CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

680 LOWER CARDIOVASCULAR DISEASE RISK ASSOCIATED WITH INTEGRASE INHIBITORS Jane A. O’Halloran , John Sahrmann, Anne M. Butler, Margaret A. Olsen, William Powderly Washington University in St Louis, St Louis, MO, USA Background: Several antiretroviral therapy (ART) classes have been associated with increased myocardial infarction (MI) risk. No studies have examined cardiovascular disease (CVD) in people living with HIV (PLWH) on integrase strand transfer inhibitors (INSTI). We examine the risk of CVD in PLWH on INSTI- based regimens. Methods: Using Truven Health Analytics MarketScan® databases for commercially insured and Medicaid covered adults, we identified PLWH newly initiated on ART between Jan 1, 2008 and Dec 30, 2015. New users were those without ART claims in the 6 months prior to study inclusion. The primary outcome, major adverse cardiac event (MACE), was a composite of acute MI, ischemic stroke, coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI) assessed through Dec 30, 2016. We excluded PLWH with MACE events 6 months prior to the first stable regimen start. We identified cardiac outcomes and covariates associated with risk of cardiac events using ICD-9-CM diagnosis and procedure codes and CPT-4 codes. Calendar-time specific inverse-probability-weighted Cox proportional hazards models were used to estimate hazard ratios (HR) and 95% confidence intervals (CI) for association between INSTI use and MACE. Propensity score models included potential predictors of CVD and INSTI use. Censoring occurred for the earliest of MACE events during the first 6 months of a stable regimen, 90 days post-ART switch, health plan disenrollment, death and study end. Results: 20,459 new ART initiators were identified. 5,128 (25%) PLWH initiated INSTI-based regimens (raltegravir 33%, elvitegravir 49%, dolutegravir 18%). 11,191 (55%) initiated non-nucleoside reverse transcriptase inhibitors and 4,145 (20%) protease inhibitors. Median duration of follow-up was 561 (348, 985) days. Mean age was 40.6 years, 79%were male, and 17%were Medicaid insured. Hypertension was present in 9.5% of INSTI users vs 7.4% non-users; lipid lowering treatment in 19.8% vs 17.9%; diabetes in 6% vs 4.8% and smoking in 13.5% vs 10.2%. 161 MACE events occurred; acute MI 11 (0.21%) vs 55 (0.36%), stroke 14 (0.27%) vs 48 (0.31), CABG 1 (0.02%) vs 6 (0.04%), PCI 5 (0.1%) vs 21 (0.14%) of INSTI users vs. non-users. INSTI-based ART was associated with significantly lower risk of MACE events (HR 0.57; 95% CI 0.45, 0.73) compared to non-INSTI based regimens. Conclusion: INSTI-based regimens were associated with a 43% decreased risk of CVD in this cohort. Validation of these findings in cohorts with longer follow up is needed. Jordan E. Lake 1 , Carlee Moser 2 , Maxine Olefsky 2 , Kristine M. Erlandson 3 , Ann Scherzinger 3 , James H. Stein 4 , Judith S. Currier 5 , Todd T. Brown 6 , Grace A. McComsey 7 1 University of Texas at Houston, Houston, TX, USA, 2 Harvard University, Cambridge, MA, USA, 3 University of Colorado Anschutz Medical Campus, Aurora, CO, USA, 4 University of Wisconsin–Madison, Madison, WI, USA, 5 University of California Los Angeles, Los Angeles, CA, USA, 6 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 7 Case Western Reserve University, Cleveland, OH, USA Background: Adipose tissue (AT) disturbances are common in people living with HIV (PLWH), and changes in AT quality may occur independently of changes in AT quantity. Decreases in AT density, a marker of AT quality, suggest disrupted adipocyte function and lipid accumulation. We previously reported that subcutaneous AT (SAT) density on computed tomography (CT) reflects biopsy-quantified SAT adipocyte size in PLWH, and that AT quantity increases on antiretroviral therapy (ART). In this exploratory analysis, we assessed changes in AT density after ART initiation and associations with immuno-metabolic parameters. Methods: ACTG A5257 randomized ART-naïve, adult PLWH to raltegravir (RAL) or ritonavir-boosted atazanavir (ATV/r) or darunavir (DRV/r), each with tenofovir disoproxil fumarate and emtricitabine, for 96 weeks. The subset with Weeks 0 and 96 (W0 and W96) abdominal CT scans and W96 HIV-1 RNA <50 copies/mL were included. Linear regression models compared W0, W96 and 96-week changes in SAT and visceral AT (VAT) density (in Hounsfield units, HU), adjusting for AT area and clinical/demographic parameters. Partial Spearman’s correlations adjusting for AT area assessed relationships between AT density and immuno-metabolic parameters. 681 CHANGES IN FAT DENSITY AFTER ART INITIATION

patients. Further studies are needed to determine whether alterations in gut integrity and microbiota with GB play a role in these improvements.

679 DOLUTEGRAVIR AND INSULIN RESISTANCE

Janet Lo 1 , James Oyee 2 , Melissa Crawford 2 , Richard Grove 2 , Ralph DeMasi 3 , Lloyd Curtis 2 , Anna Fettiplace 2 , Vani Vannappagari 3 , Nassrin Payvandi 4 , Michael Aboud 4 , Jean van Wyk 4 1 Massachusetts General Hospital, Boston, MA, USA, 2 GlaxoSmithKline, Uxbridge, UK, 3 ViiV Healthcare, Research Triangle Park, NC, USA, 4 ViiV Healthcare, Brentford, UK Background: HIV infection has been independently associated with insulin resistance (IR), potentially through chronic immune activation/inflammation, however this effect is not necessarily mitigated through successful antiretroviral therapy (ART). ART has been associated with IR through varying mechanisms, however, in the context of combination ART, increased obesity, and an aging HIV-infected population, these potential associations are difficult to interpret. We investigated potential risk factors associated with HOMA-IR (homeostasis model of assessment – insulin resistance) and the potential effect of dolutegravir (DTG) on IR over time. Methods: Data from 4 DTG clinical trials (SPRING-1, STRIIVING, SWORD-1 and -2) with fasting insulin and glucose measurements available, were included; subjects with diabetes were excluded. IR was determined by HOMA mathematical model and defined as a HOMA-IR value ≥ 2; additional cut-offs of 3 and 4 were also explored. Analysis of relationship between baseline (BL) risk factors and HOMA-IR was completed. Change in HOMA-IR over time and relative to controls were assessed with logistic regression and ANCOVA models, respectively. Results: HOMA-IR data was available at BL, week 24 and week 48 for 824, 304 and 543 DTG-exposed subjects and 713, 219 and 460 control subjects, respectively. At BL, subjects were mostly male (81%), white (76%) and had a median age of 43yrs; 50%were overweight/obese; 70% had a HOMA-IR>2. Results are shown in the table. There were similar modest increases in HOMA-IR between DTG and control groups over time (24 and 48 weeks). Overall, there was no difference in the odds of HOMA-IR>2 between treatment groups at 48 weeks. An association between BL HOMA-IR and increasing age, geographic region, increased BMI/weight, the presence of metabolic or cardiac disorders, lipids, and elevated liver function tests (ALT, ALP and albumin) was observed. Risk factors for IR (HOMA-IR>2) at week 48 were BL HOMA-IR, Sex, BMI, AIDS CDC category, smoking history, and elevated ALT. Conclusion: There was no association between treatment and insulin resistance observed in this analysis over a 48 week period, however IR modestly increased over time in all groups. In general, risk factors identified as being associated with IR at Week 48 were consistent with known risk factors for diabetes/IR. These results should be interpreted with caution as the studies were not primarily designed to assess effects of DTG exposure on insulin resistance.

Poster Abstracts

CROI 2019 260