J Mol Neurosci DOI 10.1007/s12031-015-0550-5
Microstructural White Matter Abnormalities and Cognitive Dysfunction in Subcortical Ischemic Vascular Disease: an Atlas-Based Diffusion Tensor Analysis Study Lin Lin 1 & Yunjing Xue 1 & Qing Duan 1 & Bin Sun 1 & Hailong Lin 1 & Xiaodan Chen 1 & Ling Luo 2 & Xiaofan Wei 2 & Zhongping Zhang 3
Received: 14 January 2015 / Accepted: 23 March 2015 # Springer Science+Business Media New York 2015
Abstract Recent studies in subcortical ischemic vascular disease (SIVD) suggest the involvement of white matter (WM) abnormalities underlying the pathogenesis of cognitive function impairment. Here, we performed magnetic resonance diffusion tensor imaging (DTI) on detecting WM damage and to investigate the correlations between DTI measures and cognitive dysfunction in SIVD patients. Fifty right-handed SIVD patients were recruited and divided into vascular cognitive impairment on dementia (VCIND) group and normal cognition (NC) group. Twenty-two VCIND patients and 28 NC patients underwent DTI scanning and neuropsychological assessment. Atlas-based analysis (ABA) was performed on each subject for extracting FA and MD measures from supratentorial tracts. Among VCIND, as compared to NC patients, decreased FA and increased MD were observed in all projection fibers (bilateral anterior, posterior limb, and retrolenticular part of internal capsule, anterior, superior, and posterior corona radiata and posterior thalamic radiation), association fibers (bilateral sagittal stratum, external capsule, cingulum, fornix, and stria terminalis, superior longitudinal fasciculus, superior fronto-occipital fasciculus, and uncinate fasciculus), and commissural fibers (genu, body, splenium, and bilateral tapetum of corpus callosum). Furthermore, we also found that MoCA scores correlated with DTI values in all supratentorial WM tracts. The results suggested that SIVD * Yunjing Xue [email protected] 1
Department of Radiology, Fujian Medical University Union Hospital, Fuzhou, Fujian, China
2
Department of Neurology, Fujian Medical University Union Hospital, Fuzhou, Fujian, China
3
GE Healthcare China, Beijing, China
patients demonstrated abnormal WM connectivity in all supratentorial regions. Moreover, the severity of damage in WM tracts correlated with cognitive dysfunction. Keywords Subcortical ischemic vascular disease . Diffusion tensor imaging . Atlas-based analysis . Vascular cognitive impairment
Introduction Subcortical ischemic vascular disease (SIVD is characterized by white matter hyperintensity (WMH) and lacunes in regions of subcortical whiter matter and basal ganglia on magnetic resonance imaging (MRI). It is a more homogeneous subtype of cerebral vascular disease and considered as one of the major causes for vascular cognitive impairment (VCI) (Erkinjuntti et al. 2000; Roman et al. 2002). VCI has been classified into three subtypes: vascular cognitive impairment with no dementia (VCIND), vascular dementia (VaD), and mixed Alzheimer’s disease (AD) with VaD (Moorhouse and Rockwood 2008). VCIND has been proposed as a prodromal stage of VaD (Frisoni et al. 2002; Galluzzi et al. 2005; Seo et al. 2010), which suggested potentially reversible through recent reports (Jak et al. 2009; Ravaglia et al. 2006). So, early detection of VCI is critical for the timely treatment and improvement of prognosis. Diffusion tensor imaging (DTI), a more recent MRI-based technique, is a noninvasive method for delineating specific white matter (WM) structures in vivo. DTI detects microstructural alterations in white matter tracts by measuring the directionality of molecular diffusion (fractional anisotropy, FA) and the average motion of water molecules (mean diffusivity, MD) (Xu et al. 2010). Although the white matter damage is crucial in patients with SIVD, few research have been reported for
J Mol Neurosci
investigating the relationship between the cognitive function and abnormalities of DTI parameters (O'Sullivan et al. 2001, 2004). Moreover, most DTI studies have used manual tracing of region of interest (ROI) and thus included only limited regions rather than assessing whole brain areas (O'Sullivan et al. 2004). To examine the integrity of the entire WM regions, voxel-based group analysis (VBA) is an effective initial method used in SIVD (Della Nave et al. 2007; Kim et al. 2011). However, the VBA often suffers from low statistical power because of high noise and partial volume effects. In our current study, we utilized a novel DTI analysis method—atlas-based whole brain white matter analysis (ABA) (Oishi et al. 2009), to assess the microstructure (including FA and MD measures) of a large number of white matter tracts in 50 SIVD patients with or without cognitive impairment. Because the supratentorial white matter is a common location for WMH and lacunars in SIVD patients (Munoz et al. 1993), we hypothesized that FA would decrease and MD would increase in the supratentorial WM, including projection (e.g., internal capsule and corona radiata), association (e.g., superior longitudinal, superior, and inferior fronto-occipital fasciculi), and commissural (e.g., corpus callosum) fibers. Moreover, as the white matter integrity is the fundament of global cognitive function, we also speculated that the reduced FA and increased MD values of the WM would be associated with the poor neuropsychological scores .
Material and Methods Participants From January 2012 to March 2013, 50 right-handed SIVD patients were consecutively enrolled from the Department of Neurology. All patients had experienced a clinical lacunar stroke at least 3 months prior to recruitment. We used the SIVD imaging criteria suggested by Galluzzi et al. (2005), which require at least moderate white matter hyperintensities on T2-weighted MRI plus lacunar infarct(s) in the periventricular and deep WM structures. These criteria are based on a modification of those for subcortical VaD described by Erkinjuntti et al. (2000), which aims to include milder cases. The diagnosis of vascular cognitive impairment on dementia (VCIND) was performed by two experienced neurologists in consensus according to criteria (Erkinjuntti et al. 2000; Moorhouse and Rockwood 2008; Petersen 2004; Roman et al. 2002) that included the following: (1) subjective cognitive complaints reported by the participant or his/her caregiver; (2) objective cognitive impairments, but not meeting the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) criteria for dementia; (3) normal or nearnormal performance of general cognitive functioning and no
or minimum impairments of daily life activities; (4) Clinical Dementia Rating Scale (CDR) score=0.5; (5) Mini-Mental State Examination (MMSE) score >24. The exclusive criteria for included VCIND (Roman et al. 2002; Zhou and Jia 2009): (1) memory deficit and other cognitive impairment in the absence of focal lesions on imaging; (2) cognitive impairments caused by nonvascular reasons, such as tumor, trauma, psychiatric disease, systemic disease; (3) signs of large vessel disease, such as cortical and/or corticosubcortical nonlacunar territorial infarcts and watershed infarcts or hemorrhages; (4) leukoencephalopathy as a result of other causes, such as normal pressure hydrocephalus, multiple sclerosis, brain irradiation, and metabolic diseases. In addition, other SIVD patients were included as the normal cognition (NC) group if they had no cognitive complaints, impairments of daily life activities, and abnormalities in above neuropsychological scales. As a result, 50 SIVD patients recruited were subdivided into NC group (n=28) and VCIND group (n=22). The hospital research ethics committee approved the study protocol, and written informed consent was obtained from all subjects. All participants underwent a standardized clinical evaluation, which included a general neurological examination, a global cognitive level test (i.e., CDR, MMSE, MoCA) and MR scan (i.e., T1, T2, T2FLAIR, DWI, DTI). Neuropsychological Evaluations Participants underwent neuropsychological evaluations, including measures of visual-spatial ability, executive function, name, language, memory, attention, and general intellectual ability. Tests included the Clinical Dementia Rating (CDR) scale, the Mini-Mental State Examination (MMSE) (Folstein et al. 1975), and the Montreal Cognitive Assessment (MoCA) (Nasreddine et al. 2005). All neuropsychological evaluations were performed by two attending neurologists at our hospital. DTI Acquisition and Processing Patients were performed on a 3 T MR scanner (GE discovery 750) equipped with an eight channel high-resolution head coil. The diffusion-weighted imaging acquired as the following parameters: single-shot EPI, spin echo sequence, 40 diffusion weighting directions, matrix size=128×128, repetition time=8325 ms, echo time=87.6 ms, field of view=256× 256 mm2, slice thickness=4 mm, slice number=35 slices. The first volumes were b0 volumes, and the b value of other volumes was 1000 s/mm2. After acquisitions, the data were downloaded from the scanner, transferred, and processed using a pipeline tool, PANDA (Cui et al. 2013) (http://www. nitrc.org/projects/panda/). The main procedure of PANA included preprocessing and producing diffusion metrics. First, The preprocess steps (including Converting Dicom files into Nifti images, Estimating the brain mask, Cropping
J Mol Neurosci Demographic and clinical information for NC and VCIND
the raw images, Correcting for the eddy-current effect, Calculating diffusion tensor metrics) were executed one by one. Then, we used the tool to normalize diffusion metrics (FA, MD) into the MNI space and calculated regional diffusion metrics by averaging the values within each region of the ICBM DTI-81 atlas (Mori et al. 2008). For further analyses, we focused on FA and MD values of all available projection, association, and commissural tract ROIs as hypothesized previously (Tables 1 and 2, Fig. 1).
Table 2 subjects
Statistical Analysis
Data are presented as mean±SD, percentages, n (%), or median (ranges)
All analyses were conducted using IBM SPSS Statistics software (version 19.0). The Shapiro–Wilk test of normality was used to investigate the distribution of demographic variables, cognitive performance, diffusion parameters. Demographic variables and cognitive performances were compared between the NC and VCIND group with independent sample t test for continuous variables, Mann-Whitney U test for nonparametric data, and χ2 test for sex proportion. As the distribution was not normal in diffusion values of several tracts, we conducted the nonparametric analysis (Mann-Whitney U test) to compare FA and MD values between groups. The mean and standard deviation, or the median and the interquartile range, were Table 1
NC
VCIND
P
n Age Sex (% male) Education (years)
28 70.9±8.2 57.1 12(9–12)
22 72.8±6.8 63.6 9.5(6–13)
0.39 0.64 0.37
MMSE MoCA
29(29–30) 26.8±1.6
26(24–27) 20.2±3.5
Microstructural White Matter Abnormalities and Cognitive Dysfunction in Subcortical Ischemic Vascular Disease: an Atlas-Based Diffusion Tensor Analysis Study Lin Lin 1 & Yunjing Xue 1 & Qing Duan 1 & Bin Sun 1 & Hailong Lin 1 & Xiaodan Chen 1 & Ling Luo 2 & Xiaofan Wei 2 & Zhongping Zhang 3
Received: 14 January 2015 / Accepted: 23 March 2015 # Springer Science+Business Media New York 2015
Abstract Recent studies in subcortical ischemic vascular disease (SIVD) suggest the involvement of white matter (WM) abnormalities underlying the pathogenesis of cognitive function impairment. Here, we performed magnetic resonance diffusion tensor imaging (DTI) on detecting WM damage and to investigate the correlations between DTI measures and cognitive dysfunction in SIVD patients. Fifty right-handed SIVD patients were recruited and divided into vascular cognitive impairment on dementia (VCIND) group and normal cognition (NC) group. Twenty-two VCIND patients and 28 NC patients underwent DTI scanning and neuropsychological assessment. Atlas-based analysis (ABA) was performed on each subject for extracting FA and MD measures from supratentorial tracts. Among VCIND, as compared to NC patients, decreased FA and increased MD were observed in all projection fibers (bilateral anterior, posterior limb, and retrolenticular part of internal capsule, anterior, superior, and posterior corona radiata and posterior thalamic radiation), association fibers (bilateral sagittal stratum, external capsule, cingulum, fornix, and stria terminalis, superior longitudinal fasciculus, superior fronto-occipital fasciculus, and uncinate fasciculus), and commissural fibers (genu, body, splenium, and bilateral tapetum of corpus callosum). Furthermore, we also found that MoCA scores correlated with DTI values in all supratentorial WM tracts. The results suggested that SIVD * Yunjing Xue [email protected] 1
Department of Radiology, Fujian Medical University Union Hospital, Fuzhou, Fujian, China
2
Department of Neurology, Fujian Medical University Union Hospital, Fuzhou, Fujian, China
3
GE Healthcare China, Beijing, China
patients demonstrated abnormal WM connectivity in all supratentorial regions. Moreover, the severity of damage in WM tracts correlated with cognitive dysfunction. Keywords Subcortical ischemic vascular disease . Diffusion tensor imaging . Atlas-based analysis . Vascular cognitive impairment
Introduction Subcortical ischemic vascular disease (SIVD is characterized by white matter hyperintensity (WMH) and lacunes in regions of subcortical whiter matter and basal ganglia on magnetic resonance imaging (MRI). It is a more homogeneous subtype of cerebral vascular disease and considered as one of the major causes for vascular cognitive impairment (VCI) (Erkinjuntti et al. 2000; Roman et al. 2002). VCI has been classified into three subtypes: vascular cognitive impairment with no dementia (VCIND), vascular dementia (VaD), and mixed Alzheimer’s disease (AD) with VaD (Moorhouse and Rockwood 2008). VCIND has been proposed as a prodromal stage of VaD (Frisoni et al. 2002; Galluzzi et al. 2005; Seo et al. 2010), which suggested potentially reversible through recent reports (Jak et al. 2009; Ravaglia et al. 2006). So, early detection of VCI is critical for the timely treatment and improvement of prognosis. Diffusion tensor imaging (DTI), a more recent MRI-based technique, is a noninvasive method for delineating specific white matter (WM) structures in vivo. DTI detects microstructural alterations in white matter tracts by measuring the directionality of molecular diffusion (fractional anisotropy, FA) and the average motion of water molecules (mean diffusivity, MD) (Xu et al. 2010). Although the white matter damage is crucial in patients with SIVD, few research have been reported for
J Mol Neurosci
investigating the relationship between the cognitive function and abnormalities of DTI parameters (O'Sullivan et al. 2001, 2004). Moreover, most DTI studies have used manual tracing of region of interest (ROI) and thus included only limited regions rather than assessing whole brain areas (O'Sullivan et al. 2004). To examine the integrity of the entire WM regions, voxel-based group analysis (VBA) is an effective initial method used in SIVD (Della Nave et al. 2007; Kim et al. 2011). However, the VBA often suffers from low statistical power because of high noise and partial volume effects. In our current study, we utilized a novel DTI analysis method—atlas-based whole brain white matter analysis (ABA) (Oishi et al. 2009), to assess the microstructure (including FA and MD measures) of a large number of white matter tracts in 50 SIVD patients with or without cognitive impairment. Because the supratentorial white matter is a common location for WMH and lacunars in SIVD patients (Munoz et al. 1993), we hypothesized that FA would decrease and MD would increase in the supratentorial WM, including projection (e.g., internal capsule and corona radiata), association (e.g., superior longitudinal, superior, and inferior fronto-occipital fasciculi), and commissural (e.g., corpus callosum) fibers. Moreover, as the white matter integrity is the fundament of global cognitive function, we also speculated that the reduced FA and increased MD values of the WM would be associated with the poor neuropsychological scores .
Material and Methods Participants From January 2012 to March 2013, 50 right-handed SIVD patients were consecutively enrolled from the Department of Neurology. All patients had experienced a clinical lacunar stroke at least 3 months prior to recruitment. We used the SIVD imaging criteria suggested by Galluzzi et al. (2005), which require at least moderate white matter hyperintensities on T2-weighted MRI plus lacunar infarct(s) in the periventricular and deep WM structures. These criteria are based on a modification of those for subcortical VaD described by Erkinjuntti et al. (2000), which aims to include milder cases. The diagnosis of vascular cognitive impairment on dementia (VCIND) was performed by two experienced neurologists in consensus according to criteria (Erkinjuntti et al. 2000; Moorhouse and Rockwood 2008; Petersen 2004; Roman et al. 2002) that included the following: (1) subjective cognitive complaints reported by the participant or his/her caregiver; (2) objective cognitive impairments, but not meeting the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) criteria for dementia; (3) normal or nearnormal performance of general cognitive functioning and no
or minimum impairments of daily life activities; (4) Clinical Dementia Rating Scale (CDR) score=0.5; (5) Mini-Mental State Examination (MMSE) score >24. The exclusive criteria for included VCIND (Roman et al. 2002; Zhou and Jia 2009): (1) memory deficit and other cognitive impairment in the absence of focal lesions on imaging; (2) cognitive impairments caused by nonvascular reasons, such as tumor, trauma, psychiatric disease, systemic disease; (3) signs of large vessel disease, such as cortical and/or corticosubcortical nonlacunar territorial infarcts and watershed infarcts or hemorrhages; (4) leukoencephalopathy as a result of other causes, such as normal pressure hydrocephalus, multiple sclerosis, brain irradiation, and metabolic diseases. In addition, other SIVD patients were included as the normal cognition (NC) group if they had no cognitive complaints, impairments of daily life activities, and abnormalities in above neuropsychological scales. As a result, 50 SIVD patients recruited were subdivided into NC group (n=28) and VCIND group (n=22). The hospital research ethics committee approved the study protocol, and written informed consent was obtained from all subjects. All participants underwent a standardized clinical evaluation, which included a general neurological examination, a global cognitive level test (i.e., CDR, MMSE, MoCA) and MR scan (i.e., T1, T2, T2FLAIR, DWI, DTI). Neuropsychological Evaluations Participants underwent neuropsychological evaluations, including measures of visual-spatial ability, executive function, name, language, memory, attention, and general intellectual ability. Tests included the Clinical Dementia Rating (CDR) scale, the Mini-Mental State Examination (MMSE) (Folstein et al. 1975), and the Montreal Cognitive Assessment (MoCA) (Nasreddine et al. 2005). All neuropsychological evaluations were performed by two attending neurologists at our hospital. DTI Acquisition and Processing Patients were performed on a 3 T MR scanner (GE discovery 750) equipped with an eight channel high-resolution head coil. The diffusion-weighted imaging acquired as the following parameters: single-shot EPI, spin echo sequence, 40 diffusion weighting directions, matrix size=128×128, repetition time=8325 ms, echo time=87.6 ms, field of view=256× 256 mm2, slice thickness=4 mm, slice number=35 slices. The first volumes were b0 volumes, and the b value of other volumes was 1000 s/mm2. After acquisitions, the data were downloaded from the scanner, transferred, and processed using a pipeline tool, PANDA (Cui et al. 2013) (http://www. nitrc.org/projects/panda/). The main procedure of PANA included preprocessing and producing diffusion metrics. First, The preprocess steps (including Converting Dicom files into Nifti images, Estimating the brain mask, Cropping
J Mol Neurosci Demographic and clinical information for NC and VCIND
the raw images, Correcting for the eddy-current effect, Calculating diffusion tensor metrics) were executed one by one. Then, we used the tool to normalize diffusion metrics (FA, MD) into the MNI space and calculated regional diffusion metrics by averaging the values within each region of the ICBM DTI-81 atlas (Mori et al. 2008). For further analyses, we focused on FA and MD values of all available projection, association, and commissural tract ROIs as hypothesized previously (Tables 1 and 2, Fig. 1).
Table 2 subjects
Statistical Analysis
Data are presented as mean±SD, percentages, n (%), or median (ranges)
All analyses were conducted using IBM SPSS Statistics software (version 19.0). The Shapiro–Wilk test of normality was used to investigate the distribution of demographic variables, cognitive performance, diffusion parameters. Demographic variables and cognitive performances were compared between the NC and VCIND group with independent sample t test for continuous variables, Mann-Whitney U test for nonparametric data, and χ2 test for sex proportion. As the distribution was not normal in diffusion values of several tracts, we conducted the nonparametric analysis (Mann-Whitney U test) to compare FA and MD values between groups. The mean and standard deviation, or the median and the interquartile range, were Table 1
NC
VCIND
P
n Age Sex (% male) Education (years)
28 70.9±8.2 57.1 12(9–12)
22 72.8±6.8 63.6 9.5(6–13)
0.39 0.64 0.37
MMSE MoCA
29(29–30) 26.8±1.6
26(24–27) 20.2±3.5
