Reduced structural connectivity underlies dementia in Lewy Body disease
Poster No:
230
Submission Type:
Abstract Submission
Authors:
Naomi Hannaway1, Angeliki Zarkali2, Rohan Bhome3, Ivelina Dobreva3, Rimona Weil4
Institutions:
1University College London, LONDON, United Kingdom, 2University College London, London, NA, 3University College London, London, London, 4University College London, London, United Kingdom
First Author:
Co-Author(s):
Introduction:
There is continued debate over whether Dementia with Lewy Bodies (DLB) and Parkinson's Dementia (PDD) should be considered as a disease spectrum within Lewy Body Dementia (LBD) or as separate diseases. DLB and PDD have shared pathology of Lewy bodies containing alpha synuclein and shared symptoms of parkinsonism, hallucinations and fluctuations.
Diffusion MRI has potential to examine differences between these groups, if present, as axonal changes are amongst the earliest changes in LBD. Diffusion weighted imaging has shown reduced fractional anisotropy in visual association, posterior temporal and posterior cingulate areas for DLB compared to PDD [Lee et al., 2010]. To date, structural connectivity has not been compared between PDD and DLB.
Diffusion MRI has potential to examine differences between these groups, if present, as axonal changes are amongst the earliest changes in LBD. Diffusion weighted imaging has shown reduced fractional anisotropy in visual association, posterior temporal and posterior cingulate areas for DLB compared to PDD [Lee et al., 2010]. To date, structural connectivity has not been compared between PDD and DLB.
Methods:
We performed diffusion MRI and clinical assessments in 39 PD, 14 PDD and 31 DLB patients and in 21 age-matched controls.
Diffusion MRI images were pre-processed using Mrtrix3 including denoising, removal of ringing artefacts, eddy current correction, motion correction and bias-field correction. Diffusion-weighted images were upsampled to a spatial resolution of 1.3mm3 [Andersson et al., 2016].
Fibre orientation distributions for each participant were computed using multishell 3-tissue-constrained spherical deconvolution using the group-average response function for each tissue type.
Anatomically constrained tractography was performed with 10 million streamlines [Smith at al., 2012] and filtered using SIFT to reduce bias. The resulting tractogram was converted into a connectivity matrix, with 232 regions of interest generated by segmenting the participant's T1 weighted imaged using the 200 cortical [Schaefer et al., 2018] and 32 subcortical regions [Tian et al., 2020].
Network-based statistics (NBS) [Zalesky et al., 2010] was used to test differences between groups: a general linear model was constructed, with PD/LBD, control/LBD and PDD/DLB as contrasts of interest. Associations with cognitive and motor scores were also tested. Permutation testing (5000 permutations) with unpaired t-tests was performed, and a test statistic calculated for each connection. Each comparison was age-corrected and thresholded at T = 3.1, PFWE <.05
Diffusion MRI images were pre-processed using Mrtrix3 including denoising, removal of ringing artefacts, eddy current correction, motion correction and bias-field correction. Diffusion-weighted images were upsampled to a spatial resolution of 1.3mm3 [Andersson et al., 2016].
Fibre orientation distributions for each participant were computed using multishell 3-tissue-constrained spherical deconvolution using the group-average response function for each tissue type.
Anatomically constrained tractography was performed with 10 million streamlines [Smith at al., 2012] and filtered using SIFT to reduce bias. The resulting tractogram was converted into a connectivity matrix, with 232 regions of interest generated by segmenting the participant's T1 weighted imaged using the 200 cortical [Schaefer et al., 2018] and 32 subcortical regions [Tian et al., 2020].
Network-based statistics (NBS) [Zalesky et al., 2010] was used to test differences between groups: a general linear model was constructed, with PD/LBD, control/LBD and PDD/DLB as contrasts of interest. Associations with cognitive and motor scores were also tested. Permutation testing (5000 permutations) with unpaired t-tests was performed, and a test statistic calculated for each connection. Each comparison was age-corrected and thresholded at T = 3.1, PFWE <.05
Results:
The ages of the PDD (mean = 73.6 (6.9), 10 male), DLB (mean = 71.5 (5.5), 28 male) and control (mean age = 73.3 (5.8), 11 male) groups did not differ significantly, but the PD group (mean = 67.9 (5.4), 16 male), were younger than the PDD and DLB groups (p=.001). The PDD and DLB groups contained more men than both the PD (p <.001) and control groups (p= .002).
MoCA score did not differ between PDD (23.1) and DLB (21.3). As expected, the MoCA score was reduced for the combined LBD group compared to PD (28.6) and controls (28.8, p<.0001 for both). UPDRS-III did not differ significantly between PDD (32.9), DLB (34.4) and PD (26.2) groups.
No differences in structural connectivity were observed between PDD and DLB. Using network-based statistics, a combined LBD group, consisting of PDD and DLB, showed reduced connectivity compared to PD in a network consisting of 89 nodes and 118 edges (PFWE = .009, Figure 1A). Reduced connectivity was also shown for LBD relative to controls in a network of 160 nodes and 273 edges (PFWE = .008, Figure 1B).
Across all patient groups, a network of 459 nodes and 196 edges showed a significant association with MoCA score (PFWE = .02, Figure 2). There were no differences in structural connectivity associated with UPDRS-III score across patient groups.
MoCA score did not differ between PDD (23.1) and DLB (21.3). As expected, the MoCA score was reduced for the combined LBD group compared to PD (28.6) and controls (28.8, p<.0001 for both). UPDRS-III did not differ significantly between PDD (32.9), DLB (34.4) and PD (26.2) groups.
No differences in structural connectivity were observed between PDD and DLB. Using network-based statistics, a combined LBD group, consisting of PDD and DLB, showed reduced connectivity compared to PD in a network consisting of 89 nodes and 118 edges (PFWE = .009, Figure 1A). Reduced connectivity was also shown for LBD relative to controls in a network of 160 nodes and 273 edges (PFWE = .008, Figure 1B).
Across all patient groups, a network of 459 nodes and 196 edges showed a significant association with MoCA score (PFWE = .02, Figure 2). There were no differences in structural connectivity associated with UPDRS-III score across patient groups.
· Figure 1. Network showing reduced connectivity for A) Lewy Body Dementia compared to Parkinson’s disease and B) Lewy Body Dementia compared to controls. Analysis is adjusted for age. Network based st
·Figure 2. Network showing connectivity associated with MoCA score. Analysis is adjusted for age. Network based statistics were thresholded at T = 3.1, FWE corrected p<.05. Nodes and edges with signifi
Conclusions:
We report a widespread network of reduced connectivity in LBD, compared to both PD and controls. Cognitive, but not motor scores were also associated with structural connectivity changes. We found no differences between PDD and DLB groups but may have lacked power to detect these in the current analysis. In future, structural and functional connectivity could be examined in combination, to further our understanding of connectivity changes in LBD.
Disorders of the Nervous System:
Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) 1
Modeling and Analysis Methods:
Connectivity (eg. functional, effective, structural) 2
Diffusion MRI Modeling and Analysis
Keywords:
Cognition
Memory
Movement Disorder
White Matter
Other - Dementia
1|2Indicates the priority used for review
Provide references using author date format
Lee, J. E. (2010). A. comparative analysis of cognitive profiles and white-matter alterations using voxel-based diffusion tensor imaging between patients with Parkinson's disease dementia and dementia with Lewy bodies. Journal of Neurology, Neurosurgery & Psychiatry; 81(3), 320-326.
Smith, R.E. (2012). Anatomically-constrained tractography: improved diffusion MRI streamlines tractography through effective use of anatomical information. Neuroimage; 62(3):1924-38.
Schaefer, A. (2018). Local-Global Parcellation of the Human Cerebral Cortex from Intrinsic Functional Connectivity MRI. Cerebral Cortex; 28(9):3095-114.
Tian, Y. (2020). Topographic organization of the human subcortex unveiled with functional connectivity gradients. Nature Neuroscience; 23(11):1421-32.
Zalesky, A. (2010). Network-based statistic: identifying differences in brain networks. Neuroimage; 53(4):1197-207.
Zarkali, A. (2020). Differences in network controllability and regional gene expression underlie hallucinations in Parkinson’s disease. Brain; 143(11):3435-48.