CROI 2024 Abstract eBook

Abstract eBook

Poster Abstracts

ICPs may play a protective role against β cell stress in young CPHIV. The figure, table, or graphic for this abstract has been removed. Metabolomic Signature in Youth With Perinatally Acquired HIV and Non-Alcoholic Fatty Liver Disease Silvia Chafino 1 , Laura Tarancón-Diez 2 , Jara Hurtado-Gallego 3 , Sonia Alcolea 3 , Antonio Olveira 3 , María Luisa Navarro 2 , Salvador Fernández-Arroyo 4 , Consuelo Viladés 1 , María Luisa Montes 5 , Francesc Vidal 4 , Joaquim Peraire 1 , Anna Rull 1 , Talia Sainz 5 1 Hospital Universitario de Tarragona Joan XXIII, Tarragona, Spain, 2 Hospital General Universitario Gregorio Marañón, Madrid, Spain, 3 La Paz University Hospital, Madrid, Spain, 4 Rovira i Virgili University, Tarragona, Spain, 5 Hospital La Paz Institute for Health Research, Madrid, Spain Background: Non-alcoholic fatty liver disease (NAFLD) is characterized by accumulated fat producing hepatocellular inflammation and injury. Persistent immune activation and chronic inflammation in response to HIV infection may be factors underlying the development of NAFLD which is recognized as a cause of liver disease although currently diagnosis and management are a challenge. Specifically, metabolomic changes and biomarkers associated with this pathology are poorly studied in perinatally acquired HIV people (PHIV). Methods: Plasmatic lipidomic and bile acids were analysed by LC-QTOF and HPLC-MS/MS in the study cohort consisted of 29 youth living with HIV under viral suppression by ART, of whom 10 presented NAFLD and 19 did not (control). Shear wave ultrasound and/or controlled attenuation parameter (CAP) ≥248 dB/m were used to determine NAFLD condition Results: The PHIV study cohort was composed mainly of women with a median age of 18 (13.5-24) years. Body mass index was higher in NAFLD group (23.25 Kg/m 2 ) than control (19 Kg/m 2 ) (P=0.051). Interestingly, 30% of NAFLD patients was overweight and 10% presented obesity, whereas only 5.3% was overweight in the control. The hepatic steatosis index (HSI) was significantly increased in the NAFLD group (33.95 (28.475-39.72)) compared to control (27.95 (26.84-32.05)) (P=0.017). Circulating concentrations of ursodeoxycholic acid (UDCA) and eight different lipid species (diglyceride (DG) 36:3 and triglycerides (TG) 53:4, TG54:4, TG54:5, TG52:4, TG52:5, TG52:3, and TG56:7) were significantly increased in PHIV presenting NAFLD. Random forest analyses identified UDCA, TG56:7, and TG54:5 as the plasmatic metabolites that better differentiate both groups although the TG56:7 obtained the best discriminatory power to identify NAFLD with an AUC of 0.898. Additionally, regression model combining UDCA, TG56:7, and TG54:5 with the clinical HSI parameter increased the discriminatory power of the model up to an AUC of 0.920. Of interest, principal compound analysis based on bile acid levels identified two possible subpopulations within the control group. One subpopulation was differentiated from the NAFLD group (n=6) and the other one (n=13) showed an intermediate profile, suggesting that the biomarker panel could identify patients at risk for progression to NAFLD Conclusion: A panel containing UDCA, TG56:7 and TG54:5 in combination with HSI could be a good predictive biomarker to identify individuals at risk of NAFLD among PHIV Low Bone Density Accrual and Increased Inflammation in Zimbabwean Adolescents With Perinatal HIV Lisha Jeena 1 , Victoria Simms 2 , Rukuni Ruramayi 2 , Andrea M. Rehman 2 , Celia Gregson 3 , Rashida A. Ferrand 2 , Sarah Rowland-Jones 1 , Anthony Hsieh 1 1 University of Oxford, Oxford, United Kingdom, 2 London School of Hygiene & Tropical Medicine, London, United Kingdom, 3 London School of Hygiene & Tropical Medicine, Bristol, United Kingdom Background: Living with HIV is associated bone density deficits, increasing fracture risk. We investigated the effect of HIV on bone density accrual over one year and inflammatory biomarkers relevant to osteoclast formation in peripubertal children and adolescents with and without perinatal HIV. Methods: From 2018 to 2020, clinically healthy children with HIV (CWH) on antiretroviral treatment (ART) for at least 2 years, aged 8-16 years, were recruited from HIV clinics as were children of similar ages without HIV (CWOH) from nearby schools in Harare, Zimbabwe. Dual X-ray absorptiometry was performed at baseline and at 12months quantifying height-adjusted total-body less-head bone mineral content for lean mass (TBLH-BMC) and size-adjusted lumbar spine bone mineral apparent density (LS-BMAD) Z-scores. Change in Z-scores accounted for follow-up time. Baseline plasma levels for CRP, sCD14, TNFα, IL-6, IL-17α, IL-18, IL-10 and IFNλ were measured using Luminex. Bone outcome measurements were univariately plotted with age, with linear regression used to compare longitudinal changes in bone density by HIV status, adjusting for age, pubertal stage and baseline bone density. Principal components analysis was used to group inflammatory biomarkers. Linear

cumulative CD4, C-reactive protein [CRP], and Interleukin-6 [IL-6] with DNA methylation were adjusted for ART type and age at ART initiation. Results: Overall, median age was 11 and 17 years at timepoints 1 and 2, respectively. Groups were balanced by sex (51% male) and race/ethnicity (64% non-Hispanic Black). Genome-wide, there were no DM sites comparing YPHIV to YPHEU at either timepoint. For targeted genes, 1 CpG site on SPERT and 6 CpG sites on PSMB9 were DM at timepoint 1 comparing YPHIV to YPHEU. At timepoint 2, 10 CpG sites on PSMB9 and 4 CpG sites on EBF4 were DM comparing YPHIV to YPHEU. Among YPHIV, cumulative VL or CD4 count were not associated with any CpG sites. CRP was associated with 27 CpG sites (16 genes) at timepoint 1 and 19 CpG sites (15 genes) at timepoint 2, with 11 genes overlapping timepoints (Fig 1). IL-6 was associated with 6 CpG sites (4 genes) at timepoint 1 and 2 CpG sites (2 genes) at timepoint 2, with 2 genes overlapping timepoints (Fig 1). Conclusion: Associations between markers of inflammation and epigenetic signatures among genes involved in blood pressure regulation, chronic kidney disease, and fatty acid processing were detected in YPHIV at two timepoints in childhood/adolescence. These genes may provide insight into inflammatory pathways contributing to HIV-associated chronic comorbidities among YPHIV.

972

Poster Abstracts

971

Young CPHIV Have Low Proinsulin to C-Peptide Ratios That Inversely Correlate With T-Cell Exhaustion Wei Li 1 , Emily Sims Emily Sims 1 , Mussa Mwamzuka 32 , Fatma Marshed 2 , Aabid Ahmed 2 , Alka Khaitan 1 1 Indiana University School of Medicine, Indianapolis, IN, USA, 2 Bomu Hospital, Mombasa, Kenya Background: Persons with HIV have an increased risk of developing diabetes mellitus. The ratio of immature (Pro-insulin) relative to mature (C-peptide) insulin products (PI:C) in circulation serves as a biomarker of pancreatic β cell stress and predicts progression to type 1 or type 2 diabetes. In adults with HIV lower PI:C was observed in untreated persons compared to those on treatment or without HIV, suggesting immune dysregulation may preserve β cell function. Children with perinatal HIV (CPHIV) have higher rates of insulin resistance compared with healthy children, but whether β cell dysfunction contributes is unknown. We investigated PI:C ratios in CPHIV and their correlations with clinical and immune activation and exhaustion markers. Methods: We quantified plasma levels of proinsulin (TECO intact proinsulin ELISA) and C-peptide (TOSOH immunoassay) in 200 Kenyan children who were HIV unexposed (HU) and CPHIV who were treatment naïve (ART-) or virally suppressed on treatment (ART+) aged 0-5 years ("0-5y" n=28 ART-, 21 ART+, 36 HU) and 5-20 years ("5-20y" n=41 ART-, 28 ART+, 46 HU). We calculated PI:C and assessed correlations with HIV viral load, CD4%, CD4:CD8, monocyte (sCD14, sCD163), T cell (CD38+HLA-DR+), and systemic (CRP, IL-6) activation markers and immune checkpoints (ICPs: PD-1, CD160, TIM3) on memory T cells. Mann Whitney and Spearman's correlations were performed on GraphPad Prism. Results: Compared to HU, 0-5y ART- had lower proinsulin levels (p=0.005) and PI:C (p=0.004), whereas ART+ trended toward lower PI:C (p=0.06). In 5-20y, both ART- and ART+ had a higher C-peptide level (p=0.006 and p=0.04), and ART+ had lower PI:C (p=0.04). In both age groups in CPHIV, PI:C did not correlate with age, viral load, CD4 levels, or activation markers. However, in 0-5y CPHIV PI:C inversely correlated with PD-1 (p=0.05, r=-0.46) and TIM3 (p=0.01, r=-0.40) on memory CD4 T cells as well as PD-1 (p=0.03, r=-0.51) and CD160 (p=0.04, r=-0.32) on memory CD8 T cells. There were no significant correlations between PI:C with ICPs in older CPHIV. Conclusion: CPHIV had a lower PI:C compared with HU, stemming from lower proinsulin in younger CPHIV and higher C-peptide in older CPHIV. Clinical markers of HIV disease progression and inflammation were not associated with β cell stress in CPHIV. T cell exhaustion correlated with PI:C in younger CPHIV. Overall, our data show no evidence of β cell dysfunction in CPHIV and suggest

973

CROI 2024 309