

692 COMPARING STRATEGIES FOR REDUCING MYOCARDIAL INFARCTION RATES IN HIV PATIENTS Priscilla Hsue 1 , Grace A. McComsey 2 , Calvin Cohen 3 , Alejandro Sola 4 , Anne C. Beaubrun 3 1 University of California San Francisco, San Francisco, CA, USA, 2 Case Western Reserve University, Cleveland, OH, USA, 3 Gilead Sciences, Inc, Foster City, CA, USA, 4 Datakhan Analytics Solutions, New York, NY, USA Background: Studies show higher rates of myocardial infarctions (MI) with HIV and aging. Abacavir (ABC) has also been associated in some studies with an increased risk of cardiovascular (CV) events. To assess the relative impact of one intervention to reduce MI risk in HIV patients versus another, we modelled the impact of interventions that address traditional risk factors and replacing ABC on predicted MI rates. While other HIV antivirals have been associated with elevated MI risk, we used ABC as an example of the impact of changing HIV treatment in addition to a focus on traditional risk factors. Methods: Strategies for reducing MI rates in HIV patients were compared over 10 years using a decision tree model. Assumptions about the effectiveness of smoking cessation counseling, substitution of ABC with an alternative regimen, anti-hypertensive and anti-hyperlipidemia medication use were based on publications from the HIV or general population. We adjusted for sex, age, and presence of the four MI risk factors. Interventions were compared based on published data on the probability of success of changing the risk factor and the impact of changing it when successful. For smoking cessation, the impact was based on published quit rates following counseling, 36.5% after one year and 10% annual relapse rate. Results: In the base case of 50-year old HIV positive male smokers who only replaced ABC, there was a 46% reduction in the MI rate compared to those who continued ABC (0.31/ 100 vs. 0.58/ 100 PY). Men who are counseled and treated for smoking cessation which resulted in an 11%MI rate reduction versus those who did not attempt smoking cessation (0.52/ 100 vs. 0.58/ 100 PY). Over 10 years, compared to no MI intervention, ABC substitution prevented more MIs than counseling about smoking (2.64 vs 0.63 MIs per 100 persons). The impact of treating hypertension and hyperlipidemia was a 19% and 31% reduction in MI risk, respectively (see Table). Conclusion: By incorporating the impact of CV risk factor modification based on real world data, this model suggests that replacing ABC, which can be accomplished in most patients, is potentially more impactful in reducing MI risk than interventions solely on traditional risk factors. While this model does not account for all tobacco risks, findings highlight the role that ABC substitution can have on MI risk over time compared to antismoking, hypertension and lipid lowering interventions. Interventions to address all CV risk factors are warranted.

Poster Abstracts

CROI 2018 257