J Neurol (1992) 239 :441-450
Journal of
Neurology © Springer-Verlag 1992
Dementia in cerebral amyloid angiopathy: a clinicopathological study* M. Yoshimura l, H. Yamanouchi 3, S. Kuzuhara 2, H. Mori 1, S. Sugiura a, T. Mizutani 1, H. Shimada ~, M.Tomonaga 5, and Y. Toyokura 3 1Department of Neuropathology, Tokyo Metropolitan Institute of Gerontology, 35-2, Sakaecho, Itabashi-ku, Tokyo, 173 Japan
2Department of Neurology, Faculty of Medicine, University of Mie, Tsu, Japan 3Department of Neurology, Tokyo Metropolitan Geriatric Medical Centre, Tokyo, Japan 4Department of Pathology, Tokyo Medical College, Tokyo, Japan 5Department of Pathology, Brain Research Institute, Tokyo University, Tokyo, Japan Received July 31, 1991 / Received in revised form December 31, 1991 / Accepted January 21, 1992
Summary. Dementia is in addition to cerebral haemorrhage major sympton of cerebral amyloid angiopathy (CAa). In order to explore the pathological basis for dementia in CAa-related conditions, we made a clinicopathological analysis of CAa, with special attention to dementia. Among 150 patients (mean age 78.6 years) with autopsy-proven intracranial haemorrhage in Tokyo Metropolitan Geriatric Medical Center, CAa with cerebral haemorrhage accounted for 8.0% (12 cases), associated with hypertension and metastatic brain tumour. Among 38 patients with lobar haemorrhage, CAa represented the second most common cause (21.1%) of intracranial haemorrhage after hypertension. A total of 20 patients with CAa (mean age 82.5 years) were studies clinically and pathologically. Hypertension was present in 50%. Thirteen had a history of stroke and others had either ill-defined or no strokes. The average number of strokes 2.9. Fifteen patients (75%) had dementia. Based on the clinicopathological grounds for dementia, CAarelated conditions could be divided into three subtypes: "haemorrhagic", "dementia-haemorrhagic" and "dementia" type. Haemorrhagic type (30%, 6 cases) showed multiple recurrent lobar haemorrhages caused by CAa. Hypertension was present in only 1 patient. The incidence of senile plaques and neurofibrillary tangles was generally correlated with age. Only 1 patient had dementia. The dementia-haemorrhagic type (40%, 8 patients) had recurrent strokes with cerebral haemorrhage after preceding dementia. There were two different neuropathological subsets: CAa with atypical senile dementia of Alzheimer type (SDAT) and CAa with diffuse leucoencephalopathy. Patients with CAa with atypical SDAT had multiple cerebral haemorrhages caused by CAa combined with atypical Alzheimer-type pathology. * Presented at the 11th International Congress of Neuropathology, Kyoto 1990 Correspondence to: M. Yoshimura
Patients with CAa with diffuse leucoencephalopathy had cerebral haemorrhages in combination with diffuse white matter damage like Binswanger's subcortical vascular encephalopathy (BSVE). The incidence of senile changes correlated with age. Patients with the dementia type (30%, 6 patients) showed progressive dementia with or without haemorrhage. All had hypertension. They had a combined condition of Alzheimer-type pathology with conspicuous CAa with BSVE. Dementia in CAa-related conditions may be responsible for multiple factors including not Alzheimer-type degeneration, but also diffuse leucoencephalopathy like Binswanger's disease. We also found an asymptomatic type, an ischaemic type, a vasculitis type and an hereditary type in this condition. Key words: Cerebral amyloid angiopathy - Dementia Senile dementia of Alzheimer type - Leucoencephalopathy - ]3-protein
Introduction
Cerebral amyloid angiopathy (CAa) is a condition characterized by deposition of amyloid in the arteries and arterioles of the brain and meninges [14, 31]. It is now accepted as a cause of non-hypertensive cerebral haemorrhage in the elderly [2, 7, 8, 11, 21, 24, 28, 35]. CAa is also found in association with dementia [31], but the pathological basis for dementia in CAa has not been systematically investigated. In our study, we have examined the incidence of CAa causing cerebral haemorrhage among elderly patients with spontaneous intracranial haemorrhage and attempted to make a clinicopathological classification of CAa-related conditions in order to investigate the pos-
442 sible neuropathology of underlying dementia in this condition. Patients and methods
A total of 150 patients with cerebral haemorrhage (excluding ball haemorrhages and haemorrhagic infarctions) out of 1700 consecutive autopsies during the period 1983-1987 in the Tokyo Metropolitan Geriatric Medical Center were examined. After macroscopical observation of the coronally sectioned brains, multiple sections were taken from various parts, including the cerebral cortex, basal ganglia, thalamus, brain stem and cerebellum as well as haematomas. The neuropathological stainings of the formalin-fixed brain tissues were standard, and included Congo red, modified Bielshowsky, Mallory and van Gieson's elastic stains. The confirmation of amyloid was determined by positive Congo red stain with applegreen birefringence under polarized light. An ultrastructural demonstration of amyloid fibrils was carried out using formalin-fixed brain tissue from the occipital lobes of 20 cases with pathologically confirmed CAa-related conditions. The immunohistochemical demonstration of ~3-protein, for which antiserum was obtained containing synthetic peptide of residues 1-28 of [bprotein, was performed by the avidin-biotin peroxidase complex method [34] in 5 of 20 patients with CAa-related conditions. The patients diagnosed as having CAa were 12 of 150 patients with cerebral haemorrhage, and an additional 8 patients came to our Medical Centre after 1987 (Table 3). Cases 1-5, 8, 9, 13, 14 in Table 3 have been described by Yamanouchi et al. [35].
Results
Incidence of CAa with cerebral haernorrhage
Table 1. Causes of cerebral haemorrhage in elderly patients. DIC,
Disseminated intravascular coagulopathy Hypertensive cerebral haemorrhage Metastatic brain tumour Amyloid angiopathy Leukaemia DIC Intracranial bleeding due to subarachnoid haemorrhage Primary brain tumour Trauma Haemorrhagictendency Cerebral vascular malformation Unknown origin
Clinicopathological classification of CAa The clinicopathological findings in 20 patients with autopsy-proven CAa-related conditions are summarized in Table 3. Their ages ranged from 71 to 94 years (mean, 82.5 years). Nine patients were men and 11 were women. Hypertension was present in 10 patients (50%). Thirteen patients had a history of stroke, 5 showed uncertain frequency of strokes and 2 had no vascular accidents. The average number of stroks was 2.9. Fifteen patients (75%) had dementia. Of the demented patients 8 (cases 5, 8-11, 13, 14, 20 in Table 3) had a history of stroke with cerebral haemorrhage, while 7 (cases 7, 12, 15-19 in Table 3) had progressive dementia and mild paresis without a history of clear vascular accidents. Based on the clinicopathological analysis for dementia, 20 patients with CAa could be subdivided into three
(83 patients) (18 patients) (12 patients) (8 patients) (8 patients)
4.8% (7 patients) 2.7% (4patients) 2.0% (3 patients) 2.0% (3patients) 1.3% (2 patients) 1.3% (2 patients)
Total
150 patients
Table 2. Causes of cerebral lobar haemorrhage in elderly patients
Hypertensive haemorrhage Amyloid angiopathy Metastatic brain tumour Intracranial bleeding due to subarachnoid bleeding Leukaemia Cerebralvascular malformation Trauma Primary brain tumour Haemorrhagictendency Total
Among 150 patients (average age 78.6 years) with cerebral haemorrhage, hypertensive cerebral haemorrhage accounted for 55.3% (83 patients), metastatic brain turnout for 12.0% (18 patients) and CAa with haemorrhage for 8.0% (12 patients) (Table 1). Out of 150 patients with cerebral haemorrhage 25.3% (38 patients) showed massive lobar haemorrhage. Among these with lobar haemorrhage, CAa accounted for 21.1% (8 patients) and hypertensive haemorrhage for 36.8% (14 patients) (Table 2).
55.3% 12.0% 8.0% 5.3% 5.3%
36.8% (14 patients) 21.1% (8patients) 10.5% (4 patients) 10.5% 5.3% 5.3% 5.3% 2.6% 2.6%
(4 patients) (2 patients) (2patients) (2 patients) (1patient) (1 patient) 38 patients
subtypes: "haemorrhagic type", "dementia-haemorrhagic type" and "dementia type". The haemorrhagic type (case 1-6 in Table 3, 30%) showed recurrent lobar haemorrhage without hypertension, except for case 6. The number of cerebral haemorrhages ranged from 1 to 8. Dementia was found in only I patient (case 5). The average weight of the brains was 1300 g. Three patients had a single lobar haemorrhage and another had multiple lobar haemorrhages (Fig. 1A). The areas of cerebral haemorrhages in all cases in Table 3 were occipital (39%), temporal (28%), frontal (18%) and parietal (15%). Case 5 had cerebral haemorrhages in both occipital lobes, involving the visual cortex. All patients had subarachnoid bleeding over the affected lobes in varying degree with scattered cerebrocortical necrosis. The CAa-associated vasculopathies (CAa-V) [16] were widely distributed in the meningocortical regions both in the cerebrum (Fig. 1B) and cerebellum, showing a similar distribution pattern to that of cerebral haemorrhage. A few CAa-V were found in the hipocampus and the motor and visual cortices, but none in the basal ganglia, thalamus or brain stem. Among the CAa-V, there were many medium-tosmall arteries with segmental amyloid deposition in the outer layer of the tunica media and adventitia. Some
443
I
c) c'-I
+1
+l
~
+1
~
,.~
,.~ ~
~
+1 ~
.~
8 +1
+
+
+1
+
+1
+
+
+ + +
+
o -0
0
~+~
O0
ddHq
.~-]-
+
+ +._~ ++
Journal of
Neurology © Springer-Verlag 1992
Dementia in cerebral amyloid angiopathy: a clinicopathological study* M. Yoshimura l, H. Yamanouchi 3, S. Kuzuhara 2, H. Mori 1, S. Sugiura a, T. Mizutani 1, H. Shimada ~, M.Tomonaga 5, and Y. Toyokura 3 1Department of Neuropathology, Tokyo Metropolitan Institute of Gerontology, 35-2, Sakaecho, Itabashi-ku, Tokyo, 173 Japan
2Department of Neurology, Faculty of Medicine, University of Mie, Tsu, Japan 3Department of Neurology, Tokyo Metropolitan Geriatric Medical Centre, Tokyo, Japan 4Department of Pathology, Tokyo Medical College, Tokyo, Japan 5Department of Pathology, Brain Research Institute, Tokyo University, Tokyo, Japan Received July 31, 1991 / Received in revised form December 31, 1991 / Accepted January 21, 1992
Summary. Dementia is in addition to cerebral haemorrhage major sympton of cerebral amyloid angiopathy (CAa). In order to explore the pathological basis for dementia in CAa-related conditions, we made a clinicopathological analysis of CAa, with special attention to dementia. Among 150 patients (mean age 78.6 years) with autopsy-proven intracranial haemorrhage in Tokyo Metropolitan Geriatric Medical Center, CAa with cerebral haemorrhage accounted for 8.0% (12 cases), associated with hypertension and metastatic brain tumour. Among 38 patients with lobar haemorrhage, CAa represented the second most common cause (21.1%) of intracranial haemorrhage after hypertension. A total of 20 patients with CAa (mean age 82.5 years) were studies clinically and pathologically. Hypertension was present in 50%. Thirteen had a history of stroke and others had either ill-defined or no strokes. The average number of strokes 2.9. Fifteen patients (75%) had dementia. Based on the clinicopathological grounds for dementia, CAarelated conditions could be divided into three subtypes: "haemorrhagic", "dementia-haemorrhagic" and "dementia" type. Haemorrhagic type (30%, 6 cases) showed multiple recurrent lobar haemorrhages caused by CAa. Hypertension was present in only 1 patient. The incidence of senile plaques and neurofibrillary tangles was generally correlated with age. Only 1 patient had dementia. The dementia-haemorrhagic type (40%, 8 patients) had recurrent strokes with cerebral haemorrhage after preceding dementia. There were two different neuropathological subsets: CAa with atypical senile dementia of Alzheimer type (SDAT) and CAa with diffuse leucoencephalopathy. Patients with CAa with atypical SDAT had multiple cerebral haemorrhages caused by CAa combined with atypical Alzheimer-type pathology. * Presented at the 11th International Congress of Neuropathology, Kyoto 1990 Correspondence to: M. Yoshimura
Patients with CAa with diffuse leucoencephalopathy had cerebral haemorrhages in combination with diffuse white matter damage like Binswanger's subcortical vascular encephalopathy (BSVE). The incidence of senile changes correlated with age. Patients with the dementia type (30%, 6 patients) showed progressive dementia with or without haemorrhage. All had hypertension. They had a combined condition of Alzheimer-type pathology with conspicuous CAa with BSVE. Dementia in CAa-related conditions may be responsible for multiple factors including not Alzheimer-type degeneration, but also diffuse leucoencephalopathy like Binswanger's disease. We also found an asymptomatic type, an ischaemic type, a vasculitis type and an hereditary type in this condition. Key words: Cerebral amyloid angiopathy - Dementia Senile dementia of Alzheimer type - Leucoencephalopathy - ]3-protein
Introduction
Cerebral amyloid angiopathy (CAa) is a condition characterized by deposition of amyloid in the arteries and arterioles of the brain and meninges [14, 31]. It is now accepted as a cause of non-hypertensive cerebral haemorrhage in the elderly [2, 7, 8, 11, 21, 24, 28, 35]. CAa is also found in association with dementia [31], but the pathological basis for dementia in CAa has not been systematically investigated. In our study, we have examined the incidence of CAa causing cerebral haemorrhage among elderly patients with spontaneous intracranial haemorrhage and attempted to make a clinicopathological classification of CAa-related conditions in order to investigate the pos-
442 sible neuropathology of underlying dementia in this condition. Patients and methods
A total of 150 patients with cerebral haemorrhage (excluding ball haemorrhages and haemorrhagic infarctions) out of 1700 consecutive autopsies during the period 1983-1987 in the Tokyo Metropolitan Geriatric Medical Center were examined. After macroscopical observation of the coronally sectioned brains, multiple sections were taken from various parts, including the cerebral cortex, basal ganglia, thalamus, brain stem and cerebellum as well as haematomas. The neuropathological stainings of the formalin-fixed brain tissues were standard, and included Congo red, modified Bielshowsky, Mallory and van Gieson's elastic stains. The confirmation of amyloid was determined by positive Congo red stain with applegreen birefringence under polarized light. An ultrastructural demonstration of amyloid fibrils was carried out using formalin-fixed brain tissue from the occipital lobes of 20 cases with pathologically confirmed CAa-related conditions. The immunohistochemical demonstration of ~3-protein, for which antiserum was obtained containing synthetic peptide of residues 1-28 of [bprotein, was performed by the avidin-biotin peroxidase complex method [34] in 5 of 20 patients with CAa-related conditions. The patients diagnosed as having CAa were 12 of 150 patients with cerebral haemorrhage, and an additional 8 patients came to our Medical Centre after 1987 (Table 3). Cases 1-5, 8, 9, 13, 14 in Table 3 have been described by Yamanouchi et al. [35].
Results
Incidence of CAa with cerebral haernorrhage
Table 1. Causes of cerebral haemorrhage in elderly patients. DIC,
Disseminated intravascular coagulopathy Hypertensive cerebral haemorrhage Metastatic brain tumour Amyloid angiopathy Leukaemia DIC Intracranial bleeding due to subarachnoid haemorrhage Primary brain tumour Trauma Haemorrhagictendency Cerebral vascular malformation Unknown origin
Clinicopathological classification of CAa The clinicopathological findings in 20 patients with autopsy-proven CAa-related conditions are summarized in Table 3. Their ages ranged from 71 to 94 years (mean, 82.5 years). Nine patients were men and 11 were women. Hypertension was present in 10 patients (50%). Thirteen patients had a history of stroke, 5 showed uncertain frequency of strokes and 2 had no vascular accidents. The average number of stroks was 2.9. Fifteen patients (75%) had dementia. Of the demented patients 8 (cases 5, 8-11, 13, 14, 20 in Table 3) had a history of stroke with cerebral haemorrhage, while 7 (cases 7, 12, 15-19 in Table 3) had progressive dementia and mild paresis without a history of clear vascular accidents. Based on the clinicopathological analysis for dementia, 20 patients with CAa could be subdivided into three
(83 patients) (18 patients) (12 patients) (8 patients) (8 patients)
4.8% (7 patients) 2.7% (4patients) 2.0% (3 patients) 2.0% (3patients) 1.3% (2 patients) 1.3% (2 patients)
Total
150 patients
Table 2. Causes of cerebral lobar haemorrhage in elderly patients
Hypertensive haemorrhage Amyloid angiopathy Metastatic brain tumour Intracranial bleeding due to subarachnoid bleeding Leukaemia Cerebralvascular malformation Trauma Primary brain tumour Haemorrhagictendency Total
Among 150 patients (average age 78.6 years) with cerebral haemorrhage, hypertensive cerebral haemorrhage accounted for 55.3% (83 patients), metastatic brain turnout for 12.0% (18 patients) and CAa with haemorrhage for 8.0% (12 patients) (Table 1). Out of 150 patients with cerebral haemorrhage 25.3% (38 patients) showed massive lobar haemorrhage. Among these with lobar haemorrhage, CAa accounted for 21.1% (8 patients) and hypertensive haemorrhage for 36.8% (14 patients) (Table 2).
55.3% 12.0% 8.0% 5.3% 5.3%
36.8% (14 patients) 21.1% (8patients) 10.5% (4 patients) 10.5% 5.3% 5.3% 5.3% 2.6% 2.6%
(4 patients) (2 patients) (2patients) (2 patients) (1patient) (1 patient) 38 patients
subtypes: "haemorrhagic type", "dementia-haemorrhagic type" and "dementia type". The haemorrhagic type (case 1-6 in Table 3, 30%) showed recurrent lobar haemorrhage without hypertension, except for case 6. The number of cerebral haemorrhages ranged from 1 to 8. Dementia was found in only I patient (case 5). The average weight of the brains was 1300 g. Three patients had a single lobar haemorrhage and another had multiple lobar haemorrhages (Fig. 1A). The areas of cerebral haemorrhages in all cases in Table 3 were occipital (39%), temporal (28%), frontal (18%) and parietal (15%). Case 5 had cerebral haemorrhages in both occipital lobes, involving the visual cortex. All patients had subarachnoid bleeding over the affected lobes in varying degree with scattered cerebrocortical necrosis. The CAa-associated vasculopathies (CAa-V) [16] were widely distributed in the meningocortical regions both in the cerebrum (Fig. 1B) and cerebellum, showing a similar distribution pattern to that of cerebral haemorrhage. A few CAa-V were found in the hipocampus and the motor and visual cortices, but none in the basal ganglia, thalamus or brain stem. Among the CAa-V, there were many medium-tosmall arteries with segmental amyloid deposition in the outer layer of the tunica media and adventitia. Some
443
I
c) c'-I
+1
+l
~
+1
~
,.~
,.~ ~
~
+1 ~
.~
8 +1
+
+
+1
+
+1
+
+
+ + +
+
o -0
0
~+~
O0
ddHq
.~-]-
+
+ +._~ ++
