Deciphering white-matter changes in Progressive Supranuclear Palsy using free-water diffusion MRI

189 

Abstract Submission 

Amir Fazlollahi^1,2, Lena Oestreich^3,4, Peter Nestor^5,6

¹Queensland Brain Institute, The University of Queensland,, Brisbane, Australia, ²Department of Radiology, Royal Melbourne Hospital, The University of Melbourne, Melbourne, Australia, ³School of Psychology, The University of Queensland, Brisbane, Queensland, Australia, Brisbane, Australia, ⁴Australian Institute for Bioengineering and Nanotechnology, The University of Queensland, Brisbane, Australia, ⁵Queensland Brain Institute, The University of Queensland, Brisbane, Australia, ⁶Mater Hospital, Brisbane, Australia

Amir Fazlollahi  
Queensland Brain Institute, The University of Queensland,|Department of Radiology, Royal Melbourne Hospital, The University of Melbourne
Brisbane, Australia|Melbourne, Australia

Lena Oestreich  
School of Psychology, The University of Queensland, Brisbane, Queensland, Australia|Australian Institute for Bioengineering and Nanotechnology, The University of Queensland
Brisbane, Australia|Brisbane, Australia

Peter Nestor  
Queensland Brain Institute, The University of Queensland|Mater Hospital
Brisbane, Australia|Brisbane, Australia

The overlapping motor and non-motor features of progressive supranuclear palsy (PSP) with Lewy body disorders (LBD comprising Parkinson's disease ± dementia, and dementia with Lewy bodies) may lead to misdiagnosis. Previous studies have suggested white-matter changes measured by diffusion-weighted MRI are prominent in PSP in the whole brain (Sajjadi, Acosta-Cabronero et al. 2013, Talai, Sedlacik et al. 2018) and in particular in the corpus callosum (Spotorno, Hall et al. 2019, Nguyen, Cheng et al. 2021) and the internal capsule (Agosta, Pievani et al. 2012, Nguyen, Cheng et al. 2021). We aimed to assess the usefulness of diffusion metrics to reliably discriminate PSP from LBD patients and understand the pathogenesis of the underlying disease. We focused on the corpus callosum as a proof-of-concept because the confounds of partial volume effects, white-matter lesions and crossing fibres can be eliminated in this region.

Thirty-nine healthy controls, 28 LBD (including 15 with dementia) and 13 PSP patients underwent 3.0T MRI. Multi-shell diffusion-weighted images were acquired using: repetition/echo times= 4100/75ms, FA = 90°, field-of-view 244 × 244 × 136, acquisition matrix 122 × 122 × 68 slices. The acquisition includes 9 non-diffusion weighted images (b = 0 s/mm2) as well as 27 (b = 1000 s/mm2) and 62 (b = 2500 s/mm2) unique directions. A FLAIR image was also acquired to assess global white-matter lesion load using SPM 12. The pre-processing of the diffusion images included head motion correction (using rigid registration), denoising, susceptibility distortions correction and removing intensity inhomogeneities. Free-water imaging was used on the pre-processed DWI data to quantify the amount of extracellular free-water (FW) by separating the diffusion properties of brain tissue, such as white matter tracts, from the surrounding extracellular free water, such as cerebrospinal fluid (Pasternak, Sochen et al. 2009). The DWI metrics of fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) were estimated using MRtrix3 before and after free-water correction. Two regions of interest were automatically placed using an atlas-based approach on the centre of the genu (GCC), and splenium (SCC), of corpus callosum. Differences between groups were assessed using two‐tailed unpaired Wilcoxon‐rank sum tests (after correcting for age and sex).

There was no significant difference in global white-matter lesion load between PSP and LBD patients (Figure 1). FA, MD and RD were significantly different in PSP compared with controls in GCC with p=0.001, p=0.022 and p=0.0018, respectively, as well as between PSP and LBD patients (p=0.0047, p=0.05 and p=0.0043, respectively) as shown in Figure 2. None of the diffusion metrics showed a statistically significant change in SCC. When DTI metrics were corrected for free-water, no significant difference was observed between PSP and LBD patients. However, free-water component showed a significant increase in PSP compared to LBD patients (p=0.0047) and control participants (p=0.01).

White-matter changes in progressive supranuclear palsy patients in the rostral corpus callosum (genu) showed a significant focal decrease in FA and an increase in RD using conventional DTI processing. However, after applying the free-water correction, it became evident that the FA and RD changes in the genu could be attributed to an increase in the extracellular space. Previous pathological studies have noted an abundance of white-matter tau pathology in PSP (Zhukareva, Joyce et al. 2006) and the present results suggest that this tau pathology is associated with an increase in extracellular water. When integrated into a multimodal diagnostic imaging approach, diffusion MRI metrics in the genu of the corpus callosum may hold the potential to aid in distinguishing individuals with progressive supranuclear palsy from those with Lewy body disorders.

Neurodegenerative/ Late Life (eg. Parkinson’s, Alzheimer’s) ¹

Diffusion MRI Modeling and Analysis ²

Aging

Degenerative Disease

MRI

Neuron

WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC