3912

Evaluation of the relation between tau protein and white matter structural changes in Alzheimer's disease using fixel-based analysis
Takafumi Kitagawa1,2, Koji Kamagata1, Wataru Uchida1, Keigo Yamazaki1,3, Kaito Takabayashi1, Yuya Saito1, Christina Andica1,4, Akifumi Hagiwara1, Toshiaki Akashi1, Katsuhiro Sano1, Akihiko Wada1, and Shigeki Aoki1,2,4
1Department of Radiology, Juntendo University Graduate School of Medicine, Tokyo, Japan, 2Department of Data Science, Juntendo University Graduate School of Medicine, Tokyo, Japan, 3Department of Radiological Sciences, Graduate School of Human Health Sciences, Tokyo Metropolitan University, Tokyo, Japan, 4Faculty of Health Data Science, Juntendo University, Chiba, Japan

Synopsis

Keywords: Alzheimer's Disease, Alzheimer's Disease

Motivation: Studies have suggested that tau deposition-induced neurotoxicity causes progressive white matter (WM) degeneration in Alzheimer's disease (AD) and its prodromal phase; however, this is not yet fully elucidated.

Goal(s): To employ fixel-based analysis (FBA) to assess tau-related WM degeneration and its effect to cognitive function.

Approach: The WM integrity and cortical tau load were compared across AD, mild cognitive impairment, and cognitive normal subjects using FBA and tau PET.

Results: FBA metrics in WM under highly tau-deposited cortex were downward with higher effect sizes in patients. An association between entorhinal tau and cognitive function was mediated by FBA metrics in the parahippocampal cingulum.

Impact: Our findings suggest the possibility that neurons were degenerated through prion-like tau propagation along axons and may shed light on future research on mechanisms of cognitive decline in Alzheimer's disease.

INTRODUCTION

It is challenging to detect the underlying pathophysiology of Alzheimer's disease (AD), which involves amyloid, tau, and white matter (WM) degeneration, in vivo1. Recently, it has been suggested that tau deposition-induced neurotoxicity causes progressive neurostructural degeneration in AD and its prodromal phase; however, this is not yet fully elucidated2. Studies have utilized voxel-based approaches to link WM degeneration and tau deposition; however, averaging diffusion in multiple neuronal sources precludes evaluation for each fiber integrity in the voxel existing fibers in multiple orientations3. Fixel-based analysis (FBA) can model both microscopic and macroscopic neural structures in multiple directions. Thus, we examined whether FBA can be utilized to assess the association between the impact of tau deposition, structural changes in WM, and cognitive decline.

METHODS

Study participants and data acquisition
We obtained the diffusion-weighted imaging (DWI) and standardized uptake value ratio (SUVR) of 18F-flortaucipir to bind to AD-type tau4, in the cortical area derived from the Desikan–Killiany atlas5 of 58 cognitively normal (CN) subjects, 20 mild cognitive impairments (MCI), and 24 ADs from the Alzheimer's Disease Neuroimaging Initiative (ADNI) (Table 1). All participants were confirmed for tau pathology (18F-flortaucipir SUVR cutoff=1.27 with the temporal meta-ROI6), with healthy participants being tau-negative and the disease group being tau-positive. Acqusition parameters of DWIs were shown in Table 2.
FBA
DWIs were treated with the noise reduction7, Gibbs artifacts correction7, and motion correction implemented8 in the MRtrix3tissue9. First, response functions were estimated10 and utilized to estimate the fiber orientation distribution (FOD) using the single-shell 3-tissue constrained spherical deconvolution10. The fiber density (FD)9, log-transformed fiber-bundle cross-section (log-FC)9, and product of FD and FC (FDC)9 were determined in the FOD template space. In addition, the mean value of FBA metrics was calculated along the parahippocampal WM defined with the JHU Atlas11.
Statistical analysis
A permutation test (10,000 permutations) was performed for FBA using a general linear model with age, sex, and intracranial volume as covariates. An analysis of variance (Post hoc: Tukey) was conducted to compare 18F-flortaucipir SUVR in the cerebral cortex among AD, MCI, and CN groups. Moreover, since the entorhinal cortex is known as one of the primary regions of early tau involvement12, we assessed the association between entorhinal tau and cognitive function mediated by the FBA matrices in the parahippocampal WM through the bootstrapped mediation approach with 5,000 iterations using IBM SPSS Statistics version 27. The Bonferroni-corrected p < 0.05 was considered statistically significant in all analyses.

RESULTS

In a whole-brain FBA (Figure 1), upon comparing the groups with AD and CN, we found that there were significantly lower FD values in the frontal, medial temporal, and limbic WM, lower log-FC values in the medial and parietal WM, and lower FDC in the WM overlapping them. Additionally, significantly lower FD was shown mainly in the medial temporal and limbic WM in MCI compared to CN. Significantly higher levels of 18F-flortaucipir SUVR were found in the AD group compared to the CN group, mainly in the frontal, temporal, and parietal lobes. Further, significantly higher levels of 18F-flortaucipir SUVR were found in the entorhinal, fusiform, inferior temporal, insula, middle temporal, and parahippocampal areas in the MCI group (Figure 2). In the mediation analysis, the FD and FDC along the parahippocampal WM significantly mediated the relationship between entorhinal tau and cognitive function as shown in Figure 3.

DISCUSSION

The FD values and the log-FC methodologically observe the micro-structural WM changes (i.e., axonal injury or demyelination) and subsequent macroscopic bundle atrophy, respectively9. The tau propagation pattern in AD is known to begin in the entorhinal cortex, spreading through the inferolateral temporal lobes and medial parietal lobes, and eventually to wide neocortex12. The low log-FC in the medial temporal and medial parietal WM and the low FD values in the frontal and limbic WM may have captured changes related to the neurotoxicity of hyperphosphorylated tau2.
FBA metric in the WM under highly tau-deposited cortices showed higher effect sizes. Furthermore, high tau-deposited cortices and corresponding pathways in MCI were congruent with the regions that showed high effect sizes in AD, including the entorhinal and temporal cortex. Incorporating the result that the FD and FDC significantly mediated the relationship between tau deposition and cognitive function, our findings suggest the possibility that the neuron was degenerated through prion-like tau propagation along axons13 in accordance with the progression of the Braak stage12. The highest effect in FDC might be related to the FDC’s technical superiority to assess WM integrity in situations where both micro-structural and macrostructural changes existed14.

CONCLUSION

FBA can assess tau deposition-associated structural changes in cerebral WM.

Acknowledgements

This study was partially supported by the Juntendo Research Branding Project, the Sportology Center of Juntendo University Graduate School of Medicine, the Leading-edge Research Project for Sports Medicine and Science (LRP) of the Japan Sports Agency, the Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research (KAKENHI; Grant numbers JP19K17244, 23H02865), and the Brain/MINDS Beyond program of the Japan Agency for Medical Research and Development (AMED) under Grant numbers JP18dm0307004, JP19dm0307101, and JP21wm0425006.

References

  1. Witte MM, Trzepacz P, Case M, Yu P, et al. Association between clinical measures and florbetapir F18 PET neuroimaging in mild or moderate Alzheimer's disease dementia. J Neuropsychiatry Clin Neurosci. 2014 Summer;26(3):214-20.
  2. Naseri NN, Wang H, Guo J, et al. The complexity of tau in Alzheimer's disease. Neurosci Lett. 2019 Jul 13;705:183-194.
  3. Li K, Wang S, Luo X, et al. Associations of Alzheimer's Disease Pathology and Small Vessel Disease With Cerebral White Matter Degeneration: A Tract-Based MR Diffusion Imaging Study. J Magn Reson Imaging. 2023 Sep 22.
  4. Lowe VJ, Curran G, Fang P, et al. An autoradiographic evaluation of AV-1451 Tau PET in dementia. Acta Neuropathol Commun. 2016 Jun 13;4(1):58.
  5. Desikan RS, Ségonne F, Fischl B, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006 Jul 1;31(3):968-80.
  6. Ossenkoppele R, Rabinovici GD, Smith R, et al. Discriminative Accuracy of [18F]flortaucipir Positron Emission Tomography for Alzheimer Disease vs Other Neurodegenerative Disorders. JAMA. 2018 Sep 18;320(11):1151-1162.
  7. Veraart J, Fieremans E, Novikov DS. Diffusion MRI noise mapping using random matrix theory. Magn Reson Med. 2016 Nov;76(5):1582-1593.
  8. Andersson JLR, Sotiropoulos SN. An integrated approach to correction for off-resonance effects and subject movement in diffusion MR imaging. Neuroimage. 2016 Jan 15;125:1063-1078.
  9. Raffelt DA, Tournier JD, Smith RE, et al. Investigating white matter fibre density and morphology using fixel-based analysis. Neuroimage. 2017 Jan 1;144(Pt A):58-73.
  10. Dhollander T, Mito M, Raffelt, et al. Improved white matter response function estimation for 3-tissue constrained spherical deconvolution. In Proc. 27th International Society of Magnetic Resonance in Medicine 555 (2019). Montréal, Québec, Canada.
  11. MRI Atlas of Human White Matter. AJNR Am J Neuroradiol. 2006 Jun;27(6):1384–5.
  12. Braak, H. & Braak, E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol. 82, 239–259 (1991).
  13. Stancu IC, Vasconcelos B, Ris L, et al. Templated misfolding of Tau by prion-like seeding along neuronal connections impairs neuronal network function and associated behavioral outcomes in Tau transgenic mice. Acta Neuropathol. 2015 Jun;129(6):875-94.
  14. Dhollander T, Clemente A, Singh M, et al. Fixel-based Analysis of Diffusion MRI: Methods, Applications, Challenges and Opportunities. Neuroimage. 2021 Nov 1;241:118417.

Figures

Table 1. Demographic information of participants

Age, ICV, and clinical scores with p-values from the one-way analysis of variance (Post hoc: Tukey), sex from chi-squared test; ICV=intracranial volume (cm3); MMSE=mini-mental state examination; MoCA=Montreal cognitive assessment; CDRSB=clinical dementia rating sum of boxes; TAU_METAROI=18F-flortaucipir SUVR of temporal meta-ROI. Temporal meta-ROI comprises the bilateral entorhinal, amygdala, fusiform, inferior, and middle temporal cortices, forming a composite ROI.


Table 2. The acquisition parameters of Diffusion-weighted Images


Figure 1. Results of whole-brain Fixel-based Analyses

Streamlines are only shown along the parts of the pixels that exhibited significant changes in FBA metrics (FWE-corrected p<0.05). The streamlines are then color-coded to indicate their direction: green for anterior-posterior, blue for feet-head, and red for left-right.


Figure 2. Comparison of the effect size of SUVR in the cortex and the effect size of whole-brain log-FC

(a.) shows the Cohen’s d in each comparison of 18F-flortaucipir standardized uptake value ratio (SUVR) in cortices that were significantly lower in Alzheimer’s disease (AD) (upper panel) and significantly lower in mild cognitive decline (MCI) compared to CN (lower panel). The (b.) shows the streamlines encoded by the percentage effect (%) of log-transformed fiber-bundle cross-section (FC) in AD (upper panel) and MCI (lower panel) compared to CN.

Figure 3. Mediation Analysis for Fixel-based Analysis metrics

Diagram of pathways showing the unmediated effect between tau deposition (TAU) and the total mini-mental state examination score (MMSE; lower diagram) and the mediation model including pathways when fixel-based analysis metrics are mediated (upper diagram). Each effect on each path (Ba, Bb, Bc’, Bc) is noted on arrows. In the diagram of the mediation model, the indirection effect and its ratio to the direct effect are noted in the light gray box.


Proc. Intl. Soc. Mag. Reson. Med. 32 (2024)
3912
DOI: https://doi.org/10.58530/2024/3912