CROI 2019 Abstract eBook

Abstract eBook

Poster Abstracts

Results: 119 PLWH and 55 CTL were included in the study (PLWH age [mean±SD]:56±8; education:13±3; sex:81%male; CTL age:56±12; education:14±2; sex:51%male). PLWH had smaller brain volumes and poorer cognitive performance compared to the CTL group. Total WMH load and factors commonly linked with vascular disease were similar between the two groups. Older age and hypertension were significantly associated with greater WMH loads for all participants (Fig. 1A). Higher WMH load was significantly associated with reduced brain volumes and cortical thickness and worse cognitive function in all participants, independent of HIV status (Fig. 1B-D). Conclusion: We observed that the PLWH in this study did not have greater WMH load. However, WMH load was associated with reduced brain volumes and poorer cognition in the entire sample. These findings suggest that CSVD could explain some of the brain atrophy and cognitive impairment found in people living with HIV.

retained in a logistic regression model (stepwise forward selection, best model fit by Akaike information criterion (AIC)). Results: CSF NFL (r=-.53, p<.02), non-AIDS comorbidities (r=.48, p<.03), nadir CD4 (r=.41, p=.05), age (r=-.42, p<.05), and NCI (r=-.39, p=.07) were associated with greater brain injury (lower CNM). Non-AIDS comorbidities remained a significant predictor (p<.03) yielding the best model fit (AIC=111.17). Conclusion: Composite 1H MRS signal identifies currently active brain injury, well below the threshold for NCI, in regions known to be associated with HIV- related brain injury and pathological aging. This injury is dominantly driven by non-AIDS co-morbidities and expressed through an HIV-related pathway, implying that HIV is the driver of the co-morbidities.

Poster Abstracts

454 1H MRS IDENTIFIES SUBCLINICAL NEURONAL INJURY DESPITE CHRONIC VIRAL SUPPRESSION Lucette A. Cysique 1 , Thomas Gates 2 , Lauriane Juge 1 , Gemma Howdle 1 , Tory Johnson 3 , Avindra Nath 4 , Caroline Rae 1 , Bruce J. Brew 5 , for the CHIEF group 1 Neuroscience Research Australia, Randwick, NSW, Australia, 2 St. Vincent’s Hospital, Sydney, NSW, Australia, 3 Johns Hopkins University School of Medicine, Baltimore, MD, USA, 4 NIH, Bethesda, MD, USA, 5 University of New South Wales, Sydney, NSW, Australia Background: The significance of proton Magnetic Resonance Spectroscopy (1H MRS) abnormalities in chronic suppressed HIV infection is unclear. Previous studies have included unsuppressed patients and did not relate findings to markers of current neuronal injury such as neurofilament light-chain (NFL), nor to non-AIDS comorbidities. We hypothesized that 1H MRS would identify active brain injury. Methods: 22 HIV+men (aged 48.7±12.3) with plasma and CSF viral suppression (<20cp/mL) underwent 1H MRS scanning to assess in vivo brain injury in the frontal white matter (FWM), posterior cingulate cortex (PCC), and basal ganglia (BG). Brain metabolite concentrations for N-Acetyl-Aspartate (NAA), Choline (Cho), Creatine (Cr), Glutamate (Glu) and myo-Inositol (MI) were quantified in jMRUI and referenced to H2O. As MRS data are amenable to data reduction techniques to yield a single robust component, we extracted a composite neurochemical in vivo marker – “CNM” (more negative values indicate greater brain injury; Fig. 1). Participants also completed neuropsychological testing and lumbar puncture to assess CNM’s potential as a marker of active brain injury. Besides NFL, CSF biomarkers included neopterin, CCL2, and CSF tat. Neurocognitive impairment (NCI) was classified using standard criteria (37.5%, none demented). Univariate correlations with CNMwere tested for CSF biomarkers, HIV disease markers, demographics, neurocognition, psychiatric and alcohol/drug use comorbidities, current psychological distress, and non-AIDS comorbidities (cardiovascular/kidney diseases, sleep disorders, malignancies, neuropathy/pain, and loss of consciousness>30min from non-traumatic causes). Predictors at p<0.10 were

455 INFLAMMATORY PLASMA BIOMARKERS CORRELATE WITH DIFFUSION TENSOR IMAGING IN CHRONIC HIV Kevin C. Chang 1 , Thomas Premeaux 2 , Yann Cobigo 1 , Benedetta Milanini 1 , Joanna Hellmuth 1 , Shireen Javandel 1 , Lishomwa C. Ndhlovu 2 , Robert Paul 3 , Victor Valcour 1 1 University of California San Francisco, San Francisco, CA, USA, 2 University of Hawaii at Manoa, Honolulu, HI, USA, 3 University of Missouri St Louis, St Louis, MO, USA Background: Despite adherence to combination antiretroviral therapy (cART) and suppression of plasma viral RNA, a large proportion of people living with HIV experience cognitive symptoms. This study investigated correlations between plasma inflammatory biomarkers and diffusion tensor imaging (DTI) measures of brain white matter injury. Methods: Participants underwent neuropsychological testing, blood draw, and DTI scans on one of two separate 3 Tesla MRI scanners. Plasma biomarkers, sCD163, sCD14, neopterin, IP-10, and MCP-1, were quantified by Luminex/ ELISA. DTI metrics fractional anisotropy (FA) and mean diffusivity (MD) were averaged across a priori defined regions known to be affected by HIV – the corpus callosum (CC), corona radiata (CR), and superior longitudinal fasciculus (SLF). These were regressed against biomarker levels and the results of global, executive domain, and attention domain neuropsychological testing, reported as z-scores relative to standard norms (NPZ), controlling for age, duration of infection, and scanner model. Voxelwise analysis by Tract-Based Spatial Statistics (TBSS) compared FA and MD to biomarker levels, controlling for age and duration of infection. Results: 43 HIV+ participants (median age 64 [IQR 62-66] years, 91%male) enrolled, all of whomwere on cART with suppressed plasma HIV RNA and self- reported cognitive symptoms. 38 met criteria for HIV-associated neurocognitive

CROI 2019 166