Susan Morgello

People with HIV (PWH) experience higher rates of depression compared with the general population. Inflammation has been associated with depressive symptoms and may be associated with a subtype of depression, but the relationship between inflammation and depressive symptoms in PWH on antiretroviral therapy (ART) is unclear. In this study, inflammation biomarkers (IFN-γ, IL-1β, IL-6, IL-8, IL-12p70, IL-15, IP-10, MCP-1, VEGF, CRP) were measured in plasma from 195 PWH on ART and depressive symptoms were assessed using Beck Depression Inventory-II (BDI-II) scores. Logistic regression and mixed-effects models were used to examine the associations between inflammation biomarkers and depressive symptoms at baseline and over 18 months of follow-up. PWH had a median age of 52 years, with 95% being virally suppressed below 200 copies/mL, 33% having high depressive symptoms, and 62% having ≥1 medical comorbidity (HCV, cardiovascular disease, diabetes, chronic kidney disease, chronic lung disease). VEGF levels were increased in PWH with high vs. low depressive symptoms (p = 0.01). The association between VEGF and depressive symptoms remained significant in covariate-adjusted models (p = 0.005) and was augmented in PWH with medical comorbidities (p = 0.002). Other inflammation biomarkers were increased in PWH with medical comorbidities, but not significantly different between groups stratified by depressive symptoms. Among PWH with high depressive symptoms, biomarker clustering identified inflammatory and noninflammatory subgroups distinguished by levels of VEGF/MCP-1/IL-8 or IL-6/CRP and prevalence of HCV, cardiovascular disease, and diabetes. These findings suggest that VEGF is a biomarker associated with depressive symptoms in PWH on ART and may identify a subtype of depression for targeted interventions.

For people with HIV (PWH), substance use disorders (SUDs) are a prominent neurological risk factor, and the impacts of both on dopaminergic pathways are a potential point of deleterious convergence. Here, we profile, at single nucleus resolution, the substantia nigra (SN) transcriptomes of 90 postmortem donors in the context of chronic HIV and opioid/cocaine SUD, including 67 prospectively characterized PWH. We report altered microglial expression for hundreds of pro- and anti-inflammatory regulators attributable to HIV, and separately, to SUD. Stepwise, progressive microglial dysregulation, coupled to altered SN dopaminergic and GABAergic signaling, was associated with SUD/HIV dual diagnosis and further with lack of viral suppression in blood. In virologically suppressed donors, SUD comorbidity was associated with microglial transcriptional changes permissive for HIV infection. We report HIV-related downregulation of monoamine reuptake transporters specifically in dopaminergic neurons regardless of SUD status or viral load, and additional transcriptional signatures consistent with selective vulnerability of SN dopamine neurons.

HIV-associated neurocognitive impairment (HIV-NCI) affects 15%–50% of people with HIV (PWH), despite viral suppression with antiretroviral therapy (ART). HIV neuropathogenesis is mediated, in part, by transmigration of infected CD14 ⁺ CD16 ⁺ monocytes across the blood-brain barrier (BBB) into the central nervous system (CNS). In the CNS, CD14 ⁺ CD16 ⁺ monocytes contribute to infection and activation of parenchymal cells, resulting in production of neurotoxic viral and host factors that cause neuronal damage. Mechanisms by which CD14 ⁺ CD16 ⁺ monocytes contribute to HIV-NCI have not been characterized in a study population of PWH on ART without contribution from confounders that affect cognition (e.g., substance use, hepatitis C virus coinfection). We assessed cognitive function, PBMC transmigration across the BBB, and neuronal health markers in a well-defined cohort of 56 PWH on ART using stringent criteria to eliminate confounding factors. We demonstrated that PWH on ART with HIV-NCI have significantly increased transmigration of their CD14 ⁺ CD16 ⁺ monocytes across the BBB compared with those with normal cognition. We showed that hypertension and diabetes may be effect modifiers on the association between CD14 ⁺ CD16 ⁺ monocyte transmigration and cognition. This study underscored the persistent role of CD14 ⁺ CD16 ⁺ monocytes in HIV-NCI, even in PWH with viral suppression, suggesting them as potential targets for therapeutic interventions.

Structural brain abnormalities, including those in white matter (WM), remain common in people with HIV (PWH). Their pathogenesis is uncertain and may reflect multiple etiologies. Oxidative stress is associated with inflammation, HIV, and its comorbidities. The post-translational carbonylation of proteins results from oxidative stress, and circulating protein carbonyls may reflect this. In this cross-sectional analysis, we evaluated the associations between protein carbonyls and a panel of soluble biomarkers of neuronal injury and inflammation in plasma (N = 45) and cerebrospinal fluid (CSF, n = 32) with structural brain MRI. The volume of abnormal WM was normalized for the total WM volume (nAWM). In this multisite project, all regression models were adjusted for the scanner. The candidate covariates included demographics, HIV disease characteristics, and comorbidities. Participants were PWH on virally suppressive antiretroviral therapy (ART) and were mostly white (64.4%) men (88.9%), with a mean age of 56.8 years. In unadjusted analyses, more nAWM was associated with higher plasma protein carbonyls (p = 0.002) and higher CCL2 (p = 0.045). In the adjusted regression models for nAWM, the association with plasma protein carbonyls remained significant (FDR p = 0.018). Protein carbonyls in plasma may be a valuable biomarker of oxidative stress and its associated adverse health effects, including within the central nervous system. If confirmed, these findings would support the hypothesis that reducing oxidative stress could treat or prevent WM injury in PWH.

Combination antiretroviral therapy (cART) has extended lifespans of people living with HIV (PWH), increasing both the risk for age-related neuropathologies and the importance of distinguishing effects of HIV and its comorbidities from neurodegenerative disorders. The accumulation of hyperphosphorylated tau (p-tau) in hippocampus is a common degenerative change, with specific patterns of hippocampal subfield vulnerability observed in different disease contexts. Currently, associations between chronic HIV, its comorbidities, and p-tau burden and distribution in the hippocampus are unexplored. We used immunohistochemistry with antibody AT8 to analyze hippocampal p-tau in brain tissues of PWH (n = 71) and HIV negative controls (n = 25), for whom comprehensive clinical data were available. Using a morphology-based neuroanatomical segmentation protocol, we annotated digital slide images to measure percentage p-tau areas in the hippocampus and its subfields. Factors predicting p-tau burden and distribution were identified in univariate analyses, and those with significance at p ≤ 0.100 were advanced to multivariable regression. The patient sample had a mean age of 61.5 years. Age predicted overall hippocampal p-tau burden. Subfield p-tau predictors were for Cornu Ammonis (CA)1, age; for CA2 and subiculum, seizure history; for CA3, seizure history and head trauma; and for CA4/dentate, history of hepatitis C virus (HCV) infection. In this autopsy sample, hippocampal p-tau burden and distribution were not predicted by HIV, viral load, or immunologic status, with viral effects limited to associations between HCV and CA4/dentate vulnerability. Hippocampal p-tau pathologies in cART-era PWH appear to reflect age and comorbidities, but not direct effects of HIV infection.

Objectives: As lifespans increase in people with HIV (PWH), there is concern that age-related neurodegenerative disorders may contribute to cognitive decline. We asked whether brain accumulation of Alzheimer's disease (AD)-associated proteins amyloid-beta (Aβ) and hyperphosphorylated tau (p-tau) predicted cognitive performance in middle-aged PWH. Methods: In a prospectively followed, cognitively-characterized autopsy sample of 135 PWH, we used immunohistochemistry to assess Aβ plaques and neuronal p-tau in medial temporal and lateral frontal lobes. These pathologies were tested for associations with cognitive performance in seven domains: motor, speed of information processing, working memory, memory encoding, memory retrieval, verbal fluency, and abstraction/executive function. Univariate and multivariate analyses accounting for HIV-associated variables, reading level, and comorbidities were conducted. Longitudinal trajectories of memory functions were evaluated in 60 individuals with a median follow-up of 6.0 years. Results: In this population with mean age 51.4 ± 0.9 years, 58% displayed neuronal p-tau and 29% Aβ plaques. Neuronal p-tau, but not Aβ, predicted worse memory encoding and retrieval, but not other cognitive functions. With an ordinal hierarchy of neuronal p-tau locations (entorhinal, hippocampal, neocortical), decreased memory performance correlated with neocortical distribution. Memory function trajectories could not be distinguished between individuals with and without neuronal p-tau, and over 80% of the sample showed no change over time. Conclusion: In this middle-aged sample, neuronal p-tau accumulation contributes to memory deficits, but is not associated with accelerated decline in function over time. In the absence of AD-like deterioration, other etiologies for neuronal p-tau in cognitively impaired PWH must be considered.

Determination of functional capacity in cognitively impaired persons living with HIV (PLHIV) is pivotal to the accurate diagnosis of HIV-associated neurocognitive disorders (HAND). Functional data is typically collected through self-report. Reliability concerns arise with memory and executive functioning impairments, which could compromise the integrity of self-report and result in inaccurate HAND diagnoses. The current study tested the accuracy of older PLHIV functional reports through examination of concordance rates between self-report and caregiver’s (CG) report. Cross-sectional cognitive, mood, and functional status data were sampled from the Manhattan HIV Brain Bank. Participants and caregivers independently completed an Activities of Daily Living (ADL) questionnaire, producing 78 participant-caregiver dyads. Functional report concordance was operationalized by calculating differences between participant and CG ADL total scores. Assessment pairs differing by 2 or more points were considered to be discordant. Analyses revealed that one-third of the patient sample was discordant in the ADL report. ANOVA revealed that PLHIV overestimating their functional impairments, were significantly older, more educated, and more depressed than other participants. Global cognitive functioning was not associated with concordance. Thus, the majority of PLHIV were consistent with their caregivers’ ADL report, and older age and increased depressive symptomatology, but not cognitive status, were factors associated with discordance.

Few publications have documented the utility of in-home telephone-based cognitive screeners during COVID-19. This manuscript describes the adaptation of select face-to-face (FTF) neuropsychological tests to telephonic administration in a longitudinal cohort of people with HIV (PWH). Using the cohort’s pre-pandemic neuropsychological data, we explore the utility of telephonic administration in this population. Of a longitudinal cohort of 170 adult PWH, 59 completed telephonic medical and cognitive screenings with comparable pre-pandemic FTF data. Telephone screeners and FTF evaluations were compared using repeated measures ANCOVAs to examine whether test performance differed between administration types and levels of pre-pandemic cognitive performance. Individuals with pre-pandemic test scores more than a standard deviation below the demographically-corrected mean were categorized as “below average” cognitive performance (n = 23), and the remainder as “average” (n = 36). Over 90% of participants gave positive feedback about the telephone encounter. The average cognitive performance group scored higher than the below average group on all measures across both administration types. Telephone and FTF test scores did not differ significantly for measures of category fluency, letter fluency, and verbal learning. However, the below average group scored higher on a verbal memory measure administered via telephone compared with FTF. Support for telephonic adaptation of select FTF measures in longitudinal research is mixed, with verbal fluency tasks showing the strongest equivalency. When employed carefully with a clear understanding of their limitations, telephone adaptations can provide an opportunity to continue study objectives, promote equity, and monitor participant well-being during times of duress.