CROI 2020 Abstract eBook

Abstract eBook

Poster Abstracts

baseline and week 24 were measured by mass spectrometry. Changes in log 10 lipid concentrations were compared by HIV serostatus using t-tests. P-values unadjusted for multiple comparisons (unadj-p) and Benjamini-Hochberg corrected (adj-p) are reported. Results: Among PWH and controls, at baseline, there were no statistically significant differences in concentrations of total lipids. With the exercise intervention, changes in total triacylglycerol (TAG) levels significantly differed among people with and without HIV (unadj-p=0.006, adj-p=0.078): TAGs tended to decrease in PWH (% Change: -4.5 [-14.1, 6.2]), but significantly increased in controls after 24 weeks of exercise (% Change: 14.7 [6, 24.1]). Concentrations of TAG species (Table) composed of long chain fatty acids increased among uninfected controls but not PWH (unadj-p=0.001-0.036, adj-p =0.10-0.12) from baseline to week 24. Total diacylglycerols (DAGs) increased in PWH from baseline to week 24 (% Change: 6.1 [0, 12.5]), but decreased in controls (% Change: -5.1 [-12.7, 3.2]) (unadj-p=0.03, adj-p=0.2). Baseline to week 24 changes in specific DAGs composed of palmitic acid (16:0), palmitoleic acid (16:1), and stearic acid (18:0) varied by serostatus, with increases in PWH (unadj-p=0.009-0.03; adj-p 0.10-0.12) and non-significant decreases in controls (Table). The change in concentrations of lysophosphatidylcholine (LPC) species composed of saturated fatty acids (LPC FA(15:0;16:0;17:0)) also differed by serostatus, with increases in PWH and decreases among controls (unadj-p=0.02-0.05; adj-p=0.12-0.21); Table. Conclusion: Although exploratory, the effects of exercise on the plasma lipidome may differ among people with and without HIV, potentially due to underlying alterations in lipid processing and fatty acid oxidation in PWH.

(R=0.29; P=0.01) and soluble CD163 (R=0.54; P=0.04) were also positively associated with FCP in treated and naïve, resp. FCP was inversely associated with CD4 (R=-0.24; P=0.02), but not with other HIV variables, nor age, sex, or race. Conclusion: Stool concentrations of FCP are elevated in PWH. ART appears to reduce FCP but not to concentrations seen in uninfected controls. FCP concentrations are positively correlated with several markers of systemic inflammation/immune activation, and negatively with CD4. FCP may serve as a useful biomarker to monitor gastrointestinal inflammation and associated systemic inflammation/immune activation in HIV.

Poster Abstracts

240 LONG-TERM ELEVATED IL-6 AND D-DIMER AFTER DELAYED ART INITIATION IN THE START TRIAL

Jason V. Baker 1 , Birgit Grund 2 , Shweta Sharma 2 , Keith Henry 1 , Jennifer Hoy 3 , Stefan Esser 4 , Silvia Nozza 5 , Kiat Ruxrungtham 6 , Adam Rupert 7 , H. Clifford Lane 8 , Jens D. Lundgren 9 , James Neaton 2 , for the INSIGHT START Study Group 1 Hennepin Healthcare Research Institute, Minneapolis, MN, USA, 2 University of Minnesota, Minneapolis, MN, USA, 3 Monash University, Melbourne, VIC, Australia, 4 University Hospital Essen, Essen, Germany, 5 San Raffaele Scientific Institute, Milan, Italy, 6 Chulalongkorn University, Bangkok, Thailand, 7 Leidos Biomedical Research, Inc, Frederick, MD, USA, 8 NIH, Bethesda, MD, USA, 9 University of Copenhagen, Copenhagen, Denmark Background: Inflammation and coagulation are associated with disease risk among persons with HIV. ART reduces inflammation, but whether time of initiation during infection affects this reduction has not been studied experimentally. We report on the interleukin-6 (IL-6) and D-dimer trajectories in the immediate versus deferred arms of the START trial Methods: In participants randomized to immediate (CD4>500 cells/µL) vs. deferred (CD4 <350 cells/µL) ART initiation, IL-6 and D-dimer levels were measured from stored plasma specimens at baseline, month 8, and annually up to 7 years. Mean change from entry and from start of ART in log2-transformed levels were compared between the deferred versus immediate groups using longitudinal mixed models adjusted for age, sex, geographic region, baseline biomarker levels and visit. Results were presented as percent change Results: Among 2209 participants (median age 36 years, 20% female, 67% enrolled in high-income countries), the median levels at entry of IL-6 were 1.47 pg/mL, D-dimer 0.31 µg/mL, and CD4 counts 649 cells/µL. In the immediate group, 94-97% had viral load <200 cp/mL at all annual visits, whereas the deferred group suppression rates increased over time: 18%, 61%, 89%, and 95% at years 1, 3, 5, 7, respectively. In the deferred group, IL-6 and D-dimer levels remained significantly higher than the immediate group through 5 years (Fig). Over the follow-up period, treatment difference in IL-6 was 10.3% (95%CI: 7.6 to 12.9, p<0.001), and D-dimer 14.0% (95%CI: 11.5 to 16.5, p<0.001). When comparing treatment groups based on the time from ART start, biomarker levels were higher in the deferred compared to the immediate group over at least the first 2 years of ART. At 2 years on ART est. diff. 9.9% (95% CI: 4.0 to 15.8; p<0.001) for IL-6 and 10.0% (95% CI: 6.4 to 13.6, p<0.001) for D-dimer, and >96% in each group had HIV RNA <200 cp/mL Conclusion: Compared to immediate ART, deferral of ART was associated with higher levels of IL-6 and D-dimer over at least 5 years. During the first 2 years of ART treatment, despite viral suppression in both groups, biomarker levels

239 FECAL CALPROTECTIN IS ELEVATED IN HIV AND RELATED TO SYSTEMIC INFLAMMATION Allison Ross Eckard 1 , Nancy L. Hagood 1 , Heather Y. Hughes 1 , Mary Ann O'Riordan 2 , Danielle Labbato 2 , Sarah E. Scott 2 , Grace A. McComsey 2 1 Medical University of South Carolina, Charleston, SC, USA, 2 University Hospitals Cleveland Medical Center, Cleveland, OH, USA Background: Fecal calprotectin (FCP), a biomarker of gastrointestinal inflammation, is used in the diagnosis and management of inflammatory bowel disease. HIV infection severely damages gut-associated lymphoid and epithelial tissues leading to gut inflammation, microbial translocation and systemic inflammation/immune activation. We sought to investigate FCP in people with HIV (PWH) for the first time and determine its relationship to HIV-specific factors and systemic inflammation/immune activation. Methods: PWH naïve to ART, ART-treated and uninfected controls were prospectively enrolled. Stool samples were collected and FCP was measured by ELISA. Plasma biomarkers of inflammation/immune activation were also measured. FCP was evaluated as a continuous variable and by thresholds. Spearman correlations were used to investigate associations with FCP. Results: 101 PWH (83 ART-treated, 18 naïve) and 89 uninfected controls were enrolled. ART-treated were older than naïve (51 vs 31 yrs; P=0.006), but sex and race were similar (overall 78%males, 66% blacks). All but one ART-treated had HIV RNA <200 copies/mL. CD4 counts for treated and naïve were 683 and 410 cells/µL, resp. Controls had a median age of 37 yrs (78%males, 22% blacks). There was a difference (P=0.001) in FCP among the 3 groups with the highest median (25th, 75th %ile) FCP in ART-naïve [144 (33, 262) µg/g] followed by ART- treated [78 (36, 141) µg/g] and then controls [41 (21, 89) µg/g] (Fig). 56% of ART- naïve had FCP >100 µg/g vs 37% in treated and 19% in controls (P=0.0003). In PWH, high-sensitivity C-reactive protein (R=0.30; P=0.008), soluble tumor necrosis factor-II (R=0.28; P=0.006), and soluble vascular cellular adhesion molecule (R=0.21; P=0.04) were positively associated with FCP. Interleukin-6

81

CROI 2020