Glial activation colocalizes with structural abnormalities in amyotrophic lateral sclerosis

Abstract

Objective: In this cross-sectional study, we aimed to evaluate brain structural abnormalities in relation to glial activation in the same cohort of participants.

Methods: Ten individuals with amyotrophic lateral sclerosis (ALS) and 10 matched healthy controls underwent brain imaging using integrated MR/PET and the radioligand [¹¹C]-PBR28. Diagnosis history and clinical assessments including Upper Motor Neuron Burden Scale (UMNB) were obtained from patients with ALS. Diffusion tensor imaging (DTI) analyses including tract-based spatial statistics and tractography were applied. DTI metrics including fractional anisotropy (FA) and diffusivities (mean, axial, and radial) were measured in regions of interest. Cortical thickness was assessed using surface-based analysis. The locations of structural changes, measured by DTI and the areas of cortical thinning, were compared to regional glial activation measured by relative [¹¹C]-PBR28 uptake.

Results: In this cohort of individuals with ALS, reduced FA and cortical thinning colocalized with regions demonstrating higher radioligand binding. [¹¹C]-PBR28 binding in the left motor cortex was correlated with FA (r = -0.68, p < 0.05) and cortical thickness (r = -0.75, p < 0.05). UMNB was correlated with glial activation (r = +0.75, p < 0.05), FA (r = -0.77, p < 0.05), and cortical thickness (r = -0.75, p < 0.05) in the motor cortex.

Conclusions: Increased uptake of the glial marker [¹¹C]-PBR28 colocalizes with changes in FA and cortical thinning. This suggests a link between disease mechanisms (gliosis and inflammation) and structural changes (cortical thinning and white and gray matter changes). In this multimodal neuroimaging work, we provide an in vivo model to investigate the pathogenesis of ALS.

Figure 1. [¹¹C]-PBR28 uptake colocalizes with reduced fractional anisotropy (FA) in the motor cortex

(A) Tract-based spatial statistical analysis (TBSS) for 10 participants with amyotrophic lateral sclerosis (ALS) and 6 healthy controls (HC) revealed decreased FA values in ALS (p_FWE < 0.05). Blue represents decreased FA in ALS in comparison with HC. Green shows the mean FA skeleton calculated for all participants by TBSS. The statistical map and the skeleton were transformed from FMRIB58 space to Montreal Neurological Institute (MNI) 152-T1-1 mm standard space and are shown at coordinates x = 8, y = −20, z = 64. (B) Voxel-wise whole brain analysis for 10 participants with ALS and 10 controls shows differences in [¹¹C]-PBR28 binding in ALS (p_FWE < 0.05). Red to yellow represents higher [¹¹C]-PBR28 binding in ALS in comparison with HC. The statistical map is displayed in MNI 152-T1-1 mm standard space and shown at coordinates x = 8, y = −20, z = 64. (C) Boxplot representation (box contains median, 25th, and 75th percentiles) of FA values as measured in the left and right precentral gyrus (PCG). The left orange box corresponds to FA values in ALS and the right blue box to HC. The horizontal white line in each box represents the median. (D) Scatterplot shows the correlation between FA values and [¹¹C]-PBR28 SUVR as measured in the left PCG in ALS. FWE = family-wise error; SUVR = standardized uptake value ratio.

Figure 2. [¹¹C]-PBR28 uptake colocalizes with cortical thinning

(A) Statistical map of cortical thickness differences between patients with amyotrophic lateral sclerosis (ALS) and healthy controls (HC) shows cortical thinning in the left precentral gyrus (PCG) in ALS. The results are FDR corrected. (B) Projection onto the surface of voxel-wise whole brain analysis for 10 patients with ALS and 10 HC shows brain regions of higher [¹¹C]-PBR28 binding in ALS compared to controls (p_FWE < 0.05). (C) Boxplot representation (box contains median, 25th, and 75th percentiles) of cortical thickness in the left and right PCG. The left orange box shows cortical thickness in ALS and the right blue box in HC. The horizontal white line in each box represents the median. (D) Scatterplot shows the correlation between cortical thickness and [¹¹C]-PBR28 standardized uptake value ratio (SUVR) as measured in the left PCG in ALS. (E) Scatterplot shows the correlation between cortical thickness and radial diffusivity in the left PCG in ALS. FWE = family-wise error.

Figure 3. Correlations with Upper Motor Neuron Burden Scale (UMNB)

(A) Correlation between UMNB and fractional anisotropy (FA) within the bilateral precentral gyrus (PCG). (B) Correlation between UMNB and bilateral PCG thickness. (C) Correlation between UMNB and standardized uptake value ratio within the bilateral PCG. ALS = amyotrophic lateral sclerosis.