XML Template (2014) [19.12.2014–2:23pm] //blrnas3.glyph.com/cenpro/ApplicationFiles/Journals/SAGE/3B2/ACRJ/Vol00000/140224/APPFile/SG-ACRJ140224.3d
(ACR)
[1–8] [PREPRINTER stage]
Acta Radiol OnlineFirst, published on December 23, 2014 as doi:10.1177/0284185114566088
Original Article
Alteration patterns of brain glucose metabolism: comparisons of healthy controls, subjective memory impairment and mild cognitive impairment
Acta Radiologica 0(0) 1–8 ! The Foundation Acta Radiologica 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0284185114566088 acr.sagepub.com
In-Uk Song1, Eun Kyoung Choi2, Jin Kyoung Oh2, Yong-An Chung2 and Sung-Woo Chung1
Abstract Background: Some groups have focused on the detection and management of subjective memory impairment (SMI) as the stage that precedes mild cognitive impairment (MCI). However, there have been few clinical studies that have examined biomarkers of SMI to date. Purpose: To investigate the differences in glucose metabolism as a prodromal marker of dementia in patients with SMI, MCI, and healthy controls using brain F-18 fluoro-2-deoxyglucose positron emission tomography (FDG-PET). Material and Methods: Sixty-eight consecutive patients with SMI, 47 patients with MCI, and 42 age-matched healthy subjects were recruited. All subjects underwent FDG-PET and detailed neuropsychological testing. FDG-PET images were analyzed using the statistical parametric mapping (SPM) program. Results: FDG-PET analysis showed glucose hypometabolism in the periventricular regions of patients with SMI and in the parietal, precentral frontal, and periventricular regions of patients with MCI compared with healthy controls. Interestingly, hypometabolism on FDG-PET was noted in the parietal and precentral frontal regions in MCI patients compared to SMI patients. Conclusion: The results suggest that hypometabolism in the periventricular regions as seen on FDG-PET may play a role as a predictive biomarker of pre-dementia, and the extension of reduced glucose metabolism into parietal regions likely reflects progression of cognitive deterioration.
Keywords Dementia, cognition, FDG, PET Date received: 25 August 2014; accepted: 7 December 2014
Introduction Alzheimer’s disease (AD) is the leading cause of dementia in the elderly, accounting for up to 70% of dementia cases. AD is a neurodegenerative clinicopathologic syndrome that leads to progressive, irreversible loss of memory and behavioral function, and is associated with pathologic neuronal loss and accumulation of b-amyloid plaques and neurofibrillary tangles and threads (1,2). Mild cognitive impairment (MCI) is defined as subjective memory impairment, belownormal performance on memory tests, largely unimpaired activities of daily living, and the absence of dementia (3). MCI is associated with long-term
dementia risk and represents the pre-dementia stage of neurodegenerative disorders in elderly subjects (4). Several previous studies have suggested that subjects 1 Department of Neurology, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea 2 Department of Radiology, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
Corresponding author: Yong-An Chung, Department of Radiology, Incheon St. Mary’s Hospital, The Catholic University of Korea, #56 Dongsu-ro, Bupyeong-gu, Incheon, 403-720, Republic of Korea. Email: [email protected]
Downloaded from acr.sagepub.com at Univ of Illinois at Chicago Library on February 1, 2015
XML Template (2014) [19.12.2014–2:23pm] //blrnas3.glyph.com/cenpro/ApplicationFiles/Journals/SAGE/3B2/ACRJ/Vol00000/140224/APPFile/SG-ACRJ140224.3d
(ACR)
[1–8] [PREPRINTER stage]
2
Acta Radiologica 0(0)
with MCI are at an increased risk for developing AD, in the range of 1–25% per year (5–7). An earlier clinical indicator of AD, even before the evolution of MCI, may be subjective memory impairment (SMI) (8,9). SMI is defined by subjective reports of memory worsening and cognitive performance in the normal adjusted range (9). Previous studies have also shown that SMI is a risk factor for cognitive decline and have generated preliminary evidence of AD-type pathology in patients with SMI (8,10–13). The primary goals of many neuroimaging studies that have examined biomarkers of early AD have been to detect and delay clinical progression of early AD. However, recent clinical and research interests have focused on MCI as a prodromal stage of AD. Early prevention of clinical progression to AD is important since it is a progressive and irreversible neurodegenerative disease with no known treatment. Moreover, other researchers have focused on the detection and management of SMI as the stage that precedes MCI. To the best of our knowledge, there have been few clinical studies of biomarkers of SMI to date. The aim of this study was to investigate differences in glucose metabolism in patients with SMI, MCI, and healthy controls using brain F-18 fluoro-2-deoxyglucose positron emission tomography (FDG-PET) as a potential prodromal biomarker of progression to dementia. An early, accurate diagnosis of prodromal dementia may be valuable since it could allow for proper management and prevent progression to dementia.
Material and Methods Patients This study was approved by the local ethics committee and written informed consent was obtained from each patient. All subjects were prospectively recruited, and the study was conducted between March 2011 and April 2013. There were 68 patients with SMI, 47 patients with MCI, and 42 healthy controls. All subjects were matched for their gender, age, and education level. All subjects were evaluated in the dementia clinic at Incheon St. Mary’s Hospital. The evaluation procedure consisted of a detailed medical history, physical and neurologic examinations, neuropsychological assessments, and brain magnetic resonance imaging (MRI). Additionally, positron emission tomography using FDG-PET was also performed on all subjects. The patients’ past medical histories were obtained from the patients and family members or from other caregivers. All SMI patients were diagnosed according to the definition of a subjective complaint of memory decline among old people in the absence of any objective
memory disturbance. Therefore, SMI was defined for the purposes of this study as self-reported impairments accompanied by normal performance on neuropsychological tests since there is no current consensus on standard criteria. The diagnosis of MCI was consistent based on the following criteria proposed by Petersen: (i) memory complaints; (ii) normal activities of daily living; (iii) normal general cognitive function; (iv) abnormal memory for age; and (v) lack of dementia (5). Healthy controls did not have a history or symptoms of Parkinson’s disease (PD), memory impairment, or other cognitive dysfunction according to a dementia screening questionnaire. They also had no history of other neurological problems such as head trauma, epilepsy, stroke, or brain surgery. Additionally, all subjects associated with hypertension, diabetes mellitus, hypercholesterolemia, and focal neurological signs or radiological lesions that typify cerebrovascular diseases were excluded.
Neuropsychological testing and data analysis Patients’ general cognitive state and severity of dementia were evaluated using the Mini-Mental State Examination (MMSE), the extended version of the Clinical Dementia Scale (CDR), the sum of the box score of the CDR (SOB), and Geriatric Depression Scale (GDS). Several cognitive domains were assessed by a detailed neuropsychological battery of tests, including an attention test (forward digit span, backward digit span, and calculation), a language and related function test (Boston Naming Test), a visuospatial function test (the Rey Complex Figure Test), a verbal memory test (three-word registration and recall, Hopkins Verbal Learning Test [HVLT] for immediate recall, delayed recall, and recognition), a non-verbal memory test (immediate recall, delayed recall, and recognition of a Rey complex figure) and a frontal executive function test (controlled oral word association test [animal, supermarket, and letter]). Statistical analyses were performed using SPSS software version 18.0 (SPSS Inc., Chicago, IL, USA). Analyses of variance with post hoc analyses were used for comparison of continuous variables, and the Pearson chi-squared analysis was used for comparison of categorical variables. Values were expressed as means and standard deviations. Statistical significance was assumed for P values < 0.05.
Image acquisition and analysis All F-18 FDG-PET scans were conducted after fasting period of at least 4 hours as well as with blood glucose levels lower than 180 mg/dL. Images were obtained approximately 45 min after FDG injection
Downloaded from acr.sagepub.com at Univ of Illinois at Chicago Library on February 1, 2015
XML Template (2014) [19.12.2014–2:23pm] //blrnas3.glyph.com/cenpro/ApplicationFiles/Journals/SAGE/3B2/ACRJ/Vol00000/140224/APPFile/SG-ACRJ140224.3d
(ACR)
[1–8] [PREPRINTER stage]
Song et al.
3
(185–222 MBq) using a Discovery STE PET/CT scanner (GE Healthcare, Milwaukee, WI, USA). All subjects were placed in the supine position on a bed in a dark, quiet room between the FDG injection and PET/ CT scanning. The scanner collected 47 simultaneous slices with 3.27 mm spacing, encompassing an axial field of view of 157 mm. Trans-axial and axial spatial resolution was approximately 4.4 mm full-width halfmaximum (FWHM) at the slice center. All studies were done in the three-dimensional acquisition mode. A 16-Slice CT scan was performed for tissue attenuation correction prior to the FDG-PET scan. The subject’s head was gently placed in a headholder with an air-inflated plastic spacer to minimize head movement. Imaging data were analyzed using Statistical Parametric Mapping (SPM) 2 (Institute of Neurology, University College of London, London, UK) implemented using Matlab software (Mathworks Inc, Natick, MA, USA). Prior to statistical analysis, all images were spatially normalized to the Montreal Neurological Institute (MNI, McGill University, Montreal, Canada) space to correct intersubject anatomical variabilities. An affine transformation was performed to determine the 12 optimal parameters essential for registering each brain on the MNI template. Subtle differences between transformed images and the template were removed by a non-linear registration method using the weighted sum of predefined smooth basis functions used in discrete cosine transformations. With an in-house Matlab-based program, the glucose metabolism value of each voxel was normalized to the pontine value, since glucose metabolism in the pons tends to be relatively preserved in AD (14). In this program, the volume of interest for the pons was predefined in the MNI space and applied to the individual spatially normalized PET images to measure the mean pontine activity. Normalized images were smoothed by convolution using an isotropic Gaussian
kernel with 8 mm full width at half maximum to accommodate intersubject differences in gyral and functional anatomies and to increase dataset signal-to-noise ratios. Differences in glucose metabolism between patients with SMI, MCI, and healthy controls were estimated on a voxel-by-voxel basis using t-tests. The resultant set of t-values constituted the SPM(t) map. The healthy control compared with SMI and MCI t-statistic image was under the threshold of t > 5.65, corresponding to a corrected (FWE) P value < 0.01 in conjunction with a cluster filter of 100 voxels. The t-statistic results for comparison between SMI and MCI were unavailable with the same corrected threshold. Therefore, the SMI and MCI t-statistic image was thresholded under t > 3.22, corresponding to an uncorrected P value < 0.001 in conjunction with a cluster filter of 50 voxels. This combined application of a statistical threshold and cluster filter has previously been shown to substantially reduce the false positive identification of activated pixels at any given threshold (15). The t-score clusters were projected onto the standard high-resolution T1-weighted MRI (software provides) for visualization and anatomic localization.
Results The number of subjects, mean age, education level, MMSE score, CDR score, and SOB score for each group are shown in Table 1. Each group was similar with respect to gender, age, and education level. On post hoc analyses, there were no significant differences between the SMI patients and healthy controls in most domains, including the attention test, the language and related functions test, non-verbal memory, frontal executive and visuospatial function tests, and most of the verbal memory tests (three-word registration, HVLT for immediate recall, delayed recall, and recognition). In contrast, the MCI group demonstrated
Table 1. Demographic data and general cognitive functions in the SMI, MCI and HC. Variables
MCI
SMI
HC
P value
Post hoc comparison
Subjects (n) Men (n) Mean age (years) Mean education (years) MMSE CDR SOB GDS
47 12 69.55 6.65 11.86 3.69 28.45 1.12 0.5 1.36 1.04 5.64 2.99
68 25 69.94 6.44 11.72 3.65 28.19 1.22 0.14 0.23 0.21 0.45 4.79 2.34
42 12 68.02 5.44 11.26 5.00 29.33 0.78 0.08 0.19 0.32 0.43 5.79 3.36
0.508 0.284 0.766
(ACR)
[1–8] [PREPRINTER stage]
Acta Radiol OnlineFirst, published on December 23, 2014 as doi:10.1177/0284185114566088
Original Article
Alteration patterns of brain glucose metabolism: comparisons of healthy controls, subjective memory impairment and mild cognitive impairment
Acta Radiologica 0(0) 1–8 ! The Foundation Acta Radiologica 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0284185114566088 acr.sagepub.com
In-Uk Song1, Eun Kyoung Choi2, Jin Kyoung Oh2, Yong-An Chung2 and Sung-Woo Chung1
Abstract Background: Some groups have focused on the detection and management of subjective memory impairment (SMI) as the stage that precedes mild cognitive impairment (MCI). However, there have been few clinical studies that have examined biomarkers of SMI to date. Purpose: To investigate the differences in glucose metabolism as a prodromal marker of dementia in patients with SMI, MCI, and healthy controls using brain F-18 fluoro-2-deoxyglucose positron emission tomography (FDG-PET). Material and Methods: Sixty-eight consecutive patients with SMI, 47 patients with MCI, and 42 age-matched healthy subjects were recruited. All subjects underwent FDG-PET and detailed neuropsychological testing. FDG-PET images were analyzed using the statistical parametric mapping (SPM) program. Results: FDG-PET analysis showed glucose hypometabolism in the periventricular regions of patients with SMI and in the parietal, precentral frontal, and periventricular regions of patients with MCI compared with healthy controls. Interestingly, hypometabolism on FDG-PET was noted in the parietal and precentral frontal regions in MCI patients compared to SMI patients. Conclusion: The results suggest that hypometabolism in the periventricular regions as seen on FDG-PET may play a role as a predictive biomarker of pre-dementia, and the extension of reduced glucose metabolism into parietal regions likely reflects progression of cognitive deterioration.
Keywords Dementia, cognition, FDG, PET Date received: 25 August 2014; accepted: 7 December 2014
Introduction Alzheimer’s disease (AD) is the leading cause of dementia in the elderly, accounting for up to 70% of dementia cases. AD is a neurodegenerative clinicopathologic syndrome that leads to progressive, irreversible loss of memory and behavioral function, and is associated with pathologic neuronal loss and accumulation of b-amyloid plaques and neurofibrillary tangles and threads (1,2). Mild cognitive impairment (MCI) is defined as subjective memory impairment, belownormal performance on memory tests, largely unimpaired activities of daily living, and the absence of dementia (3). MCI is associated with long-term
dementia risk and represents the pre-dementia stage of neurodegenerative disorders in elderly subjects (4). Several previous studies have suggested that subjects 1 Department of Neurology, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea 2 Department of Radiology, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea
Corresponding author: Yong-An Chung, Department of Radiology, Incheon St. Mary’s Hospital, The Catholic University of Korea, #56 Dongsu-ro, Bupyeong-gu, Incheon, 403-720, Republic of Korea. Email: [email protected]
Downloaded from acr.sagepub.com at Univ of Illinois at Chicago Library on February 1, 2015
XML Template (2014) [19.12.2014–2:23pm] //blrnas3.glyph.com/cenpro/ApplicationFiles/Journals/SAGE/3B2/ACRJ/Vol00000/140224/APPFile/SG-ACRJ140224.3d
(ACR)
[1–8] [PREPRINTER stage]
2
Acta Radiologica 0(0)
with MCI are at an increased risk for developing AD, in the range of 1–25% per year (5–7). An earlier clinical indicator of AD, even before the evolution of MCI, may be subjective memory impairment (SMI) (8,9). SMI is defined by subjective reports of memory worsening and cognitive performance in the normal adjusted range (9). Previous studies have also shown that SMI is a risk factor for cognitive decline and have generated preliminary evidence of AD-type pathology in patients with SMI (8,10–13). The primary goals of many neuroimaging studies that have examined biomarkers of early AD have been to detect and delay clinical progression of early AD. However, recent clinical and research interests have focused on MCI as a prodromal stage of AD. Early prevention of clinical progression to AD is important since it is a progressive and irreversible neurodegenerative disease with no known treatment. Moreover, other researchers have focused on the detection and management of SMI as the stage that precedes MCI. To the best of our knowledge, there have been few clinical studies of biomarkers of SMI to date. The aim of this study was to investigate differences in glucose metabolism in patients with SMI, MCI, and healthy controls using brain F-18 fluoro-2-deoxyglucose positron emission tomography (FDG-PET) as a potential prodromal biomarker of progression to dementia. An early, accurate diagnosis of prodromal dementia may be valuable since it could allow for proper management and prevent progression to dementia.
Material and Methods Patients This study was approved by the local ethics committee and written informed consent was obtained from each patient. All subjects were prospectively recruited, and the study was conducted between March 2011 and April 2013. There were 68 patients with SMI, 47 patients with MCI, and 42 healthy controls. All subjects were matched for their gender, age, and education level. All subjects were evaluated in the dementia clinic at Incheon St. Mary’s Hospital. The evaluation procedure consisted of a detailed medical history, physical and neurologic examinations, neuropsychological assessments, and brain magnetic resonance imaging (MRI). Additionally, positron emission tomography using FDG-PET was also performed on all subjects. The patients’ past medical histories were obtained from the patients and family members or from other caregivers. All SMI patients were diagnosed according to the definition of a subjective complaint of memory decline among old people in the absence of any objective
memory disturbance. Therefore, SMI was defined for the purposes of this study as self-reported impairments accompanied by normal performance on neuropsychological tests since there is no current consensus on standard criteria. The diagnosis of MCI was consistent based on the following criteria proposed by Petersen: (i) memory complaints; (ii) normal activities of daily living; (iii) normal general cognitive function; (iv) abnormal memory for age; and (v) lack of dementia (5). Healthy controls did not have a history or symptoms of Parkinson’s disease (PD), memory impairment, or other cognitive dysfunction according to a dementia screening questionnaire. They also had no history of other neurological problems such as head trauma, epilepsy, stroke, or brain surgery. Additionally, all subjects associated with hypertension, diabetes mellitus, hypercholesterolemia, and focal neurological signs or radiological lesions that typify cerebrovascular diseases were excluded.
Neuropsychological testing and data analysis Patients’ general cognitive state and severity of dementia were evaluated using the Mini-Mental State Examination (MMSE), the extended version of the Clinical Dementia Scale (CDR), the sum of the box score of the CDR (SOB), and Geriatric Depression Scale (GDS). Several cognitive domains were assessed by a detailed neuropsychological battery of tests, including an attention test (forward digit span, backward digit span, and calculation), a language and related function test (Boston Naming Test), a visuospatial function test (the Rey Complex Figure Test), a verbal memory test (three-word registration and recall, Hopkins Verbal Learning Test [HVLT] for immediate recall, delayed recall, and recognition), a non-verbal memory test (immediate recall, delayed recall, and recognition of a Rey complex figure) and a frontal executive function test (controlled oral word association test [animal, supermarket, and letter]). Statistical analyses were performed using SPSS software version 18.0 (SPSS Inc., Chicago, IL, USA). Analyses of variance with post hoc analyses were used for comparison of continuous variables, and the Pearson chi-squared analysis was used for comparison of categorical variables. Values were expressed as means and standard deviations. Statistical significance was assumed for P values < 0.05.
Image acquisition and analysis All F-18 FDG-PET scans were conducted after fasting period of at least 4 hours as well as with blood glucose levels lower than 180 mg/dL. Images were obtained approximately 45 min after FDG injection
Downloaded from acr.sagepub.com at Univ of Illinois at Chicago Library on February 1, 2015
XML Template (2014) [19.12.2014–2:23pm] //blrnas3.glyph.com/cenpro/ApplicationFiles/Journals/SAGE/3B2/ACRJ/Vol00000/140224/APPFile/SG-ACRJ140224.3d
(ACR)
[1–8] [PREPRINTER stage]
Song et al.
3
(185–222 MBq) using a Discovery STE PET/CT scanner (GE Healthcare, Milwaukee, WI, USA). All subjects were placed in the supine position on a bed in a dark, quiet room between the FDG injection and PET/ CT scanning. The scanner collected 47 simultaneous slices with 3.27 mm spacing, encompassing an axial field of view of 157 mm. Trans-axial and axial spatial resolution was approximately 4.4 mm full-width halfmaximum (FWHM) at the slice center. All studies were done in the three-dimensional acquisition mode. A 16-Slice CT scan was performed for tissue attenuation correction prior to the FDG-PET scan. The subject’s head was gently placed in a headholder with an air-inflated plastic spacer to minimize head movement. Imaging data were analyzed using Statistical Parametric Mapping (SPM) 2 (Institute of Neurology, University College of London, London, UK) implemented using Matlab software (Mathworks Inc, Natick, MA, USA). Prior to statistical analysis, all images were spatially normalized to the Montreal Neurological Institute (MNI, McGill University, Montreal, Canada) space to correct intersubject anatomical variabilities. An affine transformation was performed to determine the 12 optimal parameters essential for registering each brain on the MNI template. Subtle differences between transformed images and the template were removed by a non-linear registration method using the weighted sum of predefined smooth basis functions used in discrete cosine transformations. With an in-house Matlab-based program, the glucose metabolism value of each voxel was normalized to the pontine value, since glucose metabolism in the pons tends to be relatively preserved in AD (14). In this program, the volume of interest for the pons was predefined in the MNI space and applied to the individual spatially normalized PET images to measure the mean pontine activity. Normalized images were smoothed by convolution using an isotropic Gaussian
kernel with 8 mm full width at half maximum to accommodate intersubject differences in gyral and functional anatomies and to increase dataset signal-to-noise ratios. Differences in glucose metabolism between patients with SMI, MCI, and healthy controls were estimated on a voxel-by-voxel basis using t-tests. The resultant set of t-values constituted the SPM(t) map. The healthy control compared with SMI and MCI t-statistic image was under the threshold of t > 5.65, corresponding to a corrected (FWE) P value < 0.01 in conjunction with a cluster filter of 100 voxels. The t-statistic results for comparison between SMI and MCI were unavailable with the same corrected threshold. Therefore, the SMI and MCI t-statistic image was thresholded under t > 3.22, corresponding to an uncorrected P value < 0.001 in conjunction with a cluster filter of 50 voxels. This combined application of a statistical threshold and cluster filter has previously been shown to substantially reduce the false positive identification of activated pixels at any given threshold (15). The t-score clusters were projected onto the standard high-resolution T1-weighted MRI (software provides) for visualization and anatomic localization.
Results The number of subjects, mean age, education level, MMSE score, CDR score, and SOB score for each group are shown in Table 1. Each group was similar with respect to gender, age, and education level. On post hoc analyses, there were no significant differences between the SMI patients and healthy controls in most domains, including the attention test, the language and related functions test, non-verbal memory, frontal executive and visuospatial function tests, and most of the verbal memory tests (three-word registration, HVLT for immediate recall, delayed recall, and recognition). In contrast, the MCI group demonstrated
Table 1. Demographic data and general cognitive functions in the SMI, MCI and HC. Variables
MCI
SMI
HC
P value
Post hoc comparison
Subjects (n) Men (n) Mean age (years) Mean education (years) MMSE CDR SOB GDS
47 12 69.55 6.65 11.86 3.69 28.45 1.12 0.5 1.36 1.04 5.64 2.99
68 25 69.94 6.44 11.72 3.65 28.19 1.22 0.14 0.23 0.21 0.45 4.79 2.34
42 12 68.02 5.44 11.26 5.00 29.33 0.78 0.08 0.19 0.32 0.43 5.79 3.36
0.508 0.284 0.766
