Abstract
Background Mechanisms for SLE-mediated cognitive impairment (SLE-CI) include autoantibodies and inflammatory molecules normally excluded from the brain by an intact blood brain barrier (BBB). However, the BBB permeability may be altered in response to molecular signals from the periphery and activated microglia in the brain. We have hypothesized that in SLE-CI the hippocampus is targeted by autoantibodies leading to microglia activation and neuron dysfunction. Dynamic Contrast Enhanced (DCE) MRI was used in SLE subjects concurrently assessed with FDG-PET brain imaging to determine regional BBB permeability and association with metabolism.
Methods Data from 10 SLE and 6 healthy control (HC) subjects enrolled in a larger observational study were analyzed. Important exclusion criteria for SLE included any history of neuropsychiatric disease, current psychiatric medication use, active disease, and steroid dose > 10 mg. Statistical Parametric Mapping (SPM) analysis was used to perform voxel-wise comparison of different imaging data from the same cohort. Parametric maps created from DCE-MRI data were registered to MNI templates used in SPM analysis, allowing comparison between maps in the MNI space. Voxel-wise comparison of DCE-MRI permeability parameters, Ve, Kep and Ktrans, was conducted on data from all subjects and compared to voxel-wise assessments of glucose metabolism from FDG-PET imaging. Clusters were identified by voxel-wise analysis of the parametric maps for Kep and Ve data. SPM contrasts were thresholded at p<0.05 within a hippocampal hypothesis-testing volume. Student t-test was used to compare permeability parameters between SLE and HC.
Results Abnormal increases in Kep (red) and Ve (green) were found in the SLE group hippocampus (figure 1); they overlapped with the area of increased hippocampal metabolism (yellow) reported previously.1 Hippocampal increases in Ve were smaller in both extent and magnitude compared to Kep (color scales). Ktrans, Ve and Kep are all highly related such that Ktrans = Kep x Ve. The voxelwise contrast for Ktrans was similar to Kep, and corresponding values were highly correlated (r = 0.92, p<0.0001) in the overlap region. To simplify visualization, only Kep and Ve contrasts are displayed in figure 1A. Scatterplots of permeability parameters measured in the hippocampal overlap region for the SLE and HC subjects (figure 1B).
Conclusions SPM analysis of DCE-MRI and FDG-PET data demonstrates abnormal increased hippocampal BBB permeability and glucose metabolism in SLE subjects with inactive disease and no history of neuropsychiatric disease. These data support the hypothesis that BBB permeability is altered in SLE, allowing access of inflammatory mediators to the brain.