Jialin Hu1, Miao Zhang2, Yaoyu Zhang1, Rong Guo3,4, Yudu Li3,4, Yibo Zhao3,4, Ziyu Meng1, Biao Li2, Jun Liu5, Binyin Li5, Jie Luo1, Chao Ma6, Georges El Fakhri6, Zhi-Pei Liang3,4, and Yao Li1
1School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China, 2Department of Nuclear Medicine, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China, 3Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 4Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL, United States, 5Department of Neurology and Institute of Neurology, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China, 6Gordon Center for Medical Imaging, Massachusetts General Hospital, Harvard Medical School, Boston, MA, United States
Synopsis
Beta-amyloid (Aβ) aggregation and
neurometabolic changes are biomarkers for Alzheimer’s
disease (AD) at early stage. But their underlying relationship in association
with AD pathology is still not fully understood. In this
study, we simultaneously acquired 3D high-resolution MRSI using SPICE and 18F-AV45-PET images
on a PET-MR scanner. Concurrent changes in neurometabolite
concentrations and Aβ deposition in healthy control, mild cognitive impairment,
and AD groups were compared. An increase in myo-inositol and a decrease in N-acetylaspartate
were found as dementia became more severe. Our findings may lay a foundation
for further investigation of AD pathogenesis using multimodal metabolic
imaging.
Introduction
Beta-amyloid (Aβ) aggregation
measured by PET and neurometabolic changes measured by Magnetic Resonance Spectroscopy
(MRS) have been used as pathological biomarkers for Alzheimer’s disease (AD) at
early stage.1,2 For example, reduction in N-acetylaspartate (NAA),
increase in myo-inositol (mI) and increase in Aβ
plaques were closely related with disease progression in
AD.3,4 However, the underlying relationship between these
neurometabolites changes and Aβ aggregation
associated with AD is still not fully understood. Magnetic resonance
spectroscopic imaging (MRSI) has long been recognized as a potentially powerful
tool for noninvasive neurometabolic imaging, but most existing MRSI studies in
AD were performed using single-slice MRSI or single-voxel MRS techniques with low
spatial resolution. Moreover, the concomitant Aβ and
neurometabolic changes in AD have not been investigated yet due to the lack of
imaging technology. In this study, we performed simultaneous 3D high-resolution
1H-MRSI and 18F-AV45-PET imaging on a PET-MR scanner. Neurometabolic
mapping at 2.0 x 3.0 x 3.0 mm3 nominal resolution was achieved using
the latest SPICE (SPectroscopic Imaging by exploiting spatiospectral
CorrElation) sequence.5,6 Concurrent neurometabolites and Aβ deposition
changes across healthy control (HC), mild cognitive impairment (MCI), and AD groups
were compared both globally and locally. Their relationships with cognitive
decline in AD patients were also investigated.Method
There were 26 healthy controls, 22 MCI patients and
35 AD patients recruited for this study. The PET/MRI scans were performed on a whole-body
simulatenous PET/MR system (Biograph mMR; Siemens, Germany) with IRB approved
by Ruijin Hospital, Shanghai, China. The MR scan protocols included 3D MRSI
acquired by the SPICE sequence (2.0
× 3.0 × 3.0 mm3, FOV = 240 × 240 × 72 mm3, TR/TE = 160/1.6
ms) and 3D MPRAGE imaging (0.5 × 0.5 × 1.0 mm3, FOV = 256 × 256 × 192
mm3, TR/TE = 1900/2.44 ms). A union-of-subspaces model incorporating
pre-learned spectral basis functions was performed for reconstruction of the
spatiospectral functions of metabolites.5-7 Spectral
quantification was done using an improved LCmodel-based algorithm that incorporated
both spatial and spectral priors.6,8 Regions of interest (ROIs) masks were created based on
T1-weighted images using FreeSurfer v6.0.
The
PET data were acquired at 45~60 mins post a bolus injection of 18F-AV-45
at 3.7 MBq/kg (matrix size = 344 × 344, voxel size = 2.1 × 2.1 × 2.0 mm3, 127 slices).
After correction of random events, dead time, scattering and attenuation, the reconstructed
PET images were preprocessed using a 2mm full width half-maximum (FWHM)
Gaussian filter. Regional PET data were quantified using the standard uptake
value ratio (SUVR) as the ratio to whole cerebellum mean uptake. The
global cortical SUVR was calculated as the average of a composite region containing
frontal, anterior/posterior cingulate, lateral parietal and lateral temporal
regions. We
used the global SUVR cutoff of 1.11 to determine amyloid positivity and classified the subjects into the Aβ+ and Aβ- groups.9 MRSI and processed PET images of each subject were
coregistered to the T1-weighted image using affine transformation.
Between-group
comparisons of metabolic levels were performed using univariate
general linear models, adjusting for age, sex, education and regional volume. Partial
regression analysis was used to evaluate the associations between neurometabolic
levels and MMSE scores, controlled for age, sex and education level. All the statistical
analyses were conducted using SPSS.Results and Discussion
Figure 1 shows some representative
neurometabolic maps and PET images of subjects from the HC, MCI, and AD groups,
respectively. A global NAA reduction and mI elevation could be observed along
with increased dementia severity. Figure 2 shows a comparison of spatially
resolved spectra obtained from posterior cingulate cortex (PCC)/precuneus and
hippocampus area from subjects of the HC, MCI and AD groups. Neurometabolic concentrations
were compared among different groups in Figure 3. For the
global composite regions with Aβ deposition, the mI/Cr concentration in AD
patients was significantly higher than that in MCI patients, which was higher
than that in the HC group. The NAA/Cr significantly reduced in AD group,
but was not different between the MCI and HC groups. The changes of
neurometabolic signals in PCC/precuneus followed a similar pattern to the global
comparison. A different pattern of signal change was observed in hippocampus as
a low Aβ deposition area.10 The
comparisons of different metabolic biomarkers between the Aβ-
and Aβ+ groups were shown
in Fig. 4. There was a globally elevated mI/Cr and a reduced
NAA/Cr in the Aβ+ group compared with the Aβ-
group, consistent with previous literature.11 In hippocampus, mI/Cr
elevated in the Aβ+ group although SUVR
showed no differences between the two groups. Moreover,
NAA/Cr was correlated with MMSE score in MCI/AD patients as shown in Fig. 5,
indicating its contribution to patients’ cognitive deficits.Conclusion
This study demonstrated simultaneous 3D high-resolution
MRSI and PET imaging and investigated association
of neurometabolism with Aβ aggregation in
HC, MCI and AD patients. An increase in mI and a decrease in NAA were found as
dementia severity increased. Our findings may lay a foundation for further
investigation of AD pathogenesis using multimodal metabolic imaging.Acknowledgements
Y. L. is funded by National Science
Foundation of China (No.61671292 and 81871083) and Shanghai Jiao Tong
University Scientific and Technological Innovation Funds (2019QYA12).References
- Jack CR Jr, Knopman DS, Jagust
WJ, et al. Hypothetical model of dynamic biomarkers of the Alzheimer's
pathological cascade. Lancet Neurol. 2010;9(1):119-128.
- Murray ME, Przybelski SA,
Lesnick TG, et al. Early Alzheimer's disease neuropathology detected by proton
MR spectroscopy. J Neurosci. 2014;34(49):16247-16255.
- Kantarci K, Jack CR Jr, Xu YC,
et al. Regional metabolic patterns in mild cognitive impairment and Alzheimer's
disease: A 1H MRS study. Neurology. 2000;55(2):210-217.
- Kantarci K, Knopman DS, Dickson
DW, et al. Alzheimer disease: postmortem neuropathologic correlates of
antemortem 1H MR spectroscopy metabolite measurements. Radiology.
2008;248(1):210-220.
- Lam F, Liang ZP. A subspace
approach to high-resolution spectroscopic imaging. Magn Reson Med.
2014;71(4):1349-1357.
- Ma C, Lam F, Johnson CL, et al.
Removal of nuisance signals from limited and sparse 1H MRSI data
using a union-of-subspaces model. Magn Reson Med. 2016;75(2):488-497.
- Lam F, Ma C, Clifford B, et al.
High-resolution 1H-MRSI of the brain using SPICE: data acquisition
and image reconstruction. Magn Reson Med. 2016;76(4):1059-1070.
- Li Y, Lam F, Clifford B, et al.
A subspace approach to spectral quantification for MR spectroscopic imaging.
IEEE Trans Biomed Eng. 2017;64(10):2486-2489.
- Mattsson N, Insel PS, Donohue
M, et al. Independent information from cerebrospinal fluid amyloid-β and florbetapir
imaging in Alzheimer's disease. Brain. 2015;138(Pt 3):772-783.
- Braak H, Braak E.
Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol.
1991;82(4):239-259.
- Sheikh-Bahaei N, Sajjadi SA,
Manavaki R, et al. Positron emission tomography-guided magnetic
resonance spectroscopy in Alzheimer disease. Ann Neurol. 2018;83(4):771-778.