(Stroke. 1999;30:1643-1646.)
© 1999 American Heart Association, Inc.
Original Contributions |
2 With Putative Clinical Risk Factors
From the Departments of Neuropathology (M.O.M., J.A.R.N.) and Neurology (M.O.M., I.B.), Institute of Neurological Sciences, Southern General Hospital, Glasgow; Department of Neuropathology (J.W.I.), The University of Edinburgh, Western General Hospital, Edinburgh, UK; Department of Neuropathology (S.L.), Frenchay Hospital, Bristol, UK; and Department of Medicine, Division of Neurology (M.J.A.), Duke University Medical Center, Durham, NC.
Correspondence to Dr Mark McCarron, Department of Neuropathology, Institute of Neurological Sciences, Southern General Hospital, Glasgow G51 4TF, Scotland, UK. E-mail mmc18f{at}clinmed.gla.ac.uk
| Abstract |
|---|
|
|
|---|
4 allele predisposes to cerebral amyloid angiopathy
(CAA) whereas
2 is associated with CAA-related hemorrhage
(CAAH). The clinical risk factors for other forms of intracranial
hemorrhage are a less-frequent feature of CAAH. In this study
we examined potential clinical risk factors in patients with CAAH and
assessed these with respect to APOE genotype. MethodsThirty-six patients were identified with a pathological diagnosis of CAAH. Clinical notes were reviewed to document age of hemorrhage onset, history of dementia, antiplatelet/anticoagulant medication, hypertension, minor head trauma, or transient neurological events. In a review of reported cases of CAAH, the frequency of these clinical features was also recorded. APOE genotypes were determined with use of polymerase chain reaction techniques.
ResultsThere were 24 women and 12 men; the mean age was 70.3
years. One third (n=12) had been taking antiplatelet medication,
and a similar number were demented. Nine patients were hypertensive,
and 4 had a history of recent minor head trauma. The relative frequency
of each of these clinical features was similar to that in previous
reports. Forty-four percent (16 of 36) possessed an
2 allele.
Antiplatelet or anticoagulant medication, hypertension, or minor
head trauma were significantly more frequent antecedents of CAAH in
2 carriers than in non
2 carriers (81% versus 35%,
P=0.008), antiplatelet/anticoagulant medication in
particular (P=0.038).
ConclusionsOur findings suggest that antiplatelet or
anticoagulant medication, hypertension, or minor head trauma are most
likely to precipitate cerebral hemorrhage in patients with CAA
who are also
2 carriers. This may result from isoform-specific
effects of apoE on the structure of amyloid-laden blood vessel walls.
Key Words: apolipoproteins cerebral amyloid angiopathy intracerebral hemorrhage
| Introduction |
|---|
|
|
|---|
Recent evidence implicates the apolipoprotein E gene (APOE) in
the etiology of CAAH.2 3 We previously hypothesized that
whereas the APOE
4 allele increases Aß deposition in the
cerebral vasculature,4 APOE
2 is associated with
rupture of Aß-laden blood vessels, possibly by predisposing to the
development of recognized vasculopathic complications of
CAA.2 The clinical risk profile in CAAH has not, however,
been firmly established.5 Although previous reports of
CAAH have included patients who were taking antiplatelet or
anticoagulant medication,6 7 8 had suffered minor head
trauma9 or had hypertension,10 it is unclear
whether these clinical features are risk factors for CAAH that are
independent of the effects of APOE genotype. In the present
study we have established the frequency of these putative clinical risk
factors in a large series of patients with a pathological diagnosis of
CAAH and compared our findings with those in previous reports. In
addition, we have examined the relationship between these clinical risk
factors and APOE genotype.
| Methods |
|---|
|
|
|---|
When a full postmortem examination demonstrated lobar, cortical, or corticosubcortical cerebral hemorrhage in the presence of severe CAA without other diagnostic lesions, definite CAAH was diagnosed.3 Probable CAAH with supporting pathological evidence was diagnosed when an evacuated lobar hematoma or cortical biopsy showed at least mild CAA and no other lesions.3 Of the 36 patients, 27 had a diagnosis of definite CAAH (full autopsy) and 9 had a diagnosis of probable CAAH with supporting pathology (biopsy material).
Clinical histories were reviewed, including age at hemorrhage onset, number of intracerebral hemorrhages, evidence of minor head trauma, antiplatelet and anticoagulant medications, and a clinical history of hypertension preceding the first CAAH. Dementia, defined as evidence of progressive cognitive decline, was also noted. The frequency of dementia and the putative clinical risk factors was similarly recorded from a literature search using MEDLINE (Ovid and PubMed) databases as well as references from these reports between 1977 and January 1999 in pathologically proved cases of CAAH. Clinical profiles were established only from those reports that explicitly described the absence or presence of these features.
APOE genotypes (28 already established2 ) were determined from either blood or paraffin-embedded tissue by use of the polymerase chain reaction (PCR), as previously described.2 The PCR products were digested with HhaI and separated on a polyacrylamide gel. The clinical data and APOE genotypes were analyzed with the Mann Whitney and Fisher exact tests.
| Results |
|---|
|
|
|---|
|
One third of the patients (n=12) were taking regular antiplatelet medication (aspirin in all cases) for secondary prevention of coronary heart disease and ischemic stroke; 2 of these 12 were also taking anticoagulant medication. One of the patients on anticoagulant medication had had a deep venous thrombosis and the other had recurrent transient neurological deficits thought to be associated with a 40% stenosis of the internal carotid artery. Because of continuing transient deficits, warfarin sodium (Coumadin, Du Pont) was prescribed, and while therapeutically anticoagulated the patient developed a right temporal lobe hematoma.
The reported cases of CAAH in the literature had clinical
profiles very similar to those of our series of patients, with no
statistically significant differences between the 2 groups (Table 1
).
In our group, 44% of the patients (n=16) had 1 or more APOE
2
alleles. The presence of 1 or more of the 3 putative risk factors
(antiplatelet/anticoagulant medication, hypertension, and minor
head trauma) was significantly more frequent in
2 carriers compared
with non
2 carriers (P=0.008). When analyzed
individually all of these features were more common in
2 allele
carriers (Table 2
), but only
antiplatelet/anticoagulant medication was significantly associated
with patients possessing the
2 allele compared with those
without
2: 9 of 16
2 carriers (56%) versus 4 of 20 non
2
carriers (20%; P=0.038). No such associations were found
among APOE
4 carriers. Patients carrying an
2 allele had
their first documented hemorrhage at an earlier age than did
patients without the
2 allele, although the difference was not
statistically significant (P=0.088). In APOE
4 carriers
the age of hemorrhage onset was similar to that in non
4
carriers (median age, 72.0 versus 69.0 years; P=0.877).
Clinical evidence of dementia was not significantly more frequent
in
2 carriers compared with non
2 carriers (44% versus
25%) or in
4 carriers compared with non
4 carriers (38% versus
36%).
|
| Discussion |
|---|
|
|
|---|
4 and
2 alleles have been reported to be
overrepresented in CAAH.2 3 12 It has been
proposed that whereas age and the
4 allele increase Aß
deposition in the cerebral vasculature,4 the APOE
2
allele promotes subsequent hemorrhage.2 13 Our
current findings suggest that patients who have CAA and are exposed to
antiplatelet or anticoagulant medication, hypertension, or minor
head trauma are at a higher risk of intracerebral
hemorrhage if they are
2 carriers.
There is accumulating evidence implicating the
2 allele in CAAH.
The allele occurs at a higher frequency in patients with CAAH than
in control subjects,2 12 and CAAH has been documented to
occur at a significantly younger age in
2 carriers,2 13
although we did not replicate this finding in the current study. Some
of the recognized vascular pathological complications of CAA have also
been associated with the APOE
2 allele. Greenberg et
al13 found an elevated
2 frequency in the brains of CAA
patients demonstrating a combination of vessel wall concentric
splitting (a "vessel-within-a-vessel" appearance) and evidence of
paravascular bleeding. In a systematic analysis of the vascular
complications of CAA (vessel stenosis, vessel dilatation, or
microaneurysm formation, fibrinoid necrosis, vasculitis, a
vessel-within-a-vessel appearance, and evidence of previous microscopic
hemorrhage1 14 15 ), we have recently found a
statistically significant excess of fibrinoid necrosis in patients
possessing the
2 allele compared with non
2
carriers.16 Further work is required to clarify the
underlying mechanism associated with apoE-induced vascular changes in
CAA, but whatever the mechanism, the present study suggests that
CAA patients exposed to clinical risk factors such as
antiplatelet/anticoagulant medication, hypertension, and minor head
trauma may be most at risk of lobar hemorrhage if they also
have an
2 allele. Although our present findings form the
basis of a testable hypothesis that
2 carriers with CAA may have a
differential clinical risk profile for cerebral hemorrhage than
non
2 carriers, the results probably do not justify revision of the
indications for antiplatelet and anticoagulant medications, which
have been shown to benefit patients over and above the risk of cerebral
hemorrhage.17 18 Caution should, however, be
exercised in prescribing these medications to elderly patients with
transient neurological deficits in the absence of significant
carotid stenosis, because CAA can imitate transient
ischemic attacks. This phenomenon, which may be due to focal
seizures secondary to petechial hemorrhages, is a recognized
feature of CAA,6 8 19 and some reports have implicated
aspirin7 or anticoagulant medication6 8 as
precipitants of CAAH in this setting, a feature documented in 1 of our
patients. Intriguingly, this patient had an APOE
2/
2
genotype.
Our series and the review of pathologically confirmed cases of
CAAH reported in the literature demonstrate similar clinical profiles.
The existence of a subgroup of patients with neither the APOE
2
allele nor any putative clinical risk factor (13 of 36 patients)
suggests that other (as-yet unidentified) factors may also predispose
to rupture of Aß-laden blood vessels. The possession of an APOE
4
allele has previously been implicated in CAAH4 but did
not account for the remaining hemorrhages in our series. In
addition, none of our patients had a family history of CAA-related
hemorrhage, and amyloid precursor protein mutations have never
been found in sporadic cases of CAA-related
hemorrhage.20 21 22 23 Although there is 1 published
case of "sporadic" CAA-related hemorrhage due to a cystatin
C mutation,21 we have not identified this mutation in our
series.24
There are several potential limitations to our study. Although
systematic searches ensured high ascertainment of pathological cases of
CAAH (and hence accurate diagnoses), the study's reliance on autopsy
cases (n=27) inevitably introduces selection bias. However, many of the
patients had multiple hemorrhages, and the clinical details
were confined to the first recorded intracerebral
hemorrhage. Although the retrospective nature of the study
limited the amount of clinical data, hospital admission helped in
accurate data acquisition, which was often confirmed by family members
and primary care physicians. Despite the relatively large size of this
group of pathologically confirmed CAAH patients, the small size of
subgroups with specific clinical features may have limited the
sensitivity of some of the analyses. Our findings are
preliminary and need to be interpreted cautiously, particularly those
in the small subgroups with hypertension and minor head trauma, which
individually were not significantly associated with possession of the
2 allele. Larger studies, ideally also with pathologically
verified diagnoses, are required to confirm whether hypertension and
minor head trauma are significant risk factors for CAA-related
hemorrhage in
2 carriers.
In conclusion, clinical factors associated with other forms of
intracranial hemorrhage (antiplatelet/anticoagulant
medication, hypertension, and minor head trauma) appear to be more
significant risk factors in APOE
2 carriers with CAA than in similar
patients without the APOE
2 allele. This may result from an apoE
isoform-specific effect (
2) on amyloid-laden blood vessels that
renders the vessels more vulnerable to rupture in the presence of these
clinical factors.
| Acknowledgments |
|---|
Received March 25, 1999; revision received April 22, 1999; accepted April 22, 1999.
| References |
|---|
|
|
|---|
2
allele in hemorrhage due to cerebral amyloid angiopathy.
Ann Neurol. 1997;41:716721.[Medline]
[Order article via Infotrieve]
4 and cerebral hemorrhage
associated with amyloid angiopathy. Ann Neurol. 1995;38:254259.[Medline]
[Order article via Infotrieve]
4 is associated
with the presence and earlier onset of hemorrhage in cerebral
amyloid angiopathy. Stroke. 1996;27:13331337.
2 and vasculopathy in cerebral amyloid
angiopathy. Neurology. 1998;50:961965.
2 allele and
the pathological features in cerebral amyloid angiopathy-related
haemorrhage. J Neuropathol Exp Neurol. In
press.
This article has been cited by other articles:
![]() |
C. Tzourio, H. Arima, S. Harrap, C. Anderson, O. Godin, M. Woodward, B. Neal, M-G Bousser, J. Chalmers, F. Cambien, et al. APOE genotype, ethnicity, and the risk of cerebral hemorrhage Neurology, April 15, 2008; 70(16): 1322 - 1328. [Abstract] [Full Text] [PDF] |
||||
![]() |
H C Hanger, T J Wilkinson, N Fayez-Iskander, and R Sainsbury The risk of recurrent stroke after intracerebral haemorrhage J. Neurol. Neurosurg. Psychiatry, August 1, 2007; 78(8): 836 - 840. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Rosand Editorial Comment--Epistasis Is Coming: Are We Ready? Stroke, September 1, 2005; 36(9): 1879 - 1880. [Full Text] [PDF] |
||||
![]() |
P D Leclercq, L S Murray, C Smith, D I Graham, J A R Nicoll, and S M Gentleman Cerebral amyloid angiopathy in traumatic brain injury: association with apolipoprotein E genotype J. Neurol. Neurosurg. Psychiatry, February 1, 2005; 76(2): 229 - 233. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. A. Walker, D. F. Broderick, A. L. Kotsenas, and F. A. Rubino Routine Use of Gradient-Echo MRI to Screen for Cerebral Amyloid Angiopathy in Elderly Patients Am. J. Roentgenol., June 1, 2004; 182(6): 1547 - 1550. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Rosand, E. M. Hylek, H. C. O'Donnell, and S. M. Greenberg Warfarin-associated hemorrhage and cerebral amyloid angiopathy: A genetic and pathologic study Neurology, October 10, 2000; 55(7): 947 - 951. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. L. Sacco Lobar Intracerebral Hemorrhage N. Engl. J. Med., January 27, 2000; 342(4): 276 - 279. [Full Text] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Stroke Home | Subscriptions | Archives | Feedback | Authors | Help | AHA Journals Home | Search Copyright © 1999 American Heart Association, Inc. All rights reserved. Unauthorized use prohibited. |