Altered Gray Matter Volume and Functional Connectivity in Human Immunodeficiency Virus-Infected Adults

Abstract

People living with human immunodeficiency virus (HIV) (PLWH) are at high risk of neurocognitive impairment. The pathogenesis of neurocognitive impairment remains unclear, and there is still no diagnostic biomarker. By coupling three-dimensional T1-weighted imaging and resting-state functional imaging, we explored structural and functional alterations in PLWH and examined whether such imaging alterations had the potential to denote neurocognitive function. A total of 98 PLWH and 47 seronegative controls aged 20-53 years were recruited. Structural alterations were first explored between HIV-negative controls and PLWH. Subsequently, brain regions showing gray matter alterations were used as seeds for separate whole-brain functional connectivity (FC) analysis. Finally, the relationships between imaging alterations and cognitive function were explored. PLWH suffered from thalamus, occipital lobe, and hippocampus/parahippocampus atrophy. Visual cortices in PLWH showed decreased anticorrelation with the posterior cingulate cortex and left angular gyrus of the default mode network. FC within the visual cortices (between the left calcarine and right calcarine) and in the thalamic prefrontal circuit and between the thalamus and somatosensory association cortex were also altered. In addition, FC between the left thalamus and right dorsolateral prefrontal cortex in the cognitively impaired group was significantly different from that in the cognitively normal group in PLWH. Partial correlation analysis uncorrected for multiple comparisons suggested that some imaging alterations can be associated with neurocognition. Our study supports the presence of brain atrophy and functional reconfiguration in PLWH. Imaging alterations can be associated with neurocognitive function. We hold that neuroimaging is a promising approach in evaluating PLWH and might have the potential to clarify the pathogenesis of HIV-associated neurocognitive disorder.

Keywords: HIV-associated neurocognitive disorder; functional connectivity; human immunodeficiency virus; resting-state functional MRI; structural MRI.

FIGURE 1

Gray matter volume differences between HIV-negative controls and people living with HIV (PLWH) based on voxel-based morphometry (VBM) analysis. Brain regions with significantly reduced gray matter colored by T-statistic (red/yellow). Corrected for multiple comparisons [Gaussian random field (GRF) correction, voxel level P < 0.001, cluster level P < 0.05, two-tailed].

FIGURE 2

Functional connectivity (FC) differences between HIV-negative controls and people living with HIV (PLWH). Seeds were extracted according to the Automated Anatomical Labeling (AAL) atlas. THA.L, left thalamus (purple); LING.L, left lingual (yellow); CUN.R, right cuneus (orange); CAL.R, right calcarine (red); LING.R, right lingual (green); and THA.R, right thalamus (cyan). Corrected for multiple comparisons [Gaussian random field (GRF) correction, voxel level P < 0.001, cluster level P < 0.05, two-tailed]. Color bar indicates corrected T-values.

FIGURE 3

Volume (ml) and functional connectivity (FC) (Z-score) differences among HIV-negative controls, global deficit score (GDS) normal, and GDS impaired groups. For voxel-based morphometry (VBM) analysis, the left Y-axis is the volume of gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF), and the right Y-axis is the volume of thalamus (THA), occipital, and left hippocampus/parahippocampus (HIP/PHG.L). ^∗P < 0.05; ^∗∗P < 0.01; ^∗∗∗P < 0.001. Least significant difference (LSD) for multiple comparison correction. LING.L, left lingual; PCC, posterior cingulate cortex; CAL.R, right calcarine; CAL.L, left calcarine; CUN.R, right cuneus; ANG.L, left angular; LING.R, right lingual; DLPFC, dorsal lateral prefrontal cortex; THA.L, left thalamus; THA.R, right thalamus; SA.L, left somatosensory association cortex.

FIGURE 4

partial correlation between specific cognitive domains, global deficit score (GDS), and imaging alterations. The areas between two dotted curves indicate the 95% confidence interval. All variables shown above were unstandardized residuals plus their own mean value.