- Email: [email protected]
Molecular markers of Alzheimer's Disease pathology and their relationship with default mode network integrity
S2
Alzheimer’s Imaging Consortium: IC-01: Oral 1-Clinical and Imaging Correlates of Brain Amyloidosis
Background: Beta-amyloid (Aß) is a key component of Alzheimer’s disease pathology, and previous reports suggest that w25% of cognitively healthy older adults have increased levels of amyloid. Relatively little is known about the cognitive consequences of amyloid deposition in healthy adults, or the magnitude of its effects on cognition. In the present study, we examined the relationship of amyloid burden to cognitive function in a large lifespan sample of healthy adults, beginning at age 30 up to age 89. We hypothesized that amyloid would exert a negative effect on cognition, but expected that a substantial amount of age-related decline would remain, even after effects due to amyloid deposition were controlled. Methods: 137 highly educated (mean 16.40 years), healthy adults (age 30-89, mean MMSE 29.28) screened against history of cardiovascular, neurological or psychiatric conditions were scanned on a Siemens ECAT HR scanner 50 min after injection of 370 MBq bolus of 18F-AV45(Florbetapir) to image beta-amyloid deposition. PET data were spatially normalized to an AV-45 template and SUVRs extracted from 8 ROIs (normalized to cerebellum). Two cognitive tests measured each of four domains: processing speed, working memory, fluid reasoning and crystallized intelligence. Results: In the full sample (controlling forage), as amyloid burden increased, decreases were observed in processing speed, working memory, and fluid reasoning but not crystallized intelligence. A subgroup of 25 individuals aged 60+ with marked amyloid elevation (exceeded the 95% CI) showed strong associations between level of amyloid and cognitive function: in these subjects, higher amyloid burden was associated with decreased speed (r ¼ -.59), working memory (r ¼ -.63), and reasoning (r ¼ -.54), but not crystallized ability. Despite these effects, when we removed these subjects from the sample, strong age-related decline was still observed in low amyloid participants. Conclusions: Although these data clearly implicate amyloid in age-related cognitive changes, very strong effects of aging on speed, working memory and reasoning remain after the elevated amyloid subgroup was removed. Thus, even healthy subjects without high amyloid burden experience declines in cognitive function with age. IC-01-02
THE IMAP* PROJECT: RELATIVE DEGREES OF REGIONAL GREY MATTER ATROPHY, HYPOMETABOLISM AND B-AMYLOID DEPOSITION IN ALZHEIMER’S DISEASE ASSESSED THROUGH MULTIMODAL NEUROIMAGING
Renaud La Joie1, Tauber Clovis1, Florence Mezenge2, Vincent Camus2, Audrey Perrotin3, Maria Ribeiro2, Brigitte Landeau1, Meziane Ibazizene3, Denis Guilloteau2, Louisa Barre3, Vincent de La Sayette1, Ahmed Abbas1, Francis Eustache1, Beatrice Desgranges1, Ga€el Chetelat1, 1Inserm U923, Caen, France; 2Universite Franc¸ois Rabelais de Tours, Tours, France; 3 University of Caen/Basse-Normandie. Background: Grey matter atrophy, cortical hypometabolism and b-amyloid (Ab) deposition are well-described hallmarks of Alzheimer’s disease (AD)
that involve partly overlapping brain networks; however the relationship between these different alterations is poorly understood. The aim of the present study was to assess the relative degree of regional atrophy, hypometabolism and Ab deposition using three neuroimaging modalities in the same patients. Methods: 9 patients with clinical AD and 20 healthy controls underwent structural MRI, FDG-PET and AV45-PET at the Cyceron center (Caen, France). MRI images were handled using the VBM5 toolbox. PET data were corrected for partial volume effects and spatially normalized using the corresponding MRI data. For each patient and modality, z-score maps were computed in comparison to the control group (MRI and FDG-PET scores were reversed so that positive scores indicate a pathology) and mean values were extracted from several regions of interest. Results: Regional mean zvalues in neocortical regions suggested a hierarchy in the three alterations (see Figure), with atrophy (mean neocortical z-score¼1.78) being found only in some of the hypometabolic regions, and hypometabolism (z¼2.25) being present in some of the regions that showed high Ab load (z¼3.65). The predominance of Ab deposition is especially marked in the frontal cortex. Notably, a contrasted gradient was found in the medial temporal lobe where atrophy strongly predominated (z¼2.48) while hypometabolism was mild (z¼1.58) and Ab deposition non-significant (z¼0.89). Conclusions: These findings showed that in most brain areas, the amount of local Ab deposition exceeds the degree of glucose hypometabolism which itself exceeds atrophy. This observation might be explained by the successive appearance of those three pathological features or differences in their rate of change over time, in line with recent models on the dynamic of AD pathogenesis (see Peterson et al., 2010). However, the hierarchy is clearly different in the hippocampus, which may reflect an increased susceptibility to Ab toxicity, the presence of neurofibrillary tangles and/or compensatory mechanisms. Longitudinal studies assessing earlier stages of AD with repeated measures for the three modalities will allow confirming these interpretations. IC-01-03
MOLECULAR MARKERS OF ALZHEIMER’S DISEASE PATHOLOGY AND THEIR RELATIONSHIP WITH DEFAULT MODE NETWORK INTEGRITY
Sofie Adriaanse1, Ernesto Sanz-Arigita2, Maja Binnewijzend1, Nelleke Tolboom1, Alle Meije Wink1, Ronald Boellaard1, Maqsood Yaqub1, Albert Windhorst1, Wiesje Van der Flier1, Philip Scheltens1, Adriaan Lammertsma1, Frederik Barkhof1, Bart van Berckel1, 1VU University Medical Center, Amsterdam, Netherlands; 2CITA-Alzheimer, San Sebastian, Spain. Background: Several positron emission tomography (PET) tracers have been developed as molecular markers for in vivo imaging of Alzheimer’s disease (AD) pathology, the first being [11C]PIB and [18F]FDDNP. [11C] PIB measures the amount of fibrillar Aß load, while [18F]FDDNP also binds to neurofibrillary tangles. On a functional level, connectivity in the default mode network (DMN) is decreased in AD patients. The aim of this study was to investigate the relationship of [11C]PIB and [18F]FDDNP binding with DMN functional integrity in patients with AD, Mild Cognitive Impairment (MCI) and healthy controls. Methods: Dynamic, 90 minutes [11C]PIB and [18F]FDDNP PET scans and resting-state fMRI scans were obtained in 14 AD patients, 12 MCI patients and 15 healthy controls. For [11C]PIB and [18F]FDDNP, parametric images of binding potential (BPND) were generated using a basis function method implementation of the simplified reference tissue model. To identify the DMN, Independent Component Analysis (ICA) was performed. Next, dual-regression (back-projection) analysis was done in order to yield individual maps corresponding to the DMN. The mean of all voxels within this individual map was calculated and correlated to [11C]PIB- and [18F]FDDNP BPND. Results: Consistent with previous studies, global [11C]PIB binding was higher in AD and MCI patients compared with controls (p<0.05). Global [18F]FDDNP values were increased in AD patients only (p<0.05). DMN connectivity was modestly, albeit not significantly, reduced in AD patients (p¼0.33). Correlation analysis across groups revealed that reduced DMN connectivity was not associated with global [11C]PIB binding (Pearson r¼-0.01, p¼0.95). Also no correlations between these parameters were found when examining patients only. Increased whole brain [18F]FDDNP binding was related to decreased DMN related activity (r¼-0.34, p¼0.04). When examining regional binding of [18F]FDDNP with DMN connectivity, a negative correlation with DMN connectivity was found, most evident for [18F]FDDNP binding within the temporal lobe (r¼-0.46, p¼0.003). This association remained when
Alzheimer’s Imaging Consortium-IC-02: Oral 2-Markers for Predementia and Differential Diagnosis examining patients only (r¼-0.55, p¼0.003). Conclusions: This is the first study indicating that besides amyloid accumulation tangle formation might play an important role in functional connectivity integrity in the DMN in AD. IC-01-04
CEREBRAL MICROBLEEDS ARE ASSOCIATED WITH WORSE COGNITIVE FUNCTION: THE ROTTERDAM SCAN STUDY
M. Ikram, Mari€elle Poels, Aad van der Lugt, Albert Hofman, Gabriel Krestin, Monique Breteler, Meike Vernooij, Erasmus MC University Medical Center, Rotterdam, Netherlands. Background: Vascular pathology has shown to play a prominent role in impaired cognitive function. Microbleeds are frequently found in the general elderly population and may reflect underlying vascular disease, in particular cerebral amyloid angiopathy (CAA) and hypertensive vasculopathy. Their role in cognitive function in the general elderly population, however, is unknown. We investigated the association between microbleeds and performance in multiple cognitive domains. Methods: This study is based on 3979 non-demented persons (mean age, 60.3 years) from the population-based Rotterdam Scan Study. We used linear regression models to estimate the association of microbleeds with different cognitive domains. Separate categories were made for 1 microbleed, 2-5 microbleeds, and >5 microbleeds. We stratified analyses for location of microbleeds and APOE e4 allele status. Additional adjustments were made for vascular risk factors, brain atrophy and other markers of small vessel disease. Main outcome measures were performance on MMSE-score and neuropsychological tests in the following cognitive domains: memory, executive function, information processing speed, global cognition and motor speed. Results: Presence of more than 5 microbleeds was associated with worse performance in all cognitive domains, except memory. These associations were most marked in participants with strictly lobar microbleeds, whereas associations were much weaker or absent for deep or infratentorial microbleeds. After adjustment for brain atrophy, vascular risk factors or other markers of small vessel disease the associations attenuated only marginally. Conclusions: Presence of multiple microbleeds, especially in a strictly lobar location, is associated with worse performance on tests measuring cognitive function, even after adjustments for vascular risk factors and other markers of small vessel disease. These results suggest an independent role for microbleeds in cognitive impairment. IC-01-05
EFFECTS OF SURGERY ON CORTICAL ATROPHY
Richard Kline1, Elizabeth Pirraglia2, Hao Cheng2, Yi Li2, Michael Haile2, Mony De Leon2, Alex Bekker2, 1Langone NYU Medical School, New York, New York, United States; 2NYU Langone Medical Center, New York, New York, United States. Background: There have been numerous reports of acute and delayed impairment of neurocognitive performance following surgery. Structural MRI has been used to successfully diagnose and monitor Alzheimer’s disease, and is here applied to examine the impact of surgery on baseline atrophy in elderly patients. We compared changes in cortical anatomy for defined inter-visit intervals between a surgical cohort and a frequency matched non-surgical control. Methods: Data was obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (www.loni.ucla.edu/ADNI), and included subjects either cognitively normal (NL) or diagnosed with mild cognitive impairment (MCI), and who did not progress. Ages are from 55 through 90, with minimum 6 grades of education. Putative surgical patients were selected by searching the comment fields for positive key words (e.g. surgery, thoracotomy, operative etc.) and then screening manually for confirmation. The final cohort contained 32 subjects and 79 non-surgical subjects, frequency matched by follow up time. Repeated Measures ANCOVA was used to evaluate the change in bilateral grey (GM) and white matter (WM) volume, before and after surgery, or between the last two evaluations for non-surgery group. Covariants included intracranial volume, age, and months to follow-up evaluation with diagnostic group (NL or MCI) a cofactor. Statistical significance was p < 0.05 (SPSS version 18.0; Chicago, IL). Results: Examining GM volume (left hemisphere), the 3-way interaction of evaluation, diagnosis, and surgery was significant (F1,104 ¼6.4, p<.05) indicating that slopes were different depending on diagnostic group; where MCI with surgery showed a significantly greater decrease in GM volume compared to MCI without surgery (F1,47 ¼7.9, p<.01); no significant difference for NL. Examining GM (right hemisphere), the 3-way interaction was not significant, but there was a significant decrease in GM volume (F1,105 ¼4.1, p<.05) for surgery compared to non-surgery
S3
subjects. No significant differences found for WM volume. Conclusions: In our cohort of elderly subjects, surgery resulted in a more pronounced GM volume atrophy than seen in non-surgical subjects, with MCI subjects at higher risk. An examination of neurocognitive effects of surgery, in progress, suggests a correlation between atrophy and function. JULY16, 2011 ALZHEIMER’S IMAGING CONSORTIUM-IC-02: ORAL 2-MARKERS FOR PREDEMENTIA AND DIFFERENTIAL DIAGNOSIS IC-02-01
FUNCTIONAL CONNECTIVITY IN DEMENTIA WITH LEWY BODIES AND ALZHEIMER’S DISEASE: A RESTING BOLD STUDY
Eva Kenny1, Andrew Blamire1, Michael Firbank1, John O’Brien1, 1 Newcastle University, Newcastle upon Tyne, United Kingdom. Background: Using resting-state functional magnetic resonance imaging (fMRI), spontaneous low-frequency fluctuations (SLFs) in the blood oxygenation level dependent (BOLD) signal (<0.1Hz) were measured to investigate connectivity between key brain regions hypothesised to be differentially affected in dementia with Lewy bodies (DLB) compared to Alzheimer’s disease (AD) and age-matched healthy controls. Methods: Forty-seven subjects (>60 years); 15 DLB, 16 AD and 16 controls, were scanned using a 3 Tesla MRI system. Using FSL, the mean BOLD signal time-series was extracted from seed regions; head of caudate nucleus, thalamus, putamen, hippocampus, posterior cingulate, precuneus and primary visual cortex, and cross-correlated with all other brain voxels to determine functional connectivity. Results were cluster thresholded at Z>2.3 and a corrected significance of p¼0.05. Results: In DLB, connectivity was significantly greater than controls with the head of caudate (bilateral), putamen (left), thalamus (bilateral) and posterior cingulate (right). In AD, connectivity was significantly greater than controls with the head of caudate (right), thalamus (left), hippocampus (left) and posterior cingulate (right). Between the dementia groups, DLB subjects showed significantly greater connectivity than AD in subcortical regions (putamen), whereas AD subjects showed significantly greater connectivity than DLB in cortex (posterior cingulate). Hippocampal connectivity was abnormal in AD, but normal in DLB, while precuneus and primary visual cortex connectivity was normal in DLB and AD. No regions showed significantly lower connectivity in DLB or AD versus controls. Conclusions: This is the first study to investigate cortical and subcortical connectivity using resting BOLD in DLB. We found clear differences in connectivity between DLB and AD; abnormal subcortical connectivity in DLB and abnormal cortical connectivity in AD. Consistent with the known relative preservation of memory in DLB compared to AD, hippocampal connectivity was normal in DLB. Importantly, while metabolic imaging shows functional change in primary visual cortex in DLB, with changes hypothesised to account for visual hallucinations, we found connectivity of this region to be normal. This implicates areas beyond visual sensory input level in the visual dysfunction (visuoperceptual disturbance and visual hallucinations) seen in DLB. This study also illustrates the potential of functional connectivity as a tool for better understanding neurobiological changes in dementia. IC-02-02
PREDICTING PROGRESSION TO ALZHEIMER’S DISEASE IN MCI USING COMBINED STRUCTURAL IMAGING AND CSF BIOMARKERS
Linda McEvoy1, David Heister1, Kaj Blennow2, James Brewer1, Anders Dale1, 1University of California, San Diego, La Jolla, California, United States; 2G€oteborg University, M€olndal, Sweden. Background: Cerebrospinal fluid (CSF) biomarkers and atrophy on structural magnetic resonance images (MRIs) each predict decline to Alzheimer’s disease (AD) from mild cognitive impairment (MCI). We investigated whether the combination of these measures would improve outcome prediction in patients with MCI. Methods: The 3-year risk of progressing to AD was examined in 178 MCI participants from the Alzheimer’s Disease Neuroimaging Initiative. MCI participants were stratified into low and high risk groups on the basis of structural atrophy or on CSF levels of total tau and Aß 1-42 proteins. To classify MCI patients into risk groups on the basis of atrophy, a linear discriminant model previously found to best discriminate AD from control data (McEvoy et al. Radiology, 2009) was applied to data from MCI patients. To classify participants into
Alzheimer’s Imaging Consortium: IC-01: Oral 1-Clinical and Imaging Correlates of Brain Amyloidosis
Background: Beta-amyloid (Aß) is a key component of Alzheimer’s disease pathology, and previous reports suggest that w25% of cognitively healthy older adults have increased levels of amyloid. Relatively little is known about the cognitive consequences of amyloid deposition in healthy adults, or the magnitude of its effects on cognition. In the present study, we examined the relationship of amyloid burden to cognitive function in a large lifespan sample of healthy adults, beginning at age 30 up to age 89. We hypothesized that amyloid would exert a negative effect on cognition, but expected that a substantial amount of age-related decline would remain, even after effects due to amyloid deposition were controlled. Methods: 137 highly educated (mean 16.40 years), healthy adults (age 30-89, mean MMSE 29.28) screened against history of cardiovascular, neurological or psychiatric conditions were scanned on a Siemens ECAT HR scanner 50 min after injection of 370 MBq bolus of 18F-AV45(Florbetapir) to image beta-amyloid deposition. PET data were spatially normalized to an AV-45 template and SUVRs extracted from 8 ROIs (normalized to cerebellum). Two cognitive tests measured each of four domains: processing speed, working memory, fluid reasoning and crystallized intelligence. Results: In the full sample (controlling forage), as amyloid burden increased, decreases were observed in processing speed, working memory, and fluid reasoning but not crystallized intelligence. A subgroup of 25 individuals aged 60+ with marked amyloid elevation (exceeded the 95% CI) showed strong associations between level of amyloid and cognitive function: in these subjects, higher amyloid burden was associated with decreased speed (r ¼ -.59), working memory (r ¼ -.63), and reasoning (r ¼ -.54), but not crystallized ability. Despite these effects, when we removed these subjects from the sample, strong age-related decline was still observed in low amyloid participants. Conclusions: Although these data clearly implicate amyloid in age-related cognitive changes, very strong effects of aging on speed, working memory and reasoning remain after the elevated amyloid subgroup was removed. Thus, even healthy subjects without high amyloid burden experience declines in cognitive function with age. IC-01-02
THE IMAP* PROJECT: RELATIVE DEGREES OF REGIONAL GREY MATTER ATROPHY, HYPOMETABOLISM AND B-AMYLOID DEPOSITION IN ALZHEIMER’S DISEASE ASSESSED THROUGH MULTIMODAL NEUROIMAGING
Renaud La Joie1, Tauber Clovis1, Florence Mezenge2, Vincent Camus2, Audrey Perrotin3, Maria Ribeiro2, Brigitte Landeau1, Meziane Ibazizene3, Denis Guilloteau2, Louisa Barre3, Vincent de La Sayette1, Ahmed Abbas1, Francis Eustache1, Beatrice Desgranges1, Ga€el Chetelat1, 1Inserm U923, Caen, France; 2Universite Franc¸ois Rabelais de Tours, Tours, France; 3 University of Caen/Basse-Normandie. Background: Grey matter atrophy, cortical hypometabolism and b-amyloid (Ab) deposition are well-described hallmarks of Alzheimer’s disease (AD)
that involve partly overlapping brain networks; however the relationship between these different alterations is poorly understood. The aim of the present study was to assess the relative degree of regional atrophy, hypometabolism and Ab deposition using three neuroimaging modalities in the same patients. Methods: 9 patients with clinical AD and 20 healthy controls underwent structural MRI, FDG-PET and AV45-PET at the Cyceron center (Caen, France). MRI images were handled using the VBM5 toolbox. PET data were corrected for partial volume effects and spatially normalized using the corresponding MRI data. For each patient and modality, z-score maps were computed in comparison to the control group (MRI and FDG-PET scores were reversed so that positive scores indicate a pathology) and mean values were extracted from several regions of interest. Results: Regional mean zvalues in neocortical regions suggested a hierarchy in the three alterations (see Figure), with atrophy (mean neocortical z-score¼1.78) being found only in some of the hypometabolic regions, and hypometabolism (z¼2.25) being present in some of the regions that showed high Ab load (z¼3.65). The predominance of Ab deposition is especially marked in the frontal cortex. Notably, a contrasted gradient was found in the medial temporal lobe where atrophy strongly predominated (z¼2.48) while hypometabolism was mild (z¼1.58) and Ab deposition non-significant (z¼0.89). Conclusions: These findings showed that in most brain areas, the amount of local Ab deposition exceeds the degree of glucose hypometabolism which itself exceeds atrophy. This observation might be explained by the successive appearance of those three pathological features or differences in their rate of change over time, in line with recent models on the dynamic of AD pathogenesis (see Peterson et al., 2010). However, the hierarchy is clearly different in the hippocampus, which may reflect an increased susceptibility to Ab toxicity, the presence of neurofibrillary tangles and/or compensatory mechanisms. Longitudinal studies assessing earlier stages of AD with repeated measures for the three modalities will allow confirming these interpretations. IC-01-03
MOLECULAR MARKERS OF ALZHEIMER’S DISEASE PATHOLOGY AND THEIR RELATIONSHIP WITH DEFAULT MODE NETWORK INTEGRITY
Sofie Adriaanse1, Ernesto Sanz-Arigita2, Maja Binnewijzend1, Nelleke Tolboom1, Alle Meije Wink1, Ronald Boellaard1, Maqsood Yaqub1, Albert Windhorst1, Wiesje Van der Flier1, Philip Scheltens1, Adriaan Lammertsma1, Frederik Barkhof1, Bart van Berckel1, 1VU University Medical Center, Amsterdam, Netherlands; 2CITA-Alzheimer, San Sebastian, Spain. Background: Several positron emission tomography (PET) tracers have been developed as molecular markers for in vivo imaging of Alzheimer’s disease (AD) pathology, the first being [11C]PIB and [18F]FDDNP. [11C] PIB measures the amount of fibrillar Aß load, while [18F]FDDNP also binds to neurofibrillary tangles. On a functional level, connectivity in the default mode network (DMN) is decreased in AD patients. The aim of this study was to investigate the relationship of [11C]PIB and [18F]FDDNP binding with DMN functional integrity in patients with AD, Mild Cognitive Impairment (MCI) and healthy controls. Methods: Dynamic, 90 minutes [11C]PIB and [18F]FDDNP PET scans and resting-state fMRI scans were obtained in 14 AD patients, 12 MCI patients and 15 healthy controls. For [11C]PIB and [18F]FDDNP, parametric images of binding potential (BPND) were generated using a basis function method implementation of the simplified reference tissue model. To identify the DMN, Independent Component Analysis (ICA) was performed. Next, dual-regression (back-projection) analysis was done in order to yield individual maps corresponding to the DMN. The mean of all voxels within this individual map was calculated and correlated to [11C]PIB- and [18F]FDDNP BPND. Results: Consistent with previous studies, global [11C]PIB binding was higher in AD and MCI patients compared with controls (p<0.05). Global [18F]FDDNP values were increased in AD patients only (p<0.05). DMN connectivity was modestly, albeit not significantly, reduced in AD patients (p¼0.33). Correlation analysis across groups revealed that reduced DMN connectivity was not associated with global [11C]PIB binding (Pearson r¼-0.01, p¼0.95). Also no correlations between these parameters were found when examining patients only. Increased whole brain [18F]FDDNP binding was related to decreased DMN related activity (r¼-0.34, p¼0.04). When examining regional binding of [18F]FDDNP with DMN connectivity, a negative correlation with DMN connectivity was found, most evident for [18F]FDDNP binding within the temporal lobe (r¼-0.46, p¼0.003). This association remained when
Alzheimer’s Imaging Consortium-IC-02: Oral 2-Markers for Predementia and Differential Diagnosis examining patients only (r¼-0.55, p¼0.003). Conclusions: This is the first study indicating that besides amyloid accumulation tangle formation might play an important role in functional connectivity integrity in the DMN in AD. IC-01-04
CEREBRAL MICROBLEEDS ARE ASSOCIATED WITH WORSE COGNITIVE FUNCTION: THE ROTTERDAM SCAN STUDY
M. Ikram, Mari€elle Poels, Aad van der Lugt, Albert Hofman, Gabriel Krestin, Monique Breteler, Meike Vernooij, Erasmus MC University Medical Center, Rotterdam, Netherlands. Background: Vascular pathology has shown to play a prominent role in impaired cognitive function. Microbleeds are frequently found in the general elderly population and may reflect underlying vascular disease, in particular cerebral amyloid angiopathy (CAA) and hypertensive vasculopathy. Their role in cognitive function in the general elderly population, however, is unknown. We investigated the association between microbleeds and performance in multiple cognitive domains. Methods: This study is based on 3979 non-demented persons (mean age, 60.3 years) from the population-based Rotterdam Scan Study. We used linear regression models to estimate the association of microbleeds with different cognitive domains. Separate categories were made for 1 microbleed, 2-5 microbleeds, and >5 microbleeds. We stratified analyses for location of microbleeds and APOE e4 allele status. Additional adjustments were made for vascular risk factors, brain atrophy and other markers of small vessel disease. Main outcome measures were performance on MMSE-score and neuropsychological tests in the following cognitive domains: memory, executive function, information processing speed, global cognition and motor speed. Results: Presence of more than 5 microbleeds was associated with worse performance in all cognitive domains, except memory. These associations were most marked in participants with strictly lobar microbleeds, whereas associations were much weaker or absent for deep or infratentorial microbleeds. After adjustment for brain atrophy, vascular risk factors or other markers of small vessel disease the associations attenuated only marginally. Conclusions: Presence of multiple microbleeds, especially in a strictly lobar location, is associated with worse performance on tests measuring cognitive function, even after adjustments for vascular risk factors and other markers of small vessel disease. These results suggest an independent role for microbleeds in cognitive impairment. IC-01-05
EFFECTS OF SURGERY ON CORTICAL ATROPHY
Richard Kline1, Elizabeth Pirraglia2, Hao Cheng2, Yi Li2, Michael Haile2, Mony De Leon2, Alex Bekker2, 1Langone NYU Medical School, New York, New York, United States; 2NYU Langone Medical Center, New York, New York, United States. Background: There have been numerous reports of acute and delayed impairment of neurocognitive performance following surgery. Structural MRI has been used to successfully diagnose and monitor Alzheimer’s disease, and is here applied to examine the impact of surgery on baseline atrophy in elderly patients. We compared changes in cortical anatomy for defined inter-visit intervals between a surgical cohort and a frequency matched non-surgical control. Methods: Data was obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) (www.loni.ucla.edu/ADNI), and included subjects either cognitively normal (NL) or diagnosed with mild cognitive impairment (MCI), and who did not progress. Ages are from 55 through 90, with minimum 6 grades of education. Putative surgical patients were selected by searching the comment fields for positive key words (e.g. surgery, thoracotomy, operative etc.) and then screening manually for confirmation. The final cohort contained 32 subjects and 79 non-surgical subjects, frequency matched by follow up time. Repeated Measures ANCOVA was used to evaluate the change in bilateral grey (GM) and white matter (WM) volume, before and after surgery, or between the last two evaluations for non-surgery group. Covariants included intracranial volume, age, and months to follow-up evaluation with diagnostic group (NL or MCI) a cofactor. Statistical significance was p < 0.05 (SPSS version 18.0; Chicago, IL). Results: Examining GM volume (left hemisphere), the 3-way interaction of evaluation, diagnosis, and surgery was significant (F1,104 ¼6.4, p<.05) indicating that slopes were different depending on diagnostic group; where MCI with surgery showed a significantly greater decrease in GM volume compared to MCI without surgery (F1,47 ¼7.9, p<.01); no significant difference for NL. Examining GM (right hemisphere), the 3-way interaction was not significant, but there was a significant decrease in GM volume (F1,105 ¼4.1, p<.05) for surgery compared to non-surgery
S3
subjects. No significant differences found for WM volume. Conclusions: In our cohort of elderly subjects, surgery resulted in a more pronounced GM volume atrophy than seen in non-surgical subjects, with MCI subjects at higher risk. An examination of neurocognitive effects of surgery, in progress, suggests a correlation between atrophy and function. JULY16, 2011 ALZHEIMER’S IMAGING CONSORTIUM-IC-02: ORAL 2-MARKERS FOR PREDEMENTIA AND DIFFERENTIAL DIAGNOSIS IC-02-01
FUNCTIONAL CONNECTIVITY IN DEMENTIA WITH LEWY BODIES AND ALZHEIMER’S DISEASE: A RESTING BOLD STUDY
Eva Kenny1, Andrew Blamire1, Michael Firbank1, John O’Brien1, 1 Newcastle University, Newcastle upon Tyne, United Kingdom. Background: Using resting-state functional magnetic resonance imaging (fMRI), spontaneous low-frequency fluctuations (SLFs) in the blood oxygenation level dependent (BOLD) signal (<0.1Hz) were measured to investigate connectivity between key brain regions hypothesised to be differentially affected in dementia with Lewy bodies (DLB) compared to Alzheimer’s disease (AD) and age-matched healthy controls. Methods: Forty-seven subjects (>60 years); 15 DLB, 16 AD and 16 controls, were scanned using a 3 Tesla MRI system. Using FSL, the mean BOLD signal time-series was extracted from seed regions; head of caudate nucleus, thalamus, putamen, hippocampus, posterior cingulate, precuneus and primary visual cortex, and cross-correlated with all other brain voxels to determine functional connectivity. Results were cluster thresholded at Z>2.3 and a corrected significance of p¼0.05. Results: In DLB, connectivity was significantly greater than controls with the head of caudate (bilateral), putamen (left), thalamus (bilateral) and posterior cingulate (right). In AD, connectivity was significantly greater than controls with the head of caudate (right), thalamus (left), hippocampus (left) and posterior cingulate (right). Between the dementia groups, DLB subjects showed significantly greater connectivity than AD in subcortical regions (putamen), whereas AD subjects showed significantly greater connectivity than DLB in cortex (posterior cingulate). Hippocampal connectivity was abnormal in AD, but normal in DLB, while precuneus and primary visual cortex connectivity was normal in DLB and AD. No regions showed significantly lower connectivity in DLB or AD versus controls. Conclusions: This is the first study to investigate cortical and subcortical connectivity using resting BOLD in DLB. We found clear differences in connectivity between DLB and AD; abnormal subcortical connectivity in DLB and abnormal cortical connectivity in AD. Consistent with the known relative preservation of memory in DLB compared to AD, hippocampal connectivity was normal in DLB. Importantly, while metabolic imaging shows functional change in primary visual cortex in DLB, with changes hypothesised to account for visual hallucinations, we found connectivity of this region to be normal. This implicates areas beyond visual sensory input level in the visual dysfunction (visuoperceptual disturbance and visual hallucinations) seen in DLB. This study also illustrates the potential of functional connectivity as a tool for better understanding neurobiological changes in dementia. IC-02-02
PREDICTING PROGRESSION TO ALZHEIMER’S DISEASE IN MCI USING COMBINED STRUCTURAL IMAGING AND CSF BIOMARKERS
Linda McEvoy1, David Heister1, Kaj Blennow2, James Brewer1, Anders Dale1, 1University of California, San Diego, La Jolla, California, United States; 2G€oteborg University, M€olndal, Sweden. Background: Cerebrospinal fluid (CSF) biomarkers and atrophy on structural magnetic resonance images (MRIs) each predict decline to Alzheimer’s disease (AD) from mild cognitive impairment (MCI). We investigated whether the combination of these measures would improve outcome prediction in patients with MCI. Methods: The 3-year risk of progressing to AD was examined in 178 MCI participants from the Alzheimer’s Disease Neuroimaging Initiative. MCI participants were stratified into low and high risk groups on the basis of structural atrophy or on CSF levels of total tau and Aß 1-42 proteins. To classify MCI patients into risk groups on the basis of atrophy, a linear discriminant model previously found to best discriminate AD from control data (McEvoy et al. Radiology, 2009) was applied to data from MCI patients. To classify participants into