- Email: [email protected]
P2-014: Imaging of cortical amyloid load and cerebral glucose metabolism in patients with Alzheimer's disease and mild cognitive impairment
Poster Presentations P2 tion for partial-volume effects improved on a previously derived multivariate AD-marker. Accuracy of classification was similar, while Z-values in the bootstrap estimates were more robust and localized to circumscribed network of hippocampal, occipital and basal-ganglia regions. References: 1) Asllani et al., JCBF&M, Oct24 (2007), 2) Donahue et al., MRM X (2006), 3) Asllani et al. (abstract submitted to current conference, ICAD 2008, Chicago) P2-014
IMAGING OF CORTICAL AMYLOID LOAD AND CEREBRAL GLUCOSE METABOLISM IN PATIENTS WITH ALZHEIMER’S DISEASE AND MILD COGNITIVE IMPAIRMENT
Shizuo Hatashita, Hidetomo Yamasaki, Nobuaki Shinozaki, ShonanAtsugi Hospital & clinic, Atsugi, Japan. Contact e-mail: [email protected] Background: The aim is to investigate the deposition of amyloid plaques by carbon 11-labeled Pittsburgh Compound B ([11C]-PIB) in the brains with dementia (Alzheimer’s type, AD) and mild cognitive impairment (MCI). We clarify the association between cerebral amyloid load and glucose metabolism. Methods: Thirty-four patients with dementia, 26 with MCI and 22 healthy control (HC) were studied. All 82 patients underwent 90-min dynamic [11C]-PIB PET and 20-min static [18F]-FDG PET. [11C]PIB data was acquired from 35-60 min after injection. Regions of interest (ROI) were defined on coregistered MRI and used in the analysis of the PET data. PIB distribution volume ratios (DVR) were calculated using Logan graphical analysis (cerebellar gray as reference region). [18F]-FDG PET images were extracted using 3 dimensional stereotactic surface projections (3D-SSP) by a Z-score on a pixel-by-pixel basis. Quantitative analysis for both [11C]-PIB and [18F]-FDG used the standardized uptake value ratio (SUVR) values of cortical regions. Results: A robust increase in PIB binding was observed in the anterior and posterior cingulate, precuneus, frontal, parietal, lateral temporal cortical areas in sixteen of 27 AD patients (typical PIB AD-pattern). The mean value of DVR in these cortical areas was significantly greater than in HC (2.66⫾0.56 vs 1.32⫾0.13, P⬍0.01). Of the 26 MCI patients, twelve showed typical AD-like patterns of amyloid deposition similar to AD. On [18F]-FDG PET 3D-SSP images, eleven of 16 AD patients with typical PIB AD-like pattern showed significant reduction of cortical glucose metabolism in temporoparietal, frontal, and posterior cingulate cortex and precuneus (classic metabolic AD-pattern). The other 5 patients showed hypometabolism in posteior cingulate gyrus and precuneus, but not in cortical regions. In contrast, three of 12 MCI patients with typical PIB AD-like pattern had the classic metabolic AD-pattern on 3D-SSP images. Only 3 patients showed no hypometabolism in any cortical regions. In patients with typical PIB AD- pattern, there was no correlation between PIB and FDG SUVR values in different cortical regions. Conclusions: The [11C]-PIB PET scan could potentially determine characteristic cerebral pattern of amyloid-beta plaque load. This amyloid plaque formation is not directly responsible for cerebral glucose metabolism in cortical regions. P2-015
CHARACTERIZATION OF AN ALZHEIMER’S DISEASE MOUSE MODEL Tg4510 WITH IMAGING TECHNIQUES
Mansuo Hayashi1, Jean-Cosme Dodart1, Hugh Nuthall1, Ilonka Guenther2, Alexandre Coimbra2, Rose Ann Blenman2, Michael Hutton1, Mark Shearman1, 1Merck Research Laboratories, Boston, MA, USA; 2Merck Research Laboratories, West Point, PA, USA. Contact e-mail: [email protected] Background: The early diagnosis of Alzheimer’s Disease (AD) has been aided by the use of imaging techniques such as magnetic resonance imaging/spectroscopy (MRI/MRS) and fluorodeoxyglucose- positron emission tomography (FDG-PET). Correlation of these imaging measurements with the severity of cognitive decline in patients may provide an objective way
T371
of staging AD. Methods: To examine whether such changes are detectable in an AD mouse model, we characterized rTg4510, which expresses a mutant form of tau that leads to neurodegeneration, with various imaging techniques. Results: Our preliminary results demonstrated that rTg4510, at 5 month of age, exhibited impairment in object recognition memory, as well as reduction in parenchyma volume of the brain in MRI. We are examining whether these mice also exhibit alterations in MRS/FDG-PET measurements. Notably, in rTg4510, tau expression could be suppressed by doxycycline- containing chow, which remarkably halts neurodegeneration and recovers cognition. Therefore, we will further examine whether rTg4510, following tau suppression, exhibits restoration of abnormal MRI/ MRS and FDG-PET measurements, coinciding with the blockage in neurodegeneration. Conclusions: The characterization of Tg4510 with imaging techniques should lend insignts onto biomarker development for Alzheimer’s disease. P2-016
CORTICAL BINDING OF PITTSBURGH COMPOUND B, AN ENDOPHENOTYPE FOR GENETIC STUDIES OF ALZHEIMER’S DISEASE
Anthony L. Hinrichs, Mark Mintun, Denise Head, Anne M. Fagan, David M. Holtzman, John C. Morris, Alison Goate, Washington University School of Medicine, Saint Louis, MO, USA. Contact e-mail: [email protected] Background: Little progress has been made in identifying genetic risk factors for Alzheimer’s disease (AD) during the last decade, underscoring the need for novel approaches to this problem. Because all known AD genes have been shown to influence A we hypothesized that quantitative measures of A deposition may be a useful endophenotype. The development of in vivo imaging of A deposition in the human brain using the Positron Emission Tomography (PET) ligand, Pittsburgh compound B (PIB), offers the possibility of using cortical PIB binding as a quantiative endophenotype for genetic studies of AD. Objective: To determine the heritability of cortical PIB binding and to test for association between cortical PIB binding and known risk factors for AD. Methods: Heritability can be estimated by twice the intraclass correlation coefficient (ICC) among siblings. We therefore undertook a study of elderly sibships. We ascertained 25 sibships without regard to disease status. The mean age was 72.3 with a mean of 2.2 individuals per sibship. Since the trait distribution is naturally truncated and many individuals show no deposition, the trait is significantly non-normal (p-value ⬍.0001). We therefore tested heritability using two other methods. We examined a sample of unrelated individuals in whom both PIB and CSF biomarker measurements have been made to test how much of the variance in PIB binding is explained by the biomarker levels and by APOE genotype. Results: ICC analysis using the Winer measure shows a heritability of .64. Permutation testing yields an empirical p-value of 0.03 to reject the hypothesis that the trait is non-heritable. Treating deposition as a dichotomous trait of presence (n⫽17) or absence (n⫽38) yields a heritability estimate of 0.63. While we observed that cortical PIB binding is correlated with AB42 levels in CSF, dementia severity (CDR), ApoE genotype, and gender, together these factors only explain 52% of the overall variance in PIB cortical binding. Conclusions: These results demonstrate that cortical PIB binding is a genetic trait with significant unique variability suggesting that it may be useful phenotype for large-scale genetic studies to identify risk factors for AD. P2-017
LOBAR GRAY AND WHITE MATTER ATROPHY IN ALZHEIMER’S DISEASE: PRELIMINARY RESULTS FROM THE ADNI PUBLIC DATABASE
Juebin Huang1, Alexander P. Auchus1,2, 1University of Tennessee Health Science Center, Memphis, TN, USA; 2Memphis Veterans Affairs Medical Center, Memphis, TN, USA. Contact e-mail: [email protected] Background: Regional brain volume loss beyond the medial temporal lobe has not been extensively studied in AD, particularly in the white matter.
IMAGING OF CORTICAL AMYLOID LOAD AND CEREBRAL GLUCOSE METABOLISM IN PATIENTS WITH ALZHEIMER’S DISEASE AND MILD COGNITIVE IMPAIRMENT
Shizuo Hatashita, Hidetomo Yamasaki, Nobuaki Shinozaki, ShonanAtsugi Hospital & clinic, Atsugi, Japan. Contact e-mail: [email protected] Background: The aim is to investigate the deposition of amyloid plaques by carbon 11-labeled Pittsburgh Compound B ([11C]-PIB) in the brains with dementia (Alzheimer’s type, AD) and mild cognitive impairment (MCI). We clarify the association between cerebral amyloid load and glucose metabolism. Methods: Thirty-four patients with dementia, 26 with MCI and 22 healthy control (HC) were studied. All 82 patients underwent 90-min dynamic [11C]-PIB PET and 20-min static [18F]-FDG PET. [11C]PIB data was acquired from 35-60 min after injection. Regions of interest (ROI) were defined on coregistered MRI and used in the analysis of the PET data. PIB distribution volume ratios (DVR) were calculated using Logan graphical analysis (cerebellar gray as reference region). [18F]-FDG PET images were extracted using 3 dimensional stereotactic surface projections (3D-SSP) by a Z-score on a pixel-by-pixel basis. Quantitative analysis for both [11C]-PIB and [18F]-FDG used the standardized uptake value ratio (SUVR) values of cortical regions. Results: A robust increase in PIB binding was observed in the anterior and posterior cingulate, precuneus, frontal, parietal, lateral temporal cortical areas in sixteen of 27 AD patients (typical PIB AD-pattern). The mean value of DVR in these cortical areas was significantly greater than in HC (2.66⫾0.56 vs 1.32⫾0.13, P⬍0.01). Of the 26 MCI patients, twelve showed typical AD-like patterns of amyloid deposition similar to AD. On [18F]-FDG PET 3D-SSP images, eleven of 16 AD patients with typical PIB AD-like pattern showed significant reduction of cortical glucose metabolism in temporoparietal, frontal, and posterior cingulate cortex and precuneus (classic metabolic AD-pattern). The other 5 patients showed hypometabolism in posteior cingulate gyrus and precuneus, but not in cortical regions. In contrast, three of 12 MCI patients with typical PIB AD-like pattern had the classic metabolic AD-pattern on 3D-SSP images. Only 3 patients showed no hypometabolism in any cortical regions. In patients with typical PIB AD- pattern, there was no correlation between PIB and FDG SUVR values in different cortical regions. Conclusions: The [11C]-PIB PET scan could potentially determine characteristic cerebral pattern of amyloid-beta plaque load. This amyloid plaque formation is not directly responsible for cerebral glucose metabolism in cortical regions. P2-015
CHARACTERIZATION OF AN ALZHEIMER’S DISEASE MOUSE MODEL Tg4510 WITH IMAGING TECHNIQUES
Mansuo Hayashi1, Jean-Cosme Dodart1, Hugh Nuthall1, Ilonka Guenther2, Alexandre Coimbra2, Rose Ann Blenman2, Michael Hutton1, Mark Shearman1, 1Merck Research Laboratories, Boston, MA, USA; 2Merck Research Laboratories, West Point, PA, USA. Contact e-mail: [email protected] Background: The early diagnosis of Alzheimer’s Disease (AD) has been aided by the use of imaging techniques such as magnetic resonance imaging/spectroscopy (MRI/MRS) and fluorodeoxyglucose- positron emission tomography (FDG-PET). Correlation of these imaging measurements with the severity of cognitive decline in patients may provide an objective way
T371
of staging AD. Methods: To examine whether such changes are detectable in an AD mouse model, we characterized rTg4510, which expresses a mutant form of tau that leads to neurodegeneration, with various imaging techniques. Results: Our preliminary results demonstrated that rTg4510, at 5 month of age, exhibited impairment in object recognition memory, as well as reduction in parenchyma volume of the brain in MRI. We are examining whether these mice also exhibit alterations in MRS/FDG-PET measurements. Notably, in rTg4510, tau expression could be suppressed by doxycycline- containing chow, which remarkably halts neurodegeneration and recovers cognition. Therefore, we will further examine whether rTg4510, following tau suppression, exhibits restoration of abnormal MRI/ MRS and FDG-PET measurements, coinciding with the blockage in neurodegeneration. Conclusions: The characterization of Tg4510 with imaging techniques should lend insignts onto biomarker development for Alzheimer’s disease. P2-016
CORTICAL BINDING OF PITTSBURGH COMPOUND B, AN ENDOPHENOTYPE FOR GENETIC STUDIES OF ALZHEIMER’S DISEASE
Anthony L. Hinrichs, Mark Mintun, Denise Head, Anne M. Fagan, David M. Holtzman, John C. Morris, Alison Goate, Washington University School of Medicine, Saint Louis, MO, USA. Contact e-mail: [email protected] Background: Little progress has been made in identifying genetic risk factors for Alzheimer’s disease (AD) during the last decade, underscoring the need for novel approaches to this problem. Because all known AD genes have been shown to influence A we hypothesized that quantitative measures of A deposition may be a useful endophenotype. The development of in vivo imaging of A deposition in the human brain using the Positron Emission Tomography (PET) ligand, Pittsburgh compound B (PIB), offers the possibility of using cortical PIB binding as a quantiative endophenotype for genetic studies of AD. Objective: To determine the heritability of cortical PIB binding and to test for association between cortical PIB binding and known risk factors for AD. Methods: Heritability can be estimated by twice the intraclass correlation coefficient (ICC) among siblings. We therefore undertook a study of elderly sibships. We ascertained 25 sibships without regard to disease status. The mean age was 72.3 with a mean of 2.2 individuals per sibship. Since the trait distribution is naturally truncated and many individuals show no deposition, the trait is significantly non-normal (p-value ⬍.0001). We therefore tested heritability using two other methods. We examined a sample of unrelated individuals in whom both PIB and CSF biomarker measurements have been made to test how much of the variance in PIB binding is explained by the biomarker levels and by APOE genotype. Results: ICC analysis using the Winer measure shows a heritability of .64. Permutation testing yields an empirical p-value of 0.03 to reject the hypothesis that the trait is non-heritable. Treating deposition as a dichotomous trait of presence (n⫽17) or absence (n⫽38) yields a heritability estimate of 0.63. While we observed that cortical PIB binding is correlated with AB42 levels in CSF, dementia severity (CDR), ApoE genotype, and gender, together these factors only explain 52% of the overall variance in PIB cortical binding. Conclusions: These results demonstrate that cortical PIB binding is a genetic trait with significant unique variability suggesting that it may be useful phenotype for large-scale genetic studies to identify risk factors for AD. P2-017
LOBAR GRAY AND WHITE MATTER ATROPHY IN ALZHEIMER’S DISEASE: PRELIMINARY RESULTS FROM THE ADNI PUBLIC DATABASE
Juebin Huang1, Alexander P. Auchus1,2, 1University of Tennessee Health Science Center, Memphis, TN, USA; 2Memphis Veterans Affairs Medical Center, Memphis, TN, USA. Contact e-mail: [email protected] Background: Regional brain volume loss beyond the medial temporal lobe has not been extensively studied in AD, particularly in the white matter.