CROI 2015 Program and Abstracts

Abstract Listing

Poster Abstracts

LS, TH, and FN BMD than HIV- women. After adjustment for demographic, behavioral, body composition and metabolic factors, HIV+ status was associated with lower BMD at the TH, (-0.047 g/cm 2 ), FN (-0.048 g/cm 2 ), and LS (-0.077 g/cm 2 ) (all p<.05). There was little change in the effects of HIV on BMD after additional adjustment for adiponectin levels; HIV remained associated with lower BMD at the TH (-0.045 g/cm 2 ), FN (-0.048 g/cm 2 ) and LS (-0.078 g/cm 2 ) (all results p<0.01). Similar findings were observed after additional adjustment for leptin. Among HIV+ women, in unadjusted analyses, adiponectin was associated with lower TH BMD (-0.025 g/cm 2 per 10-fold adiponectin increase, p=0.04), whereas leptin was associated with higher BMD at FN (+0.027 g/cm 2 per 10-fold leptin increase, p=0.005) and TH (+0.019 g/cm 2 , p=0.03). After multivariable adjustment, neither adiponectin nor leptin were associated with BMD at any site (all p>0.10). Conclusions: HIV infection is associated with lower TH, FN, and LS BMD among women. Serum leptin and adiponectin levels do not appear to mediate the association of HIV infection with loss of BMD, and appear to have little association with BMD among HIV infected women. 771 Long-Term Changes in Bone Mineral Density and Insulin Resistance on Statins in HIV Kristine M. Erlandson 1 ;Ying Jiang 2 ; Sara M. Debanne 2 ; Grace A. McComsey 2 1 University of Colorado, Aurora, CO, US; 2 Case Western Reserve University, Cleveland, OH, US Background: We previously demonstrated increased hip bone mineral density (BMD) and worsening in insulin resistance after 48 weeks of rosuvastatin in HIV-infected adults on antiretroviral therapy (ART). Here, we present the final 96 week results. Methods: The Stopping Atherosclerosis and Treating Unhealthy bone with RosuvastatiN in HIV (SATURN-HIV) trial is a randomized, double-blind, placebo-controlled trial to evaluate the effect of rosuvastatin on markers of cardiovascular risk and skeletal health in HIV over 96 weeks. Subjects were on stable ART with HIV-1 RNA <1000 copies/mL, LDL-cholesterol <130 mg/dL, and immune activation (CD8+CD38+HLA-DR+>19% and/or hsCRP >2 mg/L). Subjects were randomized 1:1 to daily rosuvastatin 10 mg or matching placebo, and stratified by protease inhibitor use and presence/absence of osteopenia. Bone primary endpoint was DXA-measured changes in total hip BMD; secondary endpoint was change in homeostatic model assessment of insulin resistance (HOMA-IR), calculated as fasting [insulin (uIU/mL)*glucose (mg/dL)]/450. Primary analyses were intent-to-treat. Statistical tests included multivariable linear regression. Results: 147 subjects enrolled;78%male, 70% black; median age 47 years and CD4 count 613 cells/ m L; 78% had HIV-1 RNA<50 copies/mL. At baseline, study arms were well balanced; 23%were osteopenic at the hip and 22% at the spine. No participants had diabetes, but 18% in statin and 15% in placebo had fasting glucose >100 mg/dL. At week 96, no significant differences in hip BMD (p=0.52) or spine BMD (p=0.96) were seen between study arms. Although changes in fasting glucose by week 96 were not significantly different between the rosuvastatin and placebo (p=0.14), 20 (32%) in the rosuvastatin and 12 (5%) in the placebo arm had glucose >100 mg/dL and 1 (2%) in the rosuvastatin and 2 (3%) in the placebo arm had fasting glucose >125 mg/dL. The rosuvastatin arm had significantly greater 96 week increases in insulin (mean 49.8 [SE 86.4]% versus 20.2 [138.4]%; p=0.009) and in HOMA-IR (72.6 [109.9]% vs. 33.0 [158.3]%; p=0.009). In multivariate models, 96 week changes in HOMA-IR were associated with statin use but not age, race, family history of diabetes, hepatitis C, or change in body composition. Conclusions: Despite an initial beneficial effect of rosuvastatin on hip BMD, no longer term benefit was observed. The detrimental impact on insulin resistance persisted. Further data is needed to determine the risk/benefit of statins in HIV-infected persons. 772 Immunologic Predictors of Bone Loss in a Contemporary HIV Cohort Edgar T. Overton 1 ; Katherine H. Hullsiek 2 ; Gerome Escota 3 ; Kenneth A. Lichtenstein 4 ; Lois Conley 5 ; Pragna Patel 5 ; JohnT. Brooks 5 ; Irini Sereti 6 ; JasonV. Baker 2 the CDC SUN (Study to Understand the Natural History of HIV/AIDS in the Era of EffectiveTherapy) Investigators 1 University of Alabama at Birmingham, Birmingham, AL, US; 2 University of Minnesota, Minneapolis, MN, US; 3 Washington University School of Medicine, St. Louis, MO, US; 4 National Jewish Health, Denver, CO, US; 5 Division of HIV/AIDS Prevention, Atlanta, GA, US; 6 National Institute of Allergy and Infectious Disease (NIAID), Bethesda, MD, US Background: HIV is a recognized risk factor for osteoporosis. The role of soluble markers and cellular immune parameters on HIV-related bone loss remains unclear. Methods: We evaluated inflammatory markers and total hip bone mineral density (BMD) changes in a prospective HIV cohort enrolled during 2004-2006 (SUN Study). We abstracted clinical data & measured BMD by dual-energy X-ray absorptiometry (DXA) at baseline & annually for 4 years. We characterized soluble inflammatory biomarkers & cellular immunophenotypes by multi-color ow cytometry of baseline cryopreserved peripheral blood mononuclear cells, assessed the correlation between baseline BMD & immune parameters, and identified predictors of percent change in BMD from baseline by longitudinal regression after adjusting for traditional & HIV-related risk factors. Results: Baseline characteristics (n=637): median age 41 yrs, 77%male, median CD4 ct 471c/mm 3 , and 74% plasma HIV RNA level <400 cp/mL. There were 554 subjects with 1887 yrs of follow up (median 4 yrs) and a median of 4 DXAs per participant. At baseline and final DXA, 387 (61%) and 393 (62%) had low BMD (T score <-1.0), respectively. sCD14, but not IL-6, was inversely correlated with total hip BMD at baseline (p<0.01 & 0.08, respectively). No T-cell phenotypes correlated with baseline hip BMD while several monocyte phenotypes correlated either positively (CX3CR1+; CD14 dim CD16+; CD14 var CD16+) or negatively (CCR2+; CD14+CD16-) with baseline hip BMD (p<0.01 for all). Associations between baseline soluble biomarkers and frequencies of immunophenotypes with changes in total hip BMD were adjusted for traditional (model 1) and HIV-related risk factors (model 2) (see table). Having a higher prevalence of memory CD28+ CD4+ and CD8+ T-cells was associated with increasing total hip BMD (p<0.001 for both). IL-6, replicatively senescent CD4 T-cells, and monocytes expressing CCR5+ or tissue factor were associated with total hip BMD decline (p=0.01, 0.03, & 0.05, respectively). For T-cell activation phenotypes and other monocyte phenotypes, no apparent association with change in total hip BMD was identified.

Poster Abstracts

472

CROI 2015