The Significance of Small Cerebral Bleeds in Neurodegenerative Dementia Syndromes

Abstract

Small cerebral bleeds are frequently observed in brains of patients with Alzheimer disease (AD) and cerebral amyloid angiopathy (CAA). However, they are also observed in patients with other neurodegenerative dementias and in persons without cognitive impairment. The aim of this survey is to compare the bleeding load in brains with different dementia syndromes and in age-matched controls. Hundred sixty-five brains were examined. The prevalence and the severity of the different cerebrovascular lesions were examined. Quantification of the number of mini-bleeds allowed to determinate the bleeding load in different cerebral regions. Micro-bleeds were considered as small macroscopically visible lesions while mini-bleeds were defined as small perivascular accumulations of red blood cells or siderophages only visible on microscopic examination. Several types of cerebrovascular lesions prevailed in AD brains with CAA, compared to the controls. White matter changes prevailed in frontotemporal lobar degeneration. Mini-bleeds were significantly more frequent in the cerebral cortex of AD and Lewy body dementia brains. They also prevailed around the dentate nucleus of the cerebellum and in the tegmentum pontis of patients with progressive supranuclear palsy. On the other hand the bleeding load in frontotemporal lobar degeneration and in corticobasal degeneration was similar to that in age-matched control brains. Cerebrovascular lesions, including micro-bleeds, predominated in AD brains with CAA. Mini-bleeds, on the other hand, were more related to the neurodegenerative process itself and reflected associated disruption of the blood-brain barrier.

Small cerebral bleeds are increasingly recognized on T2*-weighted gradient-echo magnetic resonance imaging of patients with cerebrovascular disease and dementia [1]. They are mainly related to cerebral micro-angiopathy [2]. Cerebral amyloid angiopathy (CAA) and lipohyalinosis due to arterial hypertension are the main causes [3]. The topographic distribution of these small bleeds is quite different: in CAA they are more frequent in the cortico-subcortical regions, whilst mainly present in the central grey matter in patients with advanced arterial hypertension [4]. CAA is frequently associated to Alzheimer dementia (AD) [5]. Small bleeds are, however, also frequently associated with the clinical manifestation and biochemical hallmarks of AD without associated CAA, implying their possible involvement in the pathogenesis of AD [6]. The impact of small cerebral bleeds on the cognitive decline in AD patients is unknown [7]. Small bleeds are also observed in brains of demented patients with other neurodegenerative diseases and in elderly persons without cognitive disturbances [8].

The present survey compares the frequency, the impact and the distribution of small cerebral bleeds, infarcts and white matter changes in post-mortem brains of patients with different neurodegenerative dementia syndromes and age-matched controls. Complete clinical data from the Memory Clinic of Lille were available in 65% of the patients.

PATIENTS AND METHODS

The brain tissue samples were first used for diagnosis and afterward integrated in the Lille Neuro-Bank, dependant from the Lille University and co-federated by the “Centre des Resources Biologiques”, acting as institutional review board.

Table 1 labels the number of examined brains, according to the neuropathological diagnosis. Hundred sixty-five brains were examined according to the classical histological criteria. The post-mortem diagnosis of AD was made according to stages V and VI of the classification of Braak and Braak [9], of Lewy body dementia according to the McKeith and Kosaka criteria [10, 11], of frontotemporal lobar degeneration (FTLD) according to the Consortium for Frontotemporal Lobar Degeneration [12], of progressive supranuclear palsy (PSP) according to Sakamoto et al. [13] and of corticobasal degeneration (CBD) according to Lantos [14]. The CERAD criteria were used for the evaluation of the severity of CAA [15].

Table 1.

Neuropathological diagnosis and number of brains with neurodegenerative dementias and age-matched controls.

Neuropathological Diagnosis	Number of brains
Alzheimer Disease (AD)	90
Lewy Body Disease (LBD)	20
Frontotemporal Lobar Degeneration (FTLD)	22
Progressive Supranuclear Palsy (PSP)	12
Corticobasal Degeneration (CBD)	6
Age-matched controls (C)	15

A large number of brains had overlapping pathology (Table 2).

Table 2.

Number of brains with overlapping pathology (percentage between brackets)

Neuropathological Diagnosis	Number (%)
Alzheimer disease with severe cerebral amyloid angiopathy	40 (44%)
Alzheimer disease with Lewy body features	14 (16%)
Lewy body disease with Alzheimer features and amyloid angiopathy	6 (30%)
Frontotemporal lobar degeneration with mild Alzheimer features	4 (18%)
Progressive supranuclear palsy with mild Alzheimer features	1 (8%)

The prevalence and the severity of CAA, white matter changes, lacunar infarcts, territorial infarcts, cortical micro-infarcts, cerebral hematomas, micro- and mini-bleeds were examined. Micro-bleeds were considered as small, 1–3 millimeter large, macroscopically visible lesions (Fig.1A). Mini-bleeds were not visible on naked eye examination but only detected around small arteries and capillaries on microscopic examination of the brains (Fig. 2B). The prevalence of the different cerebrovascular lesions was defined as the percentage of brains affected, while the severity was determined on a semi-quantitative scale, ranking from 0 to 3, according to a previous described method [16]. The white matter changes were restricted to the periventricular regions (R1), scattered in the centrum semiovale (R2) or forming confluent lesions (R3). For the micro-infarcts and micro- and mini-bleeds R1 corresponded to 1 to 2, R2 up to 5 and R3 to more than 5 lesions. Also the regional distribution of the mini-bleeds was evaluated according to their location in the cerebral cortex and cortical-subcortical junction, the deep white matter, the striatum, the thalamus, the pons and the cerebellar hemispheres. They were far more frequently observed than micro-bleeds and allowed to determine the “bleeding load” in a particular brain [17].

Figure 1.

Small cerebral bleeds in a patient with Alzheimer disease and cerebral amyloid angiopathy. (A) Cortical micro-bleed (arrow). (B) Cortical mini-bleed consisting of perivascular red blood cells on microscopic examination after staining with haematoxylin-eosin.

The statistical analysis compared separately the items of the different groups of neurodegenerative dementia syndromes with the control group. Univariate comparisons of unpaired groups were done with the Fisher’s exact test for categorical data. The non-parametric Mann-Whitney U-test was used to compare continuous variables. The significance level was set at 0.05, two-tailed.

RESULTS

Although none of the age-matched controls had a history of stroke or of significant cognitive impairment the overall incidence of cerebral white matter changes was 10%. Micro-infarcts were present in 10% and mini-bleeds in 50%.

The vascular risk factors such as arterial hypertension, hypercholesterolemia, diabetes, and smoking were low and similar to those observed in the dementia patients. The use of antithrombotic agents was 20% and also similar to those with a neurodegenerative disease.

Cerebrovascular lesions were significantly prevalent in AD brains with CAA compared to AD brains without CAA: respectively 45% compared to13% for territorial infarcts and 45% compared to 10% for lobar hematomas.

In LBD brains only mini-bleeds were more prevalent with 80% compared to 50% in age-matched controls. Also from all cerebrovascular lesions only brains with mini-bleeds were more frequent as a whole with a significant predominance in the cerebral cortex.

Cerebrovascular lesions were rare in brains with FTLD. Only white matter changes were prevalent in FTLD (44%), compared to age-matched controls (10%).

The presence of cerebrovascular lesions was very low in PSP brains and comparable to the age-matched control brains.

In CBD brains no cerebrovascular lesions were observed except for mild white matter changes and mini-bleeds. The prevalence of white matter changes in CBD was 32% compared to 10% in the controls. The prevalence of mini-bleeds was 32% compared to 50% in the age-matched controls. These differences were statistically not significant.

The severity scores (R 0–3) of the different cerebrovascular lesions in the brains with neurodegenerative dementia syndromes compared to the controls are labeled in table 3.

Table 3:

Average severity scores of the cerebrovascular lesions, ranking from 0 to 3. Comparison between the different neurodegenerative diseases and the controls.

Items	C	AD	AD+CAA	LBD	FTLD	CBD	PSP
	n=15	n =50	n=40	n=20	n=22	n=6	n=12
White matter changes	0.2	0.7	1.7*	0.8	0.9*	0.8	0.6
Lacunar infarcts	0.1	0.3	0.2	0.2	0.1	0.0	0.1
Territorial infarcts	0.0	0.1	0.7*	0.1	0.0	0.0	0.1
Micro-infarcts	0.1	0.1	0.7*	0.2	0.1	0.0	0.1
Hematomas	0.0	0.1	0.5*	0.0	0.0	0.0	0.1
Micro-bleeds	0.0	0.1	0.5*	0.0	0.1	0.0	0.0
Mini-bleeds
Total	0.5	1.3*	2.1*	1.5*	0.9	0.4	1.3
Cortical	0.3	0.6	1.3*	0.9*	0.8	0.8	0.6
White matter	0.5	0.6	1.1	0.6	0.5	0.8	0.8
Striatum	0.1	0.2	0 .4	0.4	0.1	0.0	0.3
Thalamus	0.1	0.0	0.1	0.2	0.1	0.0	0.1
Brainstem	0.0	0.0	0.1	0.4	0.1	0.2	0.5*
Cerebellum	0.2	0.4	1.0	0.9	0.2	0.0	1.1*

Significance level at equal or more than 0.05 two-tailed is indicated by^*.

The severity of white matter changes and the amount of territorial infarcts, micro-infarcts, and micro- and mini-bleeds was significantly increased in the AD brains with CAA compared to those without CAA and controls. Mini-bleeds predominated in the cerebral cortex of ADCAA brains.

In LBD brains only mini-bleeds were more frequent as a whole with a significant predominance in the cerebral cortex.

The total amount of mini-bleeds was similar in the FTLD and in the control brains.

Although the global bleeding load was the same as in the controls, mini-bleeds were more frequently observed in the cerebellum around the dentate nucleus and in the tegmentum of the pons of PSP brains (P = 0.02).

The severity of the white matter changes and the bleeding load were similar in the CBD brains and the controls.

DISCUSSION

The prevalence and the severity of cerebrovascular lesions are different according to the type of neurodegenerative dementia syndrome. The prevalence of hemorrhagic as well as of ischemic lesions is particularly high in AD, although mainly linked to the prominent association with CAA rather than to the severity of the AD features [5]. Small bleeds in the cerebral cortex and white matter occur preferentially in local regions of concentrated amyloid deposits in leptomeningeal and cortical vessels [18]. Our previous post-mortem 7.0-Tesla magnetic resonance study has shown differences in the distribution of micro-bleeds in AD and AD-CAA brains. In AD the micro-bleeds are mainly observed in the coronal section at the level of the mamillary body, while in AD-CAA they are more widespread and present as well in the frontal, central as in the occipital section [19]. However, 7.0-Tesla T2*-magnetic resonance imaging (MRI) shows the neuropathologically detected micro-bleeds as well as mini-bleeds [20].

For the other neurodegenerative dementia syndromes only the bleeding load, determined by the number of mini-bleeds, shows differences compared to age-matched controls. The other cerebrovascular lesions are rather scarce and do not seem to have an impact on the disease process in contrast with AD.

The bleeding load is increased in LBD, due to the high incidence of mini-bleeds in the cerebral cortex [16].

Although not overall increased, mini-bleeds predominate around the dentate nucleus of the cerebellum and the tegmentum of the pons in PSP (not yet published result). As the examined section for the quantification was through the pons, it was not evaluated whether mini-bleeds were frequent at the level of the mesencephalon, where the most prominent neuronal changes are observed [21].

FTLD and CBD share a low bleeding load with control brains [22]. However our recent post-mortem 7.0-Tesla magnetic resonance study has shown that micro-bleeds are prevalent in the deep cortical layers of the frontal cortex, where the most prominent degenerative changes occur [23].

While small cerebral bleeds predominate in the cortico-subcortical structures in AD and LBD dementia [5,16,24], it is remarkable that none of them are observed in the cerebellar cortex of the brains with different neurodegenerative dementia syndromes. This can perhaps be explained by the difference in arterial angioarchitecture between the cerebral and the cerebellar cortex [25].

Micro-bleeds appear to be related to rupture of small arteries and arterioles, while mini-bleeds are more related to the type of neurodegenerative disease. There is a topographic relation between the involved structures in some degenerative diseases, such as in AD, LBD, FTLD and PSP, and the location of mini-bleeds. These findings indicate disturbances of the blood-brain barrier in these neurodegenerative dementia syndromes rather than the expression of associated cerebrovascular diseases. In FTLD and CBD these disturbances are not more prominent than in the normal ageing brain.