CROI 2024 Abstract eBook

Abstract eBook

Poster Abstracts

using PRSice. Regression analysis (adjusted for age, sex, and race) was used to examine relationships between PRS and cognitive test and domain Z-scores, inflammatory markers (sCD14, sCD163), brain volumes, cerebral blood flow (CBF), and HIV variables (CD4 t-cell count, CD8 t-cell count, CD4:CD8 ratio, duration of infection). Results: Higher AD-PRS values were associated with worse cognitive test scores for letter number sequencing (p=0.015), animal fluency (p=0.039), verb fluency (p=0.001) and language domain (p=0.024). There was a negative correlation between AD-PRS and inflammatory markers sCD163 (p=0.043) and CD14 (p=0.03), and AD-PRS and total gray volume (p=0.032). A significant positive association between the AD-PRS and CBF in the caudate (p=0.047) and putamen volume (p=0.017) was found. No association was found between AD-PRS and HIV variables. Conclusion: While no change in brain integrity was observed within cognitive domains or brain regions typically affected by AD, an elevated genetic risk measured by AD-PRS did associate with some worse cognitive performance. This mirrors patterns in the general population, implying that certain cognitive deficits among older PWH may result from aging rather than HIV-related factors. AD-PRS also had no association with HIV clinical markers, supporting previous work suggesting that HIV and AD affect brain integrity via different processes. Further Evidence of Hypertrophy in Acute Infection Identified Using Machine Learning Jacob Bolzenius 1 , Napapon Sailasuta 2 , Phillip Chan 3 , Carlo P. Sacdalan 4 , Julie Ake 5 , Somchai Sriplienchan 4 , Lydie Trautmann 6 , Sandhya Vasan 6 , Trevor A. Crowell 7 , Ferron F. Ocampo 4 , Serena S. Spudich 3 , Victor Valcour 8 , Kilian Pohl 9 , Robert Paul 1 , for the SEARCH010/RV254 Study Team 1 University of Missouri St Louis, St Louis, MO, USA, 2 University of Hawaii, Honolulu, HI, USA, 3 Yale University, New Haven, CT, USA, 4 SEARCH, Bangkok, Thailand, 5 Walter Reed Army Institute of Research, Silver Spring, MD, USA, 6 US Military HIV Research Program, Silver Spring, MD, USA, 7 Henry M Jackson Foundation, Bethesda, MD, USA, 8 University of California San Francisco, San Francisco, CA, USA, 9 Stanford University, Stanford, CA, USA Background: A recent study of structural brain volumes during acute HIV infection (AHI) revealed evidence of hypertrophy among individuals in Fiebig stages III-V compared to Fiebig I-II. The present study incorporated multiple neuroimaging modalities (structure and function), blood markers of inflammation and immune dysregulation, and psychosocial determinants of health using ensemble machine learning to more comprehensively characterize brain integrity in late vs. early AHI and discover potential explanatory factors. Methods: 112 Thai males with AHI (age 20-46) enrolled in RV254/SEARCH010 were included in the analysis. Features classifying individuals into early (Fiebig I-II; n=32) vs. late (Fiebig III-V; n=80) AHI were ranked using gradient-boosted multivariate regression (GBM) with repeated cross validation. Predictors included demographic, HIV disease, cognitive, neurologic exam, mental health, substance use, plasma immune/inflammatory, and multimodal 3T neuroimaging (i.e., volumes, resting state connectivity, diffusion, and spectroscopy) indices collected at enrollment. Results: The GBM identified a combination of variables that classified individuals into early vs. late Fiebig with an average F1 score (a metric of model performance) of .82. Individuals in late Fiebig stages exhibited larger regional volumes in the olfactory cortex, putamen, raphe nucleus, and pallidum. Smaller volumes among late Fiebig individuals were noted in the ventral tegmental area, along with lower blood neutrophil percent and high-density lipoproteins, greater functional connectivity between the right executive control network and visuospatial networks, and lower use of erectile dysfunction medication. Conclusion: Findings demonstrate that brain alterations in late vs. early stages of AHI associate with distinct multidimensional parameters: larger regional volumes observed together with lower immune levels (i.e., neutrophils) suggest that viral-immune dysregulation already underway in late Fiebig stages manifests as hypertrophy via neuroinflammation. Additionally, we reveal evidence of brain-behavior relationships via recent substance use and ventral tegmental area volume, a region implicated in addictive behaviors. Resting state connectivity differences may also reflect disrupted structural/functional integration. Psychosocial and immune processes operative during the first weeks of AHI and their effects on brain integrity may explain variability in long term outcomes for those starting ART at later stages.

591

Beta-Amyloid PET Positivity Among Cognitively-Impaired People With HIV Over Age 60 Samuel Wilson 1 , Andjelika Milicic 2 , Shireen Javandel 2 , Claire Yballa 2 , Benedetta Milanini 3 , Kilian Pohl 4 , Robert Paul 5 , Victor Valcour 2 1 Thomas Jefferson University, Philadelphia, PA, USA, 2 University of California San Francisco, San Francisco, CA, USA, 3 Inovigate GmbH, Basel, Switzerland, 4 Stanford University, Stanford, CA, USA, 5 University of Missouri St Louis, St Louis, MO, USA Background: Amyloid positron emission tomography (PET) imaging can stratify risk for Alzheimer's disease (AD). Limited data exist among people with HIV (PWH) and HIV-associated neurocognitive disorder (HAND) in older age groups. In this analysis, we sought to characterize frequency of amyloid PET positivity among older cognitively impaired PWH compared to cognitively healthy people without HIV (controls). Methods: Virally suppressed PWH were enrolled in a study of HAND where PET positivity was used to exclude the possibility of AD (exclusion criteria). Participants underwent a standardized neuropsychological battery to diagnose HAND. Healthy controls were identified and matched by age from a separate cohort at our site. No participant from either group showed clinical signs and symptoms in a pattern that would be concerning for AD. We completed amyloid PET imaging using florbetapir F-18 among both groups which were visually read as amyloid positive (PET+) or negative (PET-). Results: Compared to controls (n=65), the PWH group (n=74) was predominantly male (94.6% vs. 72.3%, p<0.001), of non-Hispanic white ethnicity (74.3% vs. 83.1%, p=0.211) and reported lower educational attainment (16.2 vs. 17.4 years, p<0.001). Among them, 6 (8.1%) had PET+ scans compared to 14 of 65 matched healthy controls (21.5 %, p=0.024). Within the PWH group, we did not identify differences in the neuropsychological testing pattern by PET status (all p-values >0.05). Conclusion: Relative to healthy controls, this group of cognitively impaired PWH did not show increased frequency of amyloid PET+ compared to controls nor differences in patterns of cognitive performance. Differential Expression Analysis Reveals Pervasive Transcriptional Changes in the Brains of PWH Kriti Agrawal, Jay S Stanley , Junchen Yang, Nicholas C. Jacobs, Haowei Wang, Le Zhang, Mark Gerstein, Yuval Kluger, Serena Spudich, for Yale Single Cell Opioid Responses in the Context of HIV (Y-SCORCH) Yale University, New Haven, CT, USA Background: Central nervous system (CNS) dysfunction is a common consequence in people with HIV (PWH). Understanding of the impact of HIV on the brain remains limited to select cell types and regions. We describe preliminary efforts to establish cell-type specific gene expression patterns across multiple regions in PWH using single cell transcriptomics. Methods: Nuclei were isolated from frozen post-mortem tissue from insular cortex (INS), prefrontal cortex (PFC), and ventral striatum (VST) from 6 PWH and 6 HIV-negative control (CTR) donors. We profiled transcriptomes and chromatin accessibility profiles of these 36 samples using 10x Genomics Chromium Single Cell Multiome ATAC+Gene Expression. In total, we analyzed single-cell transcriptomes of 144,918 cells from PWH and 153,868 cells from CTR. An integrated single cell t-SNE embedding was created for each region, and cells were clustered using Louvain. These clusters were then manually annotated to identify major cell types. We also performed pseudobulk differential expression (DE) analysis between PWH and CTR within major cell types and regions using EdgeR. Results: PWH and CTR donors were similar in age (median 45 and 39 years) and gender (33% vs 17% female) (p>.1). Canonical neuronal and non-neuronal cell types were detected in all regions. However, a population of putative T-like cells expressing CD247 and SKAP1 was identified in INS of 6 CTR and 5 HIV donors, with >80% of cells annotated as T-like originating from PWH. Consistent with their role as resident immune cells of the CNS, microglia exhibited the

590

Poster Abstracts

592

CROI 2024 163