- Email: [email protected]
OLFACTORY PROCESSING IS HIGHLY COGNITIVELY DEMANDING: SENSITIVE FUNCTIONAL MARKER FOR COGNITIVE DEFICITS AND DEMENTIA IN AD
P1118
Poster Presentations: Tuesday, July 18, 2017
Background: Neuro-inflammation plays a significant role in the pathogenesis of Alzheimer’s disease and progressive supranuclear palsy. Here we test whether the intensity and regional distribution of neuro-inflammation differs between these disorders and controls; and whether neuro-inflammation relates to disease severity. Methods: We used the radiotracer [11C]PK11195 with positron emission tomography and kinetic modeling to compare regional [11C]PK11195 binding in 16 patients with Alzheimer’s disease pathology (including amyloid-positive mild cognitive impairment), 16 with progressive supranuclear palsy, and 13 controls. We correlated [11C]PK11195 binding with clinical variables and C-reactive protein. Results: [11C]PK11195 binding in the medial temporal lobe and occipital-parietal cortex was increased in Alzheimer’s disease patients, relative to both progressive supranuclear palsy patients and controls. Progressive supranuclear palsy patients showed elevated [11C]PK11195 binding in the thalamus, putamen, and pallidum relative to controls. [11C]PK11195 binding in the pre-cuneus correlated negatively with episodic memory in Alzheimer’s disease, while [11C]PK11195 binding in the pallidum, midbrain, and pons correlated positively with disease severity in progressive supranuclear palsy. Conclusions: The magnitude and distribution of neuro-inflammation, indexed by [11C]PK11195, differed between Alzheimer’s disease and progressive supranuclear palsy, and mirrored the established neuropathological distribution for each disease. In both Alzheimer’s disease and progressive supranuclear palsy, disease severity correlated with neuro-inflammation in the regions most closely associated with principal neuropathological markers including tau aggregates. Immunotherapeutic strategies targeting neuro-inflammation may be a useful strategy in slowing the progression of these neurodegenerative disorders.
P3-404
Figure. 1. The odor-visual association paradigm. Odor presentation was synchronized with visual cues. Subjects were trained to keep the respiration rate and volume consistent prior to fMRI scanning.
Figure. 2. Olfactory and DMNs found during the odor stimulation paradigm in Healthy Controls (HC), MCI and AD subjects. The odor network encompasses the Primary olfactory cortex, amygdala and hippocampus. The DMN encompasses the posterior cingulate cortex, inferior parietal and ventral medial prefrontal cortex. Time courses for the odor networks are also shown.
OLFACTORY PROCESSING IS HIGHLY COGNITIVELY DEMANDING: SENSITIVE FUNCTIONAL MARKER FOR COGNITIVE DEFICITS AND DEMENTIA IN AD
Prasanna Karunanayaka, Brittany Martinez, Paul J. Eslinger, Qing X. Yang, The Pennsylvania State University College of Medicine, Hershey, PA, USA. Contact e-mail: [email protected] Background: A recent study provided novel evidence that olfactory processing deactivates the Default Mode Network (DMN), a wellknown resting state network implicated in cognition. In this study, we combined olfactory fMRI with independent component analysis (ICA) to investigate the relationships between the olfactory and DMN networks in AD and to propose olfactory fMRI as an objective marker for the early diagnosis of cognitive deficits and dementia in AD. Methods: Twelve AD subjects (mean age¼ 73.7 yrs), 19 aMCI (mean age¼ 72.8 yrs) and 31 cognitive normal (CN, mean age¼ 69.5 years) took part in the study. Their smell function was tested using the University of Pennsylvania Smell Identification Test (UPSIT). The olfactory fMRI paradigm consisted of the visual cue “smell?” synchronized with interleaved odor+visual and visual-only trial conditions (Fig. 1). The group ICA was performed by combining all three groups. The responsiveness of each network to stimulation events was also calculated over the olfactory paradigm. Results: Fig. 2a shows group ICA map of the olfactory network, encompassing primary olfactory cortex (POC), amygdala, hippocampus, insula, striatum, and putamen. Fig. 2b shows the DMN during the same olfactory fMRI
Figure. 3(a). Behavioral scores for the smell function (UPSIT scores) in HC, MCI and AD.
Figure. 4. The olfactory network responsiveness in respective cohorts for odor+visual and visual- only trial conditions.
Poster Presentations: Tuesday, July 18, 2017
Figure. 5. Estimated hemodynamic response functions (HRFs) for the olfactory and DMNs using IC time-courses shown in Fig.2c .
paradigm. The temporal behavior of the olfactory network is shows a high degree of variability in AD (Fig. 2c). Both ICA maps showed markedly less fMRI activation in MCI and AD groups, which follow a similar descending trend as the UPSIT scores and the measurement of olfactory network responsiveness (Fig. 3 and 4). In HC, the DMN and the olfactory network was opposite in phase during odor+visual and visual-only trial conditions suggesting competition for cognitive resources (Fig. 5.). These results indicates that olfactory network activity and DMN activity during odor stimulation is clearly different between HC and MCI which has reached to a level that is comparable to AD. Conclusions: A striking deterioration in the responsiveness of the olfactory and DMN is demonstrated in MCI and AD. By delineating the neural substrate of AD clinical presentations, olfactory fMRI may provide an objective marker for the early diagnosis of cognitive deficits and dementia in AD subjects.
P1119
levels. Magnetic resonance imaging (MRI) scans were acquired using Alzheimer Disease Neuroimaging Initiative (ADNI) protocol. Brain volumes were estimated using Freesurfer software suite (http://surfer. nmr.mgh.harvard.edu). WMH volumes (total, deep and periventricular) were estimated using W2MHS toolbox (http://pages.cs.wisc.edu/ wvamsi/w2mhs.html). To control for the confounding effect of brain atrophy, WMH values were calculated as a percentage of estimated intracranial volume. Results:Periventricular WMH positively correlated with age (r¼0.368, p¼0.018). Stroop colour-word interference time positively correlated with total WMH (r¼0.420, p¼0.009), deep WMH (r¼0.390, p¼0.016) and periventricular WMH (r¼0.473, p¼0.003) indicating higher the WMH poorer the performance on stroop test. There was no relation between serum vitamin B12 levels, homocysteine levels and WMH. Conclusions: These results further add to the existing body of literature that WMH are associated with cognitive impairment in particular executive function. Even though the preliminary findings do not indicate relation between vitamin B12 and WMH, this needs to be examined in larger population considering the high prevalence of vitamin B12 deficiency in Indian population.
P3-406
AUTOMATIC QUANTIFICATION OF BRAIN MRI TO IMPROVE THE DIAGNOSTIC WORKUP OF DEMENTIA IN MEMORY CLINIC PATIENTS
Rebecca Steketee1, Meike W. Vernooij2, Bas Jasperse1,3, Marcel Koek1, Henri A. Vrooman1, M. Arfan Ikram2, Janne M. Papma1, John C. van Swieten1, Aad van der Lugt1, Wiro J. Niessen1,4, Marion Smits1, 1Erasmus MC-University Medical Center Rotterdam, Rotterdam, Netherlands; 2 Erasmus MC University Medical Center Rotterdam, Rotterdam, Netherlands; 3Netherlands Cancer Institute, Amsterdam, Netherlands; 4 Delft University of Technology, Delft, Netherlands. Contact e-mail: [email protected] Background: Magnetic resonance imaging (MRI) can provide
P3-405
RELATION BETWEEN COGNITIVE FUNCTION AND WHITE MATTER HYPERINTENSITIES: PRELIMINARY FINDINGS FROM TATA LONGITUDINAL STUDY OF AGING
Ranjini Garani Ramesh1, Simran Purokayastha2, Mahendra Javali3, Suvarna Alladi4, Naren Prahalada Rao4, 1Indian Institute of Science, Bengaluru, India; 2Centre for Neuroscience, Indian Institute of Science, Bengaluru, India; 3Department of Neurology, MS Ramaiah Medical College & Memorial Hospital, Bengaluru, India; 4National Institute of Mental Health and Neurosciences, Bengaluru, India. Contact e-mail: jinigarani@ gmail.com Background: Cerebral white matter changes, in the form of white matter hyperintensities (WMH) are etiologically related to small vessel disease. They are considered substrates of cognitive impairment pertaining to executive functions and memory. Serum vitamin B12 deficiency and consequent elevated homocysteine is a risk factor for WMH. Hence in this study we aimed to examine the relation between cognitive functions and WMH and to examine relation between vitamin B12 levels and WMH. We examined WMH using automated technique to control for the rater bias. Methods: We examined 35 healthy volunteers (HV, male¼18) and 6 individuals with mild cognitive impairment (MCI, male¼3) from Tata longitudinal study of aging, Bangalore. The mean age was 62.7168.73 years; mean years of education was 16.364.2. All participants underwent tests for executive functions (Controlled Oral Word Association Test, digit span, Stroop test and Trail Making Test), serum vitamin B12 levels, serum homocysteine
important supportive evidence for (differential) diagnosis of dementia. However, disease-specific brain changes on conventional MRI are often difficult to detect in the early disease stages. Quantification of brain structures may be more sensitive for detecting early changes and can therefore improve diagnostic accuracy. We have developed software (Figure 1) that automatically provides the radiologist with quantitative information on brain volumes, compared to age- and sex-specific reference data from a dementia-free aging population (n¼4,915, aged>45 years). We assessed whether such automated quantification of structural brain changes improves diagnostic accuracy in a memory clinic setting. Methods: New patients from our memory clinic (December 2009-September 2011) who underwent MRI at 3T as part of clinical workup were assessed by two experienced neuro-radiologists based on 1) only visual interpretation of conventional MRI; 2) only quantitative information on total and regional brain volume; 3) combined visual and quantitative information. We determined diagnostic accuracy for AD and FTD diagnosis as [true positives+true negatives/all cases] for each strategy and rater, and differences in accuracies using McNemar tests. Results: In our sample of 42 memory clinic patients (Table 1) with Alzheimer’s disease (AD, N¼21), frontotemporal dementia (FTD, N¼15) and mild cognitive impairment (MCI, N¼6), quantitative information alone did not improve diagnostic accuracy for AD or FTD compared to visual assessment only (Table 2). For FTD, accuracy was significantly worse with quantitative assessment. Visual assessment alone performed slightly but not significantly better than combined assessment; with high diagnostic
Poster Presentations: Tuesday, July 18, 2017
Background: Neuro-inflammation plays a significant role in the pathogenesis of Alzheimer’s disease and progressive supranuclear palsy. Here we test whether the intensity and regional distribution of neuro-inflammation differs between these disorders and controls; and whether neuro-inflammation relates to disease severity. Methods: We used the radiotracer [11C]PK11195 with positron emission tomography and kinetic modeling to compare regional [11C]PK11195 binding in 16 patients with Alzheimer’s disease pathology (including amyloid-positive mild cognitive impairment), 16 with progressive supranuclear palsy, and 13 controls. We correlated [11C]PK11195 binding with clinical variables and C-reactive protein. Results: [11C]PK11195 binding in the medial temporal lobe and occipital-parietal cortex was increased in Alzheimer’s disease patients, relative to both progressive supranuclear palsy patients and controls. Progressive supranuclear palsy patients showed elevated [11C]PK11195 binding in the thalamus, putamen, and pallidum relative to controls. [11C]PK11195 binding in the pre-cuneus correlated negatively with episodic memory in Alzheimer’s disease, while [11C]PK11195 binding in the pallidum, midbrain, and pons correlated positively with disease severity in progressive supranuclear palsy. Conclusions: The magnitude and distribution of neuro-inflammation, indexed by [11C]PK11195, differed between Alzheimer’s disease and progressive supranuclear palsy, and mirrored the established neuropathological distribution for each disease. In both Alzheimer’s disease and progressive supranuclear palsy, disease severity correlated with neuro-inflammation in the regions most closely associated with principal neuropathological markers including tau aggregates. Immunotherapeutic strategies targeting neuro-inflammation may be a useful strategy in slowing the progression of these neurodegenerative disorders.
P3-404
Figure. 1. The odor-visual association paradigm. Odor presentation was synchronized with visual cues. Subjects were trained to keep the respiration rate and volume consistent prior to fMRI scanning.
Figure. 2. Olfactory and DMNs found during the odor stimulation paradigm in Healthy Controls (HC), MCI and AD subjects. The odor network encompasses the Primary olfactory cortex, amygdala and hippocampus. The DMN encompasses the posterior cingulate cortex, inferior parietal and ventral medial prefrontal cortex. Time courses for the odor networks are also shown.
OLFACTORY PROCESSING IS HIGHLY COGNITIVELY DEMANDING: SENSITIVE FUNCTIONAL MARKER FOR COGNITIVE DEFICITS AND DEMENTIA IN AD
Prasanna Karunanayaka, Brittany Martinez, Paul J. Eslinger, Qing X. Yang, The Pennsylvania State University College of Medicine, Hershey, PA, USA. Contact e-mail: [email protected] Background: A recent study provided novel evidence that olfactory processing deactivates the Default Mode Network (DMN), a wellknown resting state network implicated in cognition. In this study, we combined olfactory fMRI with independent component analysis (ICA) to investigate the relationships between the olfactory and DMN networks in AD and to propose olfactory fMRI as an objective marker for the early diagnosis of cognitive deficits and dementia in AD. Methods: Twelve AD subjects (mean age¼ 73.7 yrs), 19 aMCI (mean age¼ 72.8 yrs) and 31 cognitive normal (CN, mean age¼ 69.5 years) took part in the study. Their smell function was tested using the University of Pennsylvania Smell Identification Test (UPSIT). The olfactory fMRI paradigm consisted of the visual cue “smell?” synchronized with interleaved odor+visual and visual-only trial conditions (Fig. 1). The group ICA was performed by combining all three groups. The responsiveness of each network to stimulation events was also calculated over the olfactory paradigm. Results: Fig. 2a shows group ICA map of the olfactory network, encompassing primary olfactory cortex (POC), amygdala, hippocampus, insula, striatum, and putamen. Fig. 2b shows the DMN during the same olfactory fMRI
Figure. 3(a). Behavioral scores for the smell function (UPSIT scores) in HC, MCI and AD.
Figure. 4. The olfactory network responsiveness in respective cohorts for odor+visual and visual- only trial conditions.
Poster Presentations: Tuesday, July 18, 2017
Figure. 5. Estimated hemodynamic response functions (HRFs) for the olfactory and DMNs using IC time-courses shown in Fig.2c .
paradigm. The temporal behavior of the olfactory network is shows a high degree of variability in AD (Fig. 2c). Both ICA maps showed markedly less fMRI activation in MCI and AD groups, which follow a similar descending trend as the UPSIT scores and the measurement of olfactory network responsiveness (Fig. 3 and 4). In HC, the DMN and the olfactory network was opposite in phase during odor+visual and visual-only trial conditions suggesting competition for cognitive resources (Fig. 5.). These results indicates that olfactory network activity and DMN activity during odor stimulation is clearly different between HC and MCI which has reached to a level that is comparable to AD. Conclusions: A striking deterioration in the responsiveness of the olfactory and DMN is demonstrated in MCI and AD. By delineating the neural substrate of AD clinical presentations, olfactory fMRI may provide an objective marker for the early diagnosis of cognitive deficits and dementia in AD subjects.
P1119
levels. Magnetic resonance imaging (MRI) scans were acquired using Alzheimer Disease Neuroimaging Initiative (ADNI) protocol. Brain volumes were estimated using Freesurfer software suite (http://surfer. nmr.mgh.harvard.edu). WMH volumes (total, deep and periventricular) were estimated using W2MHS toolbox (http://pages.cs.wisc.edu/ wvamsi/w2mhs.html). To control for the confounding effect of brain atrophy, WMH values were calculated as a percentage of estimated intracranial volume. Results:Periventricular WMH positively correlated with age (r¼0.368, p¼0.018). Stroop colour-word interference time positively correlated with total WMH (r¼0.420, p¼0.009), deep WMH (r¼0.390, p¼0.016) and periventricular WMH (r¼0.473, p¼0.003) indicating higher the WMH poorer the performance on stroop test. There was no relation between serum vitamin B12 levels, homocysteine levels and WMH. Conclusions: These results further add to the existing body of literature that WMH are associated with cognitive impairment in particular executive function. Even though the preliminary findings do not indicate relation between vitamin B12 and WMH, this needs to be examined in larger population considering the high prevalence of vitamin B12 deficiency in Indian population.
P3-406
AUTOMATIC QUANTIFICATION OF BRAIN MRI TO IMPROVE THE DIAGNOSTIC WORKUP OF DEMENTIA IN MEMORY CLINIC PATIENTS
Rebecca Steketee1, Meike W. Vernooij2, Bas Jasperse1,3, Marcel Koek1, Henri A. Vrooman1, M. Arfan Ikram2, Janne M. Papma1, John C. van Swieten1, Aad van der Lugt1, Wiro J. Niessen1,4, Marion Smits1, 1Erasmus MC-University Medical Center Rotterdam, Rotterdam, Netherlands; 2 Erasmus MC University Medical Center Rotterdam, Rotterdam, Netherlands; 3Netherlands Cancer Institute, Amsterdam, Netherlands; 4 Delft University of Technology, Delft, Netherlands. Contact e-mail: [email protected] Background: Magnetic resonance imaging (MRI) can provide
P3-405
RELATION BETWEEN COGNITIVE FUNCTION AND WHITE MATTER HYPERINTENSITIES: PRELIMINARY FINDINGS FROM TATA LONGITUDINAL STUDY OF AGING
Ranjini Garani Ramesh1, Simran Purokayastha2, Mahendra Javali3, Suvarna Alladi4, Naren Prahalada Rao4, 1Indian Institute of Science, Bengaluru, India; 2Centre for Neuroscience, Indian Institute of Science, Bengaluru, India; 3Department of Neurology, MS Ramaiah Medical College & Memorial Hospital, Bengaluru, India; 4National Institute of Mental Health and Neurosciences, Bengaluru, India. Contact e-mail: jinigarani@ gmail.com Background: Cerebral white matter changes, in the form of white matter hyperintensities (WMH) are etiologically related to small vessel disease. They are considered substrates of cognitive impairment pertaining to executive functions and memory. Serum vitamin B12 deficiency and consequent elevated homocysteine is a risk factor for WMH. Hence in this study we aimed to examine the relation between cognitive functions and WMH and to examine relation between vitamin B12 levels and WMH. We examined WMH using automated technique to control for the rater bias. Methods: We examined 35 healthy volunteers (HV, male¼18) and 6 individuals with mild cognitive impairment (MCI, male¼3) from Tata longitudinal study of aging, Bangalore. The mean age was 62.7168.73 years; mean years of education was 16.364.2. All participants underwent tests for executive functions (Controlled Oral Word Association Test, digit span, Stroop test and Trail Making Test), serum vitamin B12 levels, serum homocysteine
important supportive evidence for (differential) diagnosis of dementia. However, disease-specific brain changes on conventional MRI are often difficult to detect in the early disease stages. Quantification of brain structures may be more sensitive for detecting early changes and can therefore improve diagnostic accuracy. We have developed software (Figure 1) that automatically provides the radiologist with quantitative information on brain volumes, compared to age- and sex-specific reference data from a dementia-free aging population (n¼4,915, aged>45 years). We assessed whether such automated quantification of structural brain changes improves diagnostic accuracy in a memory clinic setting. Methods: New patients from our memory clinic (December 2009-September 2011) who underwent MRI at 3T as part of clinical workup were assessed by two experienced neuro-radiologists based on 1) only visual interpretation of conventional MRI; 2) only quantitative information on total and regional brain volume; 3) combined visual and quantitative information. We determined diagnostic accuracy for AD and FTD diagnosis as [true positives+true negatives/all cases] for each strategy and rater, and differences in accuracies using McNemar tests. Results: In our sample of 42 memory clinic patients (Table 1) with Alzheimer’s disease (AD, N¼21), frontotemporal dementia (FTD, N¼15) and mild cognitive impairment (MCI, N¼6), quantitative information alone did not improve diagnostic accuracy for AD or FTD compared to visual assessment only (Table 2). For FTD, accuracy was significantly worse with quantitative assessment. Visual assessment alone performed slightly but not significantly better than combined assessment; with high diagnostic