|Home | About | Journals | Submit | Contact Us | Français|
The association between antemortem [11C]-Pittsburgh Compound B (PiB) retention and β-amyloid (Aβ) load, Lewy body (LB) and neurofibrillary tangle (NFT) densities were investigated in a pathologically confirmed case of dementia with LB (DLB). 76-year-old man presenting with a clinical diagnosis of DLB had undergone PiB–positron emission tomography (PET), 18F FDG-PET and MRI 18 months before death. The pathologic diagnosis was DLB neocortical-type with low-likelihood of Alzheimer's disease by NIA-Reagan criteria. Sections from regions of interest (ROI) on post-mortem examination were studied. A significant correlation was found between cortical Aβ density and PiB retention in the 17 corresponding ROIs (r=0.899; p<0.0001). Bielschowsky silver stain revealed mostly sparse neocortical neuritic plaques; whereas diffuse plaques were frequent. There was no correlation between LB density and PiB retention (r=0.13; p=0.66); nor between NFT density and PiB retention (r=−0.36; p=0.17). The ROI-based analysis of imaging and histopathological data confirms that PiB uptake on PET is a specific marker for Aβ density, but cannot differentiate neuritic from diffuse amyloid plaques in this case with DLB.
Lewy body (LB) disease is the second most common cause of neurodegenerative dementia after Alzheimer's disease (AD), and LB pathology is present in about 20-35% of cases with dementia [Galasko, et al.,1994, Hansen, et al.,1990, Joachim, et al.,1988, Schneider, et al.,2007, Zaccai, et al.,2005]. A majority of patients with dementia with Lewy bodies (DLB) are pathologically characterized by the presence of β-amyloid (Aβ) plaques along with α-synuclein LB deposits [McKeith, et al.,2005, Schneider, et al.,2007]. Antemortem diagnosis of Aβ deposition in DLB may play an important role in treatment decisions, as well as assessing responsiveness to treatments targeting disease specific pathologies [Thal, et al.,2006].
In vivo retention of the positron emission tomography (PET) tracer N-Methyl-[11C]2-(4′-methylaminophenyl)-6-hydroxybenzothiazole ([11C]PIB 6-OH-BTA-1), or the [11C]-Pittsburgh Compound B (PiB) reflects the deposition of fibrillary Aβ in patients with AD [Bacskai, et al.,2007, Ikonomovic, et al.,2008]. Increased PiB binding is typically present in DLB in up to 80% of cases [Edison, et al.,2008, Maetzler, et al.,2009], however the specificity of in vivo PiB retention in DLB is less clear [Burack, et al.,2010, Fodero-Tavoletti, et al.,2007, Maetzler, et al.,2008]. Fluorescence microscopy demonstrated in vitro binding of PiB to the fibrillar, misfolded proteins in the substantia nigra LBs [Maetzler, et al.,2008]. In contrast, PiB binding was not observed in the Aβ plaque-free brain homogenates in DLB, and LBs did not significantly contribute to the PiB fluorescence stain in tissue sections [Fodero-Tavoletti, et al.,2007]. Furthermore, the regional pattern of antemortem PiB retention was in agreement with the distribution of Aβ pathology in three Parkinson disease with dementia patients at autopsy [Burack, et al.,2010]. To the best of our knowledge no study has yet directly correlated in vivo PiB retention in DLB with postmortem Aβ load and α-synuclein LB density using quantitative methods and region of interest (ROI) analysis.
Our objective was to determine the association between antemortem PiB retention and Aβ, LB, and neurofibrillary tangle (NFT) loads in a pathologically confirmed case of DLB.
This case participated in the Mayo Clinic Alzheimer's Disease Research Center and the Rush University Alzheimer's Disease Center programs. Participants in both centers undergo approximately annual clinical examinations, routine laboratory tests, neuropsychological tests, and ancillary neuroimaging studies.
All imaging studies were performed at the Mayo Clinic. Magnetic resonance imaging (MRI) examinations were performed at 3 Tesla including a 3D-MPRAGE acquisition for anatomical segmentation and labeling. PET images were acquired using a PET/CT scanner (DRX; GE Healthcare) operating in 3D mode. The patient was injected with 432MBq [11C]PIB and 511MBq [18F]fluorodeoxyglucose (FDG) on the same day with one hour between the PiB and the 18F-FDG PET scan acquisitions Subjects were prepared for 18F-FDG in a dimly lit room, with minimal auditory stimulation. A CT image was obtained for attenuation correction. After a 40-min (PiB) or 30-min (18F-FDG) uptake period, the subjects were imaged. A 20-min PiB or an 8-min 18F-FDG scan was obtained. The PiB PET acquisition consisted of four 5-min dynamic frames, acquired from 40 to 60 min after injection, and the 18F-FDG image acquisition consisted of four 2-min dynamic frames, acquired from 30 to 38 min after injection. Standard corrections were applied. The pixel size for PET images were 1.0 mm and the slice thickness was 3.3 mm. [Lowe, et al.,2009].
Pathological examination was performed at the Rush Alzheimer Disease Center. The brain was removed 4 hours after death, and 1 cm slabs from the right hemisphere were fixed for 3 days in 4% paraformaldehyde. Tissue was dissected from the hippocampus, entorhinal cortex, anterior cingulate gyrus, midfrontal, middle temporal, inferior parietal cortex, basal ganglia, thalamus and midbrain; embedded in paraffin; and cut into 6 micron sections and stained with hematoxylin and eosin (H&E) and modified Bielschowsky silver stain [Mufson, et al.,1988], and antibodies to Aβ (1:600, 10D5, courtesy of Elan pharmaceuticals), PHF-tau (1:3000; AT8; Pierce Protein Research) and α-synuclein (LB509; 1:100; Invitrogen Corp). A board-certified neuropathologist with expertise in dementia reviewed all slides using standard diagnostic criteria [Braak and Braak,1991, McKeith, et al.,2005, McKhann, et al.,1984, Mirra, et al.,1991].
For the imaging–pathology co-localization, we chose 8 additional ROIs (total of 17) based on the known distribution of Aβ, LB and neurofibrillary tangle (NFT) pathologies in AD and DLB. We included regions that are known to have abundant Aβ (e.g. inferior precuneus), LB (e.g. amygdala) and NFT (e.g. hippocampus), and regions that are known to have lesser Aβ (e.g. hippocampus), LB (e.g. precentral gyrus) and NFT (e.g. calcarine cortex) in order to sample a wide spectrum of pathology.
All sections were mounted on electrostatically charged plus slides (Daigger and Co., Inc., Lincolnshire, IL), deparaffinized, dehydrated in alcohols and xylene and coverslipped with Permount. Immunostains were performed using the avidin-biotin-peroxidase method (ABC-Elite, Vector, Burlingame, CA) on an Automated Leica Bond immunostain (Leica Microsystems Inc., Bannockborn IL) with a pre-defined protocol [Bennett, et al.,2004, Schneider, et al.,2002]. Briefly, all sections are washed and blocked with normal serum; incubated with the primary antibody; washed and incubated with the biotinylated secondary antibody; and then the avidin-biotin-peroxidase complex (Vector laboratories). The chromagen for color development was 3,3′diaminobenzidine. Positive controls included neocortical tissue from known cases of AD (for amyloid and PHF immunostain) and Lewy body disease (α-synuclein immunostains).
The fractional area occupied by Aβ was determined using image analysis. An investigator blinded to imaging data outlined each ROI using Stereo Investigator 8.0 on a Microbrightfield Stereology System, instructed a grid to be randomly placed, and a motorized stage to stop at about 20% of the grid intersection points to obtain images for area analysis by the triangulation method using the NIH Image processing application (Image J 1.42g (http://rsbweb.nih.gov/ij/). All values for each region were averaged to obtain the mean fraction of Aβ. For modified Bielschowsky stained sections we counted neuritic plaques, diffuse plaques, and NFT in the area with the greatest density of each marker, using a graticule in one square mm area (× 100 magnification) in each region. Density (number/mm2) of LB and PHF-tau NFT was determined through manual counts with the ROI area outlined by Stereo Investigator 8.0 as noted above, and 100% of the area counted.
PIB-PET quantitative image analysis was performed using the fully automated image processing pipeline which is described in detail [Jack, et al.,2008, Senjem, et al.,2005]. Briefly, the method includes gray matter (GM) sharpening of PET images using MRI and partial volume correction of CSF and tissue compartments using SPM5[Ashburner and Friston, 2005]. PIB-PET cortical ratio images were calculated by dividing the each PiB-PET GM voxel value by the median value in the cerebellar GM region in the patient's MRI space. The global cortical PiB retention was calculated by taking the median value of the PIB-PET GM ratio from the bilateral parietal, posterior cingulate, precuneus, temporal, prefrontal, orbitofrontal, anterior cingulate GM regions in the anatomical labeling atlas [Jack, et al.,2008].
ROI-based pathology correlation was performed by reorienting the 3D MRI images to anatomically match to the photographed images of the 17 histological sections. ROIs that simulated the histological blocks were drawn on the matching MRI slice using the image analysis software Analyze 10.0 (Mayo Clinic). The median GM PiB retention ratio was calculated for each of the 17 ROIs from the PIB-PET GM ratio images and corrected for partial volume averaging of CSF in the MRI space (Figure 1). The entorhinal cortex and substantia nigra regions evaluated at autopsy were not analyzed because the sizes of the histological blocs were less than the resolution of PET in at least one dimension.
The patient began exhibiting recurrent dream enactment behavior around age 50. The frequency and severity of the behaviors gradually increased over many years, but injuries rarely occurred. His cognitive decline became apparent at the age 66 and over the following year he developed obvious signs of parkinsonism, leading to a diagnosis of DLB. By age 72 recurrent visual hallucinations, fluctuations in cognition, visual illusions, and auditory hallucinations became apparent. A comprehensive evaluation at age 76 was notable for a Kokmen Short Test of Mental Status [Kokmen, et al.,1987] score of 28/38, hypokinetic dysarthria, and moderate symmetric parkinsonism. Polysomnography demonstrated increased electromyographic tone and flailing limb movements during REM sleep, thereby confirming the presence of REM sleep behavior disorder (RBD). He showed clear deficits in attention and processing speed (unable to complete Trail Making Test A or B, semantic fluency at the 2nd percentile), and visual construction (Rey-Osterreith Complex figure was severely distorted, <1st percentile), with far less impairment or normal performance on confrontation naming and delayed recall on verbal memory measures.
MRI showed mild atrophy of the hippocampi and mild periventricular white matter hyperintensities. Abnormal glucose metabolism was present in the temporoparietal cortices and prefrontal cortex on 18F-FDG PET consistent with the typical pattern in AD. PiB retention was most significant in the frontal lobes where as the occipital lobes were relatively spared. The abundant PiB retention in the white matter was comparable to the non-specific PiB retention typically seen in the white matter, which was masked out of the cortical GM ROI analysis and global PiB retention ratio through partial volume correction of CSF and tissue compartments using the co-registered 3D- MRI [Jack, et al.,2008] (Figure 2). The global PiB retention ratio was 1.68, which would be classified as “PiB positive” according to the typically used cut-off of 1.50 [Jack, et al.,2008].
The patient underwent numerous pharmacologic manipulations with cholinesterase inhibitors, levodopa, dopamine agonists, memantine, psychostimulants, and atypical neuroleptics, but these never resulted in any obvious clinical benefits. He died at age 77.
At autopsy, the brain weighed 1020g, had moderate to severe cortical atrophy, mild to moderate hippocampal atrophy, moderate lateral ventricular enlargement, and pallor of the substantia nigra. H&E showed severe nigral degeneration but no other abnormalities. Bielschowsky silver stained showed NFT were moderate in the hippocampus and entorhinal cortices, and absent in neocortical regions; and neuritic plaques were sparse in the midfrontal, middle temporal, and inferior parietal cortices corresponding to a Braak stage III neurofibrillary pathology and low likelihood of AD by NIA-Reagan criteria (Table). LB were found in the substantia nigra, entorhinal, anterior cingulate, midfrontal, middle temporal and inferior parietal cortices, consistent with a pathological diagnosis of DLB – neocortical type.
Aβ was more sensitive to plaques, compared to Bielshowsky silver stain (Table, Figure 3). Overall, the density of Aβ was greatest in the frontal lobes and lowest in the hippocampus (Table). Amyloid angiopathy was mild in the middle hippocampus, inferior parietal and lateral temporal cortex and absent in other regions. NFT density was greatest in the posterior hippocampus followed by middle hippocampus, entorhinal cortex and amygdala. LB were also most numerous in limbic regions but unlike NFT, were also widespread in the neocortex (Figure 4).
Among the analyzed ROIs we identified the greatest PiB retention (PiB retention ratio >1.90), and Aβ density in the frontal lobe and medial parietal ROIs (Table). There was a strong correlation between PiB retention and Aβ density in the 17 ROIs that were analyzed on pathological examination using Spearman rank order correlation (r=0.899; p<0.0001) (Figure 5). There was no correlation between LB density and PiB retention (r=0.13; p=0.66) (Figure 6); nor between tau density and PiB retention (r=−0.36; p=0.17).
The case had many of the diagnostic features of DLB; all core clinical features and presence of RBD preceding his cognitive/motor/neuropsychiatric features by decades [Boeve,2009, McKeith, et al.,2005] and a typical DLB neuropsychological pattern of impairment [Ferman, et al.,2006], with the only notable or atypical clinical finding being lack of response to any agents which often provide at least modest benefit to those with DLB. At autopsy, the patient fulfilled the criteria for neocortical LB disease with low likelihood of AD, in agreement with most of the antemortem clinical data.
Contrary to the clinical and pathological diagnosis, the pattern of FDG-PET hypometabolism suggested coexisting AD and the cortical PiB retention ratio on PET was positive (1.68) according to the generally accepted cut-off of 1.50. Diffuse plaques, on average were 2 to 3 times more abundant than the neuritic plaques on modified Bielschowsky silver stain in regions that are standard for the diagnostic evaluation for AD [Mirra, et al.,1991], and significantly contributed to the PiB retention. For example, midfrontal cortex had one of the highest PiB retention ratio (1.92), with sparse neuritic plaques. Since diffuse plaques were >3 times more common compared to the neuritic plaques in this region, we conclude that diffuse plaques significantly contributed to the PiB retention. This trend was also present in the anterior cingulate gyrus and precuneus ROIs that had some of the highest PiB retention ratios (>1.90). A similar observation was reported in two cases with Parkinson disease with dementia who were PiB positive with a high burden of diffuse plaques, but had a low likelihood of AD according to NIA-Reagan criteria at autopsy. Because of the neocortical LB pathology, dementia was attributed to the LB disease rather than diffuse plaques [Burack, et al.,2010].
The specificity of PiB binding to Aβ has been confirmed with in vitro studies [Klunk, et al.,2003]. In the human brain, PiB labels both neuritic and diffuse plaques, although labeling of diffuse/amphorous plaques is less prominent than compact/cored plaques. PiB does not label the diffuse plaques in the molecular layer of the cerebellum, which is one of the reasons for using the cerebellar PiB retention as an internal reference region for quantitative analysis [Ikonomovic, et al.,2008, Lopresti, et al.,2005]. The pathological criteria however emphasizes neuritic plaques over the diffuse plaques for the diagnosis of AD [McKhann, et al.,1984]. Diffuse plaques are more common than neuritic plaques in DLB [Dickson,2002]. Diffuse plaques can be numerous even in cognitively normal individuals [Davis, et al.,1999, Dickson, et al.,1992, Knopman, et al.,2003]. In a patient with mild AD (clinical dementia rating = 0.5) who had low cortical PiB binding potential, diffuse plaques were the most prominent pathology observed at autopsy. It was noted that substantial densities of diffuse plaques are not benign and can be associated with early AD symptoms [Cairns, et al.,2009]. However this case also had Braak III neurofibrillary pathology, which may have contributed to the mild memory impairment.
The strong correlation between in vivo PiB retention and Aβ density identified in the current study is in agreement with a previous report on an AD case using a similar imaging and histology ROI matching procedure [Ikonomovic, et al.,2008]. The pathological findings however are significantly different. The case in the earlier report had advanced AD but did not have LB pathology. Whereas the predominant pathology in the current case was LB disease with low likelihood of AD pathology. Agreement between both data sets confirms the specificity of in vivo PiB retention, and indicates that the strong relationship between in vivo PiB retention and Aβ density is independent of α-synuclein LB load along the AD-DLB spectrum.
Diffuse plaques contribute significantly to PiB retention in LB diseases [Burack, et al.,2010, Dickson,2002], but the distinction between neuritic and diffuse plaques was not possible with PiB-PET in the case we studied. Increased PiB retention in a case with a low likelihood of AD emphasizes the fact that although PIB is considered as a golden standard of amyloid imaging today, the exact mechanism behind the PIB-amyloid interaction and the pathological diagnosis of AD is still not clear. Furthermore, whether diffuse plaques that contribute to PiB retention represent a preliminary stage in the pathogenesis of the neurotoxic degeneration requires further investigation [Dickson,1997, Guillozet, et al.,1997, Price, et al.,2009]. Determining the influence of Aβ on the clinical phenotype [Rowe, et al.,2007], and the relationship between diffuse plaques and cognitive impairment in DLB will be important for making treatment decisions with the availability of anti-amyloid agents.
This study was supported by the Paul B. Beeson Career Development Award in Aging K23-AG030935, P50-AG16574/Project1, P30-AG010161, RO1-AG15866, R01-AG11378, Alexander family, and the Robert H. and Clarice Smith and Abigail Van Buren Alzheimer's Disease Research Program.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
2. 1. 1 Disclosure Statements
The authors do not have any actual or potential conflicts of interests to disclose. This study was approved by the Mayo Clinic Institutional Review Board, and informed consent for participation was obtained from every subject and/or an appropriate surrogate.