8th Baltic Congress of Radiology

Neuroradiology

oral abstract

Neuroimaging in general, and proton magnetic resonance spectroscopy in particular are used as diagnostic tools for neurodegenerative disease. It is expected that structural and neurochemical alterations in the brain may reflect different aspects of the same pathology. Nonetheless, associations between neuroanatomical structural changes and neurochemical changes in the brain are underexplored. In this presentation we will introduce a multimodal approach to discuss associations between structural and neurochemical changes that could signify neurodegenerative processes related to mild cognitive impairment (MCI). We combined multiple MRI scanning techniques to examine these associations and discuss them in the context of routine MRI imaging as it is implemented in daily practice. Moreover, these combined neuroimaging features will be discussed in relation to pathological conditions that may be related to onset of disease such as systemic inflammation or neuroinflammation.

74 older adults (60-85y) underwent whole brain structural 3T MRI (T1W, T2W, DTI) and proton magnetic resonance spectroscopy (1H-MRS). 1H-MRS post-processing software packages were compared. The regions of interest for 1H-MRS measurements were dorsal posterior cingulate cortex (DPCC), left hippocampal cortex (HPC), left medial temporal cortex (MTC), left primary sensorimotor cortex (SM1), and right dorsolateral prefrontal cortex (DLPFC). Spectra from the aforementioned regions were acquired using a PRESS sequence (TR/TE = 2000/30 ms) with excitation water suppression. LCModel was used to quantify water referenced levels of N-acetyl aspartate (NAA), choline (Cho), myo-inositole (mIns) and their ratios to creatine (Cr). Only spectra with linewidths less than 0.1 ppm and signal to noise ratios (SNR) greater than 5 were included in the statistical analyses. DTI images were checked for artefacts by an experienced neuroradiologist. Tractography was performed using the Siemens “syngo.via” workstation (Siemens Healthineers, Erlangen, Germany). Circular voxels (1 cm radius) were manually placed on the regions of interest. The software automatically calculated the median values of fractional anisotropy, and number of tracts. For anatomical reference, a high-resolution T1W 3D structural image was acquired for all participants at each scanning site. Spearman correlation coefficients were used to evaluate relationships between NAA/Cr, Cho/Cr and mIns/Cr ratios in the five regions and DTI measures. Mann-Whitney test was used to examine group differences in neurometabolites and white matter structural properties between MCI and healthy controls. Peripheral inflammation was assessed with serum interleukin-6 and kynurenine, measured with ELISA and were used in a regression model to examine associations between neurometabolic expressions of neuroinflammation (i.e., elevated mIns/Cr and Cho/Cr) and blood biomarkers of peripheral inflammation.

Findings from correlation analysis revealed a trend showing that an increased level of hippocampal mIns/Cr was negatively related to decreased FA in the left external capsule of MCI patients (r = -0.399) which was not seen in non-MCI (r = 0.02). Contrarily, higher levels of hippocampal Cho/Cr were negatively related to lower FA in non-MCI (r = -0.455, p = 0.033) but not in MCI (r = 0.120). No associations were found between metabolites in the two other brain locations and FA. Overall, these observations may suggest that biochemical integrity of the left hippocampus is associated with microstructural organization of ipsilateral WM tracts connecting the hippocampus with temporal and prefrontal cortex. Elevated serum kynurenine levels were associated with signs of neuroinflammation, specifically in the DLPFC (Cho/Cr, β = 0.297, p = 0.02) and MTC (Cho/Cr, β = 0.418, p= 0.003). In addition, we found a negative correlation between IL-6 and measures of neural density in the MTC [NAA/Cr (β = -0.407, p = 0.011); and a positive correlation with Cho/Cr (β = 0.442, p=0.007)] in the same region. However, no association were observed between Kynurenine/ IL-6 and DTI measures.

Our observations suggest that decreased microstructural organization of the left external capsule could be related to local neuro-inflammation in the hippocampus. Our exploratory regression analysis showed associations between neurometabolic biomarkers and blood biomarkers of peripheral inflammation. There were no associations between peripheral inflammatory biomarkers and white matter structure. Overall, these findings suggest that elevated glial cell activity as expressed by elevated mIns/Cr might be an underlying mechanism for decreased connectivity between hippocampus and prefrontal/cingulate cortex that could lead to cognitive decline and potentially to neurodegenerative diseases. One possible cause of neurodegenerative processes that seems to have been overlooked in the research of normal aging is low-grade neuroinflammation. Pro-inflammatory processes have long been argued to play a role in conditions associated with cognitive decline and neurodegeneration in the normal aging process. However, only a limited number of studies have attempted to measure both peripheral and central biomarkers of inflammation and examined their interrelationship. Our results suggest that serum kynurenine may be used as a peripheral inflammatory marker that is associated with neuroinflammation and potentially neurodegeneration. Further research is needed to confirm these observations.

We examined links between biochemical status, brain structural properties and biomarkers of peripheral inflammation in order to establish a neurobiological model of aging. Our observations suggest that biochemical integrity of the left hippocampus is associated with microstructural organization of ipsilateral WM tracts connecting the hippocampus with temporal and prefrontal cortex. This project has received funding from the Research Council of Lithuania (LMTLT), agreement No S-MIP-21-37.

1270