Xuan Vinh To1, Hongjiang Wei2, Reshmi Rajendran1, Marta Garcia-Miralles3, Ling Yun Yeow1, Chunlei Liu2, Hong Xin1, Mahmoud A. Pouladi3, and Kai-Hsiang Chuang1
1Singapore Bioimaging Consortium, Agency for Science, Technology and Research (A*STAR), Singapore, Singapore, 2Brain Imaging and Analysis Center, School of Medicine, Duke University, Durham, NC, United States, 3Translational Laboratory in Genetic Medicine, Agency for Science, Technology and Research (A*STAR), Singapore, Singapore
Synopsis
This study looked at the potential for Quantitative Susceptibility Mapping (QSM) for longitudinal detection of demyelination and iron accumulation in YAC128 mouse model of Huntington disease. Control and YAC128 mice were scanned at 9, 12, and 15 months of age; with a number of mice sacrificed after each timepoint for histological validation and correlation (ongoing). Current results shows the potential for QSM in detecting demyelination is several white matter regions and iron accumulation in grey matter.Introduction
Huntington disease (HD) is an inherited
neurodegenerative disorder with motor, cognitive, and psychiatric deficits [1].
White matter (WM) atrophy has been reported in diffusion MR studies of HD
patients though what drove the changes is not clear [2-4]. Iron accumulation in
HD patients’ brains has been shown in postmortem brain analysis [5]. Quantitative
Susceptibility Mapping (QSM) is a novel method for estimation of magnetic
susceptibility in tissue from phase change of 3D gradient-echo MRI [6-7]. QSM
has been used to characterize microstructural change in white matter in
developing mouse brain [8], e.g., demyelination, and iron accumulation in human
HD patients [9]. In this study, we perform longitudinal QSM to characterize
white matter pathology and iron deposition in the YAC128 mouse model of HD that
expresses the full length human mutant huntingtin (mHTT) gene [10].
Method
The study was approved by the IACUC of Biomedical
Sciences Institutes, Singapore. 11 WT mice (5 males, 6 female) and 14 YAC128
mice (7 males, 7 females) were used in this study. Imaging was conducted on a
7T MRI (ClinScan, Bruker BioSpin, Germany) with a 4 channel array coil. 3D Multi-echo
gradient-echo images were acquired with the following parameters: TR=60ms, TE1/𝛥TE/TE6=6.26/7.77/45.11ms,
FOV=27x27x18mm3, matrix size=192x192x128, and 6 repetitions. Total
scan time was approximately 72 minutes. MR acquisitions were done at 9, 12, and
15 months of age. After each scan, 3-4 animals from each group were sacrificed
for histological validation. The raw phase was processed by Laplacian-based
phase unwrapping and V_SHARP background phase removal [7]. QSM maps were
reconstructed using a two-level STAR-QSM algorithm [11] for reducing streaking
artifacts.
For group analysis, magnitude image was linearly
registered to an in-house created T2-weighted fast spin echo-based anatomical
template based on YAC128 and WT mice of 9 months old. The registered images
were averaged to create age-specific template for further non-linear
registration using FSL’s FNIRT [12]. The QSM maps were warped to the template and
compared by t-test using SPM8 [13]. Regions-of-interest (ROIs) were chosen on
the statistical maps.
Results
Voxel-wise comparison (Fig.1) showed YAC128 mice having
decreased susceptibility in the cingulate cortex (Cg) and caudate putamen
(CPu), but increased susceptibility in the corpus callosum (cc), dorsal
subiculum (DS), globus palladius (GP), and inferior colliculus (IC) as early as
9 months old. ROI analysis (Fig. 2) suggested that increased susceptibility in
DS, GP and IC may be due to iron deposition while the reduced negative
susceptibility in cc may relate to demyelination.
Discussion
This study demonstrates the potential application
of QSM to characterize tissue property changes which may reflect demyelination
and iron accumulation in HD. The trend was not consistent in the voxel-wise
statistical map, which might be due to the decreased sample size over time. The
result is consistent with known pathology of YAC128 mice and HD patients. Further
histological analysis is being performed to confirm the QSM findings.
Acknowledgements
No acknowledgement found.References
[1] F. O. Walker, Lancet, vol. 369, pp. 218-28.
20 Jan 2007. [2] H. D. Rosas, et al., Mov
Disord, vol. 21, pp. 1317-25, Sep 2006. [3] K. E. Weaver, et al., Exp Neurol, vol. 216, pp. 525-9, Apr 2009. [4] M. Di Paola, et al., Cereb Cortex, vol. 22, pp. 2858-66, Dec 2012. [5] H. D. Rosas, et al., Arch Neurol, vol 69(7), pp.
887-93, July 2012. [6] N. Li, Magn.
Reson. Med., vol 46, pp. 907–916, 18 Oct 2001 [7] Li W, et al., Neuroimage, vol. 55, pp. 1645–56. 15 Apr 2011. [8] I. Argyridis, et al., NeuroImage, vol 88, pp.134-42. Nov 22 2013. [9] D. JF. Domíniguez, et al., Neurol Neurosurg Psychiatry, vol 0, pp. 1-5. Epub first 19 April
2015. [10] E. J. Slow, et al., Hum Mol Genet, vol. 12, pp. 1555-67, 1
Jul 2003. [11] H. Wei, et al., NMR Biomed, vol. 28, pp. 1294-303. 24
July 2015. [12] M. Jenkinson, et al., Neuroimage, vol. 62, pp. 782-90, Aug 15
2012. [13] W. Penny, et al., Statistical Parametric Mapping: The Analysis
of Functional Brain Images, 1st ed.: Academic Press, 2006