Gray and White Matter Involvement in Alzheimer Disease: A Diffusion Tensor and Magnetization Transfer Imaging Study In Vivo

Methods We studied 18 patients with clinically probable AD (median MiniMental State Examination [MMSE] score: 21; range=2-26), with 16 sex- and age-matched healthy subjects. Patients with 4 or more macroscopic lesions on T2-weighted (T2W) scan were excluded from the analysis. The following scans were obtained: a) dual-echo turbo spin echo (TSE) (TR/TE/NEX=3300/16-98/ 1; number of slices: 24, contigous, 5 mm thick); b) 2D gradient-echo (GE) (TR/TE/NEX=640/12/2; number of slices: 24, contigous, 5 mm thick) with and without an off-resonance radio-frequency (RF) saturation pulse (offset frequency=1.5 kHz, Gaussian envelope duration= 7.68 ms, flip angle=500o); c) pulsed-gradient spin-echo echo-planar (EPI) pulse sequence (inter-echo spacing=0.8, TE=123); number of slices: 10, contigous, 5 mm thick), with diffusion gradients applied in 8 non collinear directions. (maximum b factor: 1044 sec mm-2). For the EPI scans, the 10 slices were acquired with the same orientation of the dual-echo and GE scans positioning the second-last caudal slice in order to match exactly the central slices of the dual-echo and MT sets. This brain portion was chosen since it contains all the frontal and the occipital lobes, a large portion of the temporal lobes and the basal ganglia (globus pallidus, putamen, caudate nucleus and thalamus), and also because it is less affected by B0 distortions. Two experienced observers, unaware to whom the scans belonged, identified by consensus the pathological T2 hyperintensities in patients and controls. After co-registration of the two GE images, MTR maps were derived pixel-by-pixel. Finally, extra-cerebral tissue was removed from MTR maps using a local thresholding technique. For the subsequent analysis, only the 10 slices corresponding to the EP images were considered. After correction for eddy current induced distortion, the diffusion tensor was estimated linearly for every voxel, assuming a mono-exponential relationship between the signal attenuation and the elements of the tensor matrix. After diagonalization of the matrix, MD and FA were derived for every voxel. The b=0 step of EP images was coregistered on the T2W scans. The registration parameters were then used to transform MD and FA maps. Similarly, MTR maps were coregistered with the T2W images. WM and GM were segmented using a technique based on FA thresholding (4). MTR and MD normalized histograms were created and analyzed as previously described (5). For each subject the histograms of NAGM of the different brain lobes and from the GM component of the basal ganglia were also created, segmenting these regions on the correspoding coregistered T2 images and superimposing the obtained regions of interest (ROIs) on NAGM and GM MD and MTR maps. Parietal lobes were excluded from the analysis because they were only partially imaged by the selected slices. Results The peak heights of both MD and MTR histograms from the whole brain were significantly lower in AD patients compared to controls (p<0.0001). Average MD was also significantly higher in patients (p<0.0001). NAGM MD histogram metrics and NAGM MTR histogram peak height were all significantly different in patients and controls (Table 1). No significant differences were found in the NAWM MTR metrics, whereas an increase of the average MD (p=0.007) and a decrease of the MD peak height (p=0.01) were observed in patients with AD. When comparing the NAGM from the segmented lobes of the patients with the corresponding gray matter of the controls, MD histogram peak height from the frontal lobe was lower in AD patientsthan in controls (p=0.05). In the temporal lobes, MTR and MD peak height were significantly lower, while average MD and MD peak location were significantly higher in patients than in controls (Table 2). No differences were found either in the occipital lobes, or in the segmented deep gray matter. MMSE score was directly correlated with temporal lobes MTR peak height (r=0.5, p=0.05). Disease duration was significantly correlated with white matter MD peak height (r=-0.46, p=0.05). Discussion This study presents a novel MR approach to assess in vivo GM and WM pathology in AD. It confirms that intrinsic cortical GM changes, which are likely to be secondary to neuronal loss, can be measured reliably with MTI and DT-MRI in patients with AD.We also detected changes in the WM which might be the result of secondary wallerian degeneration. The correlation found between same of the MR metrics used in this study and the MMSE score and disease duration are encouraging and suggest the potential of this approach to monitor the evolution of AD.