507
Journal of Alzheimer’s Disease 22 (2010) 507–522 DOI 10.3233/JAD-2010-100234 IOS Press
Longitudinal Changes in Fiber Tract Integrity in Healthy Aging and Mild Cognitive Impairment: A DTI Follow-Up Study Stefan J. Teipela,b,∗, Thomas Meindlc , Maximilian Wagnerd , Bram Stieltjese , Sigrid Reutera,b , Karl-Heinz Hauensteinf , Massimo Filippig , Ulrike Ernemannh, Maximilian F. Reiserc and Harald Hampeld,i,j a
Department of Psychiatry, University Rostock, Rostock, Germany DZNE, German Center for Neurodegenerative Diseases, Rostock, Germany c Institute for Clinical Radiology, Ludwig-Maximilian University, Munich, Germany d Department of Psychiatry, Ludwig-Maximilian University, Munich, Germany e Division of Radiology, German Cancer Research Center, Heidelberg, Germany f Department of Radiology, University Rostock, Rostock, Germany g Institute of Experimental Neurology, Scientific Institute and University Hospital San Raffaele, Milan, Italy h Department of Diagnostic and Interventional Neuroradiology, University Hospital T¨ubingen, Germany i Discipline of Psychiatry and Trinity College Institute of Neuroscience (TCIN), Trinity College Dublin, The Adelaide and Meath Hospital incorporating the National Children’s Hospital (AMiNCH), Dublin, Ireland j Department of Psychiatry, University of Frankfurt, Frankfurt/Main, Germany
b
Handling Associate Editor: Uwe Friese Accepted 2 July 2010
Abstract. Cross-sectional studies using diffusion tensor imaging (DTI) suggest decline of the integrity of intracortically projecting fiber tracts with aging and in neurodegenerative diseases, such as Alzheimer’s disease (AD). Longitudinal studies on the change of fiber tract integrity in normal and pathological aging are still rare. Here, we prospectively studied 11 healthy elderly subjects and 14 subjects with amnestic mild cognitive impairment (MCI), a clinical risk group for AD, using high-resolution DTI and MRI at baseline and after 13 to 16 months follow-up. Fractional anisotropy (FA), a DTI measure of fiber tract integrity, was compared across time points and groups using a repeated measures linear model and tract based spatial statistics. Additionally, we determined rates of grey matter and white matter atrophy using automated deformation based morphometry. Healthy elderly subjects showed decline of FA in intracortical projecting fiber tracts, such as corpus callosum, superior longitudinal fasciculus, uncinate fasciculus, inferior fronto-occipital fasciculus, and cingulate bundle (p < 0.05, corrected for multiple comparisons). MCI subjects showed significant FA decline predominantly in the anterior corpus callosum (p < 0.05, corrected for multiple comparisons). Grey and white matter atrophy involved prefrontal, parietal, and temporal lobe areas in controls and prefrontal, cingulate, and parietal lobe areas in MCI subjects and agreed with the pattern of fiber tract changes. Our findings indicate that DTI allows detection of microstructural changes in subcortical fiber tracts over time that are related to aging as well as to early stages of AD type neurodegeneration. The underlying mechanisms for these changes are unknown. Keywords: Cortical connectivity, fractional anisotropy, longitudinal study Supplementary data available online: http://www.j-alz.com/issues/22/vol22-2.html#supplementarydata05
∗ Correspondence to: Stefan J. Teipel, M.D., Department of Psychiatry and Psychotherapy, University Rostock, Gehlsheimer Str. 20,
18147 Rostock, Germany. Tel.: +49 01149 381 494 9610; Fax: +49 01149 381 494 9682; E-mail:
[email protected].
ISSN 1387-2877/10/$27.50 2010 – IOS Press and the authors. All rights reserved
508
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
INTRODUCTION Neuroimaging and postmortem studies indicate a wide range of morphological brain changes with aging. Magnetic resonance imaging (MRI) studies have demonstrated cortical grey matter loss predominantly located in prefrontal regions [1], but also resembling pattern of grey matter loss in Alzheimer’s disease (AD) [2,3]. Additionally, reduction of cerebral white matter has been described in healthy aging with predominance in prefrontal lobe regions but extending to temporal and parietal lobe areas [4,5]. The extent of white matter changes in association tracts was correlated with decline of regional cortical glucose consumption [6]. Postmortem studies suggest a relatively moderate decline of neuron numbers across age groups. For example, neuron numbers were reported to decline by less than 10% over an age range from 20 to 90 years [7]. In contrast, length of subcortical fibers was reported to decline by 40% between the age of 20 and 80 years [7, 8]. Therefore, white matter changes seem to play an important role in the development of age-related changes in brain structure and function. Diffusion tensor imaging (DTI) has evolved as a powerful technique to determine fiber tract integrity in the living human brain [9]. Fractional anisotropy (FA) derived from DTI data is among the best established scalar markers of fiber directionality and integrity [10,11]. Knowledge on the underlying neurobiological basis of FA reductions in aging and neurodegeneration as determined by DTI is limited. Subcortical ischemic lesions have frequently been associated with aging and can reduce FA values [12]. Studies on brain maturation suggest that FA is sensitive to increases in axonal growth and myelination [13,14]. FA changes have also been suggested to reflect changes in axonal membranes [11]. Degradation of myelin and even axon deletion accompany normal aging [15,16]. Consistently, a broad range of cross-sectional studies has shown significant age-related changes in fiber tract integrity [17–19]. Additionally, studies in early stages of AD suggest a specific decline of subcortical fiber tract integrity [20–22]. Subjects with amnestic mild cognitive impairment (MCI) are believed to represent a predementia stage with an increased risk to develop AD during clinical follow-up [23]. Consistently, subjects with MCI show reduced fiber tract integrity in association fiber tracts [24–26]. Longitudinal studies on fiber tract changes using DTI in aging and neurodegenerative disease are still rare. Over three months follow-up fractional anisotropy in
fornix, cingulum, splenium, and cerebellar peduncle remained stable in MCI, AD, and healthy elderly subjects [27]. Over a longer follow-up period of 15months, a single case study of a subject with posterior cortical atrophy detected decline of fractional anisotropy in the white matter in the absence of substantial decline of grey matter volume [28]. Healthy elderly subjects showed significant decline of FA after 2 years of follow-up [29]. In the present study, we investigated longitudinal changes of fractional anisotropy in healthy elderly subjects and subjects with MCI over an average follow-up time of 13 months in MCI subjects and 16 months in controls. We compared changes in white matter tract integrity with regional decline of grey and white matter volumes. We used a cross-sectional analysis of FA values derived from the multi-center study of a novel physical DTI phantom [30] to establish an upper-limit for the estimate of unsystematic within-scanner variability in FA values over time. We expect that our data will help to determine the role of regional changes of white matter tract integrity both in healthy aging and in predementia stages of AD.
SUBJECTS AND METHODS Subjects We examined 14 right-handed subjects with the clinical diagnosis of amnestic MCI [mean age: 73.1 (SD 7.4) years, ranging from 60 to 88 years, 6 women, mean years of education: 11.2 (SD 2.3)] and 11 healthy elderly right-handed subjects [mean age: 67.4 (SD 7.7) years, ranging from 59 to 83 years, 4 women, mean years of education: 12.7 (SD 3.6)]. MCI subjects were recruited from the Memory Clinic at the University Munich and fulfilled the Mayo clinic criteria [31] for amnestic MCI. The healthy subjects were recruited among spouses of patients with dementia attending the Memory Clinic at the University Munich, who had no subjective memory complaints and scored within one standard deviation of the age- and education adjusted means of the Mini-Mental-Status Examination (MMSE) [32], CERAD cognitive battery [33], Clock-drawing-test [34] and the trail-making test [35]. MCI subjects and controls were not significantly different in mean age (T = −1.9, 23 df, p = 0.07), gender distribution (Chi2 = 0.11, Fisher’s exact test p = 0.53), years of education (T = 1.30, 23 df, p = 0.21), or num-
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
ber of apolipoprotein E4 (ApoE4) alleles (Chi2 = 0.12, Fisher’s exact test p = 0.55). As expected, both groups differed significantly in MMSE scores, with a mean MMSE score of 26.5 (SD 1.2) in the MCI subjects and 29.3 (SD 0.6) in the controls. All subjects underwent clinical assessment, neuropsychological testing, and MRI scanning both at baseline and follow-up. Duration of follow-up was on average 404 (SD 44, 311 to 456) days in MCI subjects and 472 (SD 188, 347 to 966) days in controls and was not significantly different between groups (T = 1.31, 23 df, p = 0.20). The clinical assessment included detailed medical history, clinical, psychiatric, neurological and neuropsychological examinations, and laboratory tests (complete blood count, electrolytes, glucose, blood urea nitrogen, creatinine, liver-associated enzymes, cholesterol, HDL, triglycerides, serum B12, folate, thyroid function tests, coagulation, serum iron). Additionally, ApoE4 genotyping was performed (see below). Selection of subjects included a semiquantitative rating of T2-weighted MRI scans [36]. To exclude subjects with significant subcortical cerebrovascular lesions, only subjects were included which had no subcortical white matter hyperintensities exceeding 10 mm in diameter or 3 in number. Subjects’ consent was obtained according to the Declaration of Helsinki. The study was approved by the ethics committee and the local authorities. Apolipoprotein E genotyping ApoE genotyping was performed using a polymerase chain reaction (PCR) kit for the Light Cycler system for real-time PCR (Roche Diagnostics, Mannheim, Germany). In the MCI group, there were 3 of 14 subjects carrying at least one ApoE4 allele, in the control group there were 3 of 11 subjects carrying at least one ApoE4 allele.
509
0.8 mm3 , repetition time 14 ms, echo time 7.61 ms, flip angle 20◦ , number of slices 160). To identify white matter lesions a 2-dimensional T2-weighted sequence was performed (fluid attenuation inversion recovery FLAIR, field-of-view 230 mm, repetition time 9000 ms, echo time 117 ms, voxel size 0.9 × 0.9 × 5.0 mm, TA 3:20 min, flip angle 180◦ , number of slices 28, acceleration factor 2). Diffusion-weighted imaging was performed with an echo-planar-imaging sequence (field-of-view 256 mm, repetition time 9300 ms, echo time 102 ms, voxel size 2 × 2 × 2 mm3 , 4 repeated acquisitions, b-value = 1000, 12 directions, noise level 10, slice thickness 2.0 mm, 64 slices, no overlap). Parallel imaging was performed with a generalized auto-calibrating partially parallel acquisition (GRAPPA, [37]) reconstruction algorithm and an acceleration factor of 2. Phantom study We used a physical phantom object manufactured by the Division of Radiology of the German Cancer Research Center, Heidelberg, Germany. The properties of this phantom have previously been described in detail [30]. The phantom consists of polyamide fibers of 15 µm diameter winded around a circular acrylic glass spindle. The phantom was measured at five different Siemens Avanto 1.5 Tesla MRI scanners (Siemens Medical Solutions, Erlangen), using echo-planar-imaging sequences (repetition time between 5100 ms and 5300 ms, echo time 85 ms, voxel size 2 × 2 × 2 mm3 , 3 repeated acquisitions, b-value = 1000) At three sites 30 gradient directions were used, at the other two sites only 12 gradient directions were available due to software restrictions. Parallel imaging was performed with a generalized auto-calibrating partially parallel acquisition (GRAPPA, [37]) reconstruction algorithm and an acceleration factor of 2. Diffusion tensor imaging analysis
MRI acquisition MRI acquisitions of the brain were conducted with a 3.0 Tesla scanner with parallel imaging capabilities (Magnetom TRIO, Siemens, Erlangen, Germany), maximum gradient strength: 45 mT/m, maximum slew rate: 200 T/m/s, 12 element head coil. For anatomical reference, a sagittal high-resolution 3-dimensional gradient-echo sequence was performed (magnetization prepared rapid gradient echo MPRAGE, field-of-view 250 mm, spatial resolution 0.8 × 0.8 ×
DTI data were analyzed using tract-based spatial statistics (TBSS) v1.2 implemented in FSL 4.1 (FMRIB Analysis Group, Oxford, UK, http://www.fmrib.ox.ac. uk/analysis/research/tbss) [38]. TBSS has been used in one previous longitudinal study on Huntington’s disease [39]. TBSS allows spatial reorientation of fractional anisotropy (FA) maps into standard space without systematic effects of spatial transformation on fiber tract directionality and without the need to select a spatial smoothing kernel that may impact the effects of
510
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
interest. The approach is based on the use of an a priori template of fiber tracts to which the individually extracted fiber tracts of each subject are coregistered. First, FA images were created by fitting a tensor model to the raw diffusion data using FDT (FMRIB’s Diffusion Toolbox integrated in the FSL software), and then brain-extracted using BET (Brain Extraction Tool) [40]. All subjects’ FA data were then aligned into a common space using the nonlinear registration tool FNIRT [41,42], which uses a b-spline representation of the registration warp field [43]. Next, the mean FA image was created and thinned to obtain a mean FA skeleton which represents the centers of all tracts common to the group. Each subject’s aligned FA data was then projected onto this skeleton and the resulting data fed into voxelwise within-subject statistics. In addition to voxel-wise analysis of fiber tract maps, we extracted mean FA values in predefined fiber tracts using a region of interest approach. We used white matter regions in MNI standard space as represented in the John Hopkins University (JHU) Fiber Tract–based Atlas of Human White Matter Anatomy [44]. This atlas template is implemented in the FSL software. We used an in-house written Matlab script to overlay the white matter atlas regions with the individual fiber tract skeletons. We extracted mean FA values for the corpus callosum genu, body and splenium, and fornix as well as right and left superior longitudinal fasciculus and cingulum, as indicated in Fig. 1. Phantom data analysis From the DTI data obtained from the physical phantom object, we reconstructed the FA maps using DTI studio version 2.4.01 (Laboratory of Brain Anatomical MRI and Center for Imaging Science at Johns Hopkins University, Baltimore, USA, available through http://www.mristudio.org). FA values then were measured at the most superior, inferior, left and right position of the circular FA map using circular regions of interest (ROI) with 3 mm diameter (Fig. 2). Structural MRI data analysis Structural MRI data were analyzed using deformation based morphometry based on diffeomorphic image registration as implemented in the DARTEL toolbox in spm8 [45], available through http://www.fil.ion.ucl.ac. uk/spm. We followed the steps proposed by a previous publication [46] and modified them according to the DARTEL framework. The modification
was inspired by a discussion on the spm archives website (available at https://www.jiscmail.ac.uk/cgibin/wa.exe?A2=ind0804&L=SPM&P=R48484). The following procedure was applied to the MRI scans: 1. Scans were manually aligned to place the anterior commissure at the origin of the three-dimensional Montreal Neurological Institute (MNI) coordinate system. The first scan of each subject then was rigidly registered to the second scan using maximizing the mutual information of the joint intensity histogram of the images [47]. 2. The first and second scans were segmented into grey matter and white matter maps [48]. The resulting grey (white) matter maps were rigidly aligned to each other. 3. The grey (white) matter maps from the first scans were warped to the grey (white) matter maps from the second scans using high-dimensional diffeomorphic image registration [45]. This step results in a deformation field that contains a mapping from each point in the first image to the corresponding point in the second image for each subject. The amount of regional expansion or contraction was extracted from this field, by taking the determinant of the gradient of the deformation at each point (Jacobian determinant) [49]. Additionally, this step yields a subject specific template across both time points. Due to the intra-individual matching of scans, data were not smoothed prior to warping. 4. The grey (white) matter maps of the first and second scan were warped with the resulting deformation field and modulated with the respective Jacobian determinant map to obtain modulated warped grey (white) matter maps that are in alignment across time points for each subject in the space of the subject specific template. 5. The subject specific templates derived from step 3 were warped to match each other across subjects using high-dimensional diffeomorphic image registration [45]. The resulting deformation fields and Jacobian determinant maps were applied to the grey (white) matter maps that were in the space of the subject specific templates from step 4. This resulted in the baseline and followup scans of each subject within a group specific common space where the effect of different time points was preserved in the modulated warped grey (white) matter maps.
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
A
511
B
Fig. 1. Regions of interest superimposed on the mean fiber tract skeleton. The regions of interest from the JHU Fiber Tract–based Atlas of Human White Matter Anatomy [44] superimposed on the averaged fiber tract skeleton (green), representing th e mean skeleton across the healthy controls and MCI subjects, in MNI space. A) Midsagittal section with genu (red), body (blue) and splenium (yellow) of the corpus callosum, and fornix (pink). B) Axial sections with cingulum (red) and fasciculus longitudinalis superior (blue).
Fig. 2. DTI phantom measurement. From the DTI scans of the physical phantom object (left hand side), we constructed color maps (middle) showing the fiber directions (red represents the direction ‘left–right’) and FA maps (right hand side). The FA values were measured from the FA maps at the most superior, inferior, left and right position using circular regions of interest with 3 mm diameter (red filled circles).
6. The group specific template defining the common space across subjects and time points was transformed to MNI standard space [50,51] using 12parameter affine transformation. The resulting parameters were applied to each scan in common space to transform scans into MNI space. Statistical analysis Neuropsychological data. We used a repeated measures linear model with time point as within subjects factor and diagnosis as between subjects factor to de-
termine changes in cognitive performance over time. The interaction of time by diagnosis was used to determine differences between groups in rates of cognitive change. DTI data We determined effects of time on values of the TBSS FA maps using a voxel-based repeated measures ANOVA model with time as repeated measures factor and diagnosis as between subjects factor. We determined the effect of time on FA values within each diagnostic group considering both possi-
512
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
ble directions, i.e., decline and increase of FA over time. Statistical significance was determined at p < 0.05, corrected for multiple comparisons using permutation test as implemented in the FSL 4.1 submodule Randomize (FMRIB Analysis Group, Oxford, UK, http://www.fmrib.ox.ac.uk/analysis/research/tbss) using 500 permutations and threshold-free cluster enhancement (TFCE) [52]. TFCE is a technique that does not need pre-statistical smoothing of images and does not depend on an initial clusterforming threshold such as t-statistic thresholding. However, TFCE needs several parameters to be set. Specifically these relate to the cluster extent (E) and peak height (H) of the statistic at a given voxel. As suggested by [52], these were set to E = 0.5 and H = 2.0 to provide a statistic that is sensitive to all levels of the signal. After TFCE voxel clusters were deemed significant at p < 0.05. The ROI data, representing mean FA in predefined fiber tracts, were compared across time points and groups using a repeated measures linear model with time as within subject factor, diagnosis and the interaction between time and diagnosis as between subjects factors. To evaluate whether differences in extent of atrophy between groups may be related to differences in the variability of fiber tracts across subjects, we determined the coefficient of variation at each single voxel within the fiber tracts as determined by TBSS. The coefficient of variation (CV) is defined as CV =
std(xi ) , mean(xi )
with xi equal to the FA value at voxel i. We determined the minimum and maximum value as well as the mean and median value of the CV across all voxel for each single group (MCI and controls) and each time point (baseline and follow-up). We also plotted the histograms and the spatial distribution of CVs for each group and time point to assess local differences in FA variability. Phantom data FA values were averaged across the four circular ROI positions within each scanner. We determined the CV across the five scanners as defined above. MRI data The modulated warped grey (white) matter maps representing absolute values of grey (white) matter density were smoothed using an isotropic Gaussian kernel at 8 mm FWHM. For statistical analysis we employed the
general linear model on a voxel basis implemented in SPM8 using repeated measures ANOVA with time as repeated measures factor and diagnosis as between subjects factor. We considered significant effects in both directions, i.e., decrease or increase of volumes over time. Results were thresholded at an uncorrected plevel < 0.001, and an extent threshold of 50 contiguous voxels was applied.
RESULTS Neuropsychological changes One of the 14 MCI subjects converted to dementia and AD during follow-up, further four subjects showed overall cognitive decline in MMSE scores but did still not meet criteria for dementia. The other nine MCI subjects remained cognitively stable over the follow-up period. None of the control subjects converted to MCI or dementia. After 13 to 16 months follow-up, there 1 was no significant change in MMSE score (F23 = 1.0, p = 0.3) across both groups and no significant group 1 by time interaction (F23 = 0.3, p = 0.6). There was significant decline in Boston naming test in the MCI subjects only, and a significant decline in word fluency across both groups. MCI subjects showed significantly lower performance at baseline compared to controls in the Clock drawing test (rated according to Shulman et al. [34]) and all subtests of the CERAD battery except the Boston naming and the Drawing test, consistent with the characterization of this group as amnestic MCI of several domains. For further details refer to Supplementary Table 1 (available online: http://www.j-alz.com/issues/22/vol22-2.html# supplementarydata05). Decrease in FA In the cross-sectional analysis of the baseline scans, there were no significant differences in FA values between MCI subjects and controls at p < 0.05 corrected for multiple comparisons. Healthy control subjects showed significant decline of FA in the corpus callosum, intrahemispherically projecting tracts such as the superior longitudinal fasciculus, the inferior frontooccipital fasciculus and the uncinate fasciculus as well as cortico-petal and cortico-fugal tracts of the corticospinal tract and the anterior thamalic radiation (Fig. 3) at p < 0.05, corrected for multiple comparisons. Patients with MCI showed FA reductions exclusively in
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
513
Table 1 FA changes in different white matter tracts Side Corpus callosum genu Corpus callosum body Corpus callosum splenium Fornix Cingulum Cingulum Superior long. fasciculus Superior long. fasciculus
R L R L
% change controls −1.53 −3 −0.89 −2.11 −2.23 −1.08 −0.44 −1.14
% change MCI −1.25 −2.5 −0.73 −10.64 −2.77 −2.33 −0.33 −0.19
1 main F23 effect time 8.93∗ 20.80∗∗∗ 6.51∗ 4.99∗ 7.93∗ 2.80 0.99 2.76
1 time F23 by diagnosis 0.17 0.42 0.05 3.35 0.02 0.11 0.01 1.59
Changes are given as percent change per year (relative to the mean value across both time points). F-values with 1 denominator and 23 nominator degrees of freedom for the main effect of time and the interaction effect of time by diagnosis. ∗ p < 0.05; ∗∗∗ p < 0.001; R – right; L – left.
Fig. 3. Reductions of fiber tract integrity in healthy controls. Reductions of FA in healthy controls determined using TBSS, p < 0.05, corrected for multiple comparisons using permutation test. Effects (red) are projected on a fiber tract template in MNI standard space (green). Axial slices go from MNI (Talairach-Tournoux) coordinate z = −12 to z = 43 (40) and are 5 mm apart in MNI space. A – anterior/P – posterior. L – left/R – right. S- superior/I – inferior. Abbreviations for tracts: ATR – anterior thalamic radiation; CCr – corpus callosum rostrum; CCs – corpus callosum splenium; CCt – corpus callosum truncus; Cg – cingulate; CST – corticospinal tract; FMi – forceps minor; FMa – forceps major; IFOF – inferior fronto-occipital fasciculus; SLF – superior longitudinal fasciculus; UF – uncinate fasciculus.
514
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
Fig. 4. Reductions of fiber tract integrity in MCI subjects. Reductions of FA in MCI subjects determined using TBSS, p < 0.05, corrected for multiple comparisons using permutation test. Effects (red) are projected on a fiber tract template in MNI standard space (green). Effects are located in corpus callosum rostrum and truncus. A – anterior/P – posterior. L – left/R – right. S – superior/I – inferior.
the corpus callosum (Fig. 4). The opposite contrast, i.e., FA increase over time, showed no effects in either of both groups. When we considered the interaction of time by diagnosis, there was no significant difference between the MCI and the control groups in rates of FA decline in either direction at a corrected p value of 0.05. Presence of an ApoE4 allele and gender had no effect on rates of FA decline at a corrected p value of 0.05. Similar to the voxelwise analysis, using ROIs to extract mean FA values from predefined fiber tracts we found significant reductions of FA in the corpus callosum genu, body and splenium, the fornix and the right cingulum. FA changes were not significant in the left cingulum and the bilateral fasciculus longitudinalis superior. Percent reduction of FA per year ranged between 0.20 and 10%. There was no significant time by diagnosis interaction, suggesting that rates of change
Table 2 Coefficient of variation of FA values, averaged across all tracts
Minimum Maximum Median Mean
Baseline Controls MCI 0.010 0.015 1.021 0.982 0.176 0.184 0.204 0.216
Follow-up Controls MCI 0.014 0.018 1.082 1.074 0.187 0.209 0.216 0.239
were not significantly different between the controls and the MCI subjects. For details refer to Table 1. We determined the coefficient of variation (CV) at each single voxel for each group and time point. The mean and median values of the CV were numerically higher at follow-up compared to baseline and higher in MCI compared to controls. However, the difference was relatively small, as can also be seen from the histograms (Table 2, Fig. 5). Local maps of the CV across the fiber tracts revealed no obvious differences between
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
515
Fig. 5. Histograms of the coefficients of variation of the FA values for both groups at follow-up. The frequency counts of the coefficients of variation (CV) are plotted for the follow-up data of the MCI group (green) and the control group (red). Although the distribution of the MCI group is slightly shifted to the right both distribution are not significantly different from each other.
MCI and control groups at both time points. CV values were lowest for the corpus callosum in both groups at both time points (data not shown). Phantom data The coefficient of variation across the five scanners was 6.1%. Decrease in cortical grey matter and in white matter density In the cross sectional analysis of the baseline scans, MCI subjects showed significantly reduced grey matter of bilateral lateral and medial temporal lobes as well as bilateral superior parietal lobes compared to controls at p < 0.001. Reductions in white matter in
MCI subjects compared to controls were restricted to a cluster in the lateral temporal lobe white matter (data not shown). Cortical grey matter decreased over time in control subjects in parietal, temporal and prefrontal association areas and in medial temporal lobes, particularly amygdala. Decrease was also detected in cerebellum, midbrain and precentral gyrus (Table 3). MCI subjects exhibited decline of grey matter density over time in more restricted cortical areas involving prefrontal lobe regions and precuneus (Table 4). White matter reduction in controls involved frontal, parietal and temporal lobe white matter and parahippocampal gyrus white matter (Table 5). Reductions in MCI subjects were spatially more restricted with predominance in frontal lobe white matter, cingulate and corpus callosum (Table 6). Additionally, cerebellum white matter was reduced. Figure 6 shows grey and white matter reductions in healthy controls and MCI subjects. The
516
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI Table 3 Decline of grey matter in controls
Table 5 Decline of white matter in controls
Coordinates (mm) Region Side BA x y z T24 Superior parietal lobule L 7 −26 −77 46 6.86 Inferior parietal lobule L 40 −46 −61 49 6.78 Precuneus R 19 16 −85 39 6.42 Cerebellum declive L −35 −63 −9 4.47 Medial frontal gyrus L 10 0 46 12 4.42 Midbrain L −2 −7 −4 4.4 Superior parietal lobule R 7 34 −67 50 5.95 Precuneus R 7 24 −70 53 5.26 Inferior parietal lobule R 39 48 −69 42 4.42 Cerebellum declive R 14 −67 −15 5.72 Lingual gyrus R 18 20 −76 −10 4.47 Cerebellum culmen R 28 −61 −23 4.36 Middle frontal gyrus R 10 39 61 0 4.94 Superior frontal gyrus R 10 30 58 5 3.71 Precentral gyrus R 6 62 3 35 4.44 Inferior parietal lobule R 40 59 −39 48 4.38 Amygdala R 33 −5 −11 4.02 The height threshold was set at p < 0.001, uncorrected for multiple comparisons. The cluster extension, representing the number of contiguous voxels passing the height threshold was set at > 50. Coordinates in bold delineate a cluster and the peak T-value (24 degrees of freedom) within the cluster. Subsequent non-bold coordinates identify further peaks within the same cluster that meet the significance level. Brain regions are indicated by Talairach and Tournoux coordinates, x, y and z [50]: x = the medial to lateral distance relative to midline (positive = right hemisphere); y = the anterior to posterior distance relative to the anterior commissure (positive = anterior); z = superior to inferior distance relative to the anterior commissure -posterior commissure line (positive = superior). R/L = right/left. BA = Brodmann area. Table 4 Decline of grey matter in MCI subjects Region Medial frontal gyrus Medial frontal gyrus Medial frontal gyrus Precuneus Precuneus Superior frontal gyrus Middle frontal gyrus
Side R R R R R R R
BA 9 10 10 19 7 9 10
Coordinates (mm) x y z 4 52 19 4 65 15 2 44 14 16 −83 41 13 −77 46 31 43 31 35 53 10
T24 8.71 5.94 5.49 5.44 4.10 5.30 4.80
For legend see Table 3.
opposite contrasts showed a small cluster of increasing grey matter in the left cerebellum and an increase of a small cluster of white matter in the right occipital lobe in controls, and no effects in MCI subjects (data not shown). These results had been obtained with the grey and white mater maps in MNI standard space. When we repeated the analysis with the grey and white matter maps remaining in the group specific native space as defined
Region Parahippocampal gyrus Superior frontal gyrus Superior frontal gyrus Medial frontal gyrus Precentral gyrus Paracentral lobulus Paracentral lobulus Medial frontal gyrus Middle frontal gyrus Lingual gyrus Cuneus Superior temporal gyrus Cuneus Middle occipital gyrus Lingual gyrus Middle temporal gyrus Middle temporal gyrus Middle temporal gyrus Superior frontal gyrus
Side L L L L L L L L L L L L R R R R R R R
Coordinates (mm) x y z −42 −18 −21 −17 56 −17 −12 64 −8 −10 63 5 −39 −1 40 −11 −24 45 −14 −13 45 −10 −18 53 −31 32 28 −19 −92 −4 −15 −96 2 −39 6 −23 24 −87 −2 24 −93 3 15 −93 −4 34 −65 20 45 −69 18 31 −75 25 17 9 57
T24 4.75 4.52 4.45 4.08 4.40 4.32 3.86 3.83 4.18 4.10 4.08 3.85 5.00 4.98 4.55 4.19 3.92 3.89 4.09
For legend see Table 3.
by the iteratively generated group template (step 5 of the structural MRI data processing), the results were identical. Additionally, results remained essentially unchanged, when grey and white matter maps were proportionally scaled to the global mean before statistical analysis.
DISCUSSION We found statistically significant decline of the integrity of subcortical fiber tracts over 13 to 16 months follow-up in healthy elderly subjects and subjects with MCI. Reductions involved predominantly the corpus callosum. Healthy elderly subjects showed additional reductions in the cingulate bundle, the inferior frontooccipital fasciculus, the superior longitudinal fasciculus and the uncinate fasciculus. Minor effects were found in the anterior thalamic radiation and the corticospinal tract. Parallel to the decline of the integrity of inter- and intra-hemsipheric projections grey matter and white matter volume decreased in parietal, temporal and prefrontal lobe areas. Our findings provide evidence that DTI data can be employed to determine changes of white matter microstructure over time both in cognitively healthy aging and in an at risk stage of AD. The spatial pattern of FA reduction in our healthy elderly subjects reflects those areas that have been found involved in normal aging in cross-sectional analysis us-
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
517
Table 6 Decline of white matter in MCI subjects Region Cingulate gyrus Cingulate gyrus Corpus callosum splenium Medial frontal gyrus Precentral gyrus Paracentral lobulus Corpus callosum genu Anterior cingulate Cerebellum nodule Cerebellum inferior semi-lunar lobule Cingulate gyrus Postcentral gyrus Medial frontal gyrus Cingulate gyrus Cingulate gyrus Corpus callosum truncus Caudate Internal capsule Superior frontal gyrus
Side L L L L L L L L L L R R R R R R R R R
Coordinates (mm) x y z −10.0 −30.0 31.0 −14.0 −40.0 41.0 −8.0 −41.0 18.0 −16.0 −10.0 49.0 −25.0 −24.0 57.0 −11.0 −24.0 47.0 −11.0 25.0 17.0 −11.0 19.0 24.0 −7.0 −56.0 −29.0 −16.0 −64.0 −36.0 14.0 −9.0 46.0 33.0 −25.0 35.0 19.0 −10.0 56.0 24.0 30.0 27.0 11.0 21.0 31.0 1.0 −22.0 20.0 6.0 7.0 0.0 6.0 1.0 5.0 14.0 53.0 29.0
T24 6.74 6.31 4.27 5.00 4.79 4.34 4.88 4.68 4.64 4.44 7.13 4.88 4.69 5.24 4.04 5.22 4.37 4.13 3.92
For legend see Table 3.
Fig. 6. Grey matter and white matter atrophy in MCI subjects and controls. Cluster of significant decline of grey matter (left) and white matter (right) projected on a surface rendering of the group specific grey and white matter template, respectively, in MNI space. Effects for controls are in red, for MCI in green. Cluster extension set at > 50 contiguous voxels passing the significance threshold of p < 0.001, uncorrected.
518
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
ing DTI, including corpus callosum,the anterior and the posterior limb of internal capsule, the posterior periventricular regions, and the deep frontal regions [17–19]. It further matches the regional pattern of FA decline in corpus callosum and posterior cingulate in a recent study on healthy subjects aged between 50 and 90 years after 2 years of follow-up [29]. Our data together with the data from this previous study [29] suggest that age associated fiber tract changes have a high temporal dynamic at 60 years of age and above that may account for the strong effects of aging found in previous crosssectional studies. At baseline, FA values were not different between MCI subjects and controls using TBSS. This is in line with a previous study, reporting significant differences in FA between AD patients and elderly controls, but not between MCI patients and controls using TBSS [53]. In contrast, we found significant reductions in grey matter of bilateral lateral and medial temporal lobes as well as bilateral superior parietal lobes in MCI subjects compared to controls, consistent with spatial pattern of grey matter atrophy in previous studies [54] and the characterization of these subjects as an at risk group of AD. In the longitudinal data, we found significant decline of fiber tract integrity in corpus callosum rostrum, genu and anterior body in MCI subjects. This is consistent with the longitudinal decline of corpus callosum area over time in early AD [55]. One might have expected to find more pronounced FA decline in MCI subjects than in controls. However, there was no statistically significant difference between groups when considering the interaction between time and diagnosis. We can not finally explain this finding. One possible factor may be the relatively small number of subjects limiting the sensitivity of our analysis. Another possible reason could be a higher variability of FA values in the MCI group compared to controls that would mask some of the longitudinal effects. The median coefficient of variation of FA values averaged across all tracts ranged from 17.6% to 20.9% across diagnostic groups and time points. This compares to a CV below 10% for measures of hippocampus or amygdala volumes across healthy elderly controls or MCI subjects [56], suggesting higher variability in FA than in volumetric measures. When we checked the overall variability of FA values, coefficients of variation were higher in the MCI group compared to the controls particularly at followup. This suggests that trajectories of change were more variable in MCI than in healthy aging which may have masked some of the overall effects. However, the differences in variation was rather moderate amounting
to a 5% to 11% higher median coefficient of variation in MCI compared to controls (Table 1) and therefore may not be sufficient to entirely explain the limited FA changes in MCI compared to controls. We complemented the voxel-based findings with rates of change of FA in selected fiber tracts. Rates of decline of FA ranged between 0.20% per year in superior longitudinal fasciculus and 10% per year in fornix. For comparison, changes in hippocampus volume over time, one of the best established imaging markers to date, range between 0.33 and 6.8% per year in aging and AD [57]. Therefore, the reductions of FA values over time are comparable in size to the reductions in well established volumetric markers. Similar to the voxel-wise analysis, there were no significant differences in rates of FA decline in the selected regions between groups. Follow-up studies using DTI in neurodegenerative diseases are rare. After 3 months follow-up FA values were not different to baseline in any of three groups comprising 25 AD patients, 25 MCI subjects, and 25 elderly controls [27]. Probably, the follow-up interval in this study was too short to detect age or neurodegeneration related changes in FA values. We found no significant effect of ApoE4 genotype on rates of FA decline across both diagnostic groups. In a cross-sectional study on 53 cognitively healthy elderly subjects, the presence of at least one ApoE4 allele had a significant but spatially limited effect on FA values in the left parahippocampal gyrus [58]. As only 6 of our 25 subjects carried at least one ApoE4 allele, the power was likely not sufficient to detect such a limited effect in our study. Likewise, gender had no significant effect on rates of FA decline. Several cross-sectional studies on brain aging using MRI found significant differences between women and men in the pattern of brain atrophy, with more pronounced age-related atrophy in frontal and temporal lobes in men and a relative preservation of these structures in women [59–61]. Longitudinal studies on gender effects on brain atrophy are rare. The lack of an effect in our sample, however, has to be interpreted with caution, since the number of subjects was relatively low so that a modest effect of gender on rates of FA decline can not be precluded based on our data. Drift in scanner performance over time has to be considered as potential confound in the analysis of longitudinal DTI data. A wide range of fiber tracts was spared from significant changes over time, including highly organized fiber tracts such as pyramidal tract and cerebellar peduncules, but also less organized fiber
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
tracts such as crossing fibers in the midbrain and cerebellum. Therefore it is unlikely that the selective involvement of highly organized fiber tracts such as the corpus callosum and less organized fiber tracts such as longitudinal fascicles is related to scanner drift effects. When we conducted the study, we had no access to a physical DTI phantom to measure effects of scanner drift over time. However, we recently scanned a novel physical DTI phantom [30] at five different 1.5 Tesla MRI scanners with similar hardware specifications and acquisition protocols. Overall, the FA values measured across the five sites showed a coefficient of variation of 6.1%. This value represents an upper limit for the variation of FA values due to scanner drift. As the multicenter data compromise differences in scanner hardware, software and acquisition parameters, the variation due to scanner drift within one single scanner using constant scan parameters and software is expected to lie below the multicenter variation. This notion is supported by a study on 10 healthy subjects who were scanned under two conditions: (i) scanned twice on the same scanner three days apart and (ii) scanned at two different scanners. The intracscanner CV of 2% for FA measurements was below the inter-scanner CV of 4.5% [62]. The interval in this previous study, however, was short (only few days), compared to the follow-up of on average more than one year in our study. The acquisition protocol of the phantom data differed from the in vivo data. Specifically, three of the five sites used 30 rather than 12 gradient directions. Although a higher number of gradients may yield more reliable estimates of FA values, findings from an experimental study suggest that within the limits that are met by a clinical protocol with 12 gradient directions, the differences in DTI contrasts due to different numbers of diffusion gradients are relatively small [63]. The issue of scanner drift effects on FA measurements deserves prospective studies using phantom measurements in a longitudinal design. Parallel to FA changes over time, we investigated the pattern of grey and white matter atrophy in MCI subjects and controls using deformation based morphometry. The healthy elderly control subjects showed decline of grey matter in superior and inferior parietal lobes as well as prefrontal cortex and medial temporal lobes. Additionally, cerebellum grey matter was involved. This pattern of grey matter atrophy is consistent with previous reports on grey matter atrophy in aging [1–3]. It further agrees with our DTI findings with involvement of the superior longitudinal fasciculus connecting parietal and prefrontal lobe areas [64],
519
the uncinate fasciculus projecting from hippocampus and amygdala to the orbitfrontal cortex [65], and the anterior and posterior corpus callosum, connecting homologous areas in bilateral prefrontal and parietal lobe cortex [66,67]. The MCI subjects showed prefrontal lobe atrophy, which is consistent with the involvement of anterior corpus callosum fiber tracts in the DTI data (see supplemental figures 1 and 2 for a projection of FA reductions and grey matter atrophy in the common MNI standard space in controls and MCI subjects, respectively). Previous studies have described a similar pattern of grey matter atrophy in MCI, however, extending into parietal and temporal lobe areas [68]. Similar to the spatial distribution of grey matter atrophy, white matter atrophy involved frontal, parietal, and temporal lobes, including mediotemporal areas, in elderly subjects. This pattern is consistent with the pattern of FA reduction in the subjects as determined from the DTI data. White matter atrophy in MCI subjects involved prefrontal lobe white matter, cingulate white matter and corpus callosum. Additionally, cerebellum and internal capsule were involved. Longitudinal studies on white matter atrophy in aging or MCI are rare. The significant decline of corpus callosum white matter agrees with ROI-based findings on corpus callosum atrophy in AD [55]. Furthermore, the involvement of prefrontal lobe, cingulate gyrus and corpus callosum in our MCI subjects agrees with the spatial pattern of white matter atrophy in a study on MCI converters and non-converters followed over 2 years [69]. All of our MCI subjects fulfilled clinical criteria for amnestic MCI which is regarded a clinical at risk group of AD [23,70]. However, only 5 of the MCI subjects showed cognitive decline over the follow-up period and only one out of these five converted to dementia and AD. This rate of conversion of 7% per year is slightly lower than the average 10% per year that has been shown in a systematic review on 19 longitudinal studies on rates of conversion in MCI [71]. The healthy control subjects remained cognitively stable over the follow-up period, with MMSE scores not dropping below 27 at follow-up and no subject developing MCI or dementia. Based on the concept of brain reserve capacity structural changes have to reach a critical threshold before they become clinically manifest [72]. Therefore, decline of fractional anisotropy and grey and white matter volume in healthy controls may suggest that structural changes precede the onset of cognitive decline. In summary, we provide evidence for long-term longitudinal changes in fiber tract integrity in intracortical projecting fiber systems over an average period of 13
520
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI
to 16 months in healthy elderly subjects and MCI subjects. The pattern of fiber tract changes was paralleled by reductions in cortical grey matter and white matter volumes whose distribution was consistent with the impairment of connecting fiber systems. These findings indicate that DTI may provide a useful biomarker of structural disconnection in healthy aging and in prodromal at risk stages of AD. Future studies need to employ larger samples of subjects and will have to extend to clinically manifest stages of AD.
[6]
[7]
[8]
[9]
[10]
ACKNOWLEDGMENTS Part of this work was supported by grants of the from the Interdisciplinary Faculty, Department “Ageing Science and Humanities”, University of Rostock, to S.J.T., of the Hirnliga e. V. (N¨urmbrecht, Germany) to S.J.T., an investigator initiated unrestricted research grant from Janssen-CILAG (Neuss, Germany) to H.H. and S.J.T., and a grant from the Bundesministerium f¨ur Bildung und Forschung (BMBF 01 GI 0102) awarded to the dementia network “Kompetenznetz Demenzen”. The study was further supported by the Science Foundation Ireland (SFI) investigator program award 08/IN.1/B1846 to H.H. There are no conflicts of interest associated with the work presented in this article. The corresponding author had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Authors’ disclosures available online (http://www.jalz.com/disclosures/view.php?id=523).
REFERENCES [1]
[2]
[3]
[4]
[5]
Driscoll I, Davatzikos C, An Y, Wu X, Shen D, Kraut M, Resnick SM (2009) Longitudinal pattern of regional brain volume change differentiates normal aging from MCI. Neurology 72, 1906-1913. Raji CA, Lopez OL, Kuller LH, Carmichael OT, Becker JT (2009) Age, Alzheimer disease, and brain structure. Neurology 73, 1899-1905 Davatzikos C, Xu F, An Y, Fan Y, Resnick SM (2009) Longitudinal progression of Alzheimer’s-like patterns of atrophy in normal older adults: the SPARE-AD index. Brain 132, 2026-2035. Salat DH, Greve DN, Pacheco JL, Quinn BT, Helmer KG, Buckner RL, Fischl B (2009) Regional white matter volume differences in nondemented aging and Alzheimer’s disease. Neuroimage 44, 1247-1258. Resnick SM, Pham DL, Kraut MA, Zonderman AB, Davatzikos C (2003) Longitudinal magnetic resonance imaging studies of older adults: a shrinking brain. J Neurosci 23, 32953301.
[11]
[12]
[13]
[14] [15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
Kochunov P, Ramage AE, Lancaster JL, Robin DA, Narayana S, Coyle T, Royall DR, Fox P (2009) Loss of cerebral white matter structural integrity tracks the gray matter metabolic decline in normal aging. Neuroimage 45, 17-28. Pakkenberg B, Pelvig D, Marner L, Bundgaard MJ, Gundersen HJ, Nyengaard JR, Regeur L (2003) Aging and the human neocortex. Exp Gerontol 38, 95-99. Marner L, Nyengaard JR, Tang Y, Pakkenberg B (2003) Marked loss of myelinated nerve fibers in the human brain with age. J Comp Neurol 462, 144-152. Teipel SJ, Meindl T, Grinberg L, Heinsen H, Hampel H (2008) Novel MRI techniques in the assessment of dementia. Eur J Nucl Med Mol Imaging 35(Suppl 1), S58-S69. Le Bihan D (2003) Looking into the functional architecture of the brain with diffusion MRI. Nat Rev Neurosci 4, 469-480. Beaulieu C (2002) The basis of anisotropic water diffusion in the nervous system – a technical review. NMR Biomed 15, 435-455. Gouw AA, Seewann A, Vrenken H, van der Flier WM, Rozemuller JM, Barkhof F, Scheltens P, Geurts JJ (2008) Heterogeneity of white matter hyperintensities in Alzheimer’s disease: post-mortem quantitative MRI and neuropathology. Brain 131, 3286-3298. Lebel C, Walker L, Leemans A, Phillips L, Beaulieu C (2008) Microstructural maturation of the human brain from childhood to adulthood. Neuroimage 40, 1044-1055. Huppi PS, Dubois J (2006) Diffusion tensor imaging of brain development. Semin Fetal Neonatal Med 11, 489-497. Aboitiz F, Rodriguez E, Olivares R, Zaidel E (1996) Agerelated changes in fibre composition of the human corpus callosum: sex differences. Neuroreport 7, 1761-1764. Meier-Ruge W, Ulrich J, Bruhlmann M, Meier E (1992) Agerelated white matter atrophy in the human brain. Ann N Y Acad Sci 673, 260-269. Westlye LT, Walhovd KB, Dale AM, Bjornerud A, DueTonnessen P, Engvig A, Grydeland H, Tamnes CK, Ostby Y, Fjell AM (2010) Life-span changes of the human brain white matter: diffusion tensor imaging (DTI) and volumetry. Cereb Cortex 20, 2055-2068. Bennett IJ, Madden DJ, Vaidya CJ, Howard DV, Howard JH, Jr. (2009) Age-related differences in multiple measures of white matter integrity: A diffusion tensor imaging study of healthy aging. Hum Brain Mapp 31, 378-390 Head D, Buckner RL, Shimony JS, Williams LE, Akbudak E, Conturo TE, McAvoy M, Morris JC, Snyder AZ (2004) Differential vulnerability of anterior white matter in nondemented aging with minimal acceleration in dementia of the Alzheimer type: evidence from diffusion tensor imaging. Cereb Cortex 14, 410-423. Stricker NH, Schweinsburg BC, Delano-Wood L, Wierenga CE, Bangen KJ, Haaland KY, Frank LR, Salmon DP, Bondi MW (2009) Decreased white matter integrity in latemyelinating fiber pathways in Alzheimer’s disease supports retrogenesis. Neuroimage 45, 10-16. Nakata Y, Sato N, Abe O, Shikakura S, Arima K, Furuta N, Uno M, Hirai S, Masutani Y, Ohtomo K, Aoki S (2008) Diffusion abnormality in posterior cingulate fiber tracts in Alzheimer’s disease: tract-specific analysis. Radiat Med 26, 466-473. Teipel SJ, Stahl R, Dietrich O, Schoenberg SO, Perneczky R, Bokde AL, Reiser MF, Moller HJ, Hampel H (2007) Multivariate network analysis of fiber tract integrity in Alzheimer’s disease. Neuroimage 34, 985-995.
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI [23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
Ravaglia G, Forti P, Maioli F, Martelli M, Servadei L, Brunetti N, Pantieri G, Mariani E (2006) Conversion of mild cognitive impairment to dementia: predictive role of mild cognitive impairment subtypes and vascular risk factors. Dement Geriatr Cogn Disord 21, 51-58. Stahl R, Dietrich O, Teipel SJ, Hampel H, Reiser MF, Schoenberg SO (2007) White matter damage in Alzheimer’s disease and in mild cognitive impairment: assessment with diffusion tensor MRI using parallel imaging techniques. Radiology 243, 483-492. Ukmar M, Makuc E, Onor ML, Garbin G, Trevisiol M, Cova MA (2008) Evaluation of white matter damage in patients with Alzheimer’s disease and in patients with mild cognitive impairment by using diffusion tensor imaging. Radiol Med 113, 915-922. Fellgiebel A, Muller MJ, Wille P, Dellani PR, Scheurich A, Schmidt LG, Stoeter P (2005) Color-coded diffusion-tensorimaging of posterior cingulate fiber tracts in mild cognitive impairment. Neurobiol Aging 26, 1193-1198. Mielke MM, Kozauer NA, Chan KC, George M, Toroney J, Zerrate M, Bandeen-Roche K, Wang MC, Vanzijl P, Pekar JJ, Mori S, Lyketsos CG, Albert M (2009) Regionally-specific diffusion tensor imaging in mild cognitive impairment and Alzheimer’s disease. Neuroimage 46, 47-55. Duning T, Warnecke T, Mohammadi S, Lohmann H, Schiffbauer H, Kugel H, Knecht S, Ringelstein EB, Deppe M (2009) Pattern and progression of white-matter changes in a case of posterior cortical atrophy using diffusion tensor imaging. J Neurol Neurosurg Psychiatry 80, 432-436. Barrick TR, Charlton RA, Clark CA, Markus HS White matter structural decline in normal ageing: a prospective longitudinal study using tract-based spatial statistics. Neuroimage 51, 565577. Laun FB, Huff S, Stieltjes B (2009) On the effects of dephasing due to local gradients in diffusion tensor imaging experiments: relevance for diffusion tensor imaging fiber phantoms. Magn Reson Imaging 27, 541-548. Petersen RC, Doody R, Kurz A, Mohs RC, Morris JC, Rabins PV, Ritchie K, Rossor M, Thal L, Winblad B (2001) Current concepts in mild cognitive impairment. Arch Neurol 58, 19851992. Folstein MF, Folstein SE, McHugh PR (1975) Mini-mentalstate: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189-198. Berres M, Monsch AU, Bernasconi F, Thalmann B, Stahelin HB (2000) Normal ranges of neuropsychological tests for the diagnosis of Alzheimer’s disease. Stud Health Technol Inform 77, 195-199. Shulman KI, Shedletsky R, Silver IL (1986) The challenge of time: Clock drawing and cognitive function in the elderly. Int J Geriatr Psychiatry 1, 135-140. Chen P, Ratcliff G, Belle SH, Cauley JA, DeKosky ST, Ganguli M (2000) Cognitive tests that best discriminate between presymptomatic AD and those who remain nondemented. Neurology 55, 1847-1853. Scheltens P, Barkhof F, Leys D, Pruvo JP, Nauta JJP, Vermersch P, Steinling M, Valk J (1993) A semiquantitative rating scale for the assessment of signal hyperintensities on magnetic resonance imaging. J Neurol Sci 114, 7-12. Griswold MA, Jakob PM, Heidemann RM, Nittka M, Jellus V, Wang J, Kiefer B, Haase A (2002) Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 47, 1202-1210.
[38]
[39]
[40] [41] [42] [43]
[44]
[45] [46]
[47]
[48] [49]
[50] [51]
[52]
[53]
[54]
[55]
[56]
521
Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE, Mackay CE, Watkins KE, Ciccarelli O, Cader MZ, Matthews PM, Behrens TEJ (2006) Tract-based spatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage 31, 1487-1505. Weaver KE, Richards TL, Liang O, Laurino MY, Samii A, Aylward EH (2009) Longitudinal diffusion tensor imaging in Huntington’s Disease. Exp Neurol 216, 525-529. Smith S (2002) Fast robust automated brain extraction. Hum Brain Map 17, 143-155. Andersson JLR, Jenkinson M, Smith S (2007) in FMRIB technical report TR07JA1 . Andersson JLR, Jenkinson M, Smith S (2007) in FMRIB technical report TR07JA2. Rueckert D, Sonoda LI, Hayes C, Hill DLG, Leach MO, Hawkes DJ (1999) Non-rigid registration using free-form deformations: Application to breast MR images. IEEE Trans Med Imaging 18, 712-721. Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S (2004) Fiber tract-based atlas of human white matter anatomy. Radiology 230, 77-87. Ashburner J (2007) A fast diffeomorphic image registration algorithm. Neuroimage 38, 95-113. Kipps CM, Duggins AJ, Mahant N, Gomes L, Ashburner J, McCusker EA (2005) Progression of structural neuropathology in preclinical Huntington’s disease: a tensor based morphometry study. J Neurol Neurosurg Psychiatry 76, 650-655. Friston K, Ashburner J, Frith CD, Poline J-B, Heather JD, Frackowiak RSJ (1995) Spatial registration and normalization of images. Hum Brain Mapp 2, 165-189. Ashburner J, Friston KJ (2000) Voxel-based morphometry-the methods. Neuroimage 11, 805-821. Freeborough PA, Fox NC (1998) Modeling brain deformations in Alzheimer disease by fluid registration of serial 3D MR images. J Comput Assist Tomogr 22, 838-843. Talairach J, Tournoux P (1988) Co-Planar Stereotaxic Atlas of the Human Brain, Thieme, New York. Evans A, Collins D, Millst S, Brown E, Kelly R, Peters T (1993) 3D statistical neuroanatomical models from 305 MRI volumes. Proc IEEE-Nuclear Science Symposium and Medical Imaging Conference, 1813-1817. Smith SM, Nichols TE (2009) Threshold-free cluster enhancement: addressing problems of smoothing, threshold dependence and localisation in cluster inference. Neuroimage 44, 83-98. Damoiseaux JS, Smith SM, Witter MP, Sanz-Arigita EJ, Barkhof F, Scheltens P, Stam CJ, Zarei M, Rombouts SA (2009) White matter tract integrity in aging and Alzheimer’s disease. Hum Brain Mapp 30, 1051-1059. Hamalainen A, Tervo S, Grau-Olivares M, Niskanen E, Pennanen C, Huuskonen J, Kivipelto M, Hanninen T, Tapiola M, Vanhanen M, Hallikainen M, Helkala EL, Nissinen A, Vanninen R, Soininen H (2007) Voxel-based morphometry to detect brain atrophy in progressive mild cognitive impairment. Neuroimage 37, 1122-1131. Teipel SJ, Bayer W, Alexander GE, Zebuhr Y, Teichberg D, Kulic L, Schapiro MB, Moller HJ, Rapoport SI, Hampel H (2002) Progression of corpus callosum atrophy in Alzheimer disease. Arch Neurol 59, 243-248. Teipel SJ, Ewers M, Wolf S, Jessen F, K¨olsch H, Arlt S, Luckhaus C, Sch¨onknecht P, Schmidtke K, Heuser I, Fr¨olich L, Ende G, Pantel J, Wiltfang J, Rakebrandt F, Peters O, Born C, Kornhuber J, Hampel H (2010) Multicentre variability of
522
[57]
[58]
[59]
[60]
[61]
[62]
[63]
[64]
[65]
S.J. Teipel et al. / Longitudinal DTI in Aging and MCI MRI-based medial temporal lobe volumetry in Alzheimer’s disease. Psychiatry Res Neuroimaging 182, 244-250. Barnes J, Bartlett JW, van de Pol LA, Loy CT, Scahill RI, Frost C, Thompson P, Fox NC (2009) A meta-analysis of hippocampal atrophy rates in Alzheimer’s disease. Neurobiol Aging 30, 1711-1723. Honea RA, Vidoni E, Harsha A, Burns JM (2009) Impact of APOE on the healthy aging brain: a voxel-based MRI and DTI study. J Alzheimers Dis 18, 553-564 Curiati PK, Tamashiro JH, Squarzoni P, Duran FL, Santos LC, Wajngarten M, Leite CC, Vallada H, Menezes PR, Scazufca M, Busatto GF, Alves TC (2009) Brain structural variability due to aging and gender in cognitively healthy Elders: results from the Sao Paulo Ageing and Health study. AJNR Am J Neuroradiol 30, 1850-1856. Alexander GE, Chen K, Merkley TL, Reiman EM, Caselli RJ, Aschenbrenner M, Santerre-Lemmon L, Lewis DJ, Pietrini P, Teipel SJ, Hampel H, Rapoport SI, Moeller JR (2006) Regional network of magnetic resonance imaging gray matter volume in healthy aging. Neuroreport 17, 951-956. Xu J, Kobayashi S, Yamaguchi S, Iijima K, Okada K, Yamashita K (2000) Gender effects on age-related changes in brain structure. Am J Neuroradiol 21, 112-118. Pfefferbaum A, Adalsteinsson E, Sullivan EV (2003) Replicability of diffusion tensor imaging measurements of fractional anisotropy and trace in brain. J Magn Reson Imaging 18, 427-433. Landman BA, Farrell JA, Jones CK, Smith SA, Prince JL, Mori S (2007) Effects of diffusion weighting schemes on the reproducibility of DTI-derived fractional anisotropy, mean diffusivity, and principal eigenvector measurements at 1.5T. Neuroimage 36, 1123-1138. Makris N, Kennedy DN, McInerney S, Sorensen AG, Wang R, Caviness VS, Jr., Pandya DN (2005) Segmentation of subcomponents within the superior longitudinal fascicle in humans: a quantitative, in vivo, DT-MRI study. Cereb Cortex 15, 854-869. Kier EL, Staib LH, Davis LM, Bronen RA (2004) MR imaging
[66]
[67]
[68]
[69]
[70]
[71]
[72]
of the temporal stem: anatomic dissection tractography of the uncinate fasciculus, inferior occipitofrontal fasciculus, and Meyer’s loop of the optic radiation. AJNR Am J Neuroradiol 25, 677-691. Sydykova D, Stahl R, Dietrich O, Ewers M, Reiser MF, Schoenberg SO, Moller HJ, Hampel H, Teipel SJ (2006) Fiber Connections between the Cerebral Cortex and the Corpus Callosum in Alzheimer’s Disease: A Diffusion Tensor Imaging and Voxel-Based Morphometry Study. Cereb Cortex 17, 22762282 De Lacoste MC, Kirkpatrick JB, Ross ED (1985) Topography of the human corpus callosum. J Neuropathol Exp Neurol 44, 578-591. Leow AD, Yanovsky I, Parikshak N, Hua X, Lee S, Toga AW, Jack CR, Jr., Bernstein MA, Britson PJ, Gunter JL, Ward CP, Borowski B, Shaw LM, Trojanowski JQ, Fleisher AS, Harvey D, Kornak J, Schuff N, Alexander GE, Weiner MW, Thompson PM (2009) Alzheimer’s disease neuroimaging initiative: a one-year follow up study using tensor-based morphometry correlating degenerative rates, biomarkers and cognition. Neuroimage 45, 645-655. Misra C, Fan Y, Davatzikos C (2009) Baseline and longitudinal patterns of brain atrophy in MCI patients, and their use in prediction of short-term conversion to AD: results from ADNI. Neuroimage 44, 1415-1422. Tabert MH, Manly JJ, Liu X, Pelton GH, Rosenblum S, Jacobs M, Zamora D, Goodkind M, Bell K, Stern Y, Devanand DP (2006) Neuropsychological prediction of conversion to Alzheimer disease in patients with mild cognitive impairment. Arch Gen Psychiatry 63, 916-924. Bruscoli M, Lovestone S (2004) Is MCI really just early dementia? A systematic review of conversion studies. Int Psychogeriatr 16, 129-140. Teipel SJ, Meindl T, Wagner M, Kohl T, Burger K, Reiser MF, Herpertz S, Moller HJ, Hampel H (2009) White matter microstructure in relation to education in aging and Alzheimer’s disease. J Alzheimers Dis 17, 571-583.
