Send Orders for Reprints to
[email protected] Current Alzheimer Research, 2013, 10, 000-000
1
Relationships Between White Matter Hyperintensities, Cerebral Amyloid Angiopathy and Dementia in a Population-based Sample of the Oldest Old M. Tanskanena, RN. Kalariab,*, I-L. Notkolac, M. Mäkeläa, T. Polvikoskib, L. Myllykangasa,d, R. Sulkavae, H. Kalimoa, A. Paetaua, P. Scheltensf, F. Barkhoff, ECW van Straatenf and T. Erkinjunttig a
Department of Pathology, Haartman Institute, University of Helsinki and HUSLAB, Helsinki, Finland; bInstitute for Aging and Health, Newcastle University, Campus for Ageing and Health, Newcastle upon Tyne NE4 5PL, United Kingdom; cFinnish Information Centre for Register Research, Helsinki, Finland; dFolkhälsan Institute of Genetics, Biomedicum Helsinki, Helsinki, Finland; eInstitute of Public Health and Clinical Nutrition, University of Eastern Finland, Kuopio, and Department of Neurology, Kuopio University Hospital, Kuopio, Finland; fDepartments of Neurology, Clinical Neurophysiology and Radiology, VU University Medical Centre, Amsterdam, the Netherlands; gDepartment of Neurology, Helsinki University Central Hospital and Department of Neurological Sciences, University of Helsinki, Finland Abstract: Previous reports suggest that brain white matter changes, a surrogate for small vessel disease, are related to cerebral amyloid angiopathy (CAA). However, this relationship has not been explored in population-based studies or in the oldest old (>85 years of age). We studied the relationships between white matter hyperintensities (WMH) determined by post-mortem magnetic resonance imaging (MRI) and neuropathologically assessed CAA in demented and nondemented subjects enrolled in the prospective community-based Finnish Vantaa 85+ Study. In this analysis, we evaluated scans and brain samples from 123 subjects (86% women) with a mean age of 90.6 years. We found CAA to be present in 63 % of the 123 subjects, whereas WMH was present in 74%, and dementia in 59 %. The presence of WMH of any severity did not relate to the presence or the degree of CAA severity, irrespective of the dementia status of the subjects. Furthermore, multivariate regression analysis showed a clear association between CAA and dementia but WMH was not related to dementia in this very old sample. We conclude that severe WMH may not be determined by CAA in this very elderly population.
Keywords: Cerebral amyloid angiopathy, magnetic resonance imaging, neuropathology, white matter hyperintensity. INTRODUCTION Cerebral white matter hyperintensities (WMH) on magnetic resonance imaging (MRI) are frequent in elderly subjects, both in the cognitively normal and in those who decline [1]. WMH is considered a surrogate marker of small vessel disease (SVD) [2], and extensive WMH involving the deep white mater (WM) is thought to be an indicator of subcortical ischaemic vascular disease (SIVD). SIVD is characterized by lacunar infarcts and arteriolosclerosis of particularly long penetrating arteries perfusing the deep WM. Arterial hypertension and diabetes are considered strong risk factors for SIVD, which is the most common type of vascular cognitive impairment [3, 4]. Cerebral amyloid angiopathy (CAA) is characterised by the deposition of amyloid beta (A) protein in small and middle-sized cerebral and cerebellar cortical and leptomeningeal vessels [5]. CAA is a consistent feature of Alzheimer’s disease (AD). CAA is routinely detected with *Address correspondence to this author at the Institute for Ageing and Health, Newcastle University, Campus for Ageing & Vitality, Newcastle upon Tyne, NE4 5PL, United Kingdom; Tel: 0191 248 1352; Fax: 0191 248 1301; E-mail:
[email protected] 1567-2050/13 $58.00+.00
standard tinctorial stains and present in more than 80% of patients with AD [6, 7]. This frequency is even higher if based on the deposition of A protein [8, 9]. Like WMH, CAA increases with age [10] and its prevalence in the oldest old population (>85 years of age) varies from nearly 50%, up to approximately 70% [11]. There appear several shared pathological processes related to WMH and CAA that infer potential associations between them. Both features represent abnormalities or disease processes of cerebral arteries; WMH in the long perforating arterioles and CAA in the more superficial segments of vessels of varying size [12]. Infarcts may also be involved in both conditions, old lacunar infarcts with WMH and cortical microinfarcts (MI) frequently occur in the presence of CAA [13, 14]. Demyelination preceding or resulting from axonal degeneration is associated with WMH [15, 16] and is also a feature in animal models simulating AD pathology [17]. Both WMH and CAA are related to dementia, but with different aetiologies e.g. SIVD against AD-type. However, WMH and CAA also bear distinct features of their own. Importantly, cortical AD-type of pathology is concomitant with CAA but may not necessarily be a feature of WMH. Dilatation of perivascular spaces largely characterizes WMH [18] whereas cortical microbleeds are frequently found in the © 2013 Bentham Science Publishers
2 Current Alzheimer Research, 2013, Vol. 10, No. 10
Tanskanen et al.
presence of CAA [19]. Several previous studies have assessed the relationships between WMH and CAA, and most have found an association between these variables [16, 20]. While all previous studies were performed in hospital-based or other highly selected samples this question has not been explored in community or population-based cohorts. Here, we investigated the relationship between WMH and CAA and their associations with dementia in a community-based sample, as part of the larger prospective and populationbased Vantaa 85+ Study [21].
try of Social Affairs and Health of Finland approved the use of the health and social work records and death certificates. Blood samples were collected after the subjects or their relatives gave informed consent. Consent for the autopsy was obtained from the nearest relative. The National Supervisory Authority for Welfare and Health (VALVIRA) has approved the collection of the tissue samples at autopsy as well as their use for research.
MATERIALS AND METHODS
At autopsy, the brains were weighed and the cerebrum was cut in 1 cm thick coronal and the cerebellum in 5 mm thick sagittal slices. The presence, location and number of macroscopic brain infarctions of any size in the cerebral cortex, white matter, basal ganglia region, cerebellum and brainstem were recorded. Post-mortem MRI was performed on a 1.5-T magnet (Siemens Vision) as previously described [23], with whole brain formalin fixed specimens (pons, cerebellum and medulla oblongata removed). Also coronal and axial spin echo T2 and proton density (TR 3000 ms, TE 15/90 ms, slice thickness 5 mm), and coronal 3D turbo spin echo (TR 3000 ms, TE 120 ms, slice thickness 1 mm) were acquired. The brains were marked with plastic pieces to synchronize the planes used in MRI scanning and pathological sections. We excluded samples with a space-occupying lesion or other impeding lesions such as radiation leukoencephalopathy, (n = 10), infarct in the hippocampal area (n = 3) and incomplete scan series (n = 1).
Subjects and Clinical Evaluation The Vantaa 85+ study comprised original assessment of 601 community dwelling individuals, all at least 85 years of age, who were living in the town of Vantaa (Southern Finland) on April 1, 1991 [22]. During a 10 year follow-up from 1991 to 2001, we accrued 304 autopsies, which all included neuropathological examination. The clinical diagnosis of dementia was based on the Diagnostic and Statistical Manual of Mental Disorders, 3th edition, revised (DSM III-R) criteria and required the consensus of two neurologists. The diagnosis of hypertension was based on the use of antihypertensive medication, and the diagnosis of diabetes on the use of blood glucose lowering medication. Clinical follow-up assessments were carried out in 1994, 1996, 1999 and 2001. The mean interval between the last clinical examination and death was 8 months. Of the 304 brains which were examined at autopsy, 132 were scanned by MR between 3 months and 4 years after fixation in phosphate-buffered 10% formalin solution. Of these, complete clinical data were available for 123 subjects (106 women and 17 men, mean age at death 90.6 years; range 85.1 - 104.8 years) who comprised the basis of this study. (Table 1) provides the demographic details and vascular disease of this sample in comparison to the whole study sample. Ethical Permission The Vantaa 85+ study was approved by the Ethics Committee of the Health Centre of the City of Vantaa. The MinisTable 1.
Gross Examination and Post-mortem MRI
WMH was analysed using the Age-related White Matter Changes (ARWMC) rating scale [24]. These lesions were assessed within the right frontal and parieto-occipital subcortical areas (T2-weighted MR images), corresponding to the brain regions used for the neuropathological evaluation of CAA (see below). WMH was rated in each of two areas using the scale of 0 to 3 (none visible, mild, moderate and severe), essentially as detailed previously [21]. The scores from the two brain regions were added up and used as numeric variables for the statistical analysis. They were designated in two groups: “severe WMH” equalling values 4-6 versus “mild” as values 0-3 [21, 24]. In addition, “extensive WMH” was designated to equal the combined score of 6, corresponding to the 90 percentile value for severity of WMH in each of the two brain regions.
Demographics and Clinical Characteristics of the Study Sample in Comparison with the Whole Clinically Examined Population of the Vantaa 85+ Study This Study Sub-sample N = 123
Autopsied N = 300a
Clinically Studied N = 553
Age at death, yrs (mean ± SD)
90.6 ± 3.6
92.4 ± 3.7
92.8 ± 3.8
Female sex, N (%)
106 (86.2)
248 (82.7)
440 (79.6)
Dementia, N (%)
72 (58.5)
195 (65.0)
314 (52.2)
Hypertension , N (%)
34 (27.9)
74 (24.7)
142 (25.8)
Diabetesc, N (%)
33 (26.8)
65 (21.7)
113 (20.5)
b
d
APOE 4 allele carriers, N (%)
d
35 (28.4)
e
88 (31.7)
155f (25.8)
Explanatory notes: aAutopsied subjects with clinical data available; bDiagnosis based on medical records and the use of antihypertensive medication; cDiagnosis based on the use of glucose lowering medication; dAPOE genotyping performed on 104; eAPOE genotyping performed on 278; fAPOE genotyping performed on 532. Abbreviations: APOE, apolipoprotein E, MRI, magnetic resonance imaging, SD, standard deviation.
Relationships Between White Matter Hyperintensities, Cerebral Amyloid Angiopathy & Dementia
Neuropathological Analyses Presence and Severity of CAA and Comparison with WMH CAA was assessed in the right frontal and parietooccipital areas, essentially as described previously [11]. Briefly, CAA was detected in the Congo red stained sections and confirmed by immunohistochemistry (IHC) with antibodies against A at dilutions of 1:1000-3000 (clone 4G8;Senetec PLC, Maryland Heights, MO, USA and Biodesign International, Saco, ME, USA). The severity of CAA was determined as the percentage of the Congo red positive and A -immunoreactive small and medium-sized leptomeningeal and parenchymal vessels of the total blood vessels. Two raters (MT and AP) reviewed six brain regions including the right frontal, temporal, and parietal lobe, right cerebellar lobe, left occipital lobe and hippocampus, as described previously for cortical and leptomeningeal CAA [11]. We selected the frontal and parietal regions for assessment because the prevalence of CAA was highest in the parietal (57.8%) and the frontal lobes (56.5%) and these two regions showed the greatest severity (the median CAA score of 1% for each, the upper quartile (UQ) value of 6% for the frontal and 5% for the parietal lobe). Moreover, WMH results were available from the same regions for comparison. We used the combined severity index (intracortical and leptomeningeal CAA in the frontal and parietal lobe) of CAA as a numeric variable in the statistical analyses. In addition, we created three bivariate variables: (1) the presence of CAA (subjects with any severity of CAA vs. those without CAA), and (2) “severe CAA” that was defined as the severity of CAA the upper quartile (Q3) value (= 6.5%) and of the combined severity index for CAA in these two regions. This severity value was analogous to WMH rating in which the cut-off value for severe WMH (= 4) also represented the Q3 of the WMH. In addition (3), “extensive CAA” meant the severity of CAA the 90 percentile value (= 19.4%) of the combined severity index for CAA. Cortical Microscopic Infarcts and Haemorrhages The number of cortical microinfarcts (MI) was determined in 6m-thick Haematoxylin and Eosin (H&E) stained tissue sections. Cortical MIs were defined as focal lesions
