Brain diffusion alterations in patients with COVID-19 pathology and neurological manifestations

Abstract

Background and objective: COVID-19 neurological manifestations have been progressively recognized. Among available MRI techniques, diffusion weighted imaging (DWI) shows promise to study microstructure, inflammation, and edema. Previous DWI studies reported alterations in brain diffusivity in COVID-19 patients, as assessed by morphologic evaluation of brain DWI scans only. The aim of this study was to assess and quantify brain diffusion alterations in COVID-19 patients with neurological manifestations.

Methods: 215 COVID-19 patients with neurological manifestations (olfactory and/or other neurological disorders) and 36 normal controls were compared and studied with DWI and T1-weighted MRI scans. MRI scans were processed by a semi-automatic processing procedure specifically developed for the purpose of this study, and the Apparent Diffusion Coefficient (ADC) was quantified in different brain tissues and individual white matter (WM) and gray matter (GM) regions. Differences in ADC values were assessed between COVID-19 patients and normal controls, as well as in the COVID-19 patient population grouped by hospitalization and neurological symptoms.

Results: Among COVID-19 patients (median [IQR] = 52 [42 - 60] years of age, 58 % females), 91 were hospitalized and 26 needed intensive care. 84 patients had hyposmia/ageusia only, while 131 ones showed other neurological disorders. COVID-19 patients showed significantly increased ADC values in the WM and in several GM regions (p < 0.001). ADC values were significantly correlated with MRI time from disease onset (p < 0.05). Hospitalized patients showed significantly higher ADC alteration than non-hospitalized patients in all brain tissues; similarly, COVID-19 patients with neurological disorders showed significantly higher ADC values than those with olfactory loss only. ADC alteration was highest in patients with cognitive or memory disorder and in those with encephalitis or meningitis. ADC values were neither associated with the duration of hospitalization nor with the need for intensive care.

Conclusion: Current findings suggest DWI potential as a non-invasive marker of neuroinflammation in COVID-19, and the transient nature of the same. Future longitudinal studies are needed to confirm our findings.

Keywords: ADC; COVID-19; Diffusion weighted imaging; Neuroinflammation.

Graphical abstract

Fig. 1

Diagram summarising the brain diffusion weighted imaging (DWI) processing procedure developed and used in the study. First, T1-weighted brain MRI scans are coregistered to the DWI (b = 0) sequence (step 1). The resulting images are segmented, obtaining probability maps of the different brain tissues (step 2). Gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF) probability maps are binarized and then added up to create the whole brain mask (step 3). DWI signal acquired at different b-values is interpreted by a mono-exponential model, and the apparent diffusion coefficient (ADC) whole brain map is generated by fitting the model in each voxel of the brain mask (step 4). The ADC brain map is restricted to the GM and WM. The ADC map is also restricted to individual brain regions by using available brain atlases, and further masked by the patient-specific GM or WM binary masks (step 5). ADC summary values are finally computed in the whole brain as well as in the different brain tissues and regions (step 6).

Fig. 2

White matter regions with significant diffusion differences in the 215 COVID-19 patients included in the study. WM regions of the JHU WM atlas (Oishi et al., 2009) were colour-coded based on the statistical significance (Bonferroni corrected p-value < 0.05) of the increase (red-yellow) or decrease (blue-green) in ADC in COVID-19 patients as compared with the normal controls. P-values were computed by non-parametric Wilcoxon rank sum test. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3

ADC distribution in COVID-19 patients, subgrouped by neurological complication. The distribution of WM (A) and GM (B) ADC values in COVID-19 patients with olfactory disorders only (OD, n = 84), neuromuscular disorders (NM, n = 30), cognitive and memory disorders (CM, n = 61), or encephalitis and meningitis as main neurological complication (EM, n = 7) is compared with the distribution of corresponding values in in the control group (CTRL, n = 36). Pairwise p-values were assessed by Wilcoxon test or t-test, based on normal distribution of the data, while overall p-values were assessed by Kruskal-Wallis test. * indicates the level of significance (** p < 0.001, * p < 0.05). Abbreviations: WM = white matter, GM = gray matter, ADC = apparent diffusion coefficient, OD = olfactory disorders, NM = neuromuscular disorders, CM = cognitive and memory disorders, EM = encephalitis and meningitis, CTRL = controls.