Mapping activated microglial density in individuals with sickle cell disease utilizing time-dependent diffusion-weighted MRI and correlation with cognitive performance Free

Introduction: Sickle cell disease (SCD) causes progressive cognitive deficits that begin in childhood. Other factors beyond neurovascular changes are now thought to contribute to their development. Activated microglia, a component of neuroinflammation, has been associated with cognitive deficits in numerous neurodegenerative disorders. A loss of selective synaptic pruning and a surge in phagocytic activity results in unregulated neuronal death. SCD is by itself a systemic, proinflammatory disorder with elevated production of inflammatory cytokines at steady state, which may influence the brain microenvironment via microglial and astrocyte activation. Prior work by our lab demonstrated that aged, humanized SCD mice show memory and cognitive deficits compared to both young SCD and aged control mice. In addition, aged SCD mice had significantly higher densities of activated microglia and peripherally-derived immune cells in the hippocampus/dentate gyrus (vital for episodic memory) compared to control mice on immunohistochemistry. Based on these data, we hypothesized that humans with SCD will have a greater density of activated microglia in the anterior hippocampus and frontal gray matter compared to matched controls and that the activated microglial density will negatively correlate with performance on a shortened cognition battery.

Methods: We enrolled individuals with SCD ages 18-45 years with the higher risk genotypes of HbSS and HbS-beta⁰thalassemia to undergo a one-time time-dependent diffusion-weighted MRI (td-dMRI) and the NIH Toolbox: Cognition Battery, a shortened neuropsychological battery administered via tablet application with automated scoring (mean = 100, standard deviation = 15). Individuals with a history of overt stroke or other known neurological disorders and those on chronic transfusions were excluded. Controls were frequency-matched for race, age, sex, and socioeconomic status using the area deprivation index (ADI). Each participant underwent an MRI using a GE Signa Architect 3.0T Scanner with T1, T2, T2-FLAIR, and time-dependent echo planar diffusion MRI. Td-dMRI allows fitting a model first described by Garcia-Hernandez et al. based on the changes in microglial cell shape upon activation, which includes an increase in cell body size and retraction of cellular processes. These changes alter the diffusion of water molecules around and within the microglia (the orientation dispersion index, ODI). Descriptive statistics were completed to compare the results of the SCD versus control groups in the hippocampus and frontal cortex. The results of the ODI and performance on the Toolbox sections were correlated in the SCD group only.

Results: We completed Td-dMRI on 5 individuals with SCD and 4 healthy controls (HC). The median age of individuals with SCD was 30 years versus 26 years for HC (p =0.45) with no difference in sex between the groups (p = 0.99); all participants identified as of non-Hispanic ethnicity and of Black race. The amount of neighborhood level disadvantage was also not significantly different (68th versus 57th percentile of disadvantage compared to national norms, for SCD versus controls, respectively, p = 0.4). There was no difference in the Toolbox overall cognition score between the two groups (SCD = 91 versus controls = 88, p = 0.8), which were both below the nationalized mean of 100. We found a significantly higher ODI (density of activated microglia) specifically within the anterior hippocampus in the those with SCD compared to the control group (0.15 in SCD versus 0.07 in HC, p = 0.002). Further, a higher ODI in the anterior hippocampus showed a moderate negative correlation with performance on the episodic memory subtest in the individuals with SCD (r = -0.54, p = 0.36, unadjusted). Similarly, we found a higher ODI in the frontal gray matter in the SCD group versus HC (0.26 in SCD versus 0.09 in HC, p <0.001), which was weakly negatively correlated with measures of executive attention, category switching, and processing speed (r = -0.26 to -0.3, p = 0.6, unadjusted). Conversely, there was a significantly higher ODI in the posterior hippocampus and frontal white matter in the HC versus SCD.

Conclusions: Individuals with SCD may have a greater density of activated microglia in the frontal gray matter and anterior hippocampus, which may be negatively correlated with measures of cognitive function. A larger sample size is necessary to support these findings.